Plasma-assisted catalytic reforming of toluene to hydrogen rich syngas

Article abstract of DOI:10.1039/c7cy00970d

The combination of non-thermal plasma (NTP) and transition metal catalysts is an alternative method for reducing tar derived from biomass gasification, converting it to a mix of mainly H₂ and CO (syngas). In this study, Ni- and Fe-based steam reforming catalysts supported on CaO, SiO₂, γ-Al₂O₃ and ZSM-5 were combined with a coaxial dielectric barrier discharge (DBD) plasma reactor for converting toluene, selected as a simple tar model compound. The catalysts supported by ZSM-5 showed the best potential for increasing toluene conversion, promoting the formation of syngas, reducing the production of undesirable liquid by-products (such as benzene and ethylbenzene) and corroborating the resistance to catalyst deactivation caused by sintering and carbon deposition. Toluene cracking has been investigated in a plasma-alone (PA), a post-plasma catalysis (PPC) and an in-plasma catalysis (IPC) system. The IPC system exhibited the highest toluene conversion efficiency and selectivity for syngas formation, compared with the PA and PPC systems. A maximum toluene conversion of 86.5% was achieved in the IPC system over Ni/ZSM-5, together with a maximum gas selectivity (63.3% for H₂, 18.2% for CO, and 4.6% for CH₄), and a minimum selectivity to the undesirable liquid by-products (2.5% for benzene and 0.66% for ethylbenzene and a low carbon selectivity of 6.7%). In addition, the reaction mechanism was significantly different in the IPC system from those in the PA and PPC system, since the catalyst in the IPC system might influence the discharge properties and the reaction between excited and short lifetime reactive plasma species (OH^?, O^?, N₂^?...) and toluene and its intermediates. Finally, a mechanism of toluene cracking has been proposed, after analysis of the gas and liquid products and characterization of the fresh and used catalysts.

Full text of DOI:10.1039/c7cy00970d

View Article Online
View Journal
Catalysis
Sc ience &
Technology
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: L. Liu, Q. Wang, J.
Song, S. Ahmad, X. Yang and Y. Sun, Catal. Sci. Technol., 2017, DOI: 10.1039/C7CY00970D.
Volume 6 Number 1 7 January 2016 Pages 1–308
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
Catalysis
Science &
Technology
www.rsc.org/catalysis
Accepted Manuscripts are published online shortly after
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
author guidelines.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the ethical guidelines, outlined
in our author and reviewer resource centre, still apply. In no
event shall the Royal Society of Chemistry be held responsible
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.
ISSN 2044-4753
PAPER
Qingzhu Zhang et al.
Catalytic mechanism of C–F bond cleavage: insights from QM/MM
analysis of fluoroacetate dehalogenase
rsc.li/catalysis

Page 1 of 34
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C7CY00970D
1
2
3
4
5
6
7
Plasmaꢀassisted catalytic reforming of toluene to hydrogen rich syngas
a,b
a,b
a
c
a,b
b,c,*
Lina Liu , Qiang Wang , Jianwei Song , Shakeel Ahmad , Xiaoyi Yang , Yifei Sun
a
b
c
School of Energy and Power Engineering, Beihang University, Beijing 100191, China
Energy and Environment International Centre, Beihang University, Beijing, 100191, China
Beijing Key Laboratory of Bioꢀinspired Energy Materials and Devices, School of Space and
Environment, Beihang University, Beijing 100191, China
*
Corresponding author
Tel/ Fax: +86ꢀ10ꢀ82338120
Eꢀmail address: sunif@buaa.edu.cn (Yifei Sun)
1

Catalysis Science & Technology
Page 2 of 34
View Article Online
DOI: 10.1039/C7CY00970D
8
Abstract
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
The combination of nonꢀthermal plasma (NTP) and transition metal catalysts is
an alternative method for reducing tar derived from biomass gasification, converting it
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
2
to a mix of mainly H and CO (syngas). In this study, Niꢀ and Feꢀbased steam
reforming catalysts supported on CaO, SiO , γꢀAl O and ZSMꢀ5 were combined with
2
2
3
a coaxial dielectric barrier discharge (DBD) plasma reactor for converting toluene,
selected as a simple tar model compound. The catalysts supported by ZSMꢀ5 showed
the best potential in increasing toluene conversion, promoting the formation of syngas,
reducing the production of undesirable liquid byꢀproducts (such as benzene and
ethylbenzene) and corroborating the resistance to catalyst deactivation caused by
sintering and carbon deposition. Toluene cracking has been investigated in a
plasmaꢀalone (PA), a postꢀplasma catalysis (PPC) and an inꢀplasma catalysis (IPC)
system. The IPC system exhibited the highest toluene conversion efficiency and
selectivity of syngas formation, compared with PA and PPC system. A maximum
toluene conversion of 86.5% was achieved in the IPC system over Ni/ZSMꢀ5,
2
together with a maximum gas selectivity (63.3% for H , 18.2% for CO, and 4.6% for
CH ), and a minimum selectivity of undesirable liquid byꢀproducts (2.5% for benzene
4
and 0.66% for ethylbenzene and a low carbon selectivity of 6.7%). In addition, the
reaction mechanism was significantly different in the IPC system from that in the PA
and PPC system, since the catalyst in the IPC system might influence the discharge
properties and influence the reaction between excited and short lifetime reactive
*
plasma species (OH·, O·, N …) and toluene and its intermediates. Finally, a
2
mechanism of toluene cracking has been proposed, after analysis of gas and liquid
products and characterization of the fresh and used catalysts.
Keywords: biomass tar; toluene; catalytic steam reforming; Ni/Feꢀbased catalysts;
syngas; DBD plasmaꢀcatalysis system
2

Page 3 of 34
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C7CY00970D
3
5
1. Introduction
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
6
6
6
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
Syngas (a mixture of H and CO) production by biomass gasification is
2
considered as an attractive and alternative treatment, meeting the increasing energy
1ꢀ3
demand and relieving the crisis of fossil energy. Syngas can be used as the raw
materials for the synthesis of methanol, dimethylether (DME), ethanol and
4
FischerꢀTropsch fuels. However, the formation of tar during gasification is one of the
5
major challenges that limit the commercial application of gasification techniques. Tar
is a complex and hazardous mixture of condensable hydrocarbons, including
singleꢀring to fiveꢀring aromatic compounds along with other oxygenates and
6
,7
polycyclic aromatic hydrocarbons (PAHs). The formation of tar leads to the
8
decrease of the gasification efficiency due to the loss of chemical energy of tar.
Besides, the condensation of tar at low temperature leads to the clogging of pipeꢀlines
and contamination of process equipment such as gasifier, engine and turbine, and the
9
,10
deactivation of catalysts.
As a consequence, more maintenance cost is needed for
11
the stable operation of gasification process. Therefore, the efficient removal of tar is
essential for gaining clean biogas and increasing the process efficiency during
12
biomass gasification.
Extensive studies have been reported for the elimination and reforming of tar,
13
14
and various strategies including mechanical separation, thermal cracking and
1
5,16
catalytic reforming
have been proposed. Tar cracking and simultaneous reforming
4
is more desirable than mechanical separation by filtration and
to H , CO and CH
2
water or oil scrubbing, which lead to the secondary pollution and the chemical energy
3
,6
loss of tar. However, both thermal cracking and catalytic reforming have been
restricted due to their shortcomings. Thermal cracking is generally conducted as high
temperature (>800 ºC), leading to considerable energy consumption. Catalytic
reforming is easily declining and even terminated because of the rapid deactivation of
17
catalysts by sintering, poisoning and carbon deposition. Therefore, nonꢀthermal
plasma (NTP) shows significant potential in converting complex tar to energy gas,
due to its relative low energy consumption, high removal efficiency and universality
3

Catalysis Science & Technology
Page 4 of 34
View Article Online
DOI: 10.1039/C7CY00970D
1
8
6
6
6
6
6
6
7
7
7
7
7
7
7
7
7
7
8
8
8
8
8
8
8
8
8
8
9
9
9
9
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
for various hydrocarbons. In recent years, NTP has been widely applied in the
1
9,20
removal of hydrocarbons such as volatile organic compounds (VOCs)
and
However, NTP alone produced a great amount of undesirable
byꢀproducts, and showed less selectivity towards syngas than the catalytic steam
11,21
biomass tar.
22,23
reforming.
Therefore, coupling a NTP with heterogeneous catalysts seems a more
promising alternative, since NTP and catalysts may generate a synergistic effect,
enhancing tar conversion and syngas selectivity, inhibiting the formation of
24,25
undesirable byꢀproducts, and reducing energy consumption.
Dielectric barrier
discharge (DBD) plasma is a potential option since it is easily integrated with
26
catalysts due to less carbon deposition.
Various catalysts have been used for steam reforming of tar, including transition
metal catalysts (Niꢀ, Feꢀ, Coꢀbased catalysts…), natural minerals (olivine, dolomite
27ꢀ30
ilmenite…), noble metal catalysts (Rh, Ru, Pd, Pt…) and char.
Among these,
transition metal catalysts show high performance and low cost for biomass tar
31
reforming. The selection of support materials is an important issue since the catalyst
deactivation caused by sintering of the support surface is a great challenge for
catalytic steam reforming of tar. The most commonly used supports for steam
24
reforming of tar are alumina and its modifications. Also CaO and MgO are
32
considered as active supports for tar reforming, as main components of dolomite.
Furthermore, zeolites, such as ZSMꢀ5, SBAꢀ15…, have been used for reforming
hydrocarbons because of their high surface area, thermal stability, and high affinity as
3
3,34
adsorbent for Oꢀcontaining compounds.
Although different support materials have
been applied for catalytic steam reforming of tar, few studies systematically
investigated the effect of supporting material on tar reforming, and the synergy of
plasma and catalyst is still unclear.
Based on the analysis above, it is essential to investigate the effect of different
support materials on thermal stability, resistance to carbon deposition and catalytic
activity of catalysts for steam reforming of tar. Additionally, few studies reported tar
cracking and oriented reforming to syngas in the plasmaꢀcatalysis system, and the
synergistic mechanism of plasma and catalysts is still unclear. Therefore, DBD plasma
4

Page 5 of 34
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C7CY00970D
9
9
9
9
9
9
4
5
6
7
8
9
was coupled with catalysts in order to promote the oriented conversion of tar to
syngas and reduce the formation of undesirable liquid byꢀproducts. Niꢀ and Feꢀbased
2 2 3
catalysts with different support materials, such as CaO, SiO , γꢀAl O and ZSMꢀ5,
were employed to investigate their performance on steam reforming of toluene as a
representative tar surrogate. The fresh and used catalysts were characterized by BET,
XRD, SEM, TGꢀDSC and XPS to study the influence of supports to catalytic
performances. In addition, the performance of toluene reforming in different types of
reactors (plasmaꢀalone (PA), postꢀplasma catalysis (PPC), and inꢀplasma catalysis
(IPC) systems) was also investigated, along with the synergy of plasma and catalysts.
Finally, a reaction mechanism of toluene cracking was proposed by the analysis of
liquid and gas products in the hybrid plasmaꢀcatalysis system.
1
1
1
1
1
1
00
01
02
03
04
05
1
1
06
07
2. Materials and Methods
2.1 Preparation of catalysts
1
1
1
1
1
1
1
1
08
09
10
11
12
13
14
15
Different Niꢀ and Feꢀbased catalysts were prepared by wet impregnation:
•
3 2 2 3 2 2
1.58 g of Ni (NO ) ·6H O and 2.43g of Fe(NO ) ·9H O was dissolved in
~50 mL of deionized water. The aqueous solution was mixed with 9.7 g of
CaO, SiO , γꢀAl O and ZSMꢀ5 used as different supporting materials.
2
2
3
•
•
The suspension was stirred under ambient condition for 12 hours, and
subsequently at 333 K for 24 hours.
Then the samples were dried at 378 K overnight and calcined in the
atmosphere of air at 823 K for 6 hours in a tubular furnace (OTFꢀ1200X,
1
1
1
16
17
18
Hefei Kejing Material Technology Co. Ltd. China).
•
Therefore, the theoretical Ni and Fe loading of all catalysts was 3 wt.%.
The Niꢀ and Feꢀbased catalysts prepared were defined as Ni/CaO, Ni/SiO2,
Ni/γꢀAl , Ni/ZSMꢀ5, Fe/CaO, Fe/SiO , Fe/γꢀAl and Fe/ZSMꢀ5. The catalysts
diluted with N (100
1
1
1
19
20
21
2
O
3
2
2 3
O
were sieved to 40ꢀ60 mesh and then reduced in 5 wt.% of H
2
2
mL/min) at 650 ºC for 6 h before use.
5

Catalysis Science & Technology
Page 6 of 34
View Article Online
DOI: 10.1039/C7CY00970D
1
22
2.2 Selection of tar model compounds
Due to the complexity of tar, tar model compounds are generally selected for the
1
1
1
1
1
1
1
1
23
24
25
26
27
28
29
30
1
7
tar decomposition study. The most common model compounds are recognized as
3
5
benzene, toluene, phenol, naphthalene, etc. In this study, we selected toluene as the
representative tar surrogate based on three reasons: (1) toluene is the dominant
component of biomass tar; (2) toluene shows a simple structure and high thermal
stability is well known; (3) toluene is less harmful and its highꢀtemperature
characteristics are well known. Therefore, the utilization of toluene can help get
1
1,36
realistic information of tar reforming over different catalysts.
1
31
2.3 Experimental setup and procedure
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
The schematic diagram for biomass tar reforming is displayed in Figure 1. A 60
mmꢀlong stainless steel wire surrounding a quartz tube (Length: 600 mm, I.D.: 44 mm,
thickness: 3.0 mm) was used as grounded eletrode, and a stainless steel rod (length 60
mm and outer diameter 42 mm) was the highꢀvoltage electrode of the DBD plasma
reactor. The discharge gap was 1 mm. The reactor was vertically installed in a shaft
tubular furnace (customized, Jinhua Hengfeng Electrical Instrument Co. Ltd., China).
The experiments were conducted in Mode A and Mode B. For Mode A, the catalyst
was held by silica wool at 50 mm downstream of the discharge area of a postꢀplasma
catalysis (PPC) system. For Mode B, the catalyst was placed in the discharge area of
an inꢀplasma catalysis (IPC) system. In addition, toluene cracking was also conducted
as a blank test in a plasmaꢀalone (PA) system without catalyst. For each test 1.0 g of
new catalyst was used. Toluene was selected as surrogate of biomass gasification tar.
Toluene and water was injected into the DBD reactor by two syringe pumps (TJꢀ3A,
Baoding Longer Pump Co. Ltd., China) at a flow rate of 30 ꢁL/min for toluene and
35.5 ꢁL/min for water, respectively. The molar steam to carbon (S/C) ratio was 1.0.
Nitrogen (99.999 vol.%, Beijing Haipu Gas Co. Ltd, China) was used as carrier gas at
a constant flowrate of 100 mL/min. Toluene and water were evaporated in the upper
part of the reactor, kept at 473 K by a heating strip and then entered into the
discharging area of DBD plasma by the carrier gas. The discharge time and
6

Page 7 of 34
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C7CY00970D
1
1
1
1
1
1
1
1
1
1
51
52
53
54
55
56
57
58
59
60
temperature were 30 min and 573 K, respectively. The condensable products were
absorbed in ~200 mL of methanol using two impingers cooled in iceꢀwater bath.
Finally, nonꢀcondensable gas products were collected in a sampling gas bag (10 L,
Dalian Delin Gas Packaging Co. Ltd., China) and their total volume was recorded by
a wet gas flowmeter (LMFꢀ2, Changchun Auto Filter Co., Ltd., China). The DBD
plasma reacter was connected to an AC highꢀvoltage power supply with a maximum
peak to peak voltage of 30 kV and a frequency of 10 kHz. The discharge signals were
monitored by a digital oscilloscope (Rigol DS2102A). The average discharge power
and frequency were 60 W and 14.79 kHz, respectively.
1
1
1
61
62
63
Figure 1. Schematic diagram of the experimental setup with (a) the post plasmaꢀcatalysis
system and (b) the plasmaꢀcatalysis system.
1
1
64
65
2.4 Analysis method of the products
1
1
1
1
1
1
1
66
67
68
69
70
71
72
The nonꢀcondensable gas products were analyzed by a gas chromatograph
(GCꢀ17A, Shimadzu, Japan) equipped with a thermal conductivity detector (TCD)
and a Carboxenꢀ1010 PLOT capillary column (30 m × 0.53 mm × 30 ꢁm, Supelco
Corporation, USA). The liquid products absorbed in the methanol solution were
analyzed by a highꢀresolution gas chromatographꢀlow resolution mass spectrometer
(HRGCꢀLRMS, QP2020, Shimadzu) equipped with an HPꢀ5MS capillary column (60
m × 0.25 mm × 0.25 ꢁm; Agilent Technologies, USA).
7

Catalysis Science & Technology
Page 8 of 34
View Article Online
DOI: 10.1039/C7CY00970D
1
1
73
74
The conversion of toluene (C ), and the various selectivities of H
2
, CO
,
C
m
H
n
T
and carbon deposition were calculated as follows,
moles of toluene inputꢀmoles of toluene output
1
1
75
76
C (%)=
T
×100% (1)
×100% (2)
moles of toluene input
moles ofꢀH ꢀoutputꢀ
×moles of toluene convertedꢀꢁꢀmoles ofꢀH Oꢀinput
2
S (%)=
H
2
4
2
moles of CO output
×moles of toluene converted
1
1
77
78
S (%)=
×100% (3)
CO
7
m×moles of C H output
m
n
S (%)=
C H
m n
×100% (4)
7
×moles of toluene converted
moles of carbon output
×moles of toluene converted
1
1
1
79
80
81
SC(%)=
×100% (5)
7
2 4
The yield of gas products such as H , CO and CH was calculated as Eq. (6).
moles of gas products output
moles of toluene input
Y
ꢂ
mol/mol
ꢃ
=
(6)
1
1
82
83
The energy efficiency is defined as the ratio of the mass of toluene converted per
hour to the discharge power as shown in Eq. (7).
mass of toluene convertedꢀ(g)
1
1
84
85
η
ꢂ
g/kWh
ꢃ
=
(7)
P(W)×0.5/1000
2.5 Characterization of catalysts
1
1
1
1
1
1
1
1
1
1
1
1
1
86
87
88
89
90
91
92
93
94
95
96
97
98
The fresh and used Niꢀ and Feꢀbased catalysts were characterized as follows.
Ni and Fe loading. About 0.1g of each catalyst was digested in a microwave
3
system (Mars 6, CEM, USA) using 6 mL of 36% HCl and 2 mL of 65% HNO and
generating an aqueous solution of Ni or Fe. This solution was diluted and then
analyzed using inductively coupled plasmaꢀ mass spectrometry (ICPꢀMS, XSERIES 2,
Thermo Scientific, USA) to determine the actual Ni and Fe loadings on different
supports.
The BrunauerꢀEmmettꢀTeller (BET) surface area, pore volume and average pore
diameter of the fresh catalysts were determined by the adsorptionꢀdesorption isotherm
using a Quantachrome NOVA 2200e analyzer (USA).
XꢀRay Diffraction (XRD, D6000, Shimadzu, Japan) was used to examine the
crystal species of different catalysts before and after reforming. The scanning range
2θ was 10ꢀ90° with a speed of 5°/min.
8

Page 9 of 34
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C7CY00970D
1
2
2
2
2
2
2
2
2
99
00
01
02
03
04
05
06
07
A scanning electron microscope (SEM, Quanta FEG 250, FEI, USA) allowed to
analyze the morphology and particle size of different catalysts.
Xꢀray photoelectron spectra were recorded by an imaging Xꢀray photoelectron
spectrometer (XPS, Axis Ultra, Kratos Analytical, UK) to characterize the catalyst
surface.
The carbon content of fresh and used catalysts was characterized using
thermogravimetryꢀdifferential scanning calorimetry (TGꢀDSC, STA449F3 Jupiter,
NetzschꢀGerätebau GmbH, Germany). About 40 mg of catalysts were heated in air
from room temperature to 800°C, at a heating rate of 10 °C/min.
2
08
3. Results and Discussion
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
09
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
3.1 Catalyst characterization
Table 1 shows the Ni or Fe loading, specific surface area, pore volume, average
pore diameter and the crystallite size of different supports and catalysts. The Ni or Fe
loading was measured by ICPꢀMS as 1.91ꢀ2.54 wt.% for Niꢀbased catalysts and
1.95ꢀ2.84 wt.% for Feꢀbased catalysts, slightly lower than the 3 wt.% desired.
However, the Ni/Fe loading of almost all catalysts was about 2 wt%, indicating the
reliability of the preparation method of the catalysts. The specific surface area was
calculated by the BET method, and the pore diameter distribution was determined by
the adsorptionꢀdesorption isotherm of catalysts. The nitrogen adsorptionꢀdesorption
isotherms of different Niꢀ and Feꢀbased catalysts and the pore diameter distribution
are shown in Figure S1 and Figure S2. The results showed that the catalyst supports
(CaO, SiO , γꢀAl O and ZSMꢀ5) influence the adsorptionꢀdesorption isotherm and
2
2
3
pore diameter distribution these more than the active metal (Ni and Fe). The catalysts
with different supports displayed absolutely different adsorptionꢀdesorption isotherms.
M/CaO and M/SiO catalysts (M: Ni or Fe) exhibit the type III adsorptionꢀdesorption
2
isotherm of nonꢀporous materials (Figure S1a and b). M/γꢀAl
2 3
O show a type V
adsorptionꢀdesorption isotherm with H1 hysteresis loop, clearly related to a diameter
3
7
in the mesopore range
(Figure S1c). M/ZSMꢀ5 display a type Iꢀlike
adsorptionꢀdesorption isotherm characteristic of cellular materials (Figure S1d). The
9

Catalysis Science & Technology
Page 10 of 34
View Article Online
DOI: 10.1039/C7CY00970D
2
2
2
2
2
2
2
2
2
2
2
2
2
28
29
30
31
32
33
34
35
36
37
38
39
40
pore diameter distributions shown in Figure S2 confirmed the results. The pore
diameter of M/γꢀAl O ranges 5ꢀ15 nm, i.e. mesoporous. Most M/ZSMꢀ5 catalyst
2
3
pores are micropore in the range of 0ꢀ2 nm. However, M/ZSMꢀ5 catalysts also
contain mesopores, leading to multilayer adsorption in a N adsorptionꢀdesorption
2
isotherm. It is shown in Table 1 that the addition of both Ni and Fe led to a decrease
in the BET surface area, pore volume and average pore diameter of supporting
materials. It might be caused by the coverage of Ni and Fe particles on the surface of
supporting materials and even the migration of the metals in pores of the supporting
3
8
materials, which leads to the contraction of pores. Among all catalysts supported by
different materials, M/ZSMꢀ5 exhibited the highest BET surface area of 340.1ꢀ359.6
2
2
2
2 3 2
m /g, comparing with 145.2ꢀ156.6 m /g for M/γꢀAl O , 5.9ꢀ7.5 m /g for M/SiO and
2
16.4ꢀ19.5 m /g for M/CaO. However, the pore volume of M/γꢀAl O was 0.46ꢀ0.49
2
3
3
3
cm /g, much higher than the 0.07ꢀ0.13 cm /g of M/ZSMꢀ5. The difference might be
2
2
41
42
attributed to the abundant amounts of mesopores in M/γꢀAl
2 3
O .
Table 1 Physicochemical properties and Ni/Fe loadings of the catalysts.
Niꢀbased Catalysts
Ni/HZSMꢀ5 Ni/γꢀAl
O
2 3
Ni/SiO
5.9
2
Ni/CaO
16.4
2
BET Surface Area (m /g)
359.6
1.68
0.13
21.6
1.91
156.6
Average Pore Diameter (nm)
7.91
0.49
8.8
1.47
0.01
75.6
1.93
3.84
3
Pore Volume (cm /g)
0.04
a
Crystallite Size (nm)
56.0
Ni loading (%)
2.54
2.10
Feꢀbased Catalysts
Fe/ZSMꢀ5
340.1
0.91
Fe/γꢀAl
145.2
9.76
2
O
3
Fe/SiO
2
Fe/CaO
19.5
2
BET Surface Area (m /g)
7.5
Pore Diameter (nm)
2.47
0.02
61.2
1.95
3.82
3
Pore Volume (cm /g)
0.07
0.46
0.05
a
Crystallite Size (nm)
16.4
4.7
44.1
Fe loading (%)
1.95
2.84
2.47
Supports
ZSMꢀ5
383.0
2.26
γꢀAl O3
SiO2
10.7
3.05
0.02
CaO
20.1
7.14
0.05
2
2
BET Surface Area (m /g)
165.6
15.6
0.55
Pore Diameter (nm)
3
Pore Volume (cm /g)
0.22
a
2
43
Calculated by the Scherrer equation based on the XRD results.
10

Page 11 of 34
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C7CY00970D
2
2
2
2
2
2
2
2
2
2
2
44
45
46
47
48
49
50
51
52
53
54
Figure 2 presents the XRD patterns of freshly calcined and used Niꢀbased
catalysts. In addition, the supporting materials were also analyzed using XRD, as
shown in Figure S3. It is obvious that the presence of NiO particles showed little
effect on the diffraction patterns of supporting materials. Almost all catalysts show
clear NiO diffraction peaks at 2θ=37.8°, 43.7° and 62.7°, except for Ni/γꢀAl
difference might be attributed to two reasons: (1) the diffusion peaks of NiO falling in
the broad diffusion peak range of Al at 42.5ꢀ49.6° and 58.9ꢀ70.7°, (2) a better
dispersity of NiO particles on Al . In addition, similar results were observed for the
Feꢀbased catalysts, as presented in Figure S4. The XRD results of Fe/SiO , Fe/CaO
and Fe/ZSMꢀ5 all showed the diffusion peaks at 2θ=33.2° and 35.7°, corresponding to
the characteristic peaks of Fe . However, they were not observed in Fe/Al
2 3
O . The
2
O
3
2
O
3
2
2
O
3
2 3
O .
2
2
2
55
56
57
The average crystallite size of different catalysts was determined by the Scherrer
equation, as shown in Table 1:
d=0.89λ/βcosθ
(8)
2
2
2
2
2
2
2
58
59
60
61
62
63
64
where d represents the average size of the crystallite; λ represents the Xꢀray
wavelength; β represents peak width at half height and θ is the Bragg angle,
respectively.
The crystallite size of the Niꢀbased catalysts supported by different materials
varied from 8.8 nm to 75.6 nm, generally higher than 4.7ꢀ61.2 nm for the Feꢀbased
2
catalysts, which is in agreement with the N adsorptionꢀdesorption experiment in this
study (Table 1).
2
2
65
66
(a)
(b)
11

Catalysis Science & Technology
Page 12 of 34
View Article Online
DOI: 10.1039/C7CY00970D
2
2
67
68
(c)
(d)
2
2
2
2
2
2
2
2
2
2
2
2
69
70
71
72
73
74
75
76
77
78
79
80
Figure 2. XRD analysis of the (a) Ni/SiO
2 2 3
, (b) Ni/CaO, (c) Ni/γꢀAl O and (d) Ni/ZSMꢀ5 before
and after toluene reforming.
The SEM pictures of the fresh and used Niꢀbased catalysts are shown in Figure 3.
Niꢀbased catalysts with different supports show absolutely different particle shape and
size. Ni/SiO
catalyst shows rodꢀshaped particles aggregated into clusters. The Ni/CaO catalyst is
composed of larger particles with irregular shapes. Among all catalysts, Ni/γꢀAl
2 2 3
and Ni/γꢀAl O catalysts consist of sphereꢀlike particles and Ni/ZSMꢀ5
2
O
3
has a minimal size in nanoꢀscale. During a comparison of fresh and used catalyst,
Ni/SiO , Ni/CaO and Ni/γꢀAl O all show obvious sintering, but Ni/ZSMꢀ5 catalyst
2
2
3
presents no obvious changes before and after toluene reforming, indicating absence of
sintering.
12

Page 13 of 34
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C7CY00970D
Ni/SiO2
Ni/SiO2
Ni/CaO
Ni/CaO
Ni/γꢀAl O
Ni/γꢀAl O
3
2
3
2
Ni/ZSMꢀ5
Ni/ZSMꢀ5
2
2
81
82
(a)
(b)
2
2
83
84
Figure 3. SEM image of (a) fresh and (b) used Niꢀbased catalysts.
13

Catalysis Science & Technology
Page 14 of 34
View Article Online
DOI: 10.1039/C7CY00970D
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
3.2 Catalytic steam reforming of toluene
Transition metals, such as Ni and Fe, are promising catalysts for cracking tar,
39
due to their high efficiency, low cost, and easy regeneration. Their support is an
important influencing factor since metal catalysts are not actively supported by inert
4
0
oxides.
The conversion efficiency of toluene and the selectivity of forming benzene and
ethylbenzene catalyzed by Niꢀbased catalysts in the postꢀplasma catalysis (PPC)
system are shown in Figure 4 (a). The combination of catalyst and DBD plasma
obviously enhanced toluene cracking than the DBD plasma without catalysts. The
conversion of toluene catalyzed by Niꢀbased catalysts with different supports varied
in a range of 67.0ꢀ79.5%, i.e. 1.05ꢀ1.24 times that with plasma alone. The catalytic
activities of different Niꢀbased catalysts followed the order: Ni/ZSMꢀ5 > Ni/γꢀAl O >
2
3
2
Ni/SiO > Ni/CaO > Blank, which was positively related to the BET surface area.
Catalyst with high surface area and pore volume has higher adsorption ability for
reactants and promoters such as H O and free radicals (H·, OH·, O·…). They provide
2
more active sites for toluene cracking. Benzene and ethylbenzene are the leading
liquid byꢀproducts of toluene cracking. Their selectivity was in the range of 4.2ꢀ8.5 %
and 1.4ꢀ2.9 %, respectively. Contrary to toluene conversion, the selectivity of benzene
and ethylbenzene obtained their minimum when catalyzed by Ni/ZSMꢀ5 and their
maximum with plasma alone (Figure 4a). The results showed that the addition of
Niꢀbased catalysts, especially Ni/ZSMꢀ5, inhibited the conversion of toluene to liquid
byꢀproducts.
Apart from the liquid byꢀproducts, H
products. The selectivity and yield of H and CO is shown in Figure 4(b). Ni/ZSMꢀ5
and Ni/γꢀAl O present comparable catalytic activity on the formation of syngas,
2
and CO were the most important gas
2
2
3
higher than Ni/SiO
2
and Ni/CaO. Most Niꢀbased catalysts promote the formation of
H
H
2
2
and CO, except Ni/CaO. Niꢀbased catalysts stimulate the conversion of toluene to
41
and CO, promoting steam reforming and the waterꢀgas shift (WGS) reaction.
However, both the yield and selectivity of syngas decreased catalyzed by Ni/CaO
compared with the DBD plasma system without catalysts. In previous studies,
14

Page 15 of 34
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C7CY00970D
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
MgOꢀCaO and dolomite were effective supports in the suppression of coke and their
4
1
resistance to sulfur poisoning during biomass tar reforming. However, Ni/CaO
showed little catalytic activity in toluene cracking and syngas formation in this study,
2
which might be attributed to the consumption of H O by CaO. CaO could react with
H O to generate Ca(OH) and thus the steam reforming reaction (9) and waterꢀgas
2
2
shift reaction (10) were weakened. This assumption was supported by the
XRDꢀresults of Niꢀbased catalysts before and after reforming (Figure 2). The XRD
results of Ni/SiO
and after reforming. However, the fresh Ni/CaOꢀcatalyst contains crystals of NiO and
CaO, and the characteristic diffraction peaks of Ca(OH) at 2θ=29.0°, 34.4°, 47.5°,
51.0° and 54.6° were also observed in used catalysts, cf. Figure 2(b).
2 2 3
, Ni/γꢀAl O and Ni/ZSMꢀ5 did not show obvious difference before
2
C H +7H O
→
7CO+11H2 (9)
+H (10)
2 3
, Al O and CaO without
7
8
2
CO+H
2
O
→
CO
2
2
We also tested the catalytic activity of ZSMꢀ5, SiO
2
NiO loading on toluene reforming in the PPC system. The results indicated that the
supporting materials did not show obvious catalytic activity on toluene conversion
2
and formation of H and CO, as shown in Figure S5. It can be concluded that Ni was
the main active component for toluene conversion and syngas production.
8
8
7
7
6
6
5
5
4
4
5
0
5
0
5
0
5
0
5
0
16
Toluene
Benzene
Ethylbenzene
1
1
1
8
6
4
2
0
4
2
0
Blank
Ni/ZSMꢀ5 Ni/γꢀAl₂O₃
Ni/SiO₂
Ni/CaO
3
3
33
34
(a)
15

Catalysis Science & Technology
Page 16 of 34
View Article Online
DOI: 10.1039/C7CY00970D
1
1
1
0
0
0
.8
.5
.2
.9
.6
.3
S(H₂)
S(CO)
Y(H₂)
Y(CO)
50
40
30
20
10
0
0.0
Blank
Ni/ZSMꢀ5
Ni/γꢀAl₂O₃
Ni/SiO₂
Ni/CaO
3
3
3
3
3
35
36
37
38
39
(b)
Figure 4. Effect of Niꢀbased catalysts on the (a) conversion efficiency of toluene, selectivity of
benzene and ethylbenzene, and (b) selectivity and yield of H and CO.
2
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
The catalytic activity of Feꢀbased catalysts was also evaluated with different
supports (Figure 5). The results showed that the catalytic activity of Feꢀbased
42
catalysts with different supports showed similar trend with Niꢀbased catalysts,
following a sequence of Fe/ZSMꢀ5 > Fe/γꢀAl > Fe/SiO > Blank > Fe/CaO.
2
O
3
2
Fe/ZSMꢀ5 exhibited the highest catalytic activity with the highest toluene conversion
efficiency of 74.6% and the lowest benzene and ethylbenzene selectivity of 4.1% and
1.6%, respectively (Figure 5a).
The selectivity for H
comparable, yet much higher than for Fe/SiO
catalytic effect on syngas formation and CO was not even detected during toluene
cracking. The yield of H and CO attained a maximum of 2.25 mol/molꢀtoluene for H
and 0.92 mol/molꢀtoluene for CO.
2 2 3
and CO, catalyzed by Fe/ZSMꢀ5 and Fe/γꢀAl O , was
2
and Fe/CaO. Fe/CaO showed no
2
2
In addition, Niꢀbased catalysts exhibited relatively higher activity on the
conversion of toluene than Feꢀbased catalysts, 67.0ꢀ79.5% for Niꢀbased catalysts and
63.8ꢀ74.6% for Feꢀbased catalysts. However, Feꢀbased catalysts were more potential
on syngas formation. The results can be explained by the characteristics of Niꢀ and
16

Page 17 of 34
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C7CY00970D
3
3
3
3
3
3
3
3
3
3
3
3
3
3
56
57
58
59
60
61
62
63
64
65
66
67
68
69
Feꢀbased catalysts. Niꢀbased catalysts exhibit higher catalytic activity than Feꢀbased
catalysts for toluene conversion due to the higher reactivity of CꢀC and CꢀH bonds in
43
hydrocarbons on the surface of Ni particles. However, syngas production is better
catalyzed by Feꢀbased than by Niꢀbased catalysts. Since Fe has higher oxygen affinity
44
than Ni, Feꢀbased catalysts can increase the coverage of oxygenꢀcontained
compounds (H O, O·, OH·…) on the catalyst surface during the steam reforming
2
reactions. Thus the formation of syngas is promoted by enhancing the steam
reforming reaction of toluene and the attack of free radicals to toluene. Therefore,
45
adding Fe or other oxytropic materials (OMs) such as CeO
2
in Niꢀbased catalysts,
might improve the conversion of toluene and syngas production simultaneously.
Developing modified catalysts by adding different kinds of OMs and adjusting the
OMs/Ni ratio should be an alternative choice to enhance the conversion of tar and the
formation of syngas.
8
7
7
6
6
5
5
4
4
0
5
0
5
0
5
0
5
0
Toluene Benzene Ethylbenzene
1
1
1
8
6
4
2
0
4
2
0
Blank
Fe/ZSMꢀ5
Fe/γꢀAl₂O₃
Fe/SiO₂
Fe/CaO
3
3
70
71
(a)
17

Catalysis Science & Technology
Page 18 of 34
View Article Online
DOI: 10.1039/C7CY00970D
70
60
50
40
30
20
10
0
2.4
S(H₂)
S(CO)
Y(H₂)
Y(CO)
2
1
1
0
0
.0
.6
.2
.8
.4
0.0
Blank
Fe/ZSMꢀ5
Fe/γꢀAl₂O₃
Fe/SiO₂
Fe/CaO
3
3
3
3
3
72
73
74
75
76
(b)
Figure 5. Effect of Feꢀbased catalysts on the (a) conversion efficiency of toluene, selectivity of
benzene and ethylbenzene, and (b) selectivity and yield of H and CO.
2
3
77
3.3 Toluene cracking in the PPC and IPC reactor
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
Although Niꢀ and Feꢀbased catalysts in the postꢀplasma catalysis (PPC) reactor
somewhat promoted the conversion of toluene and syngas formation than that in the
plasma system without catalysts, the promoting effect was limited. Therefore, the
catalytic steam reforming of toluene in the inꢀplasma catalysis (IPC) reactor was
studied. Considering the higher activity of Niꢀbased catalysts, they were selected for
the steam reforming of toluene in the IPC system. The comparative results of PPC and
IPC reactors are presented in Figure 6. It can be seen from Figure 6(a) that the
catalytic activity of Niꢀbased catalysts exhibited similar trend in the IPC reactor with
that in the PPC reactor. However, the catalytic activity of Niꢀbased catalysts was
obviously improved in the IPC reactor. The conversion efficiency of toluene and
energy efficiency showed higher values of 66.6ꢀ86.5% and 17.3ꢀ22.5 g/kWh in the
IPC reactor, compared with 67.0ꢀ79.5% and17.4ꢀ20.7 g/kWh in the PPC reactor.
The selectivity of benzene and ethylbenzene both decreased in the IPC reactor
than the PPC reactor catalyzed by Ni/SiO
2
, Ni/γꢀAl
2
O
3
and Ni/ZSMꢀ5, as shown in
are displayed in Figure
Figure 6(b). The selectivity and yield of H
2
, CO and CH
4
18

Page 19 of 34
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C7CY00970D
3
3
3
3
3
3
3
93
94
95
96
97
98
99
6(c). The main gas products are H
reforming in the IPC reactor compared with that in the PPC system. As the minor gas
product, the selectivity of CH was in the range of 3.2ꢀ4.6%, much lower than that of
and CO. We can find from Figure 6(c) that the formation of H , CO and CH is
greatly promoted in the IPC reactor. Ni/ZSMꢀ5 is the most promising catalyst also in
2 4
and CO, and CH is newly detected during toluene
4
H
2
2
4
2
the IPC reactor, with the highest gas selectivity of 63.3% for H , 18.2% for CO, and
4.6% for CH
4
.
1
00
23
C(T)ꢀPPC (A)
C(T)ꢀIPC (B)
ηꢀPPC (A)
2
2
9
9
8
8
7
7
6
6
5
0
5
0
5
0
5
0
21
ηꢀIPC (B)
2
1
1
1
1
1
0
9
8
7
6
5
14
Ni/ZSMꢀ5
Ni/γꢀAl₂O₃
Ni/SiO₂
Ni/CaO
4
4
00
01
(a)
9
S(B)ꢀPPC (A)
S(B)ꢀIPC (B)
S(E)ꢀPPC (A)
S(E)ꢀIPC (B)
8
7
6
5
4
3
2
1
0
Ni/ZSMꢀ5
Ni/γꢀAl₂O₃
Ni/SiO₂
Ni/CaO
4
4
02
03
(b)
19

Catalysis Science & Technology
Page 20 of 34
View Article Online
DOI: 10.1039/C7CY00970D
70
60
50
40
30
20
10
0
3.0
S(H₂)ꢀPPC (A)
S(CO)ꢀPPC (A)
S(CH₄)ꢀPPC (A)
Y(H₂)ꢀPPC (A)
Y(CO)ꢀPPC (A)
Y(CH₄)ꢀPPC (A)
S(H₂)ꢀIPC (B)
S(CO)ꢀIPC (B)
S(CH₄)ꢀIPC (B)
Y(H₂)ꢀIPC (B)
Y(CO)ꢀIPC (B)
Y(CH₄)ꢀIPC (B)
2
2
1
1
0
.5
.0
.5
.0
.5
0.0
Ni/ZSMꢀ5
Ni/γꢀAl₂O₃
Ni/SiO₂
Ni/CaO
4
4
4
4
4
4
04
05
06
07
08
09
(c)
Figure 6. (a) Conversion of toluene and energy efficiency, (b) selectivity of benzene and
ethylbenzene, (c) selectivity and yield of H , CO and CH in the postꢀplasma and inꢀplasma
system.
2
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
10
11
12
13
14
15
16
17
18
19
20
21
22
23
In order to demonstrate the effect mechanism of catalyst on toluene reforming,
XPS technique was used to characterize the chemical changes on the surface of
Ni/ZSMꢀ5 and Ni/SiO catalysts before and after toluene reforming. Figure 7 shows
2
2
the XPS spectra of Ni 2p. Ni/ZSMꢀ5 and Ni/SiO displayed similar Ni 2p spectra.
2
Both the Ni 2p3/2 spectra of the fresh Ni/ZSMꢀ5 and Ni/SiO catalysts showed two
0
main peaks, confirming the presence of Ni with a peak binding energy of ~852.9 eV,
2
+
and the presence of Ni with a peak binding energy of ~856.4 eV, and its
4
6
corresponding shakeup resonance at ~861.8 eV. The presence of satellite peaks
47
confirmed that partial nickel atoms were present in oxygen environment. Therefore,
0
the main chemical states of Ni were zeroꢀvalent metallic nickel and NiO. Ni was
41
proved to be the dominant active component for toluene reforming, and the ratios of
0
0
2+
Ni /Ni +Ni were 30% for Ni/ZSMꢀ5, higher than 27.9% for Ni/SiO
Table 2. In addition, Ni/ZSMꢀ5 contained a higher Ni ratio of 0.79 than 0.70 for
Ni/SiO . These phenomena both lead to a higher catalytic activity of Ni/ZSMꢀ5 than
2
, as presented in
2
20

Page 21 of 34
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C7CY00970D
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
Ni/SiO
~855.5 eV, along with its shakeꢀup peaks at ~861.3 eV for the used Ni/ZSMꢀ5 and
Ni/SiO catalysts. The results showed that the main Ni specie presented in the used
Ni/ZSMꢀ5 and Ni/SiO catalysts was NiO.
2
. Differently, the Ni 2p3/2 spectra exhibited only one main peak centered at
2
2
The O species is one the key factor influencing toluene cracking. Table 2
showed the element composition on the surface of the catalysts. It can be seen from
Table 2 that the amount of surface O species in Ni/ZSMꢀ5 catalyst is higher than
2 2
Ni/SiO . This suggests the higher affinity of ZSMꢀ5 towards O species than SiO .
This property is beneficial to promote the steam reforming reaction by reducing
4
8
carbon depositions and enhancing WGS reaction during toluene steam reforming. In
addition, the O 1s spectra of the fresh and used Ni/ZSMꢀ5 and Ni/SiO catalysts are
2
shown in Figure 8. Three components (OꢀI, OꢀII, OꢀIII) are identified by
deconvoluting the main peak at ~533 eV. OꢀI component located at ~531.0 eV might
o
49
be Cassigned to NiꢀO bonds in NiO. The dominating OꢀII component at ~533.3 eV
2–
50
was obviously related to O species in the support SiO2. Both the OꢀI and OꢀII
components were defined as lattice oxygen (Olat). The OꢀIII component at ~534.0 eV
may be attributed to the SiꢀOH groups with O 1s at BE ~533.9 eV or adsorbed oxygen
50
(Oads). It can be seen from Figure 8 and Table 2 that the adsorbed oxygen in the
fresh Ni/ZSMꢀ5 decreased sharply after toluene reforming, and the ratio of
O
ads/Oads+Olat decreased from 24.2% to 8.23%. However, the lattice oxygen showed
no obvious change before and after toluene reforming. It can be concluded that the
adsorbed oxygen species is the dominant promoter for the steam reforming reaction
compared with lattice oxygen. Moreover, the ratio of Oads on the surface of Ni/SiO
2
catalyst showed much less decrease than Ni/ZSMꢀ5, which lead to a lower catalytic
activity of Ni/SiO catalyst.
2
21

Catalysis Science & Technology
Page 22 of 34
View Article Online
DOI: 10.1039/C7CY00970D
4
4
49
50
(a)
4
51
4
4
4
4
52
53
54
55
(b)
Figure 7. Typical Ni 2p1/2 and 2p3/2 XPS spectra of fresh and used (a) Ni/ZSMꢀ5 and (b) Ni/SiO
2
catalysts.
4
56
Table 2 Surface composition of elements presented in fresh and used Ni/ZSMꢀ5 and Ni/SiO2
Surface element composition (%)
Catalysts
0
0
2+
2+
0
2+
Ni
O
Si
Al Ni /Ni +Ni
Ni /Ni +Ni
O /O +O
ads ads
lat
Fresh Ni/HZSMꢀ5 0.79 62.2 27.6 0.19 30 (852.9eV) 70 (856.4 eV) 24.2 (533.8eV)
Used Ni/HZSMꢀ5
0.44 53.1 23.4 0.19 100 (856.5 eV) 8.23 (534.0eV)
Fresh Ni/SiO
27.9 (852.3eV) 72.1 (856.9 eV) 16.2 (533.1eV)
Used Ni/SiO2
100 (856.6 eV) 10.2 (533.0eV)
0
0.70 57.4 21.8
0.40 49.5 29.8
0
0
2
0
22

Page 23 of 34
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C7CY00970D
4
4
4
4
57
58
59
60
(a)
(b)
2
Figure 8. Typical O 1s XPS spectra of fresh and used (a) Ni/ZSMꢀ5 and (b) Ni/SiO catalysts.
4
61
3.4 Characterization of carbon deposition
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
Carbon deposition on metal catalysts is a troublesome issue limiting their
largeꢀscale application. The nature and amount of carbon deposition on the catalyst
surface was determined by TGꢀDSC analysis of the used Niꢀbased and Feꢀbased
catalysts. The used catalysts were heated up to 800°C and their TGꢀDSC curves are
shown in Figure 9. It can be seen that the catalysts supported by the same kind of
materials showed similar TGꢀDSC curve. The TGꢀcurves showed a oneꢀstage weight
loss between 300ꢀ600°C except for CaOꢀsupported catalysts. The total mass loss in
the temperature range of RTꢀ800°C was 0.27ꢀ1.41% for SiO
5.49ꢀ8.95% for ZSMꢀ5ꢀsupported catalysts, 9.22ꢀ10.3% for γꢀAl
catalysts and 28.8ꢀ30.1% for CaOꢀsupported catalysts. No obvious weight loss was
observed for SiO ꢀsupported catalysts. The DSC profile of catalysts supported by
γꢀAl and ZSMꢀ5 showed obvious exothermic peaks at around 400ꢀ500°C,
2
ꢀsupported catalysts,
2
3
O ꢀsupported
2
2
O
3
5
1
corresponding to the oxidation of amorphous carbon. However, CaOꢀsupported
catalysts exhibited a different TGꢀDSC profile. The TG curve of used CaOꢀsupported
catalysts presented twoꢀstages of mass loss at 390ꢀ460°C and 560ꢀ740°C, respectively.
2
The mass loss in the first stage was due to Ca(OH) dehydration at 370ꢀ465 ºC, and
23

Catalysis Science & Technology
Page 24 of 34
View Article Online
DOI: 10.1039/C7CY00970D
4
4
4
4
4
4
4
4
4
4
4
4
4
78
79
80
81
82
83
84
85
86
87
88
89
90
3
the second stage loss was caused by the calcination of CaCO present in the
52
catalysts. These mass losses were confirmed by the two endothermic peaks at
around 440°C and 700°C in the DSC curve. In addition, the consumption amount of
H
2
O by reacting with CaO could be estimated by the TG results, approximately
14.2ꢀ16.2% of the overall feed of H O.
2
The yield and selectivity of carbon deposition on the used Niꢀ and Feꢀbased
catalysts is shown in Table 3. It can be seen that the yield and selectivity of carbon
deposition on Feꢀbased catalysts was generally higher than that for the Niꢀbased
catalysts in the PPC system. In addition, the catalysts in the IPC system showed
higher resistance to carbon deposition than in the PPC system, especially for
2
Ni/ZSMꢀ5. Ni/SiO and Ni/ZSMꢀ5 were the most potential in the resistance to carbon
51,53,54
deposition, much better than the metal catalysts reported in previous studies.
4
4
91
92
(a)
(b)
4
93
4
4
4
4
94
95
96
97
(c)
Figure. 9. TGꢀDSC of the used Niꢀ and Feꢀbased catalysts supported by (a) SiO
(c)γꢀAl and (d) ZSMꢀ5.
(d)
2
, (b) CaO,
2
O
3
24

Page 25 of 34
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C7CY00970D
4
4
5
5
98
99
00
01
Table 3 Selectivity of carbon deposition on the catalysts
PPC system
IPC system
Catalyst
Ni/SiO
0.003
0.57
2
Ni/γꢀAl
2
O
3
Ni/ZSMꢀ5 Fe/SiO
2
Fe/γꢀAl
2
O
3
Fe/ZSMꢀ5
0.089
Ni/SiO
0.003
0.59
2
Ni/γꢀAl
2
O
3
Ni/ZSMꢀ5
0.055
Yield (g/gꢀcat)
SC (%)
0.092
17.6
0.066
11.8
0.014
3.0
0.103
19.8
0.098
18.0
16.9
8.9
5
02
03
5
3.5 Mechanism of toluene reforming
The reaction mechanism of toluene cracking in DBD plasmaꢀcatalysis system is
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
55
complex, containing multiple electronic, ionic and free radical reactions. The liquid
byꢀproducts were analyzed for a better understanding of the reaction pathway and
mechanism of toluene cracking, as shown in Figure 10. It can be seen from Figure 10
(a) and (b) that the liquid byꢀproducts from toluene cracking in the plasmaꢀalone (PA)
and the postꢀplasma catalysis (PPC) system contained almost the same components.
Benzene and ethylbenzene were identified as the dominant liquid byꢀproducts,
2
however, the selectivity of them was much less than that of H and CO. Some other
hydrocarbons, such as oꢀxylene, styrene, cumene, nꢀpropylbenzene, isobutylbenzene,
1, 2ꢀdiphenylethane and 2ꢀmethylꢀ3ꢀphenylbutane, were also detected as the minor
composition in the liquid byꢀproducts. The formation of Nꢀcontaining benzonitrile
suggested that the N radicals also participate in the reaction. However, the amount of
2
benzonitrile was limited, which indicated that the cleavage of N was difficult and
*
36
most N was kept as a stable molecular state or nitrogen excited state (N ). The
2
2
formation of oxygenates (dimethyl phthalate and diethyl phthalate) could be attribute
to the presence of H O, OH· and O· during toluene cracking. Besides, the chain
hydrocarbons, 1, 6ꢀHeptadienꢀ3ꢀyne and 1, 6ꢀheptadiyne, were also detected due to
2
21
the cleavage of the benzene ring followed by the hydrogenation. The only difference
between the liquid byꢀproducts in the PA system and PPC system seems to be that the
peak area of liquid byꢀproducts in the PPC system was much less than that in the PA
system. This result indicated that the addition of catalysts in the downstream of the
discharge zone decreased the formation of liquid byꢀproducts, however, it had little
25

Catalysis Science & Technology
Page 26 of 34
View Article Online
DOI: 10.1039/C7CY00970D
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
effect on the reaction mechanism of toluene cracking.
Figure 10 (c) presented the GCꢀMS spectrogram of liquid byꢀproducts in the IPC
system with Ni/ZSMꢀ5 catalyst. The formation of liquid byꢀproducts was undesirable,
5
6
which would lead to the secondary pollution and weaken the discharge. It can be
seen that the amount of the liquid byꢀproducts in the IPC system decreased sharply
and the composition was quite different from that in the PO and PPC systems. The
dominant liquid byꢀproducts were identified as benzene, ethylbenzene and 1,
2ꢀdiphenylethane, similar with that in the PO and PPC systems. However, other
hydrocarbon species, such as pꢀxylene, biphenyl, diphenylmethane, (E)ꢀstilbene and 2,
3ꢀdiphenylbutane, were newly formed in the IPC system with Ni/ZSMꢀ5. In addition,
the species of oxygenꢀcontaining compounds in the IPC system (benzyl benzoate and
mesoꢀhydrobenzoin) were also different, and their amounts were minor compared
with the PO and PPC systems. All the differences might be attributed to that the
catalysts in the IPC might influence the reaction mechanism of toluene cracking. In
the IPC system, the oriented formation of H and CO, through the reaction of the
2
*
active species (H
2
O, OH·, O·, N
2
…) with toluene and its cracked intermediates, was
enhanced in the effect of Niꢀbased catalysts in the IPC system. Although the same
active free radicals might also be formed in the PO and PPC systems, the formation of
2
3
products was uncontrollable in PA system without catalysts. For the PPC system, the
*
short lifetime reactive plasma species, such as OH·, O· and N
2
, were not able to
22
reach the catalysts installed in the downstream of the plasma zone. Therefore, the
reaction mechanism in the PPC system was almost not changed compared with that in
the PA system. This result suggested that the IPC system with Ni/ZSMꢀ5 was the most
potential in inhibiting the formation of undesirable liquid byꢀproducts and promoting
the formation of H and CO.
2
26

Page 27 of 34
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C7CY00970D
1
2
5
15
1
3
10
1
4
9
11 12
7
6
8
3
4
5
5
52
53
(a)
1
2
5
15
1
3
1
0
7
9
11
12
1
4
6
8
5
5
54
55
(b)
OH
13
1
2
OH
5
1
8 11
1
0
7
17
20
21
16
23
1
9
2
2
5
5
56
57
(c)
5
5
5
5
5
5
5
58
59
60
61
62
63
64
Figure 10. GCꢀMS spectrogram of liquid byꢀproducts in the (a) plasmaꢀonly system without
catalysts, (b) the PPC system and (c) IPC system with Ni/ZSMꢀ5.
(1. benzene, 2. toluene, 3. 1,6ꢀHeptadienꢀ3ꢀyne, 4. 1,6ꢀheptadiyne, 5. ethylbenzene, 6. oꢀxylene, 7.
styrene, 8. cumene, 9. dimethyl phthalate, 10. nꢀpropylbenzene, 11. benzonitrile, 12. isobutylbenzene,
13. 1, 2ꢀdiphenylethane, 14. 2ꢀmethylꢀ3ꢀphenylbutane, 15. diethyl phthalate, 16. pꢀxylene, 17. biphenyl,
18. benzyl benzoate, 19. phenol, 20. diphenylmethane, 21. mesoꢀhydrobenzoin, 22. (E)ꢀstilbene, 23. 2,
3ꢀdiphenylbutane)
5
65
27

Catalysis Science & Technology
Page 28 of 34
View Article Online
DOI: 10.1039/C7CY00970D
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
The possible reaction pathway of toluene in the IPC system with Ni/ZSMꢀ5 was
shown in Figure 11. The major reactions of toluene reforming occurred in the plasma
were activated by the highꢀenergy electrons, free radicals (i.e., OH·, O·, N·) and
*
57
2
excited species (i.e., N ). The toluene destruction mainly contained two processes:
direct destruction and secondary oxidation. The direct destruction of toluene
proceeded as route IꢀV and VII. Compared with the CꢀH bonds in the benzene ring
(4.3 eV), CꢀC bones between the benzene ring and methyl (4.4 eV), the CꢀC bones
(5.0–5.3 eV) and C=C bones (5.5 eV) in the benzene ring, the CꢀH bonds in the
58
methyl are easier (3.7 eV) to be cleavage. Therefore, the H atoms in the methyl were
firstly removed to form benzyl (C CH ·), benzylidene (C CH=) and benzenyl
6
H
5
2
6 5
H
6 5 3 6 5
(C H C≡). The benzyl combined with a methyl (CH ·), phenyl (C H ·) and hydroxyl
(OH·) to generate ethylbenzene, diphenylmethane and benzyl alcohol. The produced
benzyl alcohol and ethylbenzene would autoꢀcombined to produce
mesoꢀhydrobenzoin and 2, 3ꢀdiphenylbutane. Ethylbenzene could also react with a
methyl or benzyl to form nꢀpropylbenzene or 1, 2ꢀdiphenylethane. Although the 1,
2ꢀdiphenylethane could also be produced by the recombination of nꢀpropylbenzene
and phenyl, nꢀpropylbenzene and phenyl were more difficult to be obtained and their
amounts were also limited compared with ethylbenzene and benzyl as shown in
Figure 10. Therefore, 1, 2ꢀdiphenylethane, as the one of dominant liquid byꢀproducts,
was tended to be produced by the combination of ethylbenzene and benzyl.
Benzonitrile in the liquid products comes from the combination of benzenyl and N
free radical. The CꢀC bone between methyl and benzene ring was easy to be broken
*
by N
2
and energetic electrons to produce benzyl. Benzyl radicals reacted with H·,
OH· and benzyl, forming benzene, phenol and biphenyl. The H atom on the
paraꢀposition of methyl could also be activated and replaced by a methyl, generating
pꢀxylene. The secondary oxidation as shown in route VI and VIII might be explained
as the oxidation of toluene in the presence of O· and OH·, and form benzyl benzoate
5
7
eventually. The formation of CO could also be attributed to the oxidation of toluene
21
and its intermediates by O· and OH·.
28

Page 29 of 34
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C7CY00970D
5
5
96
97
Figure 11. Reaction pathway of toluene reforming over Ni/ZSMꢀ5 (inꢀplasma catalysis system).
5
98
5
99
4. Conclusion
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
The oriented conversion of model tar compound toluene to H ꢀrich syngas was
2
conducted in a hybrid DBD plasmaꢀcatalysis system. Niꢀ and Feꢀbased catalysts with
different supports (CaO, SiO
catalytic activity was evaluated. The catalytic activity decreased in sequence of
M/ZSMꢀ5 > M/γꢀAl > M/SiO > M/CaO (M represents Ni or Fe), which was
2 2 3
, γꢀAl O and ZSMꢀ5) were characterized and their
2
O
3
2
positively related to the BET surface area. Catalysts supported by ZSMꢀ5 showed the
most potential in converting toluene to syngas, inhibiting the formation of undesirable
liquid byꢀproducts and carbon deposition, and resistance of deactivation caused by
sintering and carbon deposition. M/CaO had little effect on toluene conversion, and
the formation of syngas was almost decreased in the presence of M/CaO. It was
attributed to that the support CaO reacted with H O, forming Ca(OH) , and thus the
2
2
steam reforming reaction and waterꢀgas shift reaction were weakened. Niꢀbased
catalysts exhibited higher catalytic activity than Feꢀbased catalysts on toluene
cracking, due to the higher activity of CꢀC and CꢀH bonds in hydrocarbons on the
surface of Ni particles. However, the syngas production was higher catalyzed by
29