Interaction of the reaction intermediates in co-reforming of acetic acid and ethanol impacts coke properties

Article abstract of DOI:10.1016/j.mcat.2021.111461

Bio-oil is a mixture of organics and its steam reforming for production of hydrogen has attracted extensive attention. During steam reforming, the reaction intermediates derived from the varied organics in bio-oil might interact with each other, impacting composition/properties of the products. In this study, steam reforming of single or mixture of acetic acid and ethanol, the representatives of carboxylic acid and alcohol, were conducted with Ni/KIT-6 as the catalyst, aiming to probe the potential interaction of the reaction intermediates on the reforming reaction. The results showed that the interactions of the derivatives from acetic acid and ethanol did exist, enhancing conversion of reactants, formation of gaseous by-products and affecting coke properties. The co-reforming of ethanol and acetic acid generates more reaction intermediates bearing hydroxyl group that are difficult to be gasified and also forms more coke, leading to more significant catalyst deactivation. The coke formed from the co-reforming is not only prone to oxidation, but also showed distinct morphologies. Carbon nanotubes with coarse surface and hillocks was the main form of coke from steam reforming of ethanol or acetic acid, while coke from the co-reforming was mainly in the form of carbon nanofiber decorated with the hangnail-like substance or rope-like structures with no cavity and micrometers of length.

Full text of DOI:10.1016/j.mcat.2021.111461

Molecular Catalysis 504 (2021) 111461
Contents lists available at ScienceDirect
Molecular Catalysis
journal homepage: www.journals.elsevier.com/molecular-catalysis
Interaction of the reaction intermediates in co-reforming of acetic acid and
ethanol impacts coke properties
Zhiran Gaoa, Guoming Gaoa, Chao Lia, Hongli Tiana, Qing Xua, Shu Zhangb, Leilei Xuc,
,
Xun Hua *
a School of Material Science and Engineering, University of Jinan, Jinan, 250022, PR China
b College of Materials Science and Engineering, Nanjing Forestry University, Nanjing, 210037, PR China
c School of Environmental Science & Technology, Nanjing University of Information Science & Technology, Nanjing, 210044, PR China
A R T I C L E I N F O
A B S T R A C T
Keywords:
Bio-oil is a mixture of organics and its steam reforming for production of hydrogen has attracted extensive
attention. During steam reforming, the reaction intermediates derived from the varied organics in bio-oil might
interact with each other, impacting composition/properties of the products. In this study, steam reforming of
single or mixture of acetic acid and ethanol, the representatives of carboxylic acid and alcohol, were conducted
with Ni/KIT-6 as the catalyst, aiming to probe the potential interaction of the reaction intermediates on the
reforming reaction. The results showed that the interactions of the derivatives from acetic acid and ethanol did
exist, enhancing conversion of reactants, formation of gaseous by-products and affecting coke properties. The co-
reforming of ethanol and acetic acid generates more reaction intermediates bearing hydroxyl group that are
difficult to be gasified and also forms more coke, leading to more significant catalyst deactivation. The coke
formed from the co-reforming is not only prone to oxidation, but also showed distinct morphologies. Carbon
nanotubes with coarse surface and hillocks was the main form of coke from steam reforming of ethanol or acetic
acid, while coke from the co-reforming was mainly in the form of carbon nanofiber decorated with the hangnail-
like substance or rope-like structures with no cavity and micrometers of length.
Steam reforming
Acetic acid
Ethanol
Reaction intermediate
Coke
1. Introduction
Bio-oil is composed by the condensable volatiles from pyrolysis of
the renewable biomass [25,26], which can be abundantly produced and
Hydrogen is clean fuel for fuel cells and its production has always
been a hot research topic since decades ago [1,2]. Hydrogen can be
produced from various carbonaceous feedstocks via the processes such
as steam reforming [3,4], partial oxidation [5,6], oxidative reforming
(gasification) [7,8], etc. The carbonaceous feedstock can be fossil fuels
(coal, petroleum-based fuels, natural gas) [9,10], small organics like
ethanol [11,12], acetic acid [13,14] and methanol [15,16], the organic
mixture like bio-oil [17–19], etc. The availability of the feedstock de-
termines feasibility and sustainability of the process. Fossil fuels are
nonrenewable and attention for the production of hydrogen has shifted
to the feedstock such as the small organics mentioned above [20–22].
However, the small organics like ethanol is an important raw feedstock
for the synthesis of other fine chemicals and can also be directly used as
the fuel for vehicles [23,24]. Other renewable while abundantly avail-
able feedstock with the low production cost is highly demanded for
hydrogen production.
available in most of the regions with the growth of biomass. Therefore,
bio-oil has been considered as an important feedstock for hydrogen
production and steam reforming of bio-oil has been extensively inves-
tigated [27–30]. However, inheriting from the complex nature of
biomass, the organics in bio-oil are rather complicated in terms of
structures, functionalities, molecular sizes, etc [31–33]. Most impor-
tantly, the organics in bio-oil tend to polymerize or crack at the elevated
temperature, resulting in formation of coke and rapid deactivation of
catalysts [34–36]. The coke formation relates to the reaction behaviors
of the individual organic compound in bio-oil [37,38]. Nevertheless,
when a mixture of organics is formed simultaneously on the same
catalyst, a natural question is raised. Do these molecules of different
structures and properties interact with each other? If they do, do the
interactions affect their coking tendency and the formation of the re-
action intermediates. It is known under the typical reforming conditions,
the radicals could form. The varied feedstock would generate the
* Corresponding author.
E-mail address: Xun.Hu@outlook.com (X. Hu).
https://doi.org/10.1016/j.mcat.2021.111461
Received 29 December 2020; Received in revised form 26 January 2021; Accepted 9 February 2021
Available online 25 February 2021
2468-8231/© 2021 Elsevier B.V. All rights reserved.

Z. Gao et al.
Molecular Catalysis 504 (2021) 111461
Fig. 1. The characterization of physical and chemical properties of calcined catalyst. (a): the XRD characterization of the calcined catalyst and the reduced catalyst;
(b): the isothermal curves of BJH desorption hole over the calcined catalyst, the symbol of circle means adsorption isotherm curves and the square means desorption
isotherm; (c): volume distribution diagram of BJH desorption hole; (d): the TPR characterization of the calcined catalyst.
radicals of different structures. In theory, these radicals could react with
each other and hence the interaction of the reactants of different
structures seems inevitable. However, this is still a knowledge gap to be
filled to understand the complex reaction network during steam
reforming of bio-oil that generally contains hundreds of organics
[39–41].
at room temperature for 24 h, and then in an oven of 60 ◦C for another
24 h. After drying, the solid sample was calcined by heating in static air
to 600 ◦C at the rate of 20 ◦C/min and holding for 2 h in a muffle furnace
to obtain the calcined catalyst.
2.2. Catalytic experiment
As a proof of concept, in this study, acetic acid and ethanol, which
could represent the carboxylic acids and alcohols in bio-oil, are selected
as the model compounds to investigate their potential interactions.
Particularly, the potential interactions of the reaction intermediates in
the co-reforming of acetic acid and ethanol on the properties of the coke
formed and the reaction intermediates produced were focused. Since
catalyst development is not the aim of this study, a typical reforming
catalyst, Ni/ KIT-6 was used for reforming of the organics. The results
indicated that the interaction between the reaction intermediate pro-
duced from steam reforming of acetic acid and ethanol did exist which
significantly impacted the distribution of the products and the proper-
ties/morphologies of the coke formed.
The steam reforming of acetic acid, ethanol and their mixture was
carried out in a fixed bed continuous flow reactor. The configuration of
the reactor could refer to our previous work [43]. 0.5 g of calcined
catalyst was loaded in a quartz reactor and was placed in the center of
the furnace. Before the start of the reforming reaction, the calcined
catalyst was reduced at 600 ◦C at the heating rate of 20 ◦C/min for 1 h in
the mixed atmosphere of H2/N2 (H2 = 60 mL/min; N2 = 60 mL/min).
The reforming experiment was then started when the reaction temper-
ature was decreased to the preset value. The reactant(s) was pumped
into the reaction tube at a flow rate (0.12 mL/min) through a peristaltic
pump. In the co-reforming of acetic acid and ethanol, the steam to
carbon ratio (S/C) was adjusted to the same as that in steam reforming of
the single feedstock. In the mixture, the molar ratios between ethanol
and acetic acid was 1 : 1. In the steam reforming, N2 with a flow rate of
60 mL/min was used as a carrier gas. After holding at each reaction
temperature for 30 min, the liquid products were collected by a
gas-liquid separator, and the gas products were collected by a gas bag
and were analyzed with Gas Chromatography (GC). After the experi-
ment, the spent catalyst was cooled to room temperature under the
protection of N2, and was then collected and further analyzed. The
calculation formulas for the conversion of the organics and the yields of
the gases were as follows.
2. Experiment
2.1. Preparation of catalyst
KIT-6 was prepared with triblock copolymer (P123) as template,
tetraethyl silicate as silicon source and n-butanol as coexisting solute,
according to that specified in literature [42]. In the preparation of
Ni/KIT-6 catalyst, nickel was loaded on the carrier KIT-6 by an incipient
impregnation method. Nickel nitrate was selected as nickel source. A
certain amount of nickel nitrate was dissolved in deionized water to
obtain nickel nitrate solution. After that, it was mixed with KIT-6, aged
2

Z. Gao et al.
Molecular Catalysis 504 (2021) 111461
Am ount of acetic acid/ethanol in feedstock ꢀ Am ount of acetic acid/ethanol
in residual
Conversion =
× 100%
Am ount of acetic acid/ethanol in feedstock
Moles of CO/CH4/CO2produced
2 × Moles of acetic acid/ethanol
YieldCO/CH
=
× 100%
/CO2
4
Moles ofH2produced
Theoretical ofH2from the acetic acid/ethanol fed in
Yield H
=
× 100%
2
2.3. Characterization and analytical methods
Brunauer Emmet Teller (BET) surface areas of the catalysts were
measured by using the Biaode SSA 6000 instrument, and the phase
characteristics of catalysts were analyzed by Powder X ray diffraction
(XRD) of Rigaku Ultima IV. The reduction behaviors of nickel oxide were
measured (temperature-programmed reduction in hydrogen, H2-TPR)
by using a chemisorption analyzer of Biaode PCA-1200.
The amount of coke formed was measured by using thermogravi-
metric analysis (TG) with HCT-1 (Henven) instrument. The functional-
ities of coke were analyzed by Fourier transform infrared (FT-IR) with
the Nicolet IS 50 instrument. The micromorphology of coke and the
spent catalyst was obtained by the characterization of with a trans-
mission electron microscope (TEM, JEOL2010plus). The details for the
above characterizations were described in our previous studies [44].
3. Result and discussion
3.1. Characterization of catalyst
3.1.1. XRD characterization
The XRD patterns of the calcined catalyst and reduced Ni/KIT-6
catalyst were shown in Fig. 1a. In Fig. 1a, the diffraction peaks with
2θ of 37.4◦, 43.4◦, 56.4◦, 63.1◦ and 75.6◦ belonged to the characteristic
diffraction peaks of NiO (JCPD number 65–5745), and the diffraction
peaks with 2θ angles of 44.5◦, 51.9◦ and 76.5◦ belonged to the charac-
teristic diffraction peaks of metallic nickel (JCPD number 65–2865). The
carrier KIT-6 existed in the form of amorphous SiO2, and the charac-
teristic diffraction peak was located at 22.7◦. According to Scherrer
formula, the grain size of nickel oxide in the calcined catalyst was 12.9
nm and that of nickel metal in the reduced catalyst was 10.3 nm. In
theory, if no migration of metallic nickel took place, the particle size of
nickel in the metallic form should be even smaller than that of NiO as the
removal of oxygen would remarkably decrease the particle sizes. How-
ever, the migration and aggregation of nickel species almost cannot be
avoided, which relates to both the pore structure of the catalyst and the
interaction between the carrier and nickel species. The pore structures of
the catalyst were firstly analyzed.
3.1.2. BET characterization
The isothermal curve of calcined catalyst and the volume distribu-
tion of BJH desorption pore were shown in Fig. 1b and c, respectively.
The specific surface area for the calcined catalyst was 360.6 m2/g. The
average mean radius of the particles was 32.7 Å, and the pore volume
was 0.59 cm3/g. The results in Fig. 1a indicated that the adsorption
branch measured with the increase of the equilibrium pressure and the
desorption branch measured with the decrease of the pressure did not
coincide in a certain relative pressure range, and the type of hysteresis
loop belongs to H1 type. The formation of H1 type hysteresis loop was
the behavior of cylindrical pores with uniform diameter distribution at
both ends in the adsorption and desorption process, which could be
Fig. 2. The yield of the main products in the steam reforming of different
substrates with the 18 %Ni/KIT-6 as the catalysts. Reaction condition: Catalyst
loading: 0.5 g; LHSV =6.9 hꢀ 1; S/C = 5; P =1 atm.
3

Z. Gao et al.
Molecular Catalysis 504 (2021) 111461
Fig. 3. The In-situ DRIFTS characterization of steam reforming of different reaction substrates. (a): acetic acid; (b): ethanol; (c): mixture of acetic acid and ethanol.
observed in mesoporous materials with uniform size distribution [45].
From the volume distribution diagram, it could be verified again that the
pore size of the catalyst was mainly concentrated in the range of 30ꢀ 35
Å, which belongs to mesoporous materials. Mesoporous materials with
high specific surface area could enhance the interaction between the
carrier and metal active materials and improved the dispersion of metal,
which could maximize the utilization of active materials. In order to
further study the interaction between carrier and the nickel species, the
TPR characterization was carried out subsequently.
6 as the catalyst in the steam reforming of single ethanol, acetic acid and
their mixture were shown in Fig. 2. The higher temperature enhanced
the catalytic activity of Ni/KIT-6, facilitating the conversion of acetic
acid and ethanol as well as the formation of the gaseous products.
Interestingly, acetic acid and ethanol showed the varied reaction be-
haviors when they were reformed together. An apparent phenomenon
was that the conversion of both acetic acid and ethanol (Fig. 2c) was
higher than that in steam reforming of the single feedstock. The distri-
bution of the gaseous products was also different in the co-reforming and
in the reforming of the single feedstock. For example, the yields of CH4
were higher with the increased reaction temperature in the co-reforming
reactions than that in reforming of the single feedstock. Similar phe-
nomenon was also observed for the production of CO. The production of
CH4 and CO could be originated from the methanation reaction and the
reverse water gas shift reaction [47,48], respectively. Nevertheless, they
could also be produced from the decomposition of the reactants and the
radical reactions between the reaction intermediates. The co-existence
of ethanol and acetic acid in the reaction mixture should have induced
the interactions of the reaction intermediates, which not only changed
the conversion of acetic acid and ethanol, but also significantly impacted
the formation of the gaseous products. The reaction intermediates
formed in steam reforming of the single or the mixed feedstocks were
further characterized with the in situ DRIFTS studies.
3.1.3. TPR characterization
The TPR characterization results of the calcined catalyst were shown
in Fig. 1d. Three main reduction peaks could be seen at 237 ◦C, 408 ◦C
and 590 ◦C, respectively. The reduction peak at 237 ◦C was not obvious
and existed in the form of small shoulder peak. The reduction peak at
408 ◦C had the largest area while the reduction peak at 590 ◦C was
significant in terms of the peak area. The reduction temperature of the
unsupported nickel oxide to nickel was generally around 400 ◦C [46].
The higher reduction temperature indicated stronger interaction be-
tween NiO and carrier. In addition, the mass transfer of hydrogen to the
inner pores of the catalyst also affects the reduction of NiO inside the
pores. The reaction behaviors of ethanol, acetic acid and the mixture of
ethanol/acetic acid over the Ni/KIT-6 catalyst was further investigated
to explore the interaction of the reaction intermediates in the co-steam
reforming reactions.
3.2.2. In situ DRIFTS studies of steam reforming of different reaction
substrate
The results of in situ DRIFTS studies of steam reforming of three
different reaction substrates were shown in Fig. 3. The infrared char-
acteristic peaks at 2848 cmꢀ 1 and 2932 cmꢀ 1 were ꢀ CH2, [49], which
were the decomposition products of small organic molecules. The
3.2. Catalytic performance
3.2.1. Catalysis experiment
The gaseous products versus the reaction temperatures with Ni/KIT-
4

Z. Gao et al.
Molecular Catalysis 504 (2021) 111461
each peak, Table 2 was obtained. With the increase of reaction tem-
perature, the reaction speed was accelerating, and the peak value was
also gradually increasing. For the reforming of acetic acid or ethanol, the
vibration absorption of CO2 was not found at the low temperature
(100ꢀ 200 ◦C), but began to appear at 400 ◦C. Different from the single
substrate, the vibration peak of CO2 appeared in the steam reforming
reaction of mixture at the temperature even as low as 300 ◦C, and the
peak intensity was much higher than that of single substrate. The
interaction between acetic acid and ethanol made the reforming reac-
tion possible at relatively low temperature. The vibrational absorption
–
–
C H was more significant in the reforming reaction of acetic
of olefin
–
acid, but not appeared in the reforming reaction of ethanol, Meanwhile,
the peak value was not significant in the mixture. This could be due to
that the addition of ethanol could relatively inhibit the formation of
olefins from the dissociation of acetic acid, resulting in a decrease in the
–
–
C H in the steam reforming of the mixture when
peak strength of
–
compared with that in steam reforming of acetic acid. In addition, in
steam reforming of the mixed ethanol and acetic acid, the intensity for
the hydroxyl group absorption became more significant. The absorption
peak should be largely from the organic reactant as the catalyst and the
proportion of water in the reactants remained the same. The higher
abundance of the hydroxyl group indicated that the interaction between
the reaction intermediates formed the organic species with the abundant
hydroxyl group and they were more difficult to be gasified. They could
be the precursor of coke.
The peak value of ꢀ CHx was the lowest in the steam reforming of the
mixture, which may be due to the further conversion of ꢀ CHx to CH4, as
the higher yield of CH4 was obtained in steam reforming of the mixture
(Fig. 2). Similarly, compared with the steam reforming of single reac-
tant, the peak value of vibration absorption peak of CO was higher and
the yield of CO was the highest in the mixture reforming, suggesting that
the interaction between acetic acid and ethanol probably promoted the
–
–
–
– –
O C, C O and C C all
–
formation of CO. The functional groups C
came from the decomposition and further combination of organic
compounds. The increase of reaction temperature enhanced the abun-
dance of these functional groups increased, as the steam reforming was
accelerated. When the reaction temperature was 600 ◦C, it was found
that the peak value of the three functional groups in ethanol was the
highest, while the peak value in the mixture decreased. This further
indicated the interaction between acetic acid and ethanol, and also the
interaction between the reaction intermediates modified the reaction
intermediates. In order to further explore the impacts of steam reform-
ing of the mixed ethanol and acetic acid on the formation of coke, the
steam reforming of the single and the mixed feedstock versus the pro-
longed reaction time was carried out.
3.2.3. Steam reforming versus the prolonged reaction time
The conversion of the reactants and the yields of gas products versus
the prolonged reaction time at the reaction temperature of 600 ◦C and
the S/C of 2 were shown in Fig. 4. In the steam reforming of acetic acid,
although the conversion of acetic acid was relatively stable with the
extension of time-on-stream, the percentage for the decrease of H2 at the
end of the experiment was 28.6 % (Fig. 4a). The deactivation of the
catalyst probably led to the conversion of acetic acid to the organic by-
products, diminishing the production of H2. In the steam reforming of
ethanol, deactivation of the catalyst was also observed, as evidenced by
the decrease the yield of H2 by 25 % (Fig. 4b). In the steam reforming of
ethanol and acetic acid mixture (Fig. 4c), the catalyst deactivation was
obvious, and the deactivation degree was far greater than that of single
organic compounds. After 5 h of time-on-stream, the yield of H2
decreased by 44 %. In addition, the production of CO and CH4 in the
steam reforming of the mixture was higher than that in steam reforming
of the single substrate. This once again verified the interaction between
acetic acid and ethanol or their derived reaction intermediates in the
reforming process affected the reaction network and the formation of
gas products. Furthermore, the amount of coke formed in the reforming
Fig. 4. The yield of the main products in the steam reforming of different
substrates versus the prolonged reaction time. Reaction conditions: catalyst
loading: 0.3 g; LHSV =6.9 hꢀ 1; S/C: 2; reaction temperature: 600 ◦C; reaction
time: 300 min.
infrared absorption peaks at 2307 cmꢀ 1 and 2384 cmꢀ 1 belonged to the
–
vibration absorption of CO . The stretching vibration of C O at 1821
–
2
ꢀ 1
cmꢀ 1, C H at 2876 cm and C C at 1585 cmꢀ 1 were all functional
–
–
–
–
groups formed by the decomposition of organic compounds during the
steam reforming. The absorption peak of Si-O-Si at 1000 cmꢀ 1 was
caused by the asymmetric stretching and bending vibration of Si-O-Si
framework in zeolite. Meanwhile, the infrared peak of Si-O at 770
cmꢀ 1 belonged to KIT-6. According to the statistics of the peak height of
5

Z. Gao et al.
Molecular Catalysis 504 (2021) 111461
Fig. 5. The various characterizations of the spent catalysts in the steam reforming of the different substrates versus the prolonged reaction time. (a): the TG-air
curves; (b): the DTG-air curves; (c) TPO curves; (d): the FT-IR curves; (e): the TG-N2 curves; (f): the DTG curves in the N2 atmosphere; (g) the XRD characteriza-
tions of the spent catalysts after the catalytic experiment (from Fig. 2); (h) the XRD characterizations of the spent catalysts after the 5 h of time on stream
(from Fig. 4).
of the single feedstock and the mixture should also be different, as the
catalyst deactivated to the different extent in terms of the yields of
hydrogen. This was further investigated via a series of characterization
of the spent catalysts to explore the influence of the co-reforming on
coke formation and the properties of coke.
3.3. Characterization of the spent catalyst
3.3.1. TG-air and TPO characterization
The TG-air characterizations of the three spent catalysts in an air
flow were shown in Fig. 5a. The mass of the three spent catalysts started
to increase from ca. 300 ◦C, which was attributed to the oxidation of
nickel in the reduced catalysts in the air stream. When the temperature
reached ca. 500 ◦C, the oxidation of the coke deposit started, resulting in
6

Z. Gao et al.
Molecular Catalysis 504 (2021) 111461
of some carbon nanotubes with higher graphitization degree. Similarly,
the main oxidation temperature of the coke formed from steam
reforming of the mixed ethanol and acetic acid was lower. The thermal
stability properties of coke generated from steam reforming of single
feedstock and coke formed by steam reforming of mixture gas were
different.
Table 1
The nickel metallic particle size of the different catalysts.
Catalysts
After catalytic experimenta
/
After stability testb
nm
/
nm
Ni/KIT-6
12.9c/10.3d
13.0
——
13.5
14.2
12.9
Acetic acid reforming
Ethanol reforming
The mixture
14.2
13.4
3.3.2. TG-N2 characterization
reforming
The results of thermogravimetric characterization of spent catalyst in
nitrogen atmosphere were shown in Fig. 5e and f. With the increase of
temperature, the weight loss of spent catalyst began to appear at 210 ◦C,
indicating that some components of the spent catalyst began to
decompose. When the temperature finally reached 800 ◦C, the order of
weight loss of spent catalyst was acetic acid > ethanol > mixed acetic
acid and ethanol. The weight loss rate was directly related to the sta-
bility of carbon deposition. The higher the weight loss rate is, the lower
the thermal stability of carbon deposition is. Therefore, it was consistent
with the thermal stability of coke obtained by TPO (Fig. 5c). In the
reforming of acetic acid, the proportion of vaporized coke to total coke
was 35.3 % and 27.3 % in ethanol. The difference between them was
relatively small. In the reforming of mixed acetic acid and ethanol, the
proportion of vaporized coke was 57.0 %, which was obviously different
from that of single material. This further proved that there was inter-
action between acetic acid and ethanol in the mixture, which affected
the specific gravity of vaporized coke and the structure of coke.
a
b
c
The conditions for the catalytic tests were specified in Fig. 2.
The conditions for the stability tests were specified in Fig. 4.
Means the particle size of nickel oxide in calcined catalyst.
Means the metallic nickel in reduced catalyst.
d
a sharp downward trend of the TG curves. According to the calculation,
13.3 % of coke in steam reforming of ethanol, 23.1 % coke in steam
reforming of acetic acid, and 24.9 % of coke in steam reforming of the
mixture were deposited in the spent catalysts. Assuming that there was
no interaction between acetic acid and ethanol in the reforming process,
the total carbon deposition of steam reforming of the mixture of acetic
acid and ethanol with the molar ratio of 1 : 1 should be half of that of
acetic acid and ethanol, which was 18.2 %. However, the actual amount
of coke formed in the spent catalyst obtained from the steam reforming
of the mixture was 24.9 %, which was an 36.8 % increase the theoretical
calculation value. It was confirmed herein again that, in the steam
reforming of acetic acid and ethanol mixture, the interaction of the re-
action intermediates led to the formation of significantly more coke
deposit. The DTG curves in Fig. 5b indicated that the coke deposit
formed in steam reforming of ethanol was more resistant to oxidation
and followed by that in steam reforming of acetic acid. The coke formed
in the co-reforming, however, was prone to oxidation. This again
confirmed that the properties of the coke formed in steam reforming of
the single feedstock were different from that formed in the co-reforming
of the mixture. The oxidation behaviors of the catalyst were further
investigated with the TPO characterizations, as shown in Fig. 5c. The
oxidation peak formed at lower temperature might be attributed to the
oxidation of amorphous coke. The oxidation peak formed at the higher
temperature was more significant, which might be due to the oxidation
3.3.3. XRD and FT-IR characterization
The XRD patterns of the spent catalysts collected were shown in
Fig. 5g and h, respectively. The diffraction peaks with 2θ of 44.5◦, 51.9◦
and 76.5◦ were the characteristic diffraction peaks of Ni (JCPD
65–2865). The bulge at the 2θ of ca. 21◦ was amorphous SiO2. In
addition, the bulge of carbon was observed at the position of grain size of
nickel in the spent catalysts calculated with Scherrer formula was shown
in Table 1. Comparing with the nickel particle size in the reduced
catalyst, the nickel particle size in the spent catalyst increased to some
extent, especially in steam reforming of single ethanol. However, in
steam reforming of the mixture, the increase of the nickel particle size
Table 2
The intensity value of absorption peaks in in-situ DRIFTS spectrum.a
ꢀ 2c
ꢀ 2 e
ꢀ 2 g
ꢀ 2 h
ꢀ 2 i
Reforming reaction
T/◦C
CO2/ 10ꢀ 2 b
C
CH2/ 10ꢀ 2 d
C
CO/ 10ꢀ 2 f
O/ 10
–
–
–
C
–
–
C C/ 10
–
H/ 10
– –
O C/ 10
–
O H / 10
–
100
200
300
400
500
600
100
200
300
400
500
600
100
200
300
400
500
600
——
——
——
1.3
0.1
6.8
6.9
7.0
7.4
9.0
3.9
5.2
5.6
6.0
6.4
6.7
1.0
1.4
1.6
1.9
3.0
4.5
1.3
1.6
1.9
2.1
2.3
2.3
0.85
4.0
0.8
1.1
1.7
2.1
2.6
2.9
0.3
1.0
1.3
1.4
1.4
1.5
1.1
1.4
1.7
3.1
4.5
4.9
0.07
1.2
2.6
4.0
4.9
5.7
0.7
1.8
3.2
4.0
5.0
6.1
0.43
1.4
2.2
3.3
4.7
5.7
13.0
13.6
14.5
14.9
15.0
15.2
19.9
23.7
24.4
24.6
25.1
26.4
3.3
20.4
21.8
22.7
25.22
27.6
30.9
13.5
22.9
28.9
30.0
31.6
33.4
11.5
13.9
16.9
20.0
22.2
24.4
9.3
Acetic acid
11.6
18.7
25.5
3.9
1.4
1.7
——
——
——
0.3
——
——
——
——
——
——
1.5
7.1
13.7
14.8
24.4
37.1
4.6
Ethanol
2.4
3.0
——
——
1.6
1.7
4.7
12.1
17.2
18.3
18.0
19.8
1.8
6.7
Mixture
a
b
c
d
e
f
4.1
2.1
11.4
15.7
25.1
5.0
2.6
5.1
3.7
The data was derived from the results in Fig. 3.
The absorption peak of CO2 was calculated at ca. 2384 cmꢀ 1
.
ꢀ 1
–
–
The absorption peak of
The absorption peak of C
C
H was calculated at ca. 2876 cm
.
–
The absorption peak of CH2 was calculated at ca. 2932 cmꢀ 1
.
ꢀ 1
– –
O
C was calculated at ca. 1267 cm
.
The absorption peak of CO was calculated at ca. 1921 cmꢀ 1
.
g
h
i
ꢀ 1
ꢀ 1
ꢀ 1
–
The absorption peak of C O was calculated at ca. 1821 cm
.
.
.
–
–
The absorption peak of C C was calculated at ca. 1585 cm
–
–
The absorption peak of O H was calculated at ca. 3739 cm
7

Z. Gao et al.
Molecular Catalysis 504 (2021) 111461
Fig. 6. The TEM characterization of the spent catalysts in the steam reforming of the acetic acid versus the prolonged reaction time.
Fig. 7. The TEM characterization of the spent catalysts in the steam reforming of the ethanol versus the prolonged reaction time.
8

Z. Gao et al.
Molecular Catalysis 504 (2021) 111461
Fig. 8. The TEM characterization of the spent catalysts in the steam reforming of the mixed ethanol and acetic acid versus the prolonged reaction time.
was smaller. This might also due to the coverage of nickel phase with the
coke deposit by the different extent, which was further investigated.
The information of the functional groups of the three different spent
catalysts and fresh catalysts was shown in Fig. 5d. The characteristic
peak at about 3450 cmꢀ 1 corresponded to the vibration absorption of
that covered the surface of the catalyst was always observed. Basically,
carbon nanotubes were the main form of coke in the spent catalysts in
steam reforming of single acetic acid and single ethanol, which explains
the high resistivity of the coke towards oxidation. However, as for the
coke formed in the co-reforming of the mixture of acetic acid and
ethanol, the coke deposits of the varied morphologies were observed. As
shown in Fig. 8, the carbon nanofiber-like structure was observed. It
seems the structure was composed by the pile of the hangnail-like sub-
stance. The length could be a few micrometers (Fig. 8a) with the di-
ameters of from 20 to 40 nm. In addition, some rope-like structures with
very smooth surface was also observed (Fig. 8c and d). The length could
also be several micrometers. The structures were not the carbon nano-
tubes observed in steam reforming of acetic acid or ethanol. The for-
mation mechanism of these structures was obviously different from that
carbon nanotube. The results herein clearly demonstrated that the co-
reforming of acetic acid and ethanol produced the distinct reaction in-
termediates, which led to the formation of the carbon deposits of the
varied morphologies.
OH of the coke species and/or the adsorbed water. The absorption peaks
ꢀ 1
of C H and C
C were observed at 2919 cm and 1381 cmꢀ 1
,
–
– –
O
respectively, indicating that the catalyst contained some aliphatic car-
–
–
bon structures. In addition, the C C absorption was observed at 1618
cmꢀ 1, which probably belonged to the aromatic ring structure. The
infrared peak at 1085 cmꢀ 1 was caused by the asymmetric stretching
and bending vibration of the framework of silica tetrahedron, while the
infrared peak at 805 cmꢀ 1 corresponded to the symmetric stretching
vibration of silica oxygen bond. In addition to the crystal structures and
the functionalities of the spent catalysts, the morphologies of the coke
formed were further analyzed with the TEM characterizations.
3.3.4. TEM characterization
The results for morphologies of the coke in the spent catalysts were
shown in Figs. 6–8, respectively. In the steam reforming of acetic acid,
many carbon nanotube structures can be found. Nickel particles located
on top of the nanotubes. Some of the nanotubes have a big opening,
which was probably due to the fall off of the nickels during the prepa-
ration of the samples for the characterization. Depending on the particle
sizes of metallic nickel, the diameters of the carbon nanotubes varied
remarkably. In addition, the outer surface of the carbon nanotubes was
not smooth with many hillocks. Some coke deposits did not grow to form
carbon nanotubes, which might be the non-catalytic coke deposited on
the surface of the nanotubes. In the spent catalyst in steam reforming of
ethanol, similar carbon nanotubes were formed, and nickel particles
were also formed on the top of the nanotubes. The formation mechanism
of the nanotubes should be similar. In addition, some amorphous coke
4. Conclusion
In summary, the reaction behaviors of co-reforming of acetic acid
and ethanol were compared with that of reforming of single acetic acid
or ethanol. The results showed that, in the steam reforming of the mixed
ethanol and acetic acid, the interaction of the reaction intermediates
produced from the varied feedstock not only changed the conversion of
acetic acid and ethanol, but also significantly impacted the formation of
the gaseous products. The in-situ DRIFTS studies showed that the
interaction between the reaction intermediates formed the organic
species with the abundant hydroxyl group that were difficult to be
– –
O C functionalities of the reaction
gasified. The carbonyl and the C
intermediates were also distinct from that in steam reforming of the
9

Z. Gao et al.
Molecular Catalysis 504 (2021) 111461
[9] N. Piatkowski, C. Wieckert, A.W. Weimer, A. Steinfeld, Solar-driven gasification of
carbonaceous feedstock—a review, Synth. Lect. Energy Environ. Technol. Sci. Soc.
4 (2011) 73–82, https://doi.org/10.1039/C0EE00312C.
single feedstock. The catalyst also deactivated to the highest extent in
steam reforming of the mixed ethanol and acetic acid. The coke for-
mation was 36.8 % higher than the theoretical and averaged value in
steam reforming of the single feedstock. In addition, the coke formed
from the steam reforming of the mixture was less resistant to oxidation
and had a lower thermal stability than that in the steam reforming of the
single feedstock. The interaction of the reaction intermediates derived
from acetic acid and ethanol also changed morphologies of the coke
formed. The coke formed from steam reforming of ethanol or acetic acid
was mainly in the form of the carbon nanotubes but the surface was not
smooth and coated with hillocks. In comparison, the coke formed in
steam reforming of mixed ethanol and acetic acid was in the form of
carbon nanofiber-like structure piled with the hangnail-like substance.
Some rope-like structures with no cavity in the middle but with very
smooth surface and the length of several micrometers was also observed.
The formation mechanism of these structures was distinct from those
carbon nanotubes from the steam reforming of the single feedstock, due
to the interaction of the reaction intermediates derived from the con-
version of acetic acid and ethanol under the steam reforming conditions.
˜
[10] R. Navarro, M. Pena, J. Fierro, Hydrogen production reactions from carbon
feedstocks: fossil fuels and biomass, Chem. Rev. 107 (2007) 3952–3991, https://
doi.org/10.1021/cr0501994.
[11] C. Diagne, H. Idriss, A. Kiennemann, A. Hydrogen production by ethanol reforming
over Rh/CeO2–ZrO2 catalysts, Catal. Commun. 3 (2002) 565–571, https://doi.org/
10.1016/S1566-7367(02)00226-1.
[12] V. Klouz, V. Fierro, P. Denton, H. Katz, J. Lisse, S. Bouvot-Mauduit, C. Mirodatos,
Ethanol reforming for hydrogen production in a hybrid electric vehicle: process
optimization, J. Power Sources 105 (2002) 26–34, https://doi.org/10.1016/
S0378-7753(01)00922-3.
[13] A. Basagiannis, X. Verykios, Catalytic steam reforming of acetic acid for hydrogen
production, Int J Hydrogen Energ. 32 (2007) 3343–3355, https://doi.org/
10.1016/j.ijhydene.2007.04.039.
[14] S. Sanches, J. Huertas Flores, M. Paisda Silva, Cu/ZnO and Cu/ZnO/ZrO2 catalysts
used for methanol steam reforming, Mol Catal. 454 (2018) 55–62, https://doi.org/
10.1016/j.mcat.2018.05.012.
[15] Y. Lwin, W. Daud, A. Mohamad, Z. Yaakob, Hydrogen production from
steam–methanol reforming: thermodynamic analysis, Int J Hydrogen Energ. 25
(2000) 47–53, https://doi.org/10.1016/S0360-3199(99)00013-0.
[16] C. Yao, L. Wang, Y. Liu, G. Wu, Y. Cao, W. Dai, H. He, K. Fan, Effect of preparation
method on the hydrogen production from methanol steam reforming over binary
Cu/ZrO2 catalysts, Appl Catal A-Gen. 297 (2006) 151–158, https://doi.org/
10.1016/j.apcata.2005.09.002.
[17] P. Kechagiopoulos, S. Voutetakis, A. Lemonidou, I. Vasalos, Hydrogen production
via steam reforming of the aqueous phase of Bio-Oil in a fixed bed reactor, Energ
Fuel. 20 (2006) 2155–2163, https://doi.org/10.1021/ef060083q.
[18] W. Nabgan, T. Abdullah, R. Mat, B. Nabgan, Y. Gambo, M. Ibrahim, A. Ahmad,
A. Jalil, S. Triwahyono, I. Saeh, Renewable hydrogen production from bio-oil
derivative via catalytic steam reforming: an overview, Renew Sust Energ Rev. 79
(2017) 347–357, https://doi.org/10.1016/j.rser.2017.05.069.
CRediT authorship contribution statement
Zhiran Gao: Conceptualization, Methodology, Investigation, Vali-
dation, Writing - original draft, Visualization. Guoming Gao: Data
curation, Validation, Investigation, Writing - review & editing. Chao Li:
Software, Investigation, Visualization. Hongli Tian: Software, Valida-
tion, Investigation. Qing Xu: Methodology, Investigation. Shu Zhang:
Supervision, Project administration. Leilei Xu: Supervision, Writing -
review & editing. Xun Hu: Conceptualization, Methodology, Resources,
Writing - review & editing, Project administration, Funding acquisition.
[19] S. Chattanathan, S. Adhikari, N. Abdoulmoumine, A review on current status of
hydrogen production from bio-oil, Renew Sust Energ Rev. 16 (2012) 2366–2372,
https://doi.org/10.1016/j.rser.2012.01.051.
[20] S. Yoo, S. Park, J. Song, D. Kim, Hydrogen production by the steam reforming of
ethanol over K-promoted Co/Al2O3–CaO xerogel catalysts, Mol Catal. 491 (2020),
110980, https://doi.org/10.1016/j.mcat.2020.110980.
[21] Y. Wang, C. Wang, M. Chen, Z. Tang, Z. Yang, J. Hu, H. Zhang, Hydrogen
production from steam reforming ethanol over Ni/attapulgite catalysts-Part I, Fuel
Process Technol. 192 (2019) 227–238, https://doi.org/10.1016/j.
fuproc.2019.04.031.
Declaration of Competing Interest
[22] Y. Wang, Z. Zhang, S. Zhang, Y. Wang, S. Hu, J. Xiang, X. Hu, Steam reforming of
acetic acid over Ni/biochar catalyst treated with HNO3: Impacts of the treatment
on surface properties and catalytic behaviors, Fuel. 278 (2020), 118341, https://
doi.org/10.1016/j.fuel.2020.118341.
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
´
[23] M. Siqueira, O. Perrone, G. Metzker, D. Lisboa, J. Thomeo, M. Boscolo, Highly
selective 1-butanol obtained from ethanol catalyzed by mixed metal oxides:
reaction optimization and catalyst structure behavior, Mol Catal. 476 (2019),
110516, https://doi.org/10.1016/j.mcat.2019.110516.
Acknowledgements
[24] M. Balat, H. Balat, Recent trends in global production and utilization of bio-ethanol
fuel, Appl. Energ. 86 (2009) 2273–2282, https://doi.org/10.1016/j.
apenergy.2009.03.015.
This work was supported by the National Natural Science Foundation
of China (No. 51876080), the Program for Taishan Scholars of Shandong
Province Government. Collaborative Innovation Center of Technology
and Equipment for Biological Diagnosis and Therapy in Universities of
Shandong.
[25] D. Mohan, C. Pittman Jr., P. Steele, Pyrolysis of wood/biomass for bio-oil: a critical
review, Energ Fuel. 20 (2006) 848–889, https://doi.org/10.1021/ef0502397.
[26] D. Carpenter, T. Westover, S. Czernik, W. Jablonski, Biomass feedstocks for
renewable fuel production: a review of the impacts of feedstock and pretreatment
on the yield and product distribution of fast pyrolysis bio-oils and vapors, Green
Chem. 16 (2014) 384–406, https://doi.org/10.1039/C3GC41631C.
[27] R. Trane, S. Dahl, M. Skjøth-Rasmussen, A. Jensen, Catalytic steam reforming of
bio-oil, International Journal of Hydrogen Energy, Int J Hydrogen Energ. 37 (2012)
6447–6472, https://doi.org/10.1016/j.ijhydene.2012.01.023.
References
´
[1] L. Piola, J. Fernandez-Salas, F. Nahra, A. Poater, L. Cavallo, S.P. Nolan, Ruthenium-
[28] C. Wu, Q. Huang, M. Sui, Y. Yan, F. Wang, Hydrogen production via catalytic steam
reforming of fast pyrolysis bio-oil in a two-stage fixed bed reactor system, Fuel
Process Technol. 89 (2008) 1306–1316, https://doi.org/10.1016/j.
fuproc.2008.05.018.
catalysed decomposition of formic acid: fuel cell and catalytic applications, Mol
Catal. 440 (2017) (2017) 184–189, https://doi.org/10.1016/j.mcat.2017.06.021.
[2] S. Chalk, J. Miller, Key challenges and recent progress in batteries, fuel cells, and
hydrogen storage for clean energy systems, J. Power Sources 159 (2006) 73–80,
https://doi.org/10.1016/j.jpowsour.2006.04.058.
[29] T. Chen, C. Wu, R. Liu, Steam reforming of bio-oil from rice husks fast pyrolysis for
hydrogen production, Bioresource Technol. 102 (2011) 9236–9240, https://doi.
org/10.1016/j.biortech.2011.07.033.
[3] T. LeValley, A. Richard, M. Fan, The progress in water gas shift and steam
reforming hydrogen production technologies–a review, Int J Hydrogen Energ. 39
(2014) 16983–17000, https://doi.org/10.1016/j.ijhydene.2014.08.041.
[4] D. Palo, R. Dagle, J. Holladay, Methanol steam reforming for hydrogen production,
Chem. Rev. 107 (2007) 3992–4021, https://doi.org/10.1021/cr050198b.
[5] S. Freni, G. Calogero, S. Cavallaro, Hydrogen production from methane through
catalytic partial oxidation reactions, J. Power Sources 87 (2000) 28–38, https://
doi.org/10.1016/S0378-7753(99)00357-2.
[30] H. Xie, Q. Yu, Z. Zuo, Z. Han, X. Yao, Q. Qin, Hydrogen production via sorption-
enhanced catalytic steam reforming of bio-oil, Int J Hydrogen Energ. 41 (2016)
2345–2353, https://doi.org/10.1016/j.ijhydene.2015.12.156.
[31] E. Rodríguez-Aguado, A. Infantes-Molina, D. Ballesteros-Plata, J. Cecilia,
´
I. Barroso-Martín, E. Rodríguez-Castellon, Ni and Fe mixed phosphides catalysts for
O-removal of a bio-oil model molecule from lignocellulosic biomass, Mol Catal.
437 (2017) 130–139, https://doi.org/10.1016/j.mcat.2017.05.008.
[32] S. Xiu, A. Shahbazi, Bio-oil production and upgrading research: a review, Renew
Sust Energ Rev. 16 (2012) 4406–4414, https://doi.org/10.1016/j.
rser.2012.04.028.
[6] A. Lutz, R. Bradshaw, L. Bromberg, A. Rabinoich, Thermodynamic analysis of
hydrogen production by partial oxidation reforming, Int J Hydrogen Energ. 29
(2004) 809–816, https://doi.org/10.1016/j.ijhydene.2003.09.015.
[7] G. Mitran, S. Chen, D. Seo, Role of oxygen vacancies and Mn4+/Mn3+ ratio in
oxidation and dry-reforming over cobalt-manganese spinel oxides, Mol Catal. 483
(2019), 110704, https://doi.org/10.1016/j.mcat.2019.110704.
[33] D. Chen, J. Zhou, Q. Zhang, X. Zhu, Evaluation methods and research progresses in
bio-oil storage stability, Renew Sust Energ Rev. 40 (2014) 69–79, https://doi.org/
10.1016/j.rser.2014.07.159.
˜
[8] M. Alvarez-Galvan, R. Navarro, F. Rosa, Y. Briceno, M. Ridao, J. Fierroa, Hydrogen
[34] X. Hu, Z. Zhang, M. Gholizadeh, S. Zhang, H. Chun, Z. Xiong, Y. Wang, Coke
formation during thermal treatment of bio-oil, Energ Fuel. 34 (2020) 7863–7914,
https://doi.org/10.1021/acs.energyfuels.0c01323.
production for fuel cell by oxidative reforming of diesel surrogate: Influence of
ceria and/or lanthana over the activity of Pt/Al2O3 catalysts, Fuel 87 (2008)
2502–2511, https://doi.org/10.1016/j.fuel.2008.03.003.
10

Z. Gao et al.
Molecular Catalysis 504 (2021) 111461
[35] X. Li, Z. Zhang, L. Zhang, H. Fan, L. Xu, Q. Liu, S. Wang, X. Hu, Investigation of
coking behaviors of model compounds in bio-oil during steam reforming, Fuel. 265
(2020), 116961, https://doi.org/10.1016/j.fuel.2019.116961.
performance in Knoevenagel reaction, J. Solid State Chem. 213 (2014) 250–255,
https://doi.org/10.1016/j.jssc.2014.03.010.
[43] Z. Gao, Z. Zhang, H. Tian, Q. Xu, Z. Xu, B. Li, Q. Liu, C. Li, G. Chen, X. Hu, Ordinary
clay as a support of nickel catalyst for steam reforming of acetic acid: impacts of
pretreatments of clay on catalytic behaviors, Int J Energ Res. 44 (2020)
10378–10393, https://doi.org/10.1002/er.5660.
[36] D. Yao, C. Wu, H. Yang, Q. Hu, M. Nahil, H. Chen, P. Williams, Hydrogen
production from catalytic reforming of the aqueous fraction of pyrolysis bio-oil
with modified Ni–Al catalysts, Int J Hydrogen Energ. 39 (2014) 14642–14652,
https://doi.org/10.1016/j.ijhydene.2014.07.077.
[44] X. Li, Z. Zhang, L. Zhang, G. Gao, T. Wei, Z. Zhang, S. Zhang, G. Chen, C. Li, X. Hu,
Availability of steam impacts coke properties in steam reforming of acetic acid, Int
J Hydrogen Energ (2020), https://doi.org/10.1016/j.ijhydene.2020.11.269.
[45] P. Monson, Understanding adsorption/desorption hysteresis for fluids in
mesoporous materials using simple molecular models and classical density
functional theory, Microporous Mesoporous Mater. 160 (2012) 47–66, https://doi.
org/10.1016/j.micromeso.2012.04.043.
[37] Z. Li, X. Hu, L. Zhang, S. Liu, G. Lu, Steam reforming of acetic acid over Ni/ZrO2
catalysts: effects of nickel loading and particle size on product distribution and
coke formation, Appl Catal A-Gen. 417 (2012) 281–289, https://doi.org/10.1016/
j.apcata.2012.01.002.
[38] S. Wauters, G. Marin, Kinetic modeling of coke formation during steam cracking,
Ind. Eng. Chem. Res. 41 (2002) 2379–2391, https://doi.org/10.1021/ie010822k.
[39] L. Zhang, R. Liu, R. Yin, Y. Mei, Upgrading of bio-oil from biomass fast pyrolysis in
China: a review, Renew Sust Energ Rev. 24 (2013) 66–72, https://doi.org/
10.1016/j.rser.2013.03.027.
´
´
[46] B. Jankovic, B. Adnađevic, S. Mentus, The kinetic study of temperature-
programmed reduction of nickel oxide in hydrogen atmosphere, Chem. Eng. Sci. 63
(2008) 567–575, https://doi.org/10.1016/j.ces.2007.09.043.
[40] Z. Tang, Q. Lu, Y. Zhang, X. Zhu, Q. Guo, One step bio-oil upgrading through
hydrotreatment, esterification, and cracking, Ind. Eng. Chem. Res. 48 (2009)
6923–6929, https://doi.org/10.1021/ie900108d.
[47] R. Hayes, W. Thomas, K. Hayes, A study of the nickel-catalyzed methanation
reaction, J. Catal. 92 (1985) 312–326, https://doi.org/10.1016/0021-9517(85)
90266-0.
[41] S. Ren, X. Ye, A. Borole, Separation of chemical groups from bio-oil water-extract
via sequential organic solvent extraction, J Anal Appl Pyrol. 123 (2017) 30–39,
https://doi.org/10.1016/j.jaap.2017.01.004.
[48] C. Chen, W. Cheng, S. Lin, Mechanism of CO formation in reverse water–gas shift
reaction over Cu/Al2O3 catalyst, Catal Lett. 68 (2000) 45–48. https://link.springer.
com/article/10.1023/A:1019071117449.
[42] L. Xu, C. Wang, J. Guan, Preparation of acid-base bifunctional mesoporous KIT-6
(KIT: Korea Advanced Institute of Science and Technology) and its catalytic
[49] S. Jungseob, C. Yoona, S. David, S. Carsten, In-situ ATR-IR study of surface reaction
during aqueous phase reforming of glycerol, sorbitol and glucose over Pt/γ-Al2O3,
Mol Catal. 475 (2019) 110423, https://doi.org/10.1016/j.mcat.2019.110423.
11