New synthetic route for the production of mayenite support to enhance Ni resistance to coke deposition in the reforming of tar model compounds

Article abstract of DOI:10.1016/j.apcata.2019.01.029

Four nickel/mayenite catalysts were synthesized. Mayenite was prepared from different precursors, namely boehmite (AlO(OH)) + Ca(NO₃)₂·4H₂O and gibbsite + Ca(OH)₂, for each couple a specific procedure was followed. The effect of the Ni addition method was also evaluated, comparing wet impregnation and direct inclusion of nickel precursor during mayenite preparation. The obtained catalysts were characterized by XRD, BET, SEM/EDS and TPR. The Ni/mayenite catalysts were tested in steam reforming of toluene and pyrocatechol, chosen as tar model compounds. The experiments were conducted in a fixed-bed lab-scale atmospheric reactor, evaluating carbon conversion, hydrogen selectivity and stability. Characterization of the spent catalysts after toluene steam reforming experiments by XRD and DTG-TPO were performed to assess structural changes and carbon deposition. Kinetic parameters of toluene steam reforming were determined for the different catalysts.

Full text of DOI:10.1016/j.apcata.2019.01.029

Accepted Manuscript
Title: New synthetic route for the production of mayenite
support to enhance Ni resistance to coke deposition in the
reforming of tar model compounds
Authors: M.P. Bracciale, B. de Caprariis, P. De Filippis, A.D.
Hernandez, M. Scarsella
PII:
S0926-860X(19)30051-1
DOI:
Reference:
https://doi.org/10.1016/j.apcata.2019.01.029
APCATA 16969
To appear in:
Applied Catalysis A: General
Received date:
Revised date:
Accepted date:
30 November 2018
27 January 2019
29 January 2019
Please cite this article as: Bracciale MP, de Caprariis B, De Filippis P, Hernandez AD,
Scarsella M, New synthetic route for the production of mayenite support to enhance
Ni resistance to coke deposition in the reforming of tar model compounds, Applied
Catalysis A, General (2019), https://doi.org/10.1016/j.apcata.2019.01.029
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

New synthetic route for the production of mayenite support to enhance Ni
resistance to coke deposition in the reforming of tar model compounds.
M.P. Bracciale, B. de Caprariis, P. De Filippis, A.D. Hernandez*, M. Scarsella
Department of Chemical Engineering Materials Environment, Sapienza University of Rome, via Eudossiana
18, 00184 Rome, Italy
Graphical abstract
Highlights




Mayenite carriers were synthesized by two innovative routes using different precursors,
AlO(OH) + Ca(NO3)2 and Al(OH)3 + Ca(OH)2.
Nickel addition to the carriers was performed via wet impregnation and a one-step method
during the mayenite preparation.
The catalyst prepared with the one-step method and using AlO(OH) + Ca(NO3)2 as precursors
showed promising tolerance to deactivation by carbon deposition.
Different nickel-support interactions due to the applied nickel addition method were verified
by TPR analysis.
ABSTRACT
Four nickel/mayenite catalysts were synthesized. Mayenite was prepared from different precursors,
namely boehmite (AlO(OH)) + Ca(NO3)2·4H2O and gibbsite + Ca(OH)2, for each couple a specific procedure

was followed. The effect of the Ni addition method was also evaluated, comparing wet impregnation and
direct inclusion of nickel precursor during mayenite preparation. The obtained catalysts were characterized
by XRD, BET, SEM/EDS and TPR. The Ni/mayenite catalysts were tested in steam reforming of toluene and
pyrocatechol, chosen as tar model compounds. The experiments were conducted in a fixed-bed lab-scale
atmospheric reactor, evaluating carbon conversion, hydrogen selectivity and stability. Characterization of
the spent catalysts after toluene steam reforming experiments by XRD and DTG-TPO were performed to
assess structural changes and carbon deposition. Kinetic parameters of toluene steam reforming were
determined for the different catalysts
Keywords: mayenite; steam reforming; nickel; tar model compounds; on-step synthesis; catalyst
characterization
1. Introduction
Biomass has the potential to partly replace fossil fuels via syngas generation through gasification and
subsequent conversion of the produced syngas to green-fuels, energy and chemicals [1,2]. However, the
presence of deleterious contaminant species at concentrations detrimental to most downstream
applications hinders the utility of biomass derived syngas [3]. Among the by-products of biomass
gasification tar species pose the greatest problem to end users. Tar is defined as all hydrocarbons with a
molecular weight higher than benzene, which at high temperatures (≥ 700 °C) mainly includes aromatic
compounds with a condensation temperature ranging between 150 and 300 °C [4,5]. If not removed from
the produced syngas, tar compounds could block downstream pipelines and foul engines and turbines
causing unsustainable penalties on the profitability of the plant. Therefore, the removal and/or conversion
of these contaminants represents a critical step to exploit efficiently and sustainably the valuable gases
present in the syngas.
In spite of the extensive effort devoted to reduce the concentration of tar species inside the gasifier to the
actual stringent values required for downstream applications, the results are still far from satisfactory [6].
Hence the use of cleaning and conditioning technologies downstream the gasifier is crucial to adhere to the
requirements of end users. Amid those strategies, the catalytic steam reforming reaction is a promising
solution because it has the potential to increase the amount of valuable gases such as H2, CO and CH4
present in the syngas while reducing the waste streams generated using the classic wet cold gas cleaning
technologies[7]. Additionally, it can take advantage of the high temperature of the syngas at the gasifier
exit.

The majority of commercial steam reforming plants from 1950 to 2009 exploited nickel-based catalysts [7]
and nickel is still the most common active metal used nowadays. This can be ascribed to the optimum
balance between nickel cost and catalytic reforming activity. Nevertheless, deactivation by carbon
deposition is still an unsolved issue that has negative consequences on the operational costs [8].
Many efforts for suppressing carbon formation on Ni catalysts have been undertaken, and it has been
shown that the use of thermal-stable supports featuring high amount of “free” oxygen species can enhance
the carbon deposition resistance of nickel catalysts [9–11]. A pioneering work on the production of H2 from
catalytic steam reforming of bio-oil conducted by Wang et al. [12] tested the unique O- storage and
emission behavior of mayenite (C12Al14O33). The authors speculated that the high initial activity of the
Mg/mayenite catalyst could be attributed to reactions of the vaporized bio-oil with the active O- species
present in the mayenite structure.
Mayenite has a unique cubic crystal structure composed of three-dimensionally connected sub-nanometer-
sized cages where 2 out of 12 cages are statistically occupied by O2- and O- ions to compensate the positive
charge of the cage framework having the chemical composition of [Ca24Al28O64]4+ [13]. These anions feature
a nearly free mobility within the cage. Hosono et al. [13] found many evidences of the possible
-
replacement of the free O2- with OH-, O2 ,O- by tuning the atmosphere during the heat treatment. Boysen et
al. [14] by means of a high-temperature neutron powder diffraction study showed that at high temperature
the mobility of the “free” oxygen species increased significantly in the mayenite structure due to a jump-
like process which involve the exchange of “free” oxygen with oxygen species from the mayenite
framework. Based on these findings it is believed that during the steam reforming reaction these “free”
oxygen species in the mayenite lattice will migrate to nickel site to gasify the carbon deposited on nickel
catalyst, reducing the carbon deposition on the active sites [9].
Li et al. [9] used Ni/mayenite for the steam reforming of toluene obtaining promising results in terms of
resistance to carbon formation and tolerance to sulfur containing species. Di Carlo et al. [15] studied the
steam reforming activity of a Ni/mayenite catalyst using a microreactor fed by a slipstream coming from a
bench-scale fluidized-bed biomass gasifier. The catalyst maintained a high conversion (≈0.9) of heavy
hydrocarbons for 12 h at 800 °C. The above referenced studies prepared the mayenite support using
different methods. Li et al. [9] used Ca(OH)2 and Al(OH)3 as precursors applying the solid-state reaction
method, whereas Di Carlo et al. [15] worked in slurry employing Ca(CH3COO)2 mixed with Al2O3.
In the present work, two different synthetic routes using different precursors to produce mayenite were
investigated and compared. The Ni/mayenite catalysts with a Ni loading of 10 wt% were obtained by wet
impregnation or one-step inclusion during the carrier synthesis. The catalysts were tested in the steam
reforming of tar model compounds i.e. toluene and pyrocatechol, as representatives of aromatic stable
compounds and oxygenated aromatic compounds, respectively. Additionally, characterization of the

synthesized mayenite support and catalysts was conducted to correlate the structural, morphologic and
topologic features of the samples with their steam reforming performance.
2. Experimental
2.1. Supports and catalysts syntheses
Mayenite, used as support in all the catalysts, was obtained starting from different precursors: from
boehmite (AlO(OH)) and Ca(NO3)2·4H2O (termed in this study mayenite B) and from gibbsite (Al(OH)3) and
Ca(OH)2 (termed mayenite H).
Mayenite B was synthetized starting from a stable dispersion of boehmite [16]. First a dispersion was
prepared adding commercial aluminum oxide hydroxide powder (Sigma Aldrich ≥ 99.0 %) to a 0.4 % (w/w)
HNO3 aqueous solution and then Ca(NO3)2·4H2O (Sigma Aldrich ≥ 99.0 %)was added to the dispersion under
vigorous stirring. The used amount of Ca(NO3)2·4H2O was determined based on the stoichiometric ratio in
the mayenite (Ca/Al molar ratio = 12/14). The dispersion was kept under stirring for 24 h at room
temperature until a gel was obtained. The gel was then dried overnight at 100 °C and the resultant solid
was grinded and calcined at 950 °C for 16h under static air.
For the synthesis of mayenite H, Al(OH)3 (Sigma Aldrich, reagent grade) and Ca(OH)2 (Sigma Aldrich ≥ 99.9
%) were stoichiometrically mixed in solid state, then water was added until a homogeneous dense slurry
was obtained. The slurry was dried for 24 h at 105 °C, then the obtained solid was grinded and calcined at
1250 °C for 16 hours under static air.
The calcination temperature for mayenite B and H was determined based on TGA-DSC analysis (SDTQ600,
TA instrument) performed on the precursors, where the exothermic peak of the phase transformation was
recorded at 940 °C and 1250 °C for boehmite and hydroxides precursors, respectively (Fig. S1).
Four nickel supported catalysts were prepared with two different routes for the metal addition on
mayenite B and H. The first route was by wet impregnation (wi) on the previously obtained supports
(Ni/mayenite B-wi and Ni/mayenite H-wi) sieved between 0.1-0.3 mm. These solids were mixed under
stirring with the aqueous solution of Ni precursor and then dried for 24 h at 105 °C. The powders obtained
were calcined at 900 °C for 6 h under static air atmosphere.
The second route was by a “one step” (os) procedure consisting in adding the Ni precursor directly during
the mayenite preparation procedures described above. In the case of Ni/mayenite B-os the Ni precursor
was added to the dispersion together with calcium precursor, while to prepare Ni/mayenite H-os the Ni
precursor was dissolved in the water added to the mixture of solid precursor to obtain a dense slurry.
In all the cases Ni(NO3)2·6H2O (Sigma Aldrich 99.99 %)was used as precursor and its amount was calculated
to produce a catalyst having 10 wt. % of nickel.
2.2. Experimental set-up
The experimental setup used for the steam reforming of toluene (STR) and pyrocathecol (SPR) is depicted
in Fig. 1. The detailed description of the set-up is reported in [11].

Fig. 1. Experimental set-up.
The reagents used in this study were deionized water, toluene (Sigma-Aldrich ≥ 99.5 %) and pyrocatechol
(Sigma-Aldrich ≥ 99.0 %). During the STR, toluene and water were fed separately to the reactor, whereas in
the case of SPR, pyrocatechol was dissolved in deionized water in the appropriate steam to carbon ratio
(S/C) before feeding the solution to the reactor. In both cases N2 was employed as carrier gas; the reagent
mixture was sent to an evaporator held at 300 °C before it entered to the reactor. The catalysts were held
by a small piece of quartz wool placed on top of a stainless-steel grid. In each experiment 0.1 g of catalyst
(grain size 0.1-0.3 mm) mixed with 1 g of SiO2 (grain size 0.1-0.3 mm) was used. The height of the bed was
15 mm. At the reactor exit the unreacted species and condensable products were recovered in a counter-
current condenser using water maintained at 5 °C by means of a chiller. The produced gas flow was
measured with a volumetric flow meter. CO, CO2 and CH4 were measured by an online analyzer (Siemens
Ultramat 21). A fixed bed of small Al2O3 sticks were placed before the on-line analyzers to adsorb any
residual steam content present in the gas. The concentration of H2 and possible traces of C2H4 were
determined using a GC mass spectrometer (Hiden QGA). Before the experiments the prepared catalysts
were activated in situ at 700 °C for 30 min in 0.5 NL min-1 of 16 % H2/N2 and then held at 750 °C for 1h in N2
flow 0.24 Nl min-1. At the end of each test the heating system was turned off and the catalyst was cooled
down under N2 to room temperature, then it was collected and kept in a glass sealed container.
To evaluate the activity of the SiO2 filler material on the conversion of the tar model compounds a pure
SiO2 bed was used in the reactor and experiments with the model tar compounds were performed at 750
°C keeping constant all the experimental conditions used for the catalytic tests. The concentration values of

CO and H2 measured at the reactor exit were negligible, hence the catalytic/thermal effect of the reactor
and the SiO2 particles was neglected.
The occurrence of deactivation was evaluated conducting 6 h long tests at 700 °C
The experimental conditions used are reported in Table 1. In the present work severe conditions with
respect to the inlet tar concentration were used, these conditions are comparable with the tar
concentration measured in atmospheric bubbling fluidized bed biomass gasifiers with steam as gasifier
agent [17]. The chosen S/C ratio was based on reported values of syngas composition obtained in air [18]
and steam/oxygen [17] fluidized bed gasifiers. In addition, the relatively high S/C value was applied to
prevent a fast deactivation of the catalysts due to carbon deposition during the tests conducted to obtain
the kinetics parameters
Table 1
Experimental conditions used in the experimental tests
STR
SPR
Temperature (°C)
670-750
0.41
0.24
47
670-750
0.41
0.24
65
Total gas inlet flow (NL min-1)
N2 flow rate (NL min-1)
Tar model compound concentration (g Nm-3)
GHSV (h-1 at 700 °C)
Space-time, Wcat/Ftoluene (kgcat h Nm-3)
73750
0.33
5
73900
0.28
5
S/C
2.3. Catalyst characterization
The obtained catalysts and supports were analysed by X-ray powder diffraction (XRPD) (Philips Analytical
PW1830), Brunauer-Emmett-Teller (BET) and Barrett-Joyner-Halenda (BJH) analyses. XRD patterns were
acquired using a Philips Analytical PW1830 X-ray diffractometer, equipped with Cu Kα (1.54056 Å)
radiation, in the 2θ range from 15 to 70° with a step size of 0.02° and a time for step of 3.5 s. The data were
collected with an acceleration voltage and applied current of 40 kV and 30 mA, respectively. The crystalline
phases in the resulting diffractograms were identified through the COD database (Crystallography Open
Database – an open-access collection of crystal structures) [19]. The average nickel crystallite sizes were
calculated with the Scherrer’s equation on the most intense Ni peak plane.
BET analyses were determined by N2 adsorption–desorption isotherms acquired at −196°C using a
Micromeritics Triflex analyzer (Micromeritics Instrument Corp.). The adsorption-desorption isotherms were
acquired in the p/p0 range from 0.01 to 0.99. Isotherm analyses were performed using the 3Flex Version
4.05 software. Samples were previously outgassed at 300 °C for 3 h. The BET and BJH equations were used
to determine the specific surface area, pore volume and average pore diameter, respectively.
In order to obtain morphologic and topologic information, SEM/EDS (Scanning Electron Microscopy
combined with Energy Dispersive X-ray Spectrometry) analysis were performed on the obtained catalysts.
Backscattered electron (BSE) images were recorded with a High Resolution-Field Emission Scanning

Electron Microscope (HR-FESEM, AURIGA Zeiss) operated at 15 kV. The samples were prior sputter-coated
with 10 nm thin layer of chromium using a Quorumteach Q150T sputter coater. For elemental analyses EDS
maps were acquired with QUANTAX EDS XFlash® 6 detector (Bruker Nano GmbH), providing elemental
topography of zones of external surfaces.
Ni/mayenite reducibility properties of the fresh catalysts were measured by temperature-programmed
reduction (TPR) experiments with a thermogravimetric analyzer SDTQ600 (TA Instruments, USA). The
samples were pre-treated under Ar flow at 150 °C for 60 min to remove adsorbed moisture and air. Then
the TPR profiles were recorded by heating the sample from 40 °C to 900 °C at 5 °C min-1 under a H2 flow
(5.0 % H2 in Ar, 50 mL min-1).
The carbon deposited content was determined by Thermogravimetric (TG) analysis using an SDTQ600
simultaneous TGA- DSC analyzer (TA Instruments, USA) under a gas mixture of 20 % O2 in nitrogen and with
a flux rate of 100 mL min−1. The samples were prior heated under nitrogen in an open ceramic pan up to
800 °C with a ramp rate of 20 °C min−1 in order to strip away volatile organic compounds physically
absorbed and to thermally decompose any carbonatic species produced on spent catalysts during reaction.
The samples were then equilibrated to 31 °C and the O2/N2 gas mixture was introduced. Specimens were
afterward subjected to the following thermal program: heating up to 230 °C, 30 min hold, heating to 800 °C
and 30 min hold, all ramps at 10 °C min-1.
2.4. Carbon conversion and H2 selectivity
A simplified reaction scheme which considers only the steam reforming and the water gas shift (WGS)
reactions was adopted for the steam reforming processes [20]:
STR = C7H8 + 7H2O → 7CO + 11H2
SPR = C6H6O2 + 4H2O → 6CO + 7H2
WGS = CO + H2O ↔ CO2 + H2
(1)
(2)
(3)
The conversion of the tar model compounds was calculated as the carbon conversion XC:
n
+ n
+ n
CH4
OUT
CO
CO2
OUT
OUT
XC =
(4)
molar flow of carbon in the feed
where nCO , nCO2 , and nCH4 are the molar flow of gas containing carbon at the reactor exit, i.e.
OUT
OUT
OUT
CO, CO2 and CH4, respectively. Reactions leading to carbon formation on the catalyst such as toluene
decomposition and the Boudouard reaction are expected to proceed concurrently with the steam
reforming and WGS reactions and need to be considered.
Hydrogen selectivity was calculated according to the work of Polychronopoulou et al. [21]:
y
H
2
SH
=
(5)
2
b ∗ y + c ∗ y
CO
CO
2

SH was used to identify whether reactions other than those described in the selected reaction network
2
occurred [21]. yH yCO and yCO are the measured mole fractions of hydrogen, CO and CO2, respectively,
,
2
2
at the exit of the reactor. In equation (5) the coefficients b and c represent the stoichiometric correlation of
1 mol of hydrogen to that of the produced CO and CO2, respectively, based on reactions (1) and (3) for
toluene and (2) and (3) in the case of pyrocatechol.
2.5. Kinetics: Evaluation of mass transfer limitation
Axial isothermal conditions were checked during the experiments using the adjustable thermocouple inside
the catalyst bed. Temperature and concentration gradients within the particles were assumed to be
negligible because of the low surface area of the used catalyst (see Table 2) [22]. The occurrence of
external mass transfer limitation was excluded by calculation of the Mears’ criterion [23].
A selected number of experiments with fresh catalysts and using toluene as model tar compound were
dedicated to obtaining kinetic parameter in which fresh catalyst was used in every experiment. Four
different temperatures were set in each test. The same catalyst was used for the whole duration of the
test. The experiments were performed starting from the lowest temperature. Each selected temperature
was maintained until the concentration of the main gases reached steady state conditions, usually 20 min,
after which the temperature was increased.
Plug flow condition was checked using the criteria proposed by Froment et al. [24]. The ratio of the height
of the bed to the particle size was larger than the criterion for axial dispersion i.e. catalyst bed
height/particle diameter ≥ 50. The ratio of the internal diameter of the reactor and the catalyst particle was
higher than 30 and hence the channeling criterion was fulfilled.
The apparent kinetic parameters were carried out assuming a pseudo first-order reaction with respect to
the tar model compound and being independent from H2O concentration as the latter was fed in excess.
This assumption have been widely accepted in the literature [25,26]. The reaction rate was calculated as
follow:
−rapp = kapp ∗ Ctar
(6)
Under conditions very close to plug flow, the mass balance for the tar compound gives:
−ln(1−X )
C
kapp
=
(7)
τ
where rapp is the rate of conversion (kmol m-3 h-1), kapp is the apparent kinetic constant (m3 (Tcat. N2-free)
kg-1 h-1), Ctar is the tar model compound concentration (kmol m-3), XC is the carbon conversion, and τ is
the residence time in the catalyst bed that was calculated as follows:
푊
cat
τ =
(8)
Q
(Tcat.)
in

where 푊cat is the catalyst weight (kg) and Qin(Tcat. ) is the inlet volume flow rate at the catalyst bed
temperature after subtracting the N2 flow (m3 h-1), with this units of measurement the obtained kapp is
thus calculated at the catalyst temperature and without considering the N2 flow rate. This was done to
compare the results with the reported data of Aznar et al. [26].
The apparent kinetic constant (kapp) of tar model compounds were calculated using the carbon conversion
XC (Eq. 4) and therefore the obtained kapprefers to all the reactions involved in the formation of CO, CO2
and CH4.
2.6. Thermodynamic calculations
Thermodynamic equilibrium calculations were done using Aspen plus V8.8®. The Gibbs reactor was chosen
to calculate the equilibrium composition of the simulated gas. The SRKKD thermodynamic method was set
for all the calculations. The input for the simulation was the set molar flow rates of C7H8, water and N2. The
following vol. % dry composition of the gas was obtained at 700 °C: H2 = 21.2 %, CO = 2.7 %, CO2 = 6.6 %
and CH4 = 4.0 x 10-3. Complete carbon conversion was achieved at 700 °C. Moreover, carbon conversion
was 0.99 at the lowest temperature used in this study i.e. 620 °C. The influence of the excess of steam on
the WGS reaction and thus on the equilibrium H2 vol. % value that can be obtained considering only the
steam reforming (SR) and WGS reactions was explored using the equilibrium reactor of Aspen plus V8.8®
with the same thermodynamic method as in the Gibbs reactor case. First the simulation was run setting
only the SR reaction. Next a second run was done but this time both the WGS and SR reaction were set. The
difference in the H2 vol. % of the two simulations was due to the WGS reaction induced by the excess of
steam. The share of the WGS reaction on the equilibrium H2 vol. % value decreased from 27 to 25 % when
the temperature increased from 620 to 740 °C
3. Results and discussion
3.1. Fresh catalyst characterization
XRD patterns of the synthetized catalysts are shown in Fig. 2.

Fig. 2. XRD patterns of mayenite synthetized by boehmite (A) and hydroxides (B) precursors. M (Ca12Al14O33); C
(CaO); N (NiO); CA (CaAl2O4); C3A (Ca3Al2O6); C5A3 (Ca5Al6O14).
The main crystal structures of both supports (mayenite B and mayenite H) were similar despite of the
different precursor used. The characteristic reflections of the {211} and {420} planes of crystalline
Ca12Al14O33 (18.1° and 33.3°) [27–29] and CaO (32.2°, 37.3°, 64.1°) can be distinguished. Furthermore, in the
case of mayenite H (Fig. 2B) diffraction rays corresponding to Ca3Al2O6 (C3A) at 2θ angles of 20.9°, 21.7°,
47.6° and 45.5° were clearly detected. The presence of CaO was probably due to segregation of the Ca
during the preparation [28,29], instead Ca3Al2O6 is a high temperature (930-1330 °C) intermediate of
mayenite formation. This latter exhibits wide crystals that prevent the formation of mayenite [30].
Furthermore, as previously reported, this phase is common when powder reactants that lead to a
heterogeneous mixture are used [31]. After Ni nitrate addition and calcination, neither NiAl2O4 spinel phase
nor hydrated structures Ca(OH)2 or formation of solid solution between NiO and CaO were identified. The
Ni/mayenite catalysts showed the characteristic reflections of crystalline NiO at 43.2° and 62.8°. The wet
impregnation method led in both the cases to the formation of a little amount of CaAl2O4 phase probably
due to the hydration and subsequent calcination stages used during the preparation. This phase was
detected in larger amount in the Ni/mayenite H-os sample. Instead the Ni/mayenite B-os catalyst showed
the doublet belonging to Ca5Al6O14 (C5A3) at 2θ = 30.6 and 31.0°, which is recognized in the literature as
the low temperature metastable phase (< 950 °C) in the Ca–Al–O system [31,32].
The diffraction patterns of catalysts after the in-situ activation (700 °C for 30 min in 0.5 NL min-1 of 16 %
H2/N2 and then held at 750 °C for 1h) showed well defined diffraction rays of metallic nickel (see section
3.5). The average crystallite size of metallic nickel (Table 2) is about 20–22 nm and 38-71 nm for the
samples prepared by wet impregnation and on-step method, respectively.
Table 2 shows the catalysts’ surface area. The adsorption-desorption plots of all the catalysts showed type
IV adsorption isotherms with an H1 hysteresis [33], denoting that the mesoporous structure of the support

has not been modified by the insertion of Ni (Fig. S2-S3). The surface areas of the mayenite B catalysts
(mayenite B, Ni/mayenite B-wi and Ni/mayenite B-os) were higher than those of mayenite H (mayenite H,
Ni/mayenite H-wi and Ni/mayenite H-os). The decrease in the surface area could be attributed to collapse
of the pores in the samples, which occurs with an increase in the temperature (1250 °C compared to 950
°C) that lead to a high compaction of the crystals, as demonstrated by SEM images (Fig. 3) and the lower
pore volume values (Table 2). Furthermore, the catalysts obtained by wet impregnation method
(Ni/mayenite B-wi and Ni/mayenite H-wi) exhibited surface areas relevantly bigger than their respective
support (mayenite B and mayenite H). The increase in the surface area after impregnation could be due to
the hydration stage used during the preparation. Indeed, as previously reported [34,35], the addition of the
aqueous solution of nickel nitrate may lead to the formation of hexagonal Ca(OH)2 that become porous CaO
during the subsequent calcination step at 900 °C. Furthermore, the Ni insertion led to a general small
increase of the surface area in all the catalysts, suggesting that the metallic nanoparticles did not fill the
support’s pore structure.
The obtained BET surface area values are in fair agreement with those reported in the literature ,
regardless of the preparation method used, the latter values ranging from 2.5 to 24.3 m2 g-1 [9,28,35–37].
Table 2
BET-BJH data for fresh catalysts.
Catalysts
Crystal size
(nm)a
BET Surface Area
(m2/g)b
Pore Volume
(cm3/g)c
Average Pore Diameter
(nm)d
mayenite B
-
4.13
12.92
4.53
0.24
6.82
0.60
0.015
0.027
0.009
0.0007
0.027
0.0026
1.9
Ni/mayenite B-wi
Ni/mayenite B-os
mayenite H
20.7
37.9
-
12.6
1.9
26.7
17.9
25.82
Ni/mayenite H-wi
Ni/mayenite H-os
21.7
70.5
a by XRD using Scherrer equation on Ni {111} plane; b by BET equation; c-d by BJH desorption.
The morphological study of the obtained samples was carried out with SEM-EDS (Fig. 3). The mayenite B
series consisted of well-defined laminar aggregates with nano-thickness ~50 nm. Ni/mayenite B-wi
displayed NiO crystallites, not uniformly distributed on Ca12Al14O33, as it is clear in Fig. 3. Ni/mayenite H-wi,
instead, exhibited a microstructure of Ca12Al14O33 formed by dendritical aggregates. The lower specific
surface area and poorly developed pore structure of Ni/mayenite H-os effectively inhibited the dispersion
of Ni active metal, as shown by EDS maps in Fig. 3. As a confirmation of the obtained XRD results, some
cubic crystals of CaO were detected during the SEM analysis.

mayenite B
Ni/mayenite B-wi
Ni/mayenite B-os
mayenite H
Ni/mayenite H-wi
Ni/mayenite H-os
Fig. 3. SEM micrographs showing the Ni dispersion on the synthetized catalysts. EDS maps of Ni are included as
insets. M (Ca12Al14O33); C (CaO).
TPR profiles of Ni/mayenite catalysts in Fig. 4 were used to estimate the Ni available for reduction. The
reduction degree calculated considering the theoretical nickel loading is listed in Table 3.
Fig. 4. H2 consumption during TPR of catalysts
On all the catalysts the main H2 consumption was observed in the temperature range between 150 and 650
°C. On Ni/mayenite B catalysts the main reduction peaks, observed around 230 and 615 °C, lead to a
reduction degree of 66 and 63 % for Ni/mayenite B-wi and Ni/mayenite B-os, respectively. The peak at
lower temperature could be ascribed to the reduction of isolated NiO particles [38,39] with smaller size and
very finely distributed, which weakly interacted with or was far from the support and then were easier to
reduce. For higher temperatures, the reduction peak at 615 °C (Ni/mayenite B-os) can be ascribed to a

stronger interaction state of NiO with the mayenite support and CaO [40,41]. The area corresponding to
the reduction of these latter Ni species is the largest among the peaks detected, denoting that Ni bonded
to the mayenite support are the principal Ni species in this catalyst. As previous reported, the strong
interaction of Ni2+ ions with the mayenite support reduces the mobility and agglomeration with other Ni
particles on the surface [28]. The peaks were slightly shifted to lower temperature for the Ni/mayenite B-wi
sample. Furthermore, the Ni-based reduction degree (Table 3) on the B-wi catalyst is higher than that on
the B-os catalyst. This behavior can be interpreted as a consequence of the stronger interaction between Ni
and mayenite introduced by the one step method. The highest reduction degree was obtained with the
Ni/mayenite H-wi catalyst (83.23%). In this case the area associated with the low and high reduction
temperature peaks were identical. From Table 3 it is possible to identify on this latter catalyst that the
temperature and the H2 consumption of the low temperature reduction peak were considerably higher
than the corresponding values of the Ni/mayenite B catalysts. This result could be ascribed to higher
content of isolated NiO particles in this catalyst as reported by Cabello et al. [39]. Furthermore, the high
temperature reduction peak was shifted to lower temperatures with respect to the Ni/mayenite B
catalysts, this result, as stated above, could suggest a weaker metal-support interaction. It is well
documented [42–44] that the strong nickel-support interactions are beneficial for the enhancement of the
catalysts stability and the carbon deposition tolerance. Considering the obtained TPR it is expected that the
catalyst where the strongest Ni-support interactions were present was the Ni/mayenite B-os (Table 3). On
the other hand, the reduction degree of Ni/mayenite H-os was only about 5%, probably due to the high
calcination temperatures markedly decreasing the accessibility of nickel and thus its reducibility [25,45].
From the TPR profiles of the synthetized catalysts, the reduction temperature for catalysts testing was set
at 700 °C, in order to reduce an important proportion of the Ni2+ species with weak and high interaction
with the mayenite support.
Table 3
Catalyst properties after H2 temperature-programmed reduction (TPR) experiments
Catalysts
Content
(10-3 mol gcat
Main consumption
peaks
H2 consumption
Reduction degree [11]
(%)
-1
-1
)
(10-3 mol gcat
)
(°C)
Ni/mayenite B-wi
Ni/mayenite B-os
Ni/mayenite H-wi
Ni/mayenite H-os
228; 578
0.31; 0.81
0.44; 0.64
0.71; 0.71
0.09
66.0
235; 614
253; 543
460
63.31
83.23
5.08
1.7
3.2. Steam Reforming of Toluene (STR)
Ni/mayenite H-os did not show any steam reforming activity at the experimental conditions used in the
present work, thus it was not considered in the next sections. This outcome could be attributed to its

unfavorable structural and morphological features, as discussed in section 3.1, which lead to a low
reduction degree of Ni species (Table 3).
The conversion as a function of time-on-stream (ToS) at 700 °C is shown in Fig. 5. In the case of
Ni/mayenite B-wi and Ni/mayenite H-wi a similar trend was observed. These catalysts exhibited high
conversion values with a slight deactivation towards the end of the test. On the other hand, lower but
stable values were obtained using Ni/mayenite B-os. The lower activity of the latter catalyst could be
attributed to the higher nickel crystal size and/or due its lower reduction degree which is related to a lower
amount of potential chemically actives sites and hence lower conversions. Contribution to the observed
lower conversion by virtue of the degradation of the support and the presence of NiO after the tests as
found during the characterization of the spent catalyst cannot be excluded (see section 3.5). After 6 h ToS
analogous carbon conversion values were observed for the three catalysts.
Fig. 5. Conversion as a function of time-on-stream for toluene steam reforming (700 °C; S/C =5)

In Fig. 6 SH values are depicted as a function of time-on-stream (ToS) at 700 °C. The SH obtained values
2
2
were proximate to 1 throughout the entire experiment, being slightly lower in the case of Ni/mayenite B-
os. Based on these results it is safe to say that the extent of the reactions other than the complete steam
reforming of toluene to carbon oxides and hydrogen was insignificant. Moreover, the activity decrease
Fig. 6. Hydrogen selectivity parameter to identify whether reactions other than those described in the selected
reaction network occurred for STR at 700 °C.
observed at the end of the test for Ni/mayenite B-wi and Ni/mayenite H-wi did not affect the SH achieved
2
with these catalysts.
The conversion as a function of temperature is depicted in Fig. 7. As in the conversion as a function of ToS
the Ni/mayenite B-wi and Ni/mayenite H-wi catalysts showed a comparable activity, besides the curves
seemed to follow a linear behavior. A lower activity was observed for Ni/mayenite B-os as expected from
the results obtained during the six hour-long tests. The latter catalyst needed higher temperature to equal
the conversion values achieved with the catalysts synthesized by wet impregnation.

Fig. 7. Conversion as a function of temperature for toluene steam reforming (S/C=5).
Table 4 shows the measured gas volumetric flow rate (N2-free basis), the concentration of H2, CO and CO2
and the calculated hydrogen selectivity 푆퐻 at three different temperatures for the three catalysts
2
considered. As expected, the gas flow rate and the measured concentration of non-condensable gases
increased with temperature. The highest CO2 vol% values were obtained with Ni/mayenite B-os despite the
higher temperature applied during the tests with this catalyst. This result could be ascribed to a higher
WGS reaction activity of the catalyst synthesized with the one-step procedure. 푆퐻 values were close to 1 in
2
all cases, hence, it can be assumed that in the studied operating conditions the extent of reactions forming
or consuming hydrogen other than reactions (1) and (3) were negligible. The conversion and hydrogen
selectivity data of the highest temperature is not reported in Table 4 and can be found in Fig. 5 and Fig. 6 as
a function of time of stream for Ni/mayenite B-wi and Ni/mayenite H-wi, respectively.
As stated above from the equilibrium carbon conversion calculations almost complete carbon conversion
was reached already at the lowest applied temperature of this study. From the experimental results
showed in Figure 7 is evident that the tests were performed far from the equilibrium, specially the
experiments done at temperatures ≤ 650 °C with the catalysts synthesized by the wet impregnation
method. Additionally, considering the CO2, CO and H2 equilibrium values showed in Table 4 it can be
concluded that the WGS reaction was far from equilibrium and was likely controlled by the kinetics of the
process for the whole range of temperature applied.

Table 4
Total dry gas flow rate (N2-free basis) and main non-condensable gases measured at the outlet of the reactor at
different temperatures. The calculated equilibrium values for the non-condensable gases are listed for comparison.
Ni/mayenite B-wi
623 639 675
Total dry gas flow rate, N2-free basis (Nl/min) 0.04 0.07 0.09 0.03 0.05 0.09 0.03 0.07 0.09
Ni/mayenite H-wi
Ni/mayenite B-os
Cat. Temperature (°C)
624 650 678
670 690 710
H2 vol%
7.5
12.7 17.3 7.2
12.4 17.3 7.1
12.4 17.3
H2 equilibrium vol%
CO2 vol%
21.6 21.6 21.4 21.6 21.5 21.4 21.3 21.2 21.0
1.9
7.1
1.8
2.1
2.9
6.9
3.5
2.3
4.0
6.7
4.6
2.6
1.9
7.1
1.6
2.2
2.9
6.9
3.3
2.4
4.4
6.7
4.0
2.6
2.0
6.6
1.4
2.7
3.5
6.5
2.3
2.9
5.0
6.3
3.1
3.1
CO2 equilibrium vol%
CO vol%
CO equilibrium vol%
SH
0.97 0.98 0.99 0.98 0.98 0.98 0.96 0.98 0.98
2
3.3. Steam Reforming of Pyrocatechol (SPR)
The conversion as a function of time for the SPR at 700 °C is showed in Fig. 8. Conversion values were lower
than those obtained at the beginning of the STR experiments (Fig. 5Error! Reference source not found.).
The most stable catalyst was the Ni/mayenite B-os that, after a slight activity decrease, maintained a
constant conversion of ca. 0.5 throughout the experiment. The other two catalysts presented a slow and
steady deactivation. Koike et al. [47] did a comparative study on steam reforming of benzene, toluene and
phenol over Ni and Ni-Fe alloy nanoparticles supported on a Mg/Al hydrotalcite at 600 °C and contact time
W/F = 0.01 – 0.05 g h mol-1. Lower conversions values and higher carbon deposition were observed on both
catalysts when phenol was fed to the reactor compared to toluene and benzene. Stronger adsorption of
phenol and its intermediates on the surface of the catalysts with respect to the other aromatic compounds
was thought to be the reason for the results obtained.

Fig. 8. Conversion as a function of time-on-stream for pyrocatechol steam reforming (700 °C ; S/C =5).
Hydrogen selectivity values SH are illustrated in Fig. 9. As in the STR experiments the SH curves where all
2
2
in the immediacy of 1 and thus it was assumed that the main reactions taking place during the SPR
experiments where those based on reactions (2) and (3), i.e. the total steam reforming.
Fig. 9. Hydrogen selectivity parameter to identify whether reactions other than those described in the selected
reaction network occurred for SPR (700 °C).
Interestingly, CH4 was not detected during the steam reforming experiments carried out in the present
study. Probably, the methanation reaction did not occur in the present catalytic system. Furthermore, it
can be speculated that the extent of reactions forming methane e.g. hydrodealkylation, were negligible
and/or the methane formed was converted to CO, CO2 and H2. According to Mukai et al. [48] the most
abundant intermediate in the steam reforming reaction of toluene at 600 °C on a Ni/La0.7Sr0.3AlO3-δ was
C2H4. However, in the present study C2H4 was not detected in any of the tests. The results observed in the
present study are in agreement with Świerczyński et al. [25] who observed only CO, CO2 and H2 when the
catalyst temperature was ≥ 650 °C using a Ni/olivine catalyst during the STR with a space-time of 9 kgcat h
m-3 at 25 °C. Koike et al. [47] did not detect any traces of CH4 after steam reforming of phenol at 600 °C
over a Ni-Fe/Mg/Al catalyst
3.4. Kinetics
The kapp as a function of temperature for the STR is showed in Fig. 10. In the case of Ni/mayenite B-os the
higher temperature at which the kapp values presented a sharp increment could be attributed to a
different interaction between the support and the nickel active sites in comparison with Ni/mayenite B-wi
and Ni/mayenite H-wi as evidenced by the TPR results. Although strong metal-support interactions are
linked with better stability and higher carbon deposition tolerance, some studies[46,49] stated that the

strong metal-support interactions could led to reduction of the amount of chemically active nickel and as a
consequence to lower steam reforming conversions. The kapp values observed in the present work are
comparable with the values obtained by Aznar et al. [26] who studied the steam reforming of a slip-stream
taken from a fluidized bed gasifier on commercial nickel catalysts. It should be stressed that the authors
[26] used a guard bed which reduced significantly the inlet tar concentration and likely change the
composition of the tar, leaving mainly the most recalcitrant components such as benzene and toluene
unaltered [50]. When the kapp values are computed considering the total wet gas flowrate instead of the
N2-free flowrate a reasonable agreement is found also with the values obtained by Świerczyński et al. [25]
who studied the STR in the temperature range of 560 °C to 800 °C.
Fig. 10. 푘푎푝푝 as a function of temperature for toluene steam reforming.
The kinetics parameter for Ni/mayenite B-wi, Ni/mayenite H-wi and Ni/mayenite B-os are listed in Table 5.
Representation of the obtained kapp values according to the Arrhenius equation yielded a fair
concordance, as shown in Fig. 11, with coefficient of determination values ≥ 0.98 (Table 5). The Activation
energy values obtained for the three catalyst were similar, therefore, analogous chemical nature of nickel
active sites is expected for all the catalyst. Activation energy values for the steam reforming reaction of
toluene and benzene reported in the literature span from 36 to 230 kJ mol-1 [20,51–54]. The wide range of
values are mainly due to difference in experimental conditions used as well as in the reactor model
employed in the various studies. Mukai et al [51] separate the activation energy values in two temperature
regions, namely low (≤ 530 °C) and high (≥ 530 °C) temperature. The authors [51] found activation energy
values of 113 and 36 kJ mol-1 for the low and high temperature region, respectively, using a
Ni/La0.7Sr0.3AlO3−δ calcined at 800 °C. A change in the rate-determining step was thought to be the cause of
the change in the activation energy value, the former change was attributed to the lattice oxygen
contribution which was found to be important at temperatures ≥ 550 °C. In the present study no distinction

was made between low and high temperature regions. Further kinetics studies are required to elucidate
this argument.
Table 5
Kinetic parameter for STR.
Ni/mayenite B-wi Ni/mayenite H-wi Ni/mayenite B-os
Ea (kJ mol-1)
A (m3(Tcat) kg-1 h-1) 3.2 x 1010
R2
0.98
134
137
3.9 x 1010
145
5.7 x 1010
0.99
0.98
Fig. 11. Arrhenius representation for the kapp values of toluene steam reforming.
3.5. Characterization of used catalysts
The used catalysts were analyzed by X-ray diffraction and thermogravimetric oxidation (TPO) analysis in
order to study the changes in their crystalline structure and to determine the carbon content formed
during the catalytic test. The XRD patterns of the catalysts used in the steam reforming of toluene, are
presented in Fig. 12A.

Fig. 12. (A) XRD patterns of used mayenite. M (Ca12Al14O33); L (CaO); N (Ni); C (CaCO3); C3A (Ca3Al2O6). (B) DTG-TPO
profiles of spent catalysts used in the steam reforming of toluene
The diffractograms of the spent samples are quite similar to those of the fresh samples, indicating that
there were no significant structural changes caused by the exposure of the catalysts to the experimental
conditions. Furthermore, no trace of carbon (crystalline phase) was detected on the catalysts after the
reforming test. In the XRD pattern in addition to mayenite, CaO and Ni, also CaCO3 (4.8% for Ni/mayenite B-
os, 4.3% for Ni/mayenite B-wi and 11% for Ni/mayenite H-wi ) was identified in all the catalysts, which is
due to the CO2 uptake by carbonation reaction on CaO during the steam reforming tests [41]. As previous
reported, carbon dioxide removal by adsorption on a solid provide further heat to the reforming reaction
that may lead to a higher hydrogen yield. [35]. Additionally, CO2 sorption by the support could shift the
WGS equilibrium reaction to the products side hence promoting the production of H2 [36]. Furthermore for
Ni/mayenite B-os Ca3Al2O6 (41.5%) and NiO (7.8%) can be detected, which are a degradation phase of
mayenite under humid condition [55] and unreduced Ni, respectively.
From the XRD analysis it was possible to determine the size of the metallic Ni crystallites in the used
catalysts by applying the Scherrer equation to the (1 1 1) diffraction line at 2θ = 44.5°. The obtained values
after the 6 h long test for the three catalysts are showed in Table 6
When the fresh and spent catalyst are compared, it is evident that the Ni/mayenite B-wi catalyst showed
the largest Ni crystallites increment after use, which could be related to the fact that the isolated NiO
species in this sample were reduced at the lowest temperature among the studied catalysts and this result
is generally linked with high mobility of these species that upon reaction are also the most sensitive to
agglomeration phenomenon. In the other cases, after the tests the increase of the Ni crystallite size was
notably lower or almost unchanged, therefore it can be deduced that the partial deactivation of the
Ni/mayenite H-wi during the 6-hour-long test was not related to the growth or agglomeration of Ni
crystallites.

In Fig. 12B the DTG-TPO profiles corresponding to the different catalysts used are reported. The DTG
profiles of the mayenite obtained by boehmite precursor are similar in the two used catalysts, showing
three major peaks at around 200 °C, 350 °C and 600 °C. The percentage of deposited carbon corresponding
to each peak is summarized in Table 6. As widely reported in the literature, the peak at 200 °C can be
attributed to the carbon deposited on the metal active sites located on the surface of the catalyst [56,57].
The peak at 350 °C, much higher than the previous for Ni/mayenite B-wi, could be related to the oxidation
of filamentous carbon species at the metal-support interface and the last high peak appearing at 600 °C
could be associated to the oxidation of graphitic carbon species deposited on the mayenite support [28].
The oxidation profiles of the used catalysts obtained were similar, but with some differences in the
intensity of the oxidation peaks. The Ni/mayenite B-wi catalyst showed higher oxidation peaks at 350 and
600°C compared to the Ni/mayenite B-os counterpart. Probably the bigger size of the Ni particles observed
on the Ni/mayenite B-os catalyst (Table 6) lead to a reduction of the amount of filamentous carbon species,
which typically forms on small Ni particles, as in Ni/mayenite H-wi [28]. Furthermore, in the case of
Ni/mayenite B-wi the carbon deposits were mainly graphitic and thus less active, difficult to remove and
responsible for catalyst deactivation. The Ni/mayenite H-wi showed a total amount of carbon deposited
similar to that of Ni/ mayenite B with an intense peak corresponding to the filamentous carbon formation.
Table 6
Average crystallite size of the Ni particles and integrated areas of the TPO rate curves [58] on spent catalysts.
Catalysts
Crystal size (nm)
Carbon deposited at each temperature (%)
200
350
600
0.52
0.10
0
Ni/mayenite B-wi 31.52 (+ 10.8)
Ni/mayenite B-os 37.86 (- 0.04)
Ni/mayenite H-wi 27.02 (+ 5.3)
0.19
0.23
0.10
0.36
0.14
0.92
The total (sum) carbon deposited on the catalysts in mg of carbon per g of catalyst (mg of C g of cat.-1) after
a 6-h-long test at a temperature of 700 °C using toluene as tar model compound is summarized in Table 7.
The peak at 200 °C was considered because this type of carbon can move from the metal to the interface of
metal-support where further dehydrogenation and polymerization reactions are likely to occur [56],
ultimately developing into filamentous or graphitic carbon. For comparison purposes, the obtained amount
of carbon deposited values together with the main experimental conditions as well as the catalysts used
and the nickel addition method of several studies taken from the literature are showed in Table 7. In
general, the values for the carbon deposited on the catalysts are in agreement with the obtained values in
this study as the above-mentioned catalysts were prepared with the aim of reduce the carbon deposition

on the catalyst. Notably, the reported carbon deposited on the traditional Ni/α-Al2O3 catalyst was one
order of magnitude higher than the values measured in the present study.
Table 7
Carbon deposited values in addition to the main experimental conditions applied in the corresponding experiment.
Catalysts
Nickel addition
(method and loading)
Experimental conditions
Carbon deposited
(mg of C g of cat.-1)
Reference
Ni/La0.7Sr0.3AlO3
C7H8/H2O/Ar = 3/42/55 vol. %
(total 100 ml min-1);
GHSV = 12000 h-1;
S/C = 2;
T= 600 °C;
ToS = 180 min.
57
Wet impregnation
5 wt. %
Ni/α-Al2O3
431
800
[59]
Ni/LaAlO3
C7H8 = 3000 ppm;
cat. = 0.5 g;
ToS = 100 min; T = 700 °C;
S/C = 1.
12.2
10.9
C7H8 = 3000 ppm;
cat. = 0.5 g;
ToS = 100 min;
T = 700 °C;
Co-precipitation
Fe3Ni8/Palygorskite Ni = 8 wt. %
Fe = 3 wt. %
[54]
S/C = 1.5
C7H8 = 3000 ppm;
cat. = 0.5 g;
ToS = 100 min;
T = 700 °C;
11.1
6.7
S/C = 2
Co-precipitation
during the support
synthesis (one step)
Ni = 12 wt. %
Ni-Fe/Mg/Al
Ni/Mg/Al
C7H8/H2O/Ar = 0.75/8.9/26.8
mmol min-1;
Fe = 3.1 wt. %
T = 600 °C;
[47]
ToS = 80 min;
Co-precipitation
during the support
synthesis (one step)
Ni = 12 wt. %
W/F = 0.014 g h mol-1;
S/C = 1.7
79.4
Ni/mayenite B-wi
Ni/mayenite H-wi
10.7
10.2
C7H8 = 47 g Nm-3;
GHSV = 73750 h-1;
W/F = 0.33 g h mol-1;
S/C = 5;
ToS = 360 min;
T = 700 °C
Wet impregnation
10 wt. %
The
present
study
During the support
synthesis (one-step)
10 wt. %
Ni/mayenite B-os
4.7
4. Conclusion
Mayenite and Ni/mayenite catalysts were prepared using new procedures. Mayenite was synthetized
starting from a slurry using hydroxides as precursors or alternatively from a gel of boehmite and calcium

nitrate, each calcined at the suitable temperature determined by TGA-DSC. The slurry route appears of
particular interest due to the lower cost of the precursors and the easier and faster preparation, however
its calcination temperature was higher (ΔT= 300 °C) than in the case of gel route. Both procedures yielded
materials with similar crystal structure but differed in their morphological structure. Ni addition was then
synthesized both by wet impregnation and by direct inclusion of the precursors during the mayenite
preparation. Neither NiAl2O4 spinel phase nor formation of solid solution between NiO and CaO were
identified on the obtained catalysts. The Ni/mayenite catalysts were then tested on the steam reforming of
toluene and pyrocatechol as tar model compounds. The catalyst obtained by direct inclusion of Ni
precursor in the slurry route led to a poor dispersion of nickel and very low specific surface area values,
resulting therefore in the absence of catalytic activity. Addition of nickel through the wet impregnation
method led to analogous performances in all the conditions studied, despite morphological and structural
differences. The lower activity observed with Ni/mayenite obtained by direct inclusion of Ni precursor in
the gel route was attributed to its lower reduction degree, calculated after TPR characterization on the
fresh catalyst. Degradation of the support during the experiments evidenced by the presence of Ca3Al2O6 in
the XRD analysis and/or the higher initial nickel crystal size calculated for this catalyst could also have
contributed to the observed lower conversion values. Deactivation during toluene steam reforming of the
Ni/mayenite catalysts produced by wet impregnation was ascribed to filamentous and graphitic carbon
deposition for mayenite produced by the slurry route and the gel route respectively. Ni/mayenite obtained
by direct inclusion of Ni precursor in the gel route showed slightly lower conversion values with respect to
the catalysts prepared by the wet impregnation method but exhibited a promising carbon deposition
tolerance in long duration test. This catalyst is promising due to its higher resistance to carbon deposition
deactivation together with its simpler preparation method which foresees only one step.
Declarations of Interest
None
Acknowledgements
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-
profit sectors.

References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
P. McKendry, Bioresour. Technol. 83 (2002) 55–63.
R.M. Swanson, R.M. Swanson, Fuel 89 (2009) S11–S19.
N. Abdoulmoumine, S. Adhikari, A. Kulkarni, S. Chattanathan, Appl. Energy 155 (2015) 294–307.
L. Devi, K.J. Ptasinski, F.J.J.G. Janssen, Biomass and Bioenergy 24 (2002) 125–140.
K. Maniatis, A.A.C.M. Beenackers, Biomass and Bioenergy 18 (2000) 1–4.
W. Torres, S.S. Pansare, J.G. Goodwin, Catal. Rev. - Sci. Eng. 49 (2007) 407–456.
J. Kopyscinski, T.J. Schildhauer, S.M.A. Biollaz, Fuel 89 (2010) 1763–1783.
P. V. Aravind, W. De Jong, Prog. Energy Combust. Sci. 38 (2012) 737–764.
C. Li, D. Hirabayashi, K. Suzuki, Appl. Catal. B Environ. 88 (2009) 351–360.
[10] E. Savuto, A. Di Carlo, K. Gallucci, S. Natali, E. Bocci, Fuel 194 (2017) 348–356.
[11] B. de Caprariis, M.P. Bracciale, P. De Filippis, A.D. Hernandez, A. Petrullo, M. Scarsella, Can. J. Chem.
Eng. 95 (2017) 1745–1751.
[12] Z. Wang, Y. Pan, T. Dong, X. Zhu, T. Kan, L. Yuan, Y. Torimoto, M. Sadakata, Q. Li, Appl. Catal. A Gen.
320 (2007) 24–34.
[13] H. Hosono, K. Hayashi, K. Kajihara, P. V. Sushko, A.L. Shluger, Solid State Ionics 180 (2009) 550–555.
[14] H. Boysen, M. Lerch, A. Stys, A. Senyshyn, Acta Crystallogr. Sect. B Struct. Sci. 63 (2007) 675–682.
[15] A. Di Carlo, D. Borello, M. Sisinni, E. Savuto, P. Venturini, E. Bocci, K. Kuramoto, Int. J. Hydrogen
Energy 40 (2015) 9088–9095.
[16] M. Valentini, G. Groppi, C. Cristiani, M. Levi, E. Tronconi, P. Forzatti, Catal. Today 69 (2001) 307–314.
[17] J. Gil, J. Corella, M.P. Aznar, M.A. Caballero, Biomass and Bioenergy 17 (1999) 389–403.
[18] J. Corella, A. Orío, P. Aznar, Ind. Eng. Chem. Res. 37 (1998) 4617–4624.
[19] S. Gražulis, D. Chateigner, R.T. Downs, A.F.T. Yokochi, M. Quirós, L. Lutterotti, E. Manakova, J.
Butkus, P. Moeck, A. Le Bail, J. Appl. Crystallogr. 42 (2009) 726–729.
[20] D. Swierczynski, C. Courson, A. Kiennemann, Chem. Eng. Process. Process Intensif. 47 (2008) 508–
513.
[21] K. Polychronopoulou, J.L.G. Fierro, A.M. Efstathiou, J. Catal. 228 (2004) 417–432.
[22] G.F. Froment, Chemie Ing. Tech. 46 (1974) 374–386.
[23] H.S. FOGLER, Chem. Eng. (2006).
[24] G.F. Froment, K.B. Bischoff, J. De Wilde, Chemical Reactor Analysis and Design, Third, John Wiley &
Sons, Inc., 2011.
[25] D. Świerczyński, S. Libs, C. Courson, A. Kiennemann, Appl. Catal. B Environ. 74 (2007) 211–222.
[26] M.P. Aznar, M.A. Caballero, J. Gil, J.A. Martín, J. Corella, Ind. Eng. Chem. Res. 37 (1998) 2668–2680.
[27] A.S. Tolkacheva, S.N. Shkerin, S. V. Plaksin, E.G. Vovkotrub, K.M. Bulanin, V.A. Kochedykov, D.P.
Ordinartsev, O.I. Gyrdasova, N.G. Molchanova, Russ. J. Appl. Chem. 84 (2011) 907–911.
[28] E. Savuto, R.M. Navarro, N. Mota, A. Di Carlo, E. Bocci, M. Carlini, J.L.G. Fierro, Fuel 224 (2018) 676–

686.
[29] M.P. Bracciale, B. De Caprariis, C. Bassano, P. De Filippis, P. Deiana, A.D. Hernandez, M. Scarsella,
Chem. Eng. Trans. 65 (2018).
[30] M. Ruszak, S. Witkowski, P. Pietrzyk, A. Kotarba, Z. Sojka, Funct. Mater. Lett. 04 (2011) 183–186.
[31] M. Ruszak, S. Witkowski, P. Pietrzyk, A. Kotarba, Z. Sojka, Funct. Mater. Lett. 04 (2011) 183–186.
[32] M.G. Vincent, J.W. Jeffery, Acta Crystallogr. Sect. B Struct. Crystallogr. Cryst. Chem. 34 (1978) 1422–
1428.
[33] W.W. Russel, J. Chem. Educ. 21 (1944) 52.
[34] C.S. Martavaltzi, A.A. Lemonidou, Ind. Eng. Chem. Res. 47 (2008) 9537–9543.
[35] M.R. Cesário, B.S. Barros, C. Courson, D.M.A. Melo, A. Kiennemann, Fuel Process. Technol. 131
(2015) 247–253.
[36] J. Phromprasit, J. Powell, S. Wongsakulphasatch, W. Kiatkittipong, P. Bumroongsakulsawat, S.
Assabumrungrat, Int. J. Hydrogen Energy 41 (2016) 7318–7331.
[37] J.N. Kim, C.H. Ko, K.B. Yi, Int. J. Hydrogen Energy 38 (2013) 6072–6078.
[38] K. V. Manukyan, A.G. Avetisyan, C.E. Shuck, H.A. Chatilyan, S. Rouvimov, S.L. Kharatyan, A.S.
Mukasyan, J. Phys. Chem. C 119 (2015) 16131–16138.
[39] A. Cabello, P. Gayán, L.F. De Diego, A. Abad, M.T. Izquierdo, J. Adánez, Appl. Catal. B Environ. 147
(2014) 980–987.
[40] J. Zhang, H. Xu, X. Jin, Q. Ge, W. Li, Appl. Catal. A Gen. 290 (2005) 87–96.
[41] A. Di Giuliano, F. Giancaterino, C. Courson, P.U. Foscolo, K. Gallucci, Fuel 234 (2018) 687–699.
[42] B. Abdullah, N.A. Abd Ghani, D.V.N. Vo, J. Clean. Prod. 162 (2017) 170–185.
[43] S. Sokolov, E. V. Kondratenko, M.M. Pohl, A. Barkschat, U. Rodemerck, Appl. Catal. B Environ. 113–
114 (2012) 19–30.
[44] L. Zhang, X. Wang, B. Tan, U.S. Ozkan, J. Mol. Catal. A Chem. 297 (2009) 26–34.
[45] F. Arena, B.A. Horrell, D.L. Cocke, A. Parmaliana, N. Giordano, J. Catal. 132 (1991) 58–67.
[46] L. Pelletier, D.D.S. Liu, Appl. Catal. A Gen. 317 (2007) 293–298.
[47] M. Koike, D. Li, H. Watanabe, Y. Nakagawa, K. Tomishige, Appl. Catal. A Gen. 506 (2015) 151–162.
[48] D. Mukai, Y. Murai, T. Higo, S. Tochiya, T. Hashimoto, Y. Sugiura, Y. Sekine, Appl. Catal. A Gen. 466
(2013) 190–197.
[49] M.A. Goula, N.D. Charisiou, K.N. Papageridis, A. Delimitis, E. Pachatouridou, E.F. Iliopoulou, Int. J.
Hydrogen Energy 40 (2015) 9183–9200.
[50] A.M. Steele, S. Poulston, K. Magrini-Bair, W. Jablonski, Catal. Today 214 (2013) 74–81.
[51] D. Mukai, S. Tochiya, Y. Murai, M. Imori, Y. Sugiura, Y. Sekine, Appl. Catal. A Gen. 464–465 (2013)
78–86.
[52] N. Kaisalo, P. Simell, J. Lehtonen, Fuel 182 (2016) 696–703.
[53] H. Depner, A. Jess, Fuel 78 (1999) 1369–1377.
[54] X. Zou, T. Chen, P. Zhang, D. Chen, J. He, Y. Dang, Z. Ma, Y. Chen, P. Toloueinia, C. Zhu, J. Xie, H. Liu,

S.L. Suib, Appl. Energy 226 (2018) 827–837.
[55] J.-P. Eufinger, A. Schmidt, M. Lerch, J. Janek, Phys. Chem. Chem. Phys. 17 (2015) 6844–6857.
[56] J. Ruixia, X. Zaiku, Z. Chengfang, C. Qingling, Appl. Catal. A Gen. 250 (2003) 209–220.
[57] C.L. Li, O. Novaro, X. Bokhimi, E. Muñoz, J.L. Boldú, J. a Wang, T. López, R. Gómez, N. Batima, Catal.
Letters 65 (2000) 209–216.
[58] X. Xian, C. Ran, C. Nai, P. Yang, S. Zhao, L. Dong, Appl. Catal. A Gen. 547 (2017) 37–51.
[59] D. Mukai, S. Tochiya, Y. Murai, M. Imori, T. Hashimoto, Y. Sugiura, Y. Sekine, Appl. Catal. A Gen. 453
(2013) 60–70.