Hydrotreating of guaiacol: A comparative study of Red mud-supported nickel and commercial Ni/SiO2-Al2O3 catalysts

Article abstract of DOI:10.1016/j.apcata.2018.03.016

Upgrading of bio-oil through catalytic hydrotreating was investigated with guaiacol as a model compound. A nickel supported on red mud (Ni/RM) hydrotreating catalyst was developed and compared to the standard Ni/SiO₂-Al₂O₃

Full text of DOI:10.1016/j.apcata.2018.03.016

Accepted Manuscript
Title: Hydrotreating of guaiacol: A comparative study of Red
Mud-supported nickel and commercial Ni/SiO -Al O
2
2
3
catalysts
Authors: Hossein Jahromi, Foster A. Agblevor
PII:
S0926-860X(18)30126-1
DOI:
Reference:
https://doi.org/10.1016/j.apcata.2018.03.016
APCATA 16590
To appear in:
Applied Catalysis A: General
Received date:
Revised date:
Accepted date:
23-1-2018
7-3-2018
17-3-2018
Please cite this article as: Jahromi H, Agblevor FA, Hydrotreating
of guaiacol: A comparative study of Red Mud-supported nickel and
commercial Ni/SiO -Al O catalysts, Applied Catalysis A, General (2010),
2
2
3
https://doi.org/10.1016/j.apcata.2018.03.016
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.

Hydrotreating of guaiacol: A comparative study of Red Mud-supported nickel and
commercial Ni/SiO -Al catalysts
2
2
O
3
Hossein Jahromi, Foster A. Agblevor
USTAR Bioenergy Center, Department of Biological Engineering, Utah State University, UT
8
4322, USA
Graphical abstract
1

Highlights




Red Mud can serve as nickel catalyst support for hydrodeoxygenation reactions.
Ni/RM improves deoxygenation over hydrogenation compared to commercial Ni/SiO
Al
2
-
2
3
O .
Ni/RM catalyst moderates hydrocracking and coke formation versus commercial
Ni/SiO -Al
2
2
3
O .
Coke deposition, oxidation of nickel, and formation of iron nickel oxide contribute to
deactivate Ni/RM catalyst during HDO.
Abstract
Upgrading of bio-oil through catalytic hydrotreating was investigated with guaiacol as a model
compound. A nickel supported on red mud (Ni/RM) hydrotreating catalyst was developed and
compared to the standard Ni/SiO
Ni/RM catalyst was characterized by inductively coupled plasma atomic emission spectrometry
ICP-AES), X-ray diffraction analysis (XRD), scanning electron microscopy (SEM), BET
2
-Al
2
O
3
catalysts under similar experimental conditions. The
(
specific surface area, and temperature programmed reduction (TPR). The effects of reaction
temperature (300, 350, 400 °C) and initial hydrogen pressure (4.83 MPa (700 psi), 5.52 MPa
(800 psi), and 6.21 MPa (900 psi)) on products distribution were investigated. The major
products of hydrotreating process were catechol, anisole, phenol, cyclohexane, hexane, benzene,
toluene, and xylene. Increasing the reaction temperature and hydrogen pressure improved HDO
2

reactions. Complete HDO was achieved at reaction temperature of 400 °C and initial hydrogen
pressure of 6.21 MPa (900 psi). Under these conditions, the selectivity to cyclohexane, benzene,
toluene, and xylene over Ni/RM catalyst were 38.8, 24.5, 18.1, and 7.9% respectively, whereas
these values were 62.2, 15.9, 8.4, and 4.5% respectively over Ni/SiO
2
-Al
2
3
O . Reaction network
and the kinetics of guaiacol HDO were proposed according to analysis of the products. The
Ni/RM catalyst was more effective for deoxygenation reactions than hydrogenation while
commercial Ni/SiO
hydrogen consumption per gram of bio-oil was lower for the Ni/RM catalyst compared to the
Ni/SiO -Al . There was less hydrocracking and coke formation for the Ni/RM compared to the
2
-Al
2
O
3
was more effective for hydrogenation than deoxygenation. Thus,
2
2
O
3
commercial catalyst and this resulted in higher liquid yield for the new catalyst.
Keywords: Guaiacol; hydrodeoxygenation; red mud; nickel catalyst; pyrolysis
1
. Introduction
Biofuels are emerging as a promising substitute to fossil fuels, as they are renewable and
contribute to lower CO emissions [1-4]. Fast pyrolysis is an attractive thermochemical
2
conversion process to generate alternative liquid fuels from biomaterials [5, 6]. These pyrolysis
oils (crude bio-oils) are multicomponent mixtures of a large number of oxygenated compounds.
However, the high oxygen content of crude bio-oils, usually 20 to 50 wt. %, results in a low
heating value, poor stability, poor volatility, high viscosity and corrosiveness [7-12]. Therefore,
oxygen removal from bio-oils is necessary for further application as liquid fuels.
3

Noble metal [13-16] and transition metal sulfide [17-20] catalysts have been studied for this
purpose. Noble metals are relatively costly and they catalyze hydrogenation of aromatic rings
[
21, 22], which is undesirable in fuel production. Noble metal catalysts suffer from deactivation
due to oxidation or hydrolysis of the active phase unless a sulfiding agent is added [1, 23-25]
which increases the risk of product contamination. Thus, it is vital to find alternative catalysts for
HDO applications in bio-oil upgrading.
Red mud (RM) is a caustic by-product of processing of bauxite that is generated in the industrial
production of alumina using the Bayer process [26]. There are many problems with managing
RM slurry including costly maintenance of RM ponds, risk of caustic pollutants for all living
organisms, and leakage of alkaline contaminants into groundwater resources [26, 27]. The main
chemical components of RM are Fe
2
O
3
, Al
2
O
3
, SiO
2
, CaO, Na O, and a number of minor metal
2
oxides of Cr, V, S, Ni, Cu, Mn, Zn, Mg etc. [27]. RM has been tested as catalyst for various
applications, including pyrolysis of biomass [28, 29], hydrogenation and liquefaction of coal and
biomass [30-35], hydrodechlorination and desulfurization reactions [36-38], methanogenesis
reaction [39] and exhaust gas clean-up [40, 41]. The RM conversion into a value added product
such as heterogeneous catalyst will be a desirable technology for the final fate of the residue.
Bio-oil is a complicated mixture of more than two hundred compounds such as aromatics,
aliphatics, aldehydes, ketones, and ethers. Therefore it is difficult to interpret the HDO results of
the actual bio-oil [12]. Although there are too many cross-interactions between bio-oil
compounds, guaiacol is a suitable model compound representing the decomposition products of
lignin pyrolysis, as it contains phenolic hydroxyl and methoxy groups like many degradation
products of lignocellulose [2, 3, 42, 43]. It has been reported that incorporation of Mo and V in
Ni/TiO
2
catalyst improved the catalyst performance for the HDO of guaiacol. In aqueous
4

medium, Ni-V/TiO
2
was more effective resulting in better guaiacol conversion than Ni-Mo/TiO
2
and showing more tolerance for water [44]. Ni metal was effective for ring opening activity
while Co favored deoxygenation activity. The Ni catalyst favored the multiple hydrogenolysis of
C-C bond, leading to carbon loss via methanization. Cobalt not only facilitated the reduction and
dispersion of nickel metal oxides but also enhanced the HDO activity of bimetallic Ni–Co
catalyst [45]. Hydrodeoxygenation of guaiacol using silica alumina supported Ni-Pd and Ni-Pt
overlayer catalysts under atmospheric pressure showed that Pd and Pt active sites of overlayer
catalysts significantly enhanced deoxygenation activity and the BTX selectivity could be up to
8
0% on Ni-Pd overlayer catalyst [46]. HDO of guaiacol over NiB/SiO
Al catalysts showed that the conversion of guaiacol and the selectivity to cyclohexane were
much higher on the former than the latter [47]. However, SiO -supported Ni-Mo bimetallic
2
-Al
2
O
3
and Ni/SiO
2
-
2
O
3
2
catalyst showed high selectivity towards AR-O bond cleavage in the HDO reactions of anisole,
phenol and guaiacol, giving BTX selectivity more than 96% [48]. Lately, Mora-Vergara et al.
[
49] showed that the addition of potassium to alumina-supported NiMo and CoMo catalysts
promoted selectivity to deoxygenated compounds in the HDO of guaiacol.
A potential method of reducing the cost of HDO process, is the development of new HDO
catalysts from low cost materials. Red mud has interesting properties such as active components
and relatively large surface area. The main objective of this research work is to evaluate the
performance of RM-supported nickel catalyst for hydrodeoxygenation of bio-oils and guaiacol
was used as a model compound for biomass pyrolysis oil. In view of the stability and poisoning
resistance of RM, it can be a catalyst support for Ni in HDO process and it would be potentially
much more cost-effective in comparison with other catalyst supports such as Al
2
O
3
, SiO , and
2
5

TiO
2
that require complicated preparation procedures. In this paper the performance of the
-Al obtained from Sigma Aldrich.
Ni/RM is compared with commercial Ni/SiO
2
O
2 3
2
. Material and methods
2
.1.Material
Guaiacol, anisole, catechol, and nickel nitrate hexa hydrate (Ni(NO
3
).6H O) were purchased
2
from Alfa Aesar (Haverhill, MA, USA). Cyclohexane and toluene were bought from Pharmco-
Aaper (Brookfield, CT, USA). Commercial nickel on silica/alumina (~65% wt. % Ni loading)
catalyst powder, benzene, and phenol were purchased from Sigma Aldrich (St. Louis, MO,
USA). Hexane and xylene were obtained from Fisher Scientific (Hampton, NH, USA). All
chemicals were analytical grade and were used as received without any further purification. High
purity (99%) hydrogen (Airgas, PA, USA) was used for HDO experiments.
2
.2. Preparation and characterization of Ni/RM catalyst
Ni/RM catalysts were prepared at 40 wt. % nickel metal loading using wet impregnation method
50, 51]. The calculated amount of Ni(NO ).6H O equivalent to 40 wt. % nickel metal loading
[
3
2
was dissolved in 100 ml deionized water and then mixed with red mud (particle size<90 μm).
The mixture was heated to 70 °C and continuously stirred for 5 hours to prepare the catalyst
precursor. The catalyst precursor was dried at 105 °C for 10 hours and then calcined in air at 620
°
C in a muffle furnace (Thermo Scientific, Inc., Waltham, MA, USA) for 5 hours. The calcined
material was reduced for 6 hours at 450 °C using a reducing gas mixture of 10% H
at flow rate of 20 ml/min to obtain the final catalyst.
2
and 90% N
2
6

The Brunauer-Emmet-Teller (BET) specific surface area of the catalysts were determined on a
MS-16 BET analyzer (Quantachrome Instruments, Boynton Beach, FL, USA). About 0.1 g
catalyst sample was used in each measurement. Prior to the analysis, the catalyst samples were
milled into fine powder using a SPEX Certiprep 6750 cryogenic miller (Metuchen, NJ, USA).
All samples were degassed at 300 °C for 4 hours prior to duplicate measurements. A porous
Al
2
3
O standard was measured along with each run to ensure consistency between the different
samples.
Metal loading determination was carried out using inductively coupled plasma atomic emission
spectrometry (ICP-AES) analysis. For this purpose, 0.5 g catalyst sample was digested in nitric
acid at 95 °C for 1 h. Then 12 ml, 30 wt. % hydrogen peroxide was added to the digestion tube
and cooled to room temperature [52]. The extracts were measured using a Thermo iCAP 6300
ICP-OES Inductively-Coupled Plasma Spectrophotometer (Thermo Fisher Scientific, Waltham,
MA, USA) with Optical Emission.
Thermo gravimetric-temperature programmed reduction (TG-TPR) studies were carried out
using a TGA Q500 (TA Instruments, Lindon, UT, USA). Twenty-five milligrams of catalyst
(calcined form) was heated in a flow of 10% H
2
/ 90% N (20 ml/min) from room temperature to
2
7
00 °C at a heating rate of 10 °C/min. Hydrogen uptake was monitored by the change in sample
weight according to the plot of derivative weight vs. temperature.
X-ray powder diffraction (XRD) analyses were carried out by Hazen Research Inc. (Golden, CO,
USA). The samples were ground in a mortar and pestle with isopropyl alcohol and analyzed
using a Bruker D8 Advance with Davinci design and a Lynxeye detector. The diffraction pattern
was measured in the interval from 5 to 85 in 2Ɵ using a 0.02 step size and 40 s of counting time.
7

Scanning electron microscopy (SEM) analysis was conducted on a FEI Quanta FEG-650
Thermo Fisher Scientific, MA, USA). For preparation of the specimens a small amount of
(
catalyst was placed on a two-sided sticky tape resting on an aluminum holder and observed at
different magnifications. Images were recorded using low vacuum secondary electron (LFD)
detector.
2
.3. Hydrodeoxygenation (HDO) experiments
All experiments were conducted in a Parr Series 4560 300 mL autoclave reactor (Parr
Instruments, Moline, IL, USA). This reactor had a variable speed magnetic drive and turbine
agitator. A temperature sensor, immersed in the reactor content, was used to measure the liquid
temperature. The reaction temperature was maintained at its desired value with an accuracy of ±1
°
C. The setup had an electrically heated jacket to ensure isothermal conditions. The temperature
and speed of agitation were controlled by using a Parr 4848 controller. In each experiment, the
reactor was charged with guaiacol (30 g) and catalyst (4.5 g). The reactor was then purged with
2
N to ensure an inert atmosphere. The reactor was then charged with high purity hydrogen
supplied from a reservoir tank to desired pressures of 4.83, 5.52, or 6.21 MPa (700, 800, or 900
psi) via a pressure regulator. A gas sample was taken from a gas release valve from the gas
sampling port for gas analysis when the reactor was at room temperature. The reactor was then
heated to reaction temperature (300, 350, or 400 °C) at heating rate of 15 °C/min. The reaction
time was recorded when the set temperature was reached. The reactor content was stirred at the
desired speed of agitation (~1000 rpm) for all experiments. Seven liquid samples were collected
3
at fixed time intervals for kinetic studies. Every time, a sample volume equal to 0.3 cm was
collected. The change in volume of liquid inside the reactor was negligible. After the desired
8

reaction time (30 minute), the reactor was cooled to room temperature using the internal cooling
coil and an external air fan simultaneously. A gas sample was collected in a tedlar bag for gas
analysis when the reactor was cooled to room temperature. The reproducibility of experiments
was checked and the error in all experimental measurements was found to be less than 3%.
In a blank experiment (without catalyst) 30 g of guaiacol was charged into the reactor and the
reactor was pressurized to 6.2 MPa (900 psi) with hydrogen and allowed to react for 30 minutes
at 400 ° C. All experiments were conducted in triplicate.
2
.4. Analysis of HDO products
Hydrogen consumption, gas analysis, and product yields were determined as described in
previous work [12]. The liquid products of HDO experiments were analyzed by HPLC
(Shimadzu Scientific, Columbia, MD) using a RID-10A detector and a Kromasil 100-5-C18
column obtained from AkzoNobel (Amsterdam, Netherlands). The HPLC was equipped with a
LC-10AT pump, SCL-10Avp controller, and SIL-10A autosampler. CLASS-VP 7.3 SP1
software was used to analyze HPLC chromatograms. A CTO-10A column oven was used to
maintain the column temperature at 55 °C during the analysis. The injection volume was 0.25 μl
and acetonitrile at flow rate of 0.6 ml/min was used as the mobile phase. Data acquisition time
was 60 minutes for all analyses.
The liquid samples were analyzed for guaiacol, anisole, catechol, phenol, cyclohexane, hexane,
benzene, toluene, and xylene. To quantify the amount of each compound, five solutions of 20,
4
0, 60, 80, and 100 wt. % of each compound were prepared and injected to HPLC system and
peak area vs. concentration was plotted to obtain the calibration curve of the compounds.
9

The elemental composition of HDO products were determined using ThermoFischer Scientific
Flash 2000 organic elemental analyzer (ThermoFisher Scientific,Waltham, MA, USA), and the
oxygen content was calculated by difference according to ASTM D5291.
Conversion of guaiacol and product selectivity were calculated using equations (1) and (2)
respectively [53]:
푀ꢀ푙ꢁꢂ ꢀ푓 푔푢푎ꢃ푎푐ꢀ푙 ꢄꢁ푎푐푡ꢁ푑
퐶표푛푣푒푟푠푖표푛 % =
× 100
(1)
푀ꢀ푙ꢁꢂ 푔푢푎ꢃ푎푐ꢀ푙 푓ꢁ푑
푚ꢀ푙ꢁꢂ ꢈꢉ
퐶푝 푆푒ꢅ푒ꢆꢇ푖푣푖ꢇ푦 % = ∑
× 100
(2)
푚ꢀ푙ꢁꢂ ꢈꢉ
Where C
p
represents the content of products.
Kinetic studies using integral method was perform to estimate the reaction rate equations
according to equation (3) [54].
푑ꢈꢊ
= 푘. 퐶퐺ꢋ
−
푟퐺 =
(3)
푑푡
Where r
G
is the rate of disappearance of guaiacol, C
G
is the concentration of guaiacol at time t, k
is the reaction rate constant and n is the reaction order.
The water content of the aqueous products were determined by Karl-Fischer titration method
using Hydranal® -composite 5 solution. A Metrohm 701KF Titrino and 703 titration stand setup
(Brinkmann Instruments, Riverview, FL, USA) were used for the volumetric Karl Fischer
titration.
3
. Results and discussion
.1. Characterization of Ni/RM catalyst
3
1
0

RM (in reduced form) was tested in HDO experiment to investigate whether the support had
HDO activity or it acted as an inert support. The BET specific surface area of RM support
2
(
calcined form) was relatively low (37.5 m /g). Reduction of the calcined RM increased its
2
specific surface area to 54.3 m /g. The increase in specific surface area was attributed to
reduction of some metallic oxides.
2
The specific surface area of the Ni/RM catalyst precursor (calcined form) was 51.3 m /g while
2
after activation by reduction, the BET specific surface area increased to 79.3 m /g. The increase
in BET specific surface area after the reduction was attributed to the pores that were created to
accommodate the shrinkage in the catalyst precursor associated with reduction of NiO to Ni [55-
5
8]. These results are in agreement with those reported in literature [59].
The ICP analysis of the RM and 40%Ni/RM catalyst showed that the major metals present in
RM were Al, Ca, Fe, Na, Si and Ti and trace amounts of other metals such as As, Cd, Cu, Mo,
Zn etc. (Table 1). The Ni content of the red mud was negligible but after the impregnation the Ni
content was 40.8%, which was close to the estimated amount incorporated. The Ni loading of the
commercial Ni/SiO
2
-Al
2
O catalyst was 63.3% (Table 1) which was close to the 65 wt. % stated
3
on the sample bottle by the manufacturer.
TG-TPR was performed to investigate the reducibility of nickel oxide, RM support, and the
interaction between nickel and the RM support. The TPR profile of RM showed three major
reduction peaks at 279.9, 542.7, and 694.6 °C (Fig. S1 in supplementary material). The peak at
2
5
79.89 °C could be due to the reduction of Fe
2
O
3
to Fe
3
O
4
and the unresolved peaks centered at
to FeO and then to elemental Fe [27].
42.68 °C and 694.6 °C could be due to reduction of Fe
3
O
4
1
1

The TPR profiles of Ni/RM catalyst showed two major reduction peaks at 442.22 and 583.11 °C
Fig. S2 in supplementary material). The peak at 442.22 °C was attributed to reduction of nickel
(
oxide, but this peak also indicated a possible interaction with RM support since its intensity
increased with increasing the Ni content (data not reported) and its reduction temperature shifted
towards that of nickel oxide (347.89 °C according to Fig. S3 in supplementary data). This
observation will be in agreement with Jeangros et al. who observed interaction between SiO and
2
NiO during the reduction of NiO to Ni [57]. The peak at 583.11 °C was attributed to the
reduction of RM components (mostly Fe
increased (data not provided) Ni supported on Fe
method to investigate the effect of the presence of Fe
on the reducibility of Ni. The TPR profile of NiO/Fe
showed interaction between NiO and Fe
reduction of Fe to magnetite. These results suggested that, besides Fe
other components of RM too.
Fig. 1 shows the XRD pattern of fresh Ni/RM catalyst. Distinct sharp peaks observed at 2Ɵ
2.2°, 61° were attributed to elemental Ni from the reduction of NiO. The detection of NiO
signal at 2Ɵ 43.7°, 50.9°, and 74.7° showed that nickel oxide was not completely reduced to
elemental Ni after treatment with H (Fig. 1) which is in agreement with Jeangros et al (2013)
who reported that total NiO reduction does not occur until the reduction temperatures is above
00 C [57]. Detection of magnetite (Fe ) peaks at 2Ɵ 21.6°, 35.1°, 41.7°, and 74.3° were due
to reduction of hematite (Fe ) that was present in the original RM [28].
2
O
3
) since its intensity decreased as Ni loading
.
2
O
3
was prepared using wet impregnation
2
O
3
(which was the major RM component)
(Fig. S4 in supplementary material)
2
O
3
2
O
3
at 390.03 °C. The peak at 522.67 °C was ascribed to
, Ni interacted with
2
O
3
2
O
3
5
2
6
3
O
4
2
O
3
XRD pattern of used Ni/RM catalyst (Fig. 2) showed stronger NiO peaks compared to fresh
catalyst suggesting that Ni was partially oxidized during HDO. This could be one reason for the
1
2

partial deactivation of the catalyst. Additionally, iron nickel oxide (trevorite) (Fe
2
NiO ) peaks
4
were also detected after HDO that could be another reason for partial deactivation of active Ni
sites. The detection of weak elemental nickel peaks (2Ɵ 52.1° and 61°) in the used catalyst
suggested that the catalyst was only partially deactivated.
In order to recycle the catalyst, the used catalyst was regenerated in muffle furnace to burn off
the deposited carbonaceous compounds (coke) at 400 °C for 4 h. The XRD pattern of the
regenerated catalyst (Fig. 3) showed strong NiO diffraction peaks at 2Ɵ 43.9°, 50.8°, and 75°
which was due to the oxidation of nickel to NiO under heat treatment in air. Existence of
hematite (Fe
2
3
O ) diffraction peaks in regenerated catalyst was ascribed to oxidation of magnetite.
Furthermore, it was interesting to note that weak elemental Ni diffraction peaks were present in
the XRD pattern of the regenerated catalyst (Fig. 3). This suggested that some Ni particles in the
interior of the RM were not exposed to oxygen due to mass transfer limitations.
The regenerated catalyst was then activated by reduction in H
regenerated and activated catalyst (Fig. 4) showed similar diffraction peaks as the fresh catalyst
Fig. 1) suggesting complete regenerability of the catalyst which was in contrast to the
2
. The XRD pattern of the
(
commercial catalyst that could not be regenerated using similar procedure. More detailed
information on catalytic activity of the used catalyst and the recycled catalyst are discussed in
section 3.6.
The surface morphology of the catalyst was studied by using scanning electron microscopy
(SEM). The SEM images of the Red Mud support, Ni/RM catalyst precursor (calcined form),
fresh Ni/RM catalyst (reduced form), and used 40% Ni/RM catalyst are presented in Fig. 5. In
the case of RM support (Fig. 5a), in addition to amorphous looking particles, crossed cocentric
discs and spherical-shaped particles were observed. The SEM image of Ni/RM catalyst in
1
3

calcined form is shown in Fig. 5b. Ni particles with relatively uniform morphology and
approximate mean diameter of 90 nm were dispersed on the surface of RM components in the
case of activated Ni/RM (Fig. 5c). The SEM picture for the used catalyst (Fig. 5d) showed that
the catalyst was coated with coke after HDO process, which could inhibit the hydrogen transfer
and partially deactivate the catalyst.
3
.2. HDO products yield distribution and selectivity
In the blank experiment without catalyst, the guaiacol did not convert to any other compounds
and neither did it produce any gas nor coke, hence the reactor was not catalytically active and did
not influence the HDO experiments.
The major HDO liquid products using the nickel catalysts were catechol, anisole, phenol,
cyclohexane, hexane, benzene, toluene, and xylene. CO, CO
2
and C1—
C hydrocarbon gases as
5
well as water and coke. When RM was treated using similar reaction conditions as Ni/RM
catalyst and used in HDO experiment, the guaiacol conversion was only 2.7%. A relatively high
amount of coke was formed (18.3 wt. %) and 96.3 wt. % of the final liquid products was
unreacted guaiacol and small amounts of benzene, phenol, and anisole. The results of HDO
experiments using RM support are summarized in Table S1 in supplementary material. Therefore
the RM support contribution to HDO process was very low. The effects of reaction temperature
and hydrogen pressure are discussed in subsections 3.2.1 and 3.2.2 respectively.
3
.2.1. Reaction temperature effect
1
4

Fig. 6 shows the overall products yield distribution of guaiacol HDO as a function of reaction
temperature. At all reaction temperatures, Ni/RM catalyst produced higher organic liquid yield
compared to Ni/SiO
of Ni/SiO -Al , which could be due to the higher specific surface area and higher Ni content
of the catalyst. The aqueous liquid yield at reaction temperatures of 300, 350, and 400 °C were
2
-Al
2
O . In contrast higher amounts of gas and coke were formed in the case
3
2
2
O
3
9
1
.8, 12.4 and 16.7 % respectively for Ni/RM catalyst while those for Ni/SiO
5.6, and 21.1 % respectively. The higher aqueous liquid yield was due to increased methanation
-Al (Table 2) [12, 60]. The aqueous phase was
2
-Al
2
O
3
were 11.3,
of carbon dioxide over the commercial Ni/SiO
9.9 wt. % H O.
Pressure changes during the HDO of guaiacol at different reaction temperatures are shown in
Fig. S5 (supplementary material). The total pressure profile using Ni/SiO -Al fell below the
2
2
O
3
9
2
2
2
O
3
pressure profile of Ni/RM catalyst due to higher hydrogen consumption in the case of the
commercial catalyst (Table 2).
Fig. 7 and 8 show the HDO product selectivity and guaiacol conversion as a function of reaction
temperature for Ni/RM and Ni/SiO
2
-Al
2
3
O catalysts respectively. Increasing the reaction
temperature improved the conversion of guaiacol in all HDO experiments. At reaction
temperatures of 300 and 350 °C; catechol, anisole, and phenol were produced due to
demethylation, dehydroxylation, and demethoxylation of guaiacol respectively, however the
selectivity to these compounds were lower for Ni/RM compared to Ni/SiO
In contrast the selectivity to benzene, toluene, and xylene (BTX) was lower for Ni/SiO
suggesting that Ni/RM was more effective in hydrodeoxygenation than Ni/SiO -Al O .
2
-Al
2
O
3
(Fig. 7 and 8).
2
-Al
2
O
3
2
2
3
When the reaction temperature was increased from 300 °C to 400 °C the selectivity to
oxygenated intermediates (catechol, anisole, and phenol) decreased significantly indicating that
1
5

guaiacol was first converted to catechol, anisole, and phenol via demethylation, dehydroxylation,
and demethoxylation, and subsequently to cyclohexane, benzene, toluene, and xylene. Hexane
was also obtained in relatively lower selectivity as a result of ring opening. At reaction
temperature of 400 °C complete hydrodeoxygenation of guaiacol was achieved with both
catalysts. However, Ni/SiO
2
-Al
2
O
3
was more effective in hydrogenation of aromatic ring
compared to Ni/RM resulting in benzene, toluene, and xylene selectivities of 24.5, 17.1, and 7.9
%
respectively for Ni/RM while these values were 15.9, 8.4, and 4.5 % respectively for Ni/SiO
. Higher rates of hydrogenation reaction over Ni/SiO -Al resulted in cyclohexane
2
-
Al
2
O
3
2
2
O
3
selectivity of 62.2 % versus 38.8 % for Ni/RM at 400 °C (Fig. 7 and 8). This result is also
reflected in higher hydrogen consumption over Ni/SiO
The major gas product was methane for both catalysts.
2
-Al
2
3
O compared to Ni/RM (Table 2).
3
.2.2. Hydrogen pressure effect
The influence of hydrogen pressure on the guaiacol HDO products yield distribution is shown in
Fig. 9. Increasing the initial hydrogen pressure from 4.83 to 6.21 MPa increased the liquid
product yield for both catalysts, however the organic liquid yield was much higher over Ni/RM.
The increase in organic liquid yield could be due to an increase in concentration of adsorbed
hydrogen on the catalyst surface and an increase in the amount of dissolved H
phase, as hydrogen pressure was increased [12, 61, 62]. For Ni/RM, the gas yield decreased from
3.7% to 8.4% over this pressure range while the gas yield decreased from 41.4% to 16.6% for
Ni/SiO -Al . The coke yield decreased from 6.4% to 2.1% for Ni/RM and from 9.5% to 5.5%
using Ni/SiO -Al (Fig. 9). This would suggest that higher levels of hydrogen pressure helped
HDO reactions to predominate cracking and (re)polymerization reactions that produce gas and
2
in the organic
2
2
2
O
3
2
2
O
3
1
6

coke [63]. The increase in aqueous liquid yield by increasing hydrogen pressure could also
confirm that higher H pressures favored HDO reactions. Pressure profile during the HDO of
guaiacol at different hydrogen pressures are shown in Fig. S6 in supplementary material. The
reaction pressure in the case of Ni/SiO -Al was lower than that of Ni/RM due to higher
2
2
2
O
3
hydrogen consumption by the commercial catalyst (Table 3).
During the catalytic hydrotreatment processes, multiple reactions may occur, including
hydrogenation, hydrogenolysis, hydrodeoxygenation, hydrocracking, and polymerization [5, 9].
Hydrogenation for ring saturation, hydrocracking for gasification, and methanation made
hydrogen consumption exceed the deoxygenated stoichiometric ratio [9]. In order to reduce the
hydrogen consumption, the direct deoxygenation without ring saturation is desirable in HDO.
Figs 10 and 11 show the guaiacol conversion and evolution of products as a function of hydrogen
pressure over Ni/RM and Ni/SiO
.52 MPa, oxygenated intermediates (catechol, anisole, and phenol) were produced while these
compounds were completely deoxygenated when the initial H pressure was 6.21 MPa.
However, higher selectivity to benzene, toluene, and xylene were observed (24.5, 18.1, and 7.9%
respectively) for Ni/RM compared to Ni/SiO -Al (15.9, 8.4, and 4.5% respectively) resulting
in less hydrogen consumption in the case of Ni/RM (Table 3). This result was attributed to
higher ring hydrogenation activity of Ni/SiO -Al since this catalyst had higher Ni loading
63.3% according to Table 1) than Ni/RM (40%). Furthermore, higher catalytic activity of
Ni/SiO -Al led to higher cyclohexane selectivity of 62.2 % (Fig 11) because of the
2
-Al
2
O
3
respectively. At initial H
2
pressures of 4.83 MPa and
5
2
2
O
2 3
2
2
O
3
(
2
2
O
3
hydrogenation of the aromatic ring and demethylation reaction resulting in lower BTX
selectivity.
1
7

The major gas product for both catalysts was methane for both catalysts (Table 3) because of
demethylation, methanation and cracking reactions. It is interesting to note that at all
temperatures and pressures, the Ni/RM produced CO, whereas the Ni/SiO2-Al
2
3
O did not
produce any CO probably because of the methanation reaction that was stronger in this catalyst
than the Ni/RM.
Guaiacol HDO liquid products (organic and aqueous phases) were analyzed for Ni using ICP in
order to investigate leaching effect. The Ni content of HDO products were below detection limit
(0.05 mg/kg) suggesting that Ni did not leach into the HDO products (data not reported).
3
.3. Evaluation of HDO process
To accurately visualize and extrapolate trends from elemental analysis, the elemental
composition of the products were plotted in Van Krevelen-type diagrams. The molar O/C versus
molar H/C ratios of the feed guaiacol and the HDO products are shown in Figs. 12 and 13 at
different temperatures and initial hydrogen pressures respectively. The O/C ratio of HDO
products decreased as a function of reaction temperature for both catalysts, however, higher
hydrogenation occurred over Ni/SiO
Ni/RM was 0.18 versus 0.22 for Ni/SiO
Ni/RM and Ni/SiO -Al (0.09 and 0.10 respectively) however the H/C ratio at 300, 350, and
00 °C were 1.40, 1.46, and 1.59 respectively for Ni/RM versus 1.42, 1.63, and 1.93 respectively
for Ni/SiO -Al HDO products. Thus, there was more hydrogenation with Ni/SiO -Al than
Ni/RM. Moreover, the Van Krevelen plot of HDO products over Ni/RM catalyst at different H
pressures (Fig. 13) fell below the HDO products over Ni/SiO -Al showing that for HDO of
guaiacol, less saturation of double bonds took place over Ni/RM catalyst resulting in lower
hydrogen consumption which is desirable in HDO process. These results were mainly attributed
2
-Al
2
O
3
. At 300 °C the O/C ratio of HDO products of
2
-Al
2
O3. At 350 °C this value was relatively the same for
2
2
O
3
4
2
2
O
3
2
2
O
3
2
2
2
O
3
1
8

to higher catalytic activity of Ni/SiO
2
-Al
2
3
O because of the higher Ni loading that favored
hydrogenation and hydrocracking reactions. Also, to a lesser extent, silica and alumina can
catalyze demethylation reaction [1, 20, 64] contributing to increased hydrogen consumption in
the case of Ni/SiO
2
-Al
2
3
O .
3
.4. Reaction pathways
For the study of HDO pathways of guaiacol over Ni/RM and Ni/SiO
2
-Al
2
O , HDO of catechol,
3
anisole, phenol, cyclohexane, benzene, toluene, and xylene were carried out under the same
conditions as the HDO of guaiacol. These reaction results are shown in Table 4. Guaiacol was
first converted to catechol, anisole, and phenol via demethoxylation and dehydroxylation
reactions. The selectivity to catechol, anisole, and phenol were 1.5, 2.3, 9.5 % respectively for
Ni/RM and 3.4, 4.7, and 10.8 % respectively for the commercial catalyst after HDO of guaiacol.
Anisole underwent demethylation, demethoxylation, and alkylation to produce phenol, benzene,
and toluene respectively. Catechol was first converted to phenol via dehydroxylation and then
converted to benzene. Benzene was initially formed via demethoxylation of anisole,
dehydroxlylation of phenol, and demethylation of toluene, and then hydrogenated to
cyclohexane. At relatively lower selectivity compared to other intermediates, xylene was formed
via a transalkylation reaction [45, 46]. HDO of benzene produced cyclohexane through ring
hydrogenation at selectivity of 88.2% and 67.6% over Ni/RM and Ni/SiO
2
-Al
2
3
O respectively
but ring opening was more pronounced over the commercial catalyst (32.4% selectivity)
compared to Ni/RM (11.8% selectivity) (Table 4). A summary of the HDO pathways of guaiacol
based on the observed product distribution is proposed in Fig. 14. Pathways 1, 2, and 3 are
demethylation, demethoxylation, and dehydroxylation processes, respectively. Pathway 4 is
1
9

hydrogenation which was more pronounced over Ni/SiO
2
-Al
2
3
O than Ni/RM. Pathway 5 is a
transalkylation reaction in which the methyl group is transferred to the aromatic ring [65]. In
conclusion, similar reaction pathways over both catalysts were observed while products
selectivity were quite different as discussed in section 3.2. Overall, the conversion of oxygenated
compounds (guaiacol, catechol, anisole, and phenol) were higher for Ni/RM compared to the
commercial catalyst suggesting higher activity of Ni/RM catalyst for deoxygenation than
hydrogenation (Table 4). The Ni/SiO
BTX and ring opening compared to Ni/RM.
.5. Kinetic studies
In order to determine the reaction order and rate constant of the reaction, the activity data
concentration and time) was fitted to power-law equation [54]. As shown in Fig. 15, the slope of
2
-Al
2
O
3
showed higher conversion for hydrogenation of
3
(
a plot of 1/Cguaiacol as a function of reaction time is linear with slope of k, hence, HDO of
guaiacol global kinetics followed the second order mechanism which was in agreement with
other studies [44]. The rate constant of Ni/SiO
higher than the Ni/RM catalyst (24.4 ml/mol.s), which explains the higher activity of Ni/SiO
Al compared to Ni/RM. Ni/RM has shown better performance over Ni/SiO -Al since the
2
-Al
2
3
O (55.8 ml/mol.s) was more than two times
2
-
2
O
3
2
2
O
3
latter favored ring hydrogenation reaction. Since the support was different for both catalysts, the
activity difference could be due to both Ni loading and interaction with support. In the case of
Ni/RM the elemental Ni loading was less than the nominal 40% because the XRD pattern
showed the presence of NiO after the reduction with H
catalyst was calculated based on the Arrhenius-type plots (equation 4) [54] (Fig. 16). The
apparent activation energy for the Ni/RM and Ni/SiO -Al catalysts were 97.5 KJ/mol and
0.3 KJ/mol, respectively. These values were within the range of various values reported in
2
. Apparent activation energy for each
2
2
O
3
9
2
0

literature [44, 46, 66]. Therefore the global kinetic model over Ni/RM and Ni/SiO
catalysts had the forms of equations (5) and (6) respectively:
2
-Al
2
O
3
퐸
ln(푘) = ln(푘 ) −
(4)
ꢌ
푅푇
ꢍ9
7.ꢎ
푟퐺 = 56.26 푒( ꢏꢐ ). 퐶 ꢑ (5)
−
−
퐺
ꢍ9ꢒ.3
푟퐺 = 51.42 푒( ꢏꢐ ). 퐶 ꢑ (6)
퐺
The development of a more detailed kinetic model is out of the scope of this article.
.6. Catalyst deactivation and regeneration
As discussed in section 3.1, three mechanisms contributed to catalyst deactivation; coke
3
formation, oxidation of Ni, and formation of iron nickel oxide. In order to evaluate the catalyst
life, the recovered catalysts after HDO experiments were washed using a solution of 1:1 (wt.
ratio) toluene/methanol and consecutively used in further HDO tests without any catalyst
regeneration. The selectivity of HDO products changed as the catalyst deactivated. A summary
of the results of these experiments is shown in Table 5. Metal loading, and pore structure of
support have been proposed to affect coke deposition [9, 16]. After four consecutive runs using
Ni/RM, the catalyst completely deactivated due to coking and possible oxidation of active Ni
2
sites both of which can contribute to the reduction of BET specific surface area from 79.3 m /g
2
to 59.2 m /g. After the fifth run the catalyst did not indicate any activity and the guaiacol did not
undergo any HDO reactions. There was more coke built up on the catalyst surface that further
2
reduced the BET specific surface area to 56.4 m /g. The Ni/SiO
2
-Al
2
O
3
completely deactivated
2
after five consecutive runs and its BET specific surface area reduced from 118.4 m /g to 77.9
2
m /g. After the sixth run the catalyst did not show any activity for HDO reactions however its
2
1

2
BET specific surface area further reduced to 72.2 m /g. The reduction in BET specific surface
area suggested that coke was formed in the pores of the catalyst.
The spent catalysts were placed in muffle furnace to burn off the deposited coke at 400 °C for 4
h followed by reduction as explained in section 3.1. The regenerated/activated catalyst was used
in HDO experiments. A summary of these HDO test results are shown in Table 6. Guaiacol
conversion and HDO products selectivity over Ni/RM were similar to that of fresh catalyst
(
Table 5). These result indicated that the Ni/RM catalyst regained its activity after the
regeneration/activation process. However, the regeneration of the commercial Ni/SiO
was not possible following the same procedure and the catalyst did not show HDO activity after
regeneration/reduction. The BET specific surface area of the regenerated Ni/SiO -Al catalyst
2
-Al
2
O
3
2
2
O
3
2
2
was 88.8 m /g which was much lower than the fresh catalyst (118.4 m /g) suggesting that some
Ni particles probably formed stable compounds with the support phase during regeneration.
Additionally, when exposed to air, the reduced Ni/SiO
2
-Al
2
3
O catalyst underwent spontaneous
oxidation that could be due to oxidation of free Ni particles that were not interacting with the
support phase.
4
. Conclusion
Red mud, which is a waste material from the aluminum industry, was used as support material
for preparation of nickel hydrogenation/hydrodeoxygenation catalyst. Hydrodeoxygenation of
guaiacol as lignin model compound was investigated over red mud-supported nickel catalyst.
For comparison, commercial Ni/SiO
2
-Al
2
O
3
was tested in HDO process under similar conditions.
and NiMo/Al ), this catalyst did not require
Unlike traditional HDO catalysts (CoMo/Al
2
O
3
2
O
3
2
2

any pre-treatments such as sulfidation and it can be used as received. Also this catalyst is much
cheaper than typical noble metal catalysts that are widely used for HDO reactions. The Van-
Krevelen diagrams of HDO products at different reaction temperatures and hydrogen pressures
showed that increasing the reaction temperature and hydrogen pressure improved HDO
reactions. Complete deoxygenation of guaiacol was achieved at a reaction temperature of 400 °C
when the initial hydrogen pressure was 6.21 MPa (900 psi). Ni/RM showed higher activity for
deoxygenation while hydrogenation of the aromatic ring was more pronounced for the
commercial Ni/SiO
2
-Al
2
O
3
. Consequently, the Ni/RM consumed less hydrogen than Ni/SiO
2
-
Al . Hydrocracking and coke formation were significantly lower in the case of Ni/RM catalyst
2
O
3
which is desirable in HDO process. The kinetics of guaiacol HDO followed second order model
for both catalysts while higher rate constant and lower activation energy were determined for
Ni/SiO
2
-Al
2
O
3
catalyst compared to Ni/RM. The increased catalytic activity of Ni/SiO
2
-Al
2
O
3
was ascribed to higher Ni loading and higher BET specific surface area of the catalyst.
Coke formation was the major mechanism of catalyst deactivation, although oxidation of active
Ni sites and formation of iron nickel oxide could also contribute to catalyst deactivation. After
regeneration by burning off the coke and reducing with hydrogen, the activity of the Ni/RM
catalyst was completely restored in contrast to the commercial catalyst that was not regenerable.
This study showed that Red Mud can serve as a nickel catalyst support for HDO process.
Acknowledgement
The authors would like to acknowledge funding provided by the National Science Foundation
under contract # 1445735 and Utah Science Technology and Research (USTAR) to support this
research. We gratefully acknowledge the Microscopy Core Facility at Utah State University.
2
3

References
[
[
[
1]
2]
3]
K. Li, R. Wang, J. Chen, Energy Fuels 25 (2011) 854–863.
H.Y. Zhao, D. Li, P. Bui, S.T. Oyama, Appl. Catal. A Gen. 391 (2011) 305–310.
S. Ted Oyama, T. Onkawa, A. Takagaki, R. Kikuchi, S. Hosokai, Y. Suzuki, K.K. Bando,
Top. Catal. 58 (2015) 201–210.
[
[
[
[
[
[
4]
5]
6]
7]
8]
9]
Z. He, X. Wang, Catal. Sustain. Energy 1 (2012) 28–52.
J. Wildschut, I. Melián-Cabrera, H.J. Heeres, Appl. Catal. B Environ. 99 (2010) 298–306.
S. Raza, Y. Uemura, S. Bt, J. Anal. Appl. Pyrolysis 106 (2014) 57–62.
S. Czernik, A. V. Bridgwater, Energy Fuels 18 (2004) 590–598.
H. Wang, J. Male, Y. Wang, ACS Catal. 3 (2013) 1047-1070.
P.M. Mortensen, J. Grunwaldt, P.A. Jensen, K.G. Knudsen, A.D. Jensen, Appl. Catal. A
Gen. 407 (2011) 1–19.
10] P. Weerachanchai, M. Horio, C. Tangsathitkulchai, Bioresour. Technol. 100 (2009)
419–1427.
11] C.S. Lira, F.M. Berruti, P. Palmisano, F. Berruti, C. Briens, A.A.B. Pécora, J. Anal. Appl.
Pyrolysis 99 (2013) 23–31.
[
1
[
[
[
(
[
[
12] H. Jahromi, F.A. Agblevor, Energy 141 (2017) 2186–2195.
13] A. Gutierrez, R.K. Kaila, M.L. Honkela, R. Slioor, A.O.I. Krause, Catal. Today 147
2009) 239–246.
14] M.Á. González-Borja, D.E. Resasco, Energy Fuels 25 (2011) 4155–4162.
15] C.A. Fisk, T. Morgan, Y. Ji, M. Crocker, C. Crofcheck, S.A. Lewis, Appl. Catal. A Gen.
3
58 (2009) 150–156.
2
4

[
[
16] X. Zhu, L.L. Lobban, R.G. Mallinson, D.E. Resasco, J. Catal. 281 (2011) 21–29.
17] V.N. Bui, D. Laurenti, P. Afanasiev, C. Geantet, Appl. Catal. B Environ. 101 (2011)
2
39–245.
18] V.N. Bui, D. Laurenti, P. Delichère, C. Geantet, Applied Catal. B, Environ. 101 (2011)
46–255.
[
2
[
[
(
[
19] Y. Romero, F. Richard, S. Brunet, Appl. Catal. B Environ. 98 (2010) 213–223.
20] C. Sepulveda, R. Garcia, N. Escalona, D. Laurenti, L. Massin, M. Vrinat, Catal. Lett. 141
2011) 987–995.
21] D. Shen, C. Cheng, N. Liu, R. Xiao, Production of Biofuels and Chemicals from Lignin,
Springer, 2016, pp. 289-320.
22] A.L. Jongerius, Catalytic Conversion of Lignin for the Production of Aromatics, Gilderprint
Drukkerijen, 2013 pp .8-15.
[
[
[
23] E. Furimsky, Appl. Catal. A Gen. 199 (2000) 147–190.
24] E. Furimsky, Ind. Eng. Chem. Prod. Res. Dev. 22 (1983) 34–38.
[
25] Y. Yoshimura, T. Sato, H. Shimada, N. Matsubayashi, A. Nishijima, Appl. Catal. 73 (1991)
5
5–63.
[
[
[
[
[
[
26] A. Agrawal, K.K. Sahu, B.D. Pandey, Res. Con. Rec. 42 (2004) 99–120.
27] S. Sushil, V.S. Batra, Appl. Catal. B. Environ. 81 (2008) 64–77.
28] B.K. Yathavan, F.A. Agblevor, Energy Fuels 27 (2013) 6858–6865.
29] X. Lim, A. Sanna, J.M. Andresen, Fuel 119 (2014) 259–265.
30] B. Klopries, W. Hodek, F. Bandermann, Fuel 69 (1990) 448-455.
31] K.C. Pratt, V. Christoverson, Fuel 61 (1982) 460-462.
2
5

[
32] A. Eamsiri, W.R. Jackson, K.C. Pratt, V. Christoverson, M. Marshall, Fuel 71 (1992)
49-453.
4
[
[
[
[
[
[
33] J.J. Llano, R. Rosal, H. Sastre, F.V. Diez, Fuel 73 (1994) 688-694.
34] S. Ordóñez, H. Sastre, F. V. Díez, Appl. Catal. B Environ. 29 (2001) 263–273.
35] J. Alvarez, Appl. Catal. A. Gen. 128 (1995) 259–273.
36] S. Ordóez, H. Sastre, F. V. Díez, J. Hazard. Mater. 81 (2001) 103–114.
37] M. Martino, R. Rosal, H. Sastre, F. V Dõ, Appl. Catal. B. Environ. 20 (1999) 301-307.
38] L.C.A. De Oliveira, A.P. Heitmann, L.D. Almeida, S. Herman, A.A. Mansur, C.S. De
Castro, U.F. De Alfenas, J. Environ. Sci. 57 (2015) 3–5.
[
[
[
39] O.D. Mante, F.A. Agblevor, Green Chem. 16 (2014) 3364-3377.
40] J.R. Paredes, S. Ordóñez, A. Vega, Appl. Catal. B. Environ. 47 (2004) 37–45.
41] S. Djerad, M. Crocoll, S. Kureti, L. Tifouti, W. Weisweiler, Catal. Today 113 (2006)
2
08–214.
[
[
[
[
42] S.K. Wu, P.C. Lai, Y.C. Lin, Catal. Lett. 144 (2014) 878–889.
43] J. Filley, C. Roth, J. Molecular Catal. A 139 (1999) 245–252.
44] A. Aqsha, L. Katta, N. Mahinpey, Catal. Lett. 145 (2015) 1351–1363.
45] N.T.T. Tran, Y. Uemura, S. Chowdhury, A. Ramli, Appl. Catal. A Gen. 512 (2016) 93–
1
00.
[
[
[
46] Q. Lai, C. Zhang, J.H. Holles, Appl. Catal. A Gen. 528 (2016) 1–13.
47] L. Liu, Y. Liu, X. Gao, R. Zhang, Y. Zhai, J. Fuel Chem. Technol. 45 (2017) 932–938.
48] T. He, X. Liu, Y. Ge, D. Han, J. Li, Z. Wang, J. Wu, Catal. Commun. 102 (2017) 127–
1
30.
2
6

[
49] I.D. Mora-Vergara, L. Hernández Moscoso, E.M. Gaigneaux, S.A. Giraldo, V.G.
Baldovino-Medrano, Catal. Today 302 (2017) 125–135.
[
[
(
[
50] Z. Xinghua, W. Tiejun, M. Longlong, W. Chuangzhi, Fuel 89 (2010) 2697–2702.
51] X. Zhang, L. Chen, W. Kong, T. Wang, Q. Zhang, J. Long, Y. Xu, L. Ma, Energy 84
2015) 83–90.
52] R. Gavlak, R. Horneck, R.O. Miller, J. Kotuby-Amacher, Soil, plant and water reference
methods for the western region. Western Coordinating Committee on Nutrient Management,
000, pp. 129–134.
2
[
[
53] X. Zhang, T. Wang, L. Ma, Q. Zhang, T. Jiang, Bioresour. Technol. 127 (2013) 306–311.
54] J. Wiley, K. Hepburn, O. Levenspiel, Chemical Reaction Engineering, John Wiley and
Sons, 1964, pp. 13-75.
55] I. Coronado, M. Stekrova, L. García Moreno, M. Reinikainen, P. Simell, R. Karinen, J.
Lehtonen, Biomass Bioenergy 106 (2017) 29–37.
[
[
[
56] S. Chenna, R. Banerjee, P.A. Crozier, Chem. Cat. Chem., 3 (2011) 1051–1059.
57] Q. Jeangros, T.W. Hansen, J.B. Wagner, C.D. Damsgaard, R.E. Dunin-Borkowski, C.
Hébert, J. Van Herle, A. Hessler-Wyser, J. Mater. Sci. 48 (2013) 2893–2907.
58] J. Pu, K. Nishikado, N. Wang, T.T. Nguyen, T. Maki, E.W. Qian, Appl. Catal. B Environ.
24 (2018) 69–79.
59] C. Boscagli, C. Yang, A. Welle, W. Wang, S. Behrens, K. Raffelt, J.D. Grunwaldt, Appl.
Catal. A Gen. 544 (2017) 161–172.
[
2
[
[
(
[
60] S. Danaci, L. Protasova, J. Lefevere, L. Bedel, R. Guilet, P. Marty, Catal. Today 273
2016) 234–243.
61] N. Joshi, A. Lawal, Chem. Eng. Sci. 74 (2012) 1–8.
2
7

[
[
62] H. Wan, R. V. Chaudhari, B. Subramaniam, Top. Catal. 55 (2012) 129–139.
63] S. Asphaug. Catalytic hydrodeoxygenation of bio-oils with supported MoP-catalysts.
Norwegian University of Science and Technology, 2013, pp. 50-56.
[
[
64] T.-R. Viljava, R.S. Komulainen, A.O.I. Krause, Catal. Today 60 (2000) 83–92.
65] Z. Cai, F. Wang, X. Zhang, R. Ahishakiye, Y. Xie, Y. Shen, Mol. Catal. 441 (2017) 28–
3
4.
66] M.D. Skoglund, C.L. Jackson, K.J. Mckim, H.J. Olson, S. Sabirzyanov, J.H. Holles,
Applied Catal. A, Gen. 467 (2013) 355–362.
[
Fig. 1: XRD pattern of fresh Ni/RM catalyst.
2
8

Fig. 2: XRD pattern of used Ni/RM catalyst.
2
9