Hydrodeoxygenation of methyl palmitate over sulfided Mo/Al 2O3, CoMo/Al2O3 and NiMo/Al 2O3 catalysts

Article abstract of DOI:10.1039/c3ra46164e

The catalytic properties of sulfided Mo/Al₂O₃, CoMo/Al₂O₃ and NiMo/Al₂O₃ catalysts in the hydrodeoxygenation of methyl palmitate as a model compound for triglyceride feedstock were studied at 300°C and 3.5 MPa in the batch reactor using n-tetradecane, m-xylene and hydrotreated straight-run gas oil (HT-SRGO). The comparison of catalyst's performance in n-tetradecane allowed us to see that the sulfided Mo/Al₂O₃, CoMo/Al ₂O₃ and NiMo/Al₂O₃ catalysts revealed the same rate of the methyl palmitate conversion but the rate of the intermediate oxygenates conversion decreased in order: CoMoS/Al ₂O₃ > NiMoS/Al₂O₃ > MoS ₂/Al₂O₃. A mixture of linear saturated and unsaturated C₁₅ and C₁₆ hydrocarbons was produced when the oxygenates were fully consumed. The main products obtained over the Mo/Al ₂O₃ and CoMo/Al₂O₃ catalysts were C₁₆ hydrocarbons (C₁₆/C₁₅-16.1 and 2.79, respectively); however, C₁₅ hydrocarbons were preferentially formed over the NiMo/Al₂O₃ catalyst (C₁₆/C ₁₅-0.65), highlighting the different contributions of the hydrodeoxygenation (HDO) and decarboxylation/decarbonylation (DeCO_x) pathways during the hydroconversion of methyl palmitate over these catalysts. Investigating the solvent's influence on the activity of the CoMo/Al ₂O₃ and NiMo/Al₂O₃ catalysts in the methyl palmitate HDO revealed that the reaction rate was decreased in the following order: n-tetradecane > HT-SRGO > m-xylene. The aromatic compounds did not retard the methyl palmitate transformation, but inhibited the conversion of the intermediate oxygenates. Decreased C₁₆/C ₁₅ ratios were observed over both catalysts when m-xylene was used as the reaction medium instead of n-tetradecane.

Full text of DOI:10.1039/c3ra46164e

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Hydrodeoxygenation of methyl palmitate over
sulfided Mo/Al2O3, CoMo/Al2O3 and NiMo/Al2O3
catalysts
Cite this: RSC Adv., 2014, 4, 2242
ab
Evgenia N. Vlasova,a Alexey L. Nuzhdin,a Evgeny Yu. Gerasimovab
*
Irina V. Deliy,
and Galina A. Bukhtiyarovaa
The catalytic properties of sulfided Mo/Al2O3, CoMo/Al2O3 and NiMo/Al2O3 catalysts in the
hydrodeoxygenation of methyl palmitate as a model compound for triglyceride feedstock were studied
at 300 ꢀC and 3.5 MPa in the batch reactor using n-tetradecane, m-xylene and hydrotreated straight-run
gas oil (HT-SRGO). The comparison of catalyst's performance in n-tetradecane allowed us to see that
the sulfided Mo/Al2O3, CoMo/Al2O3 and NiMo/Al2O3 catalysts revealed the same rate of the methyl
palmitate conversion but the rate of the intermediate oxygenates conversion decreased in order:
CoMoS/Al2O3 > NiMoS/Al2O3 > MoS2/Al2O3. A mixture of linear saturated and unsaturated C15 and C16
hydrocarbons was produced when the oxygenates were fully consumed. The main products obtained
over the Mo/Al2O3 and CoMo/Al2O3 catalysts were C16 hydrocarbons (C16/C15 – 16.1 and 2.79,
respectively); however, C15 hydrocarbons were preferentially formed over the NiMo/Al2O3 catalyst (C16
/
C15 0.65), highlighting the different contributions of the hydrodeoxygenation (HDO) and
–
decarboxylation/decarbonylation (DeCOx) pathways during the hydroconversion of methyl palmitate
over these catalysts. Investigating the solvent's influence on the activity of the CoMo/Al2O3 and NiMo/
Al2O3 catalysts in the methyl palmitate HDO revealed that the reaction rate was decreased in the
following order: n-tetradecane > HT-SRGO > m-xylene. The aromatic compounds did not retard the
methyl palmitate transformation, but inhibited the conversion of the intermediate oxygenates. Decreased
C16/C15 ratios were observed over both catalysts when m-xylene was used as the reaction medium
instead of n-tetradecane.
Received 28th October 2013
Accepted 26th November 2013
DOI: 10.1039/c3ra46164e
www.rsc.org/advances
A renewable diesel fuel concept based on the complete
hydrodeoxygenation of triglyceride-based feedstock to produce
Introduction
linear hydrocarbons has become more popular over the last
decade.6–10 The hydroprocessing of common vegetable oils
(rapeseed, sunower, soybean and palm oil) leads to an oxygen-
free mixture of C15–C18 paraffins that are known in the literature
as second-generation biodiesel, renewable (green) diesel or bio-
gasoil. The obtained product is fully compatible with fossil-
derived diesel fuel, has a high cetane number and good
stability; engine fuel economy was improved when a blend of
bio-gasoil and petroleum diesel fuel was used.4,5
Hydroprocessing of triglyceride-based feedstocks to
produce hydrocarbons with the required technical and envi-
ronmental fuel standards may be performed in stand-alone
units or via the co-hydrotreatment of renewable oils and fats
with petroleum-derived middle distillates in the existing HDS
units.8–12 In both cases conventional hydrodesulfurisation
catalysts (CoMo or NiMo) may be used to convert triglycerides
into linear alkanes.9–22 The reaction pathways for the hydro-
processing of triglyceride molecules involve the olenic
double bonds saturation, decomposition of the triglycerides
and elimination of oxygen through different competing routes:
The decrease in available fossil fuel resources has caused
ascending interest in the development of new catalytic tech-
nologies that produce liquid transportation fuels from renew-
able organic materials.1–5 Biomass feedstocks may be arranged
into three categories: carbohydrates, lignin and fats/oils
composed of triglycerides and free fatty acids. The triglyceride-
based feedstock is an efficient source for the production of
motor fuels because of their relatively high energy density and
low oxygen content. The most popular technology used to
convert these raw materials into diesel fuel is a trans-
esterication with methanol that produces fatty acid methyl
esters (FAME) called biodiesel.2 However; FAME has many
disadvantages, including poor compatibility with fossil-derived
fuels, existing engines and distribution infrastructure, as well as
high oxygen content, low cetane numbers, high acidity and poor
stability.5
aBoreskov Institute of Catalysis SB RAS, Pr. Lavrentieva 5, Novosibirsk, 630090,
Russia. E-mail: delij@catalysis.ru; Fax: +7 3833308056; Tel: +7 3833269410
bNovosibirsk State University, Pirogova Street 2, Novosibirsk, 630090, Russia
2242 | RSC Adv., 2014, 4, 2242–2250
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
RSC Advances
hydrodeoxygenation or decarboxylation/decarbonylation.9–13 presuldation procedure in the gas phase at 400 ꢀC using a gas
In the HDO pathway, alkanes with the same numbers of mixture 5 vol.% H2S/95 vol.% H2.
carbon atoms are produced via the dehydration/hydrogena-
Methyl palmitate (grade $ 97%), n-tetradecane (grade $
tion reactions, while the DeCOx reaction produce alkanes with 98%) and m-xylene (grade $ 99.5%) were supplied by Sigma-
the shorter carbon chains and COx molecules. The major Aldrich (Japan), Vekton (Russia) and Reakhim (Russia),
challenges of developing economical and energy-efficient respectively. HT-SRGO that contained 8 ppm S, 7 ppm N and
processes to produce transportation fuels sustainably from 21.1 wt% mono-, 1.89 wt% di-, 0.81 wt% polyaromatics,
biomass feedstock include the elimination of oxygen, the respectively, was obtained by hydrotreating of straight run gas
retention of carbon atoms in the products and minimising oil (SRGO) in the pilot plant using a previously developed
hydrogen consumption.
CoMoS/Al2O3 catalyst.36 Dimethyldisulde (DMDS, 99.0 wt%)
To understand the fundamental reaction network or was purchased from Acros Organics, France.
compare the relative activities of HDO catalysts differing by the
supports, chemical compositions or preparation methods the
model mixtures of oxygenates in different solvents have been
Experimental setup and procedure
The HDO reaction of methyl palmitate was carried out in a
used.23–33 Long chain fatty acids and their alkyl esters were
usually used to model triglyceride feedstock The reaction
mechanism was studied using different reagents and solvents:
methyl and ethyl heptanoate in n-dodecane,25,26 methyl hepta-
noate and hexanoate in m-xylene27,28 and methyl laurate in
n-heptane.29 The mixture of ethyldecanoate in p-xylene was used
to investigate the inuence of the impregnation order and
active component loading on the activity of carbon-supported
CoMo catalysts;30–32 the hydrodeoxygenation of carbonyl,
carboxyl and guaiacol-type groups over sulded catalysts were
comparatively studied using n-hexadecane as a solvent.34,35 It
was stated early, that the conventional CoMo and NiMo
hydrotreating catalysts provided both reaction pathways (HDO
and DeCOx) simultaneously.13,33 However, little attention was
paid to the effect of the reaction medium's composition on the
activity and selectivity of the aliphatic methyl esters hydro-
conversion over the sulded CoMo/Al2O3 and NiMo/Al2O3
catalysts.
300 ml batch reactor (Autoclave Engineers type MagneDrive,
USA) equipped with a system for sampling of liquids during the
reaction. The experiments were carried out at 3.5 MPa, 340 C,
ꢀ
and 1000 rpm for 6 h. 0.3 g of catalyst and 100 ml of reaction
mixture were used for catalytic activity evaluation in each run.
The liquid samples were withdrawn during reaction. The reac-
tion mixtures were prepared by dissolving methyl palmitate
(6 wt%) in n-tetradecane, m-xylene or hydrotreated straight-run
gas oil with n-dodecane (grade $ 99.0%; from Acros Organics,
Germany) as the internal standard. Carbon balance in the
experiments with n-tetradecane and m-xylene exceeded 96%
and 92%, accordingly. Dimethyldisulde was added in the
reaction mixture in amount equivalent to 0.6 wt% of sulfur to
generate a constant partial pressure of H2S and maintain the
sulded state of the catalysts. The absence of external and/or
internal diffusion limitations was conrmed by additional
experiments. The time dependence of the total oxygen uptake
may be described by a polynomial:
In the current work we compared the reactivity and selec-
tivity of methyl palmitate (MP) transformations in n-tetradecane
over sulded Mo/Al2O3, CoMo/Al2O3 and NiMo/Al2O3 catalysts.
In addition, the solvent effects on the HDO of methyl palmitate
over the promoted CoMo/Al2O3 and NiMo/Al2O3 catalysts were
investigated using n-tetradecane, m-xylene and hydrotreated
straight-run gas oil (HT-SRGO) as the reaction medium.
C
¼ 1 þ at þ bt2 þ ct3;
(1)
C0
where C0 and C were the initial and current oxygen content in
the reaction mixture; a, b, c, n – the coefficients of polinominal
function. The initial oxygen uptake rates (WO0) were calculated
by differentiating the polynoms at t ¼ 0.
The kinetic curves for the methyl palmitate conversion were
described assuming a pseudo-rst order of reaction by the
following logarithmic function:
Experimental
Catalysts and chemicals
ꢀ
ꢁ
0
CMP
CMP
Commercial CoMo/Al2O3 and NiMo/Al2O3 catalysts were used in
the experiments. The non-promoted MoO3/Al2O3 catalyst was
prepared by impregnating of g-Al2O3 granules (BET surface area
– 205 m2 gꢁ1, pore volume – 0.70 cm3 gꢁ1, average pore diameter
– 13.6 nm; from JSK “Promkataliz”, Ryazan, Russia) with an
aqueous solution containing the required amounts of MoO3
(grade $ 99.0%; from Vekton, Russia), H3PO4 (grade $ 85%;
from Vekton, Russia) and citric acid (grade $ 99.8%; from
Vekton, Russia). The concentration of molybdenum in the
ln
¼ k*$t;
(2)
where CMP0 and CMP were the initial and current concentrations
of methyl palmitate in the reaction mixture, while k* was the
apparent reaction rate constant for the methyl palmitate
transformation.
Products analysis
solution was equal to 2.8 mole Lꢁ1, the molar ratio of P/Mo and The total oxygen content in the reaction mixture was deter-
citric acid/Mo was adjusted to 0.4 and 0.7, correspondingly. The mined with a CHNSO elemental analyzer Vario EL Cube (Ele-
sample was dried at 110 ꢀC for 4 h and heated under owing N2 mentar Analysensysteme GmbH, Germany). The reaction
for 4 h at 200 ꢀC. Before the reactions, the catalysts were crushed products were identied using a GC/MS technique (Agilent
and sieved to 0.14–0.25 mm and activated by the standard Technologies 7000 GC/MS Triple QQQ GC System 7890A) with a
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 2242–2250 | 2243

View Article Online
RSC Advances
Paper
Table 1 Conditions of gas chromatography analysis and parameters for atomic emission detector
Conditions for gas chromatography
Injection port temperature
AED transfer line/cavity temperature
Oven temperature program
Carrier gas
320 ꢀ
320 ꢀ
C
C
40–260 ꢀC at 5 ꢀC minꢁ1, 260–280 ꢀC at 10 ꢀC minꢁ1
He (99.9999%)
50 : 1 (carbon, sulfur)
10 : 1 (oxygen)
1 ml
Split ratio
Injection volume
Parameters for atomic emission detector
Reagent gases
Carbon (179 nm), sulfur (181 nm)
Oxygen: 7.6 psig
Hydrogen: 16.8 psig
Oxygen (171 nm)
Hydrogen: 40 psig
10% methane in nitrogen: 25 psig
Table 2 Textural characteristics and chemical composition of the
catalysts in their oxide form
HP-5 MS quartz capillary column (30 m ꢂ 0.25 mm). The liquid
samples were analyzed with a gas chromatography system
(Agilent 6890N) equipped with HP-1 MS column (60 m ꢂ 0.32
mm ꢂ 0.25 mm) and an atomic emission detector (GC-AED). The
emission line for C179 was used to follow the concentrations of
the reactants and products. To observe the concentrations of
the oxygen-containing compounds in the gas oil using AED, the
emission line for oxygen atoms at 171 nm was monitored. The
conditions for the gas chromatographic analysis and atomic
emission detection are listed in Table 1.
Mo/Al2O3
CoMo/Al2O3
NiMo/Al2O3
SBET (m2 gꢁ1
Dpore (nm)
)
183
9.5
0.45
—
—
14.1
206
8.6
0.44
2.90
—
175
10.2
0.45
—
2.65
11.2
Pore volume (cm3 gꢁ1
Co (wt%)
)
Ni (wt%)
Mo (wt%)
11.0
Results and discussion
Catalysts characterization
Catalyst characterization
The textural properties of the catalysts were determined using The chemical composition and textural properties of the cata-
nitrogen physisorption with an ASAP 2400 instrument (USA). lysts in their oxide state are listed in Table 2. According to the
The elemental analysis of the catalysts was performed using ICP-AES data, the CoMo/Al2O3 and NiMo/Al2O3 catalysts con-
inductively coupled plasma atomic emission spectroscopy (ICP- tained comparable amounts of Mo and promoter (Co or Ni). The
AES) on Optima 4300 DV (Perkin Elmer, France). The concen- textural parameters of the oxide catalysts were obtained from
ꢀ
trations of C and S in the sulded catalysts aer the reactions the N2 adsorption–desorption isotherms at ꢁ196 C. For each
were determined with a CHNSO analyzer Vario EL Cube (Ele- catalyst, the specic surface area (SBET) and pore volume were
mentar Analysensysteme GmbH, Germany).
within 175–206 m2 gꢁ1 and 0.44–0.45 cm3 gꢁ1, respectively.
The samples were studied using high resolution trans-
At present it is generally accepted that the active phase of
mission electron microscopy (HRTEM) with a JEM-2010 trans- hydrotreating catalysts consists of MoS2 nanoparticles with
mission electron microscope (JEOL, Japan) with accelerating cobalt or nickel atoms on the MoS2 edges faces, giving the so-
voltage of 200 kV and resolution of 0.14 nm. Analysis of the local called CoMoS or NiMoS phase.37 The hexagonal layered struc-
elemental composition was carried out with an energy-disper- ture of MoS2 nanoparticles can be visualized by transmission
sive EDX spectrometer equipped with a Si(Li) detector (energy electron microscopy (TEM) and this technique has become an
resolution – 130 eV). The catalyst's samples were applied to important characterization tool for the sulde catalysts.
copper gauze by dispersing an alcoholic suspension of the HRTEM micrographs depicting the typical surface fragments of
sample with an ultrasonic disperser. To obtain statistical the sulded Mo/Al2O3, CoMo/Al2O3 and NiMo/Al2O3 catalysts
information, the structural parameters of ca. 500 particles were are shown in Fig. 1(A)–(C), respectively. There are the black
measured.
thread-like fringes corresponding to the MoS2 slabs with
The crystalline structure of the catalysts was detected by different stack height and length evenly distributed on the
powder X-ray diffraction. XRD patterns were obtained with an surface of Al2O3. The formation of nickel sulde or cobalt
XTRA diffractometer (Switzerland) using CuKa radiation sulde crystallites was not observed. The average slab length
(wavelength l ¼ 1.5418 E). The measurements were carried out and stacking number of the sulde nanoparticles on the cata-
in a range of 2q from 5 to 70ꢀ with scanning step of 0.05ꢀ and an lyst's surface aer the suldation procedure are presented in
accumulation time of 3 s per point.
Table 3. The sizes of the sulde particles in the CoMo/Al2O3 and
2244 | RSC Adv., 2014, 4, 2242–2250
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
RSC Advances
Table 3 HRTEM parameters of the sulfided catalysts
Mo/Al2O3
CoMo/Al2O3
NiMo/Al2O3
Stacking number
Average slab length (nm)
1.9
3.6
1.7
2.9
2.0
3.2
XRD data didn't show the formation of cobalt or nickel
sulde phase in the CoMo/Al2O3 and NiMo/Al2O3 sulded
catalyst, Fig. 2. Aside the features of g-Al2O3, the sulded Mo/
Al2O3 (line B in Fig. 2), CoMo/Al2O3 (line D in Fig. 2) and NiMo/
Al2O3 (line C in Fig. 2) catalysts exhibited two broad poorly
resolved diffraction peaks with 2q values at around 33 and 59ꢀ
characteristic for the 100 and 110 planes of the hexagonal phase
of MoS2 [PDF 37-1492]. The diffraction peak at 2q about 14.4ꢀ,
˚
characteristic for the 002 plane (d spacing 6.1 A) was poorly
observed probably due to low stacking conrmed by TEM. The
features observed on XRD patterns of sulde catalysts pointed
out to the formation of low crystalline highly dispersed MoS2
phases.38,39
The sulfur content in the pre-sulded CoMo/Al2O3 and
NiMo/Al2O3 catalysts was 8.4 ꢃ 0.7 wt% before the experiments,
and the sulfur content was not signicantly changed during the
experiments. The carbon content of the catalysts increased up
to approximately 2–3 wt% during the experiments using n-tet-
radecane and up to 5–7 wt% during the experiments using m-
xylene or HT-SRGO.
Hydroconversion of methyl palmitate in n-tetradecane
The comparative study of non-promoted Mo/Al2O3 and
promoted Co(Ni)Mo/Al2O3 sulded catalysts in the methyl
palmitate hydroconversion was carried out in the batch reactor
using n-tetradecane as a solvent and detailed analysis of the
produced intermediates and nal products of methyl palmitate
HDO was performed. The concentration proles of the major
products formed in the hydroconversion of methyl palmitate
over the Mo/Al2O3, CoMo/Al2O3 and NiMo/Al2O3 catalysts are
presented in Fig. 3(A)–(C), respectively. At low and intermediate
Fig.
1 HRTEM micrographs of sulfide catalysts: MoS2/Al2O3 (A),
CoMoS/Al2O3 (B), NiMoS/Al2O3 (C).
NiMo/Al2O3 catalysts are similar, but Mo/Al2O3 sample contains
the particles with the larger average slab length. The stacking
number was lower for CoMo relative to the other two catalysts.
According to the TEM images, the parameters of the nano-
dispersed sulde phase were not changed signicantly aer
testing in the methyl palmitate conversion, regardless of the
solvent used.
Fig. 2 X-ray diffraction patterns of: (A) g-Al2O3, (B) MoS2/g-Al2O3, (C)
NiMoS/g-Al2O3, and (D) CoMoS/g-Al2O3 catalysts.
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 2242–2250 | 2245

View Article Online
RSC Advances
Paper
increasing of methyl palmitate conversion the ester-type
compounds disappeared quickly. The palmitic acid and 1-hex-
adecanol contents decreased gradually; only linear saturated
and unsaturated hydrocarbons with 15 or 16 carbon atoms were
detected aer 6 h of reaction. There wasn't detected any sulfur-
containing organic compounds by means of GS-AED in the
reaction mixture at the end of experiments. The products we
obtained are consistent with the products of methyl heptanoate
HDO over supported sulde catalysts.25–27 To estimate the
contributions of the HDO and DeCOx pathways, alkanes and
alkenes with the same number of carbon atoms were summed
together.
It was observed that all liquid products had 15 or 16 carbon
atoms, and the most abundant of these were palmitic acid,
1-hexadecanol, n-C15 and n-C16 alkanes as well as the corre-
sponding alkenes. This product distribution veried the exis-
tence of the two routes leading to n-C15 and n-C16 hydrocarbons.
The products associated with the decarboxylation route were
pentadecane and pentadecene, and no oxygenate intermediates
with 15 carbon atoms were observed. However, the HDO routes
leading to hexadecane and hexadecene appeared to proceed via
a more complicated mechanism because several oxygenated
intermediates with 16 carbon atoms were observed.
The different pathways could be considered for the methyl
palmitate conversion taking into account the obtained results
and literature data describing the peculiarities of hydro-
conversion of aliphatic acids and their esters over sulde cata-
lysts.25–29,41 The overall hypothetical scheme for the methyl
palmitate transformation that was based on the earlier
proposed reaction pathways for HDO of aliphatic esters25–29 is
shown in Fig. 4. First of all palmitic acid that is the predomi-
nated oxygenated products at low conversion level of palmitic
ester could be produced via methyl palmitate hydrolysis (path
1), which is the reversed route for the acid-catalysed esterica-
tion reaction. It was found early that the formation of carboxylic
acids and alcohols from aliphatic methyl esters could be cata-
lyzed by the Lewis acidic sites of the alumina support.40 The
Fig. 3 Concentration profile of the reaction products for methyl
palmitate HDO over MoS2/Al2O3 (A), CoMoS/Al2O3 (B) and NiMoS/
Al2O3 (C) catalysts at the temperature 300 ꢀC and hydrogen pressure
35 bars in n-tetradecane.
conversion, the oxygen- and sulfur-containing compounds were
identied along with the C15 and C16 hydrocarbons. The inter-
mediates of the methyl palmitate conversion over sulded
Co(Ni)Mo/Al2O3 catalysts included palmitic (hexadecanoic)
acid, 1-hexadecanol, 1-hexadecanthiol and small amount of
hexadecanal; no cyclic or aromatic products were observed
during the reaction. Small amounts of fatty ester (hexadecyl
hexadecanoate) was appeared in the initial period of experiment
Fig. 4 Possible reaction pathways for the HDO reactions of methyl
by the reaction of fatty acids with fatty alcohols, but with the palmitate over sulfided (Ni,Co)Mo/g-Al2O3 catalysts.
2246 | RSC Adv., 2014, 4, 2242–2250
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
RSC Advances
palmitic acid conversion via CO/CO2 removal over sulded NiMo
and CoMo catalysts.
Hexadecanal could be hydrogenated stepwise to form
hexadecanol and hexadecane, possibly by dehydration of
hexadecanol to yield hexadecene. Hexadecanol is not displayed
in Fig. 3(B) and (C) because its measured concentration was
negligible over both promoted sulded NiMo and CoMo cata-
lysts. The low hexadecanol content was observed due to its
quick consumption during the hydrodeoxygenation reactions.
For the non-promoted MoS2/Al2O3 catalyst (Fig. 3(A)), the pal-
mitic acid and hexadecanol were formed in signicant quanti-
ties during the methyl palmitate transformation. Apparently,
acidic sites on the alumina support and the MoS2 particles
catalyzed the hydrolysis of methyl palmitate, but further
conversion of the intermediate oxygen-containing compounds
over non-promoted MoS2 proceeded slowly.
Fig. 5 Concentration profile of the reaction products for methyl
palmitate HDO over g-Al2O3 at the temperature 300 ꢀC and hydrogen
pressure 35 bars in n-tetradecane.
Hexadecanol could be converted into hexadecene as a result
of the dehydration of alcohol, with the subsequent hydrogena-
tion of hexadecene to hexadecane. Another way for hexadecanol
transformation of methyl palmitate with the formation of pal- conversion could be the formation of hexadecanethiol by
mitic acid over the alumina support was conrmed by our the SN2 nucleophilic substitution mechanism, as the hex-
supplementary experiments performed under the same condi- adecanethiol was found in observable amount among the
tions as the HDO of methyl palmitate over the sulde catalysts reaction products, especially over the CoMo/Al2O3 and non-
(Fig. 5). The subsequent transformation of palmitic acid over promoted Mo/Al2O3 catalysts. Our observation is coinciding
the alumina support didn't proceed; none of the expected with the literature data that declared the mechanism of alka-
products of this transformation were observed aer 6 h. In nethiol formation from the primary alcohol over the Mo/Al2O3
addition, the acid-catalyzed hydrolysis of methyl palmitate can catalyst.25–29 HDS of hexadecanethiol proceeded through the
proceed over the sulded NiMo and CoMo catalysts with the well-known hydrogenolysis reaction with the formation of
participation of –SH groups.25
hexadecane and H2S.
Two other routes for the methyl palmitate transformation
The estimated kinetic parameters for the methyl palmitate
were discussed in the literature, but the evidence for these hydrodeoxygenation over the sulded Mo/Al2O3, CoMo/Al2O3
proposed routes and the nature of the active sites for this and NiMo/Al2O3 catalysts are listed in Table 4. There are no
transformation were obscured. The direct transformation of the differences between the pseudo-rst order rate constants (per
methyl ester by hydrogenation the connecting oxygen of the Mo weight unit) of methyl palmitate conversion for all the
ester and hydrogenolysis of the C–O s-bond of the carboxylic investigated catalysts. It seems that the rst stage for the methyl
group (path 2) via the intermediate formation of corresponding palmitate transformation into palmitic acid proceeded on
aldehyde or alcohol was proposed as a possible pathway leading catalyst surface acidic sites independently on the promoter.
to hexadecane.25,27,40 In accordance with the scheme proposed According to the total oxygen conversion data that are shown in
for the methyl laurate29 and methyl heptanoate40 trans- Fig. 6, complete hydrodeoxygenation of the oxygen-containing
formations, methyl palmitate could be converted via direct compounds was achieved aer 4 hours over the CoMo/Al2O3
decarboxylation (path 3) with the elimination of pentadecene, catalyst; no oxygen-containing compounds were detected aer
carbon dioxide and methane to produce C15 hydrocarbons.
the complete conversion of methyl palmitate. For the Mo/Al2O3
Palmitic acid could be hydrogenated to C16 alkane via hex- and NiMo/Al2O3 catalysts, oxygen-containing compounds (pal-
adecanal and hecadecanol as sequential intermediates or mitic acid) were observed at complete conversion of methyl
undergo a decarboxylation/decarbonylation reaction to form a palmitate and total oxygen conversion was achieved only aer 6
C
15 hydrocarbon directly, as was reported for the other aliphatic hours of reaction. Comparing the total oxygen conversion data
fatty acids.10,25–29,41 Donnis and co-workers29 proposed that reveals the promoting effect of the Co and Ni addition to the
conversion of carboxylic compounds (dodecanoate or dodecanal) non-promoted Mo/Al2O3 catalyst.
to 1-decene or 1-dodecanol could proceed by preferentially
The initial rate of the total oxygen conversion was higher in
hydrogenation of the C]O or C]C bonds of keto-enol inter- 3.2 times for CoMo/Al2O and in 1.5 times for NiMo/Al2O3 cata-
mediate adsorbed on the catalyst surface. Snare and co-workers42 lyst relative to the non-promoted Mo/Al2O3. Therefore, the
discovered that the deoxygenation of stearic acid over supported CoMo/Al2O3 catalyst exhibited higher catalytic activity in the
metal catalysts produced saturated and unsaturated hydrocar- methyl palmitate HDO than NiMo/Al2O3; both promoted cata-
bons via decarboxylation and decarbonylation reactions. The lysts displayed higher activity than Mo/Al2O3. The effect of the
same possible routes were proposed for the heptanoic acid Co and Ni promoters on the HDS activity of the molybdenum
hydroconversion over sulded NiMo and CoMo catalysts.27,41 sulde is well-established. For the HDO reaction, the synergy
Therefore, the proposed pathways cannot be excluded for the between Ni and Mo was recently observed during the HDO of
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 2242–2250 | 2247

View Article Online
RSC Advances
Paper
Table 4 The estimated kinetic parameters for methyl palmitate hydrodeoxygenation
Catalysts
Reaction medium
Parameters
MoS2
CoMoS
NiMoS
n-Tetradecane
k
MP*, mole MP per g(Mo) per s
1.40 ꢂ 10ꢁ4 ꢃ 7 ꢂ 10ꢁ6
5.8 ꢂ 10ꢁ5 ꢃ 1 ꢂ 10ꢁ6
16.1
1.45 ꢂ 10ꢁ4 ꢃ 5 ꢂ 10ꢁ6
1. 9 ꢂ 10ꢁ4 ꢃ 1 ꢂ 10ꢁ5
2.79
1.45 ꢂ 10ꢁ4 ꢃ 6 ꢂ 10ꢁ6
1.2 ꢂ 10ꢁ4 ꢃ 1 ꢂ 10ꢁ5
0.65
WO0, mole per g(Mo) per s
C16/C15
m-Xylene
k
MP*, mole MP per g(Mo) per s
1.19 ꢂ 10ꢁ4 ꢃ 7 ꢂ 10ꢁ6
2.8 ꢂ 10ꢁ5 ꢃ 1 ꢂ 10ꢁ6
1.45
1.32 ꢂ 10ꢁ4 ꢃ 4 ꢂ 10ꢁ6
4.4 ꢂ 10ꢁ5 ꢃ 1 ꢂ 10ꢁ6
0.43
WO0, mole per g(Mo) per s
C16/C15
HT-SRGO
k
MP*, mole MP per g(Mo) per s
1.12 ꢂ 10ꢁ4 ꢃ 5 ꢂ 10ꢁ6
1.72 ꢂ 10ꢁ4 ꢃ 5 ꢂ 10ꢁ7
WO0, mole MP per g(Mo) per s
6.4 ꢂ 10ꢁ5 ꢃ 1 ꢂ 10ꢁ6
5.9 ꢂ 10ꢁ5 ꢃ 1 ꢂ 10ꢁ6
hydrotreating of rapeseed oil over MoS2/Al2O3 (ref. 13) and the
methyl heptanoate hydroconversion over bulk MoS2.41 The
major products for the methyl palmitate HDO over the sulded
CoMo/Al2O3 catalyst were C16 hydrocarbons, but the contribu-
tion of DeCOx reaction pathways increased and the ratio of C16/
C
15 decreased to 2.79. The C15 hydrocarbons was preferentially
formed over NiMo/g-Al2O3 (C16/C15 ¼ 0.65), indicating that
HDO and DeCOx reaction routes contributed differently to the
methyl palmitate hydroconversion over Mo/Al2O3, CoMo/Al2O3
and NiMo/g-Al2O3 catalysts.
Effect of the reaction medium's composition
The hydroconversion of methyl palmitate over promoted
CoMoS/Al2O3 and NiMoS/Al2O3 catalysts in m-xylene produced
Fig. 6 Effect of the catalyst on the total oxygen transformation during intermediate and nal products nearly identical to reactions in
methyl palmitate HDO over MoS2/Al2O3, CoMo/Al2O3 and NiMo/Al2O3
n-tetadecane. However, the overall output of the C15 and C16
hydrocarbons was decreased due to the participation of small
amounts of alkenes in acid-catalyzed m-xylene alkylation.
catalysts at the temperature 300 ꢀC and hydrogen pressure 35 bars
using n-tetradecane.
Approximately 10 to 15% of all produced hydrocarbons were
consumed by the alkylation reaction. The alkylation was more
rapeseed oil over NiMo/Al2O3 and Mo/Al2O3 catalysts.13 The rate
pronounced over the CoMo/Al2O3 catalyst, revealing its higher
of disappearance of triglycerides was faster than the rate of
acidity. In addition, one-, two- and three-substituted
oxygen disappearance for all investigated catalysts, pointing out
that the conversion of the oxygenated intermediates was the
limiting step. The higher activity of the Ni-promoted bulk MoS2
catalyst relative to their unpromoted counterparts was observed
during the hydroconversion of methyl heptanoate.41 To better
understand the mechanism for the methyl heptanoate hydro-
conversion, the reactivity of the intermediate oxygenate
compound, heptanal, was studied both experimentally and with
DFT calculations. The evident promotion of the MoS2 phase by
the Ni atoms was explained using the DFT calculations: the
energy barriers for C]O hydrogenation and C–OH bond
breaking decreased due to Ni insertion.41
The HDO of methyl palmitate over sulded catalysts
produced a mixture of linear, saturated C15 and C16 hydrocar-
bons, as well as admixtures of the corresponding alkenes. The
ratio of C16/C15 hydrocarbons depended on the catalyst used.
The major products obtained with the non-promoted, sulded
Mo/Al2O3 catalyst were C16 hydrocarbons: the ratio of C16/C15
hydrocarbons was 16.1 (Fig. 7, Table 4). Therefore, the HDO
pathway occurred preferentially over the non-promoted Mo/
Fig. 7 Effect of the solvent on the selectivity of C16 hydrocarbons
formation during methyl palmitate HDO over sulfide Mo/Al2O3, NiMo/
Al2O3 and CoMo/Al2O3 catalysts at the temperature 300 ꢀC and
hydrogen pressure 35 bars using n-tetradecane and m-xylene as a
Al2O3 catalyst; this result was previously encountered during the solvents.
2248 | RSC Adv., 2014, 4, 2242–2250
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
RSC Advances
methylbenzenes were produced during the reaction due to Al2O3 catalysts revealed similar activities during methyl palmi-
cross-alkylation of the m-xylene molecules. These observed tate conversion when n-tetradecane was used as a solvent, but
peculiarities were not found in the earlier reports describing they converted the intermediate oxygenates with activities
the HDO reaction in aromatic solvents, such as m-xylene27,28 or decreasing in the following order:
p-xylene.30–32
CoMo/Al2O3 > NiMoS/Al2O3 > MoS2/Al2O3.
The activities of the sulded CoMo/Al2O3 and NiMo/Al2O3
catalysts that were determined in different reaction medium
were compared using their total oxygen and methyl palmitate
conversions (Table 4). The pseudo-rst order reaction rate
constants for methyl palmitate conversion were slightly affected
by the solvent for the both catalysts. However, the initial rates of
The ratio of C16/C15 hydrocarbons aer the reaction depen-
ded on the catalyst: the main products obtained over non-
promoted, sulded Mo/Al2O3 catalysts were C16 hydrocarbons
that were produced predominately using the HDO route with
total oxygen conversion decreased sharply; the order of activity
95% selectivity. The conversion of methyl palmitate over
decreased similarly for the CoMo/Al2O3 and NiMo/Al2O3
promoted CoMoS/Al2O3 and NiMoS/Al2O3 catalysts proceeded
catalysts:
using both reaction routes (HDO and DeCOx) simultaneously.
The main products for the methyl palmitate converted over
sulded CoMo/Al2O3 catalyst were C16 hydrocarbons, but C15
hydrocarbons were preferably formed over the NiMo/Al2O3
n-Tetradecane > HT-SRGO > m-xylene.
The independence of methyl palmitate transformation rates catalyst, indicating an increase in the DeCOx reaction routes'
from the reaction medium composition pointed out that the contribution.
conversion of methyl palmitate did not inhibited by aromatic
The investigation of the effects of the reaction medium on
compounds. This behaviour was not surprising taking into the CoMo/Al2O3 and NiMo/Al2O3 catalysts' activity during the
account that the rst step for the methyl palmitate conversion methyl palmitate HDO allowed us to conclude that aromatic
was acid-catalysed hydrolysis proceeding on the acidic sites of compounds did not retarded methyl palmitate transformation
the alumina support or sulde phase. The reduced rates of total but instead inhibited the conversion of the intermediate
oxygen disappearance in m-xylene and HT-SRGO, which con- oxygenates. The activity of the CoMo/Al2O3 and NiMo/Al2O3
tained appreciable amounts of aromatic compounds, reveals catalysts during the methyl palmitate HDO in different solvents
that the HDO reactions of intermediate oxygenates were decreased in the following order: n-tetradecane > HT-SRGO >
inhibited by aromatics. The strong adsorption of the aromatic m-xylene. A decreased C16/C15 ratio was observed over both
compounds from the reaction medium on the active sites of the catalysts aer changing from n-tetradecane to m-xylene, indi-
sulde phase may be a major reason for the decrease in HDO cating that the HDO route was inhibited by aromatic molecules.
activity. These strongly adsorbed species reduced the number of Therefore, the selectivity of the triglyceride hydroconversion to
active sites by competing with the reactant.43 The C16/C15 ratio hydrocarbons with the odd and even number of carbon atoms
decreased from 0.65 to 0.43 over NiMo/Al2O3 catalyst and from could be inuenced by the composition of feedstock to be
2.79 to 1.45 over CoMo/Al2O3 catalyst when m-xylene was processed.
replaced by n-tetradecane. The decreasing trend for the C16/C15
ratio aer replacing n-tetradecane with m-xylene indicated that
the aromatic compounds mostly inhibited the HDO pathways. Acknowledgements
Therefore, the distribution of the HDO products for the oils and
fats between the hydrocarbons with odd and even carbon
This research was supported by RFBR Grant no. 11-03-00611.
The authors would like to thank Utkin V. A. and Shashkov M. V.
numbers could be affected by the reaction conditions, as was
for identifying the products using GC/MS; Dr S. V. Tsibulya for
previously described,17 and by the composition of the mixture to
the assistance in the XRD data interpretation.
be hydrotreated.
Conclusions
References
A series of experiments was performed to compare non-
promoted MoS2/Al2O3 and promoted CoMoS/Al2O3 and NiMoS/
Al2O3 catalysts during the hydrodeoxygenation of methyl
palmitate, which models a triglyceride-based feedstock. When
using sulded (Ni,Co)Mo/Al2O3 catalysts, the methyl palmitate
transforms into hydrocarbons through a complicated reaction
network containing parallel and consecutive reactions, such as
hydrogenation, dehydration and decarboxylation/decarbon-
ylation. The palmitic acid was formed in the rst step from the
hydrolysis of methyl palmitate over acidic sites on the alumina
support or sulde phase. The Mo/Al2O3, CoMo/Al2O3 and NiMo/
1 D. M. Alonso, J. Q. Bond and J. A. Dumesic, Green Chem.,
2010, 12, 1493.
2 G. W. Huber, S. Iborra and A. Corma, Chem. Rev., 2006, 106,
4044.
3 J. A. Melero, J. Iglesias and A. Garsia, Energy Environ. Sci.,
2012, 5, 7393.
4 P. Simacek, D. Kubicka, G. Sebor and M. Pospisil, Fuel, 2010,
89, 611.
5 G. Knothe, Prog. Energy Combust. Sci., 2010, 36, 364.
6 T. V. Choudhary and C. B. Phillips, Appl. Catal., A, 2011,
397, 1.
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 2242–2250 | 2249

View Article Online
RSC Advances
Paper
7 J. Mikulec, J. Cvengros, L. Joricova, M. Banic and A. Kleinova, 26 E.-M. Ryymin, M. L. Honkela, T.-R. Viljava and
J. Cleaner Prod., 2010, 18, 917. A. O. I. Krause, Appl. Catal., A, 2010, 389, 114.
8 T. N. Kalnes, M. M. McCall and D. R. Shonnard, in 27 O. I. Senol, E.-M. Ryymin, T.-R. Viljava and A. O. I. Krause,
Thermochemical Conversion of Biomass to liquid Fuels and J. Mol. Catal. A: Chem., 2009, 268, 1.
Chemicals, ed. M. Crocker, RSC publishing, 2010, ch. 18, 28 O. I. Senol, T.-R. Viljava and A. O. I. Krause, Appl. Catal., A,
pp. 468–495.
2007, 326(2007), 236.
9 M. Al-Sabawi and J. W. Chen, Energy Fuels, 2012, 26, 5373.
10 D. Kubicka and V. Tucas, in Advances in Chemical
29 B. Donnis, R. G. Egeberg, P. Blom and K. G. Knudsen, Top.
Catal., 2009, 52, 229.
Engineering, ed. D. Yu. Murzin, Elsevier, Amsterdam, 2013, 30 M. Ferrari, B. Delmon and P. Grange, Carbon, 2002, 40, 497.
vol. 42, ch. 3, pp. 141–195.
11 G. W. Huber and A. Corma, Angew. Chem., Int. Ed., 2007, 46,
7184.
12 S. Melis, Catal. Courier, 2008, (73), 6.
13 D. Kubicka and L. Kaluza, Appl. Catal., A, 2010, 372, 199.
14 A. Guzman, J. E. Torres, L. P. Prada and M. L. Nunez, Catal.
Today, 2010, 156, 38.
15 M. Toba, Y. Abe, H. Kuramochi, M. Osako, T. Mochizuki and
Y. Yoshimura, Catal. Today, 2011, 164, 533.
31 M. Ferrari, B. Delmon and P. Grange, Microporous
Mesoporous Mater., 2002, 56, 279.
32 G. de la Puente, A. Gil, J. J. Pis and P. Grange, Langmuir, 1999,
15, 5800.
33 I. V. Deliy, E. N. Vlasova, A. L. Nuzhdin and
G. A. Bukhtiyarova, in Proceedings of the 2nd European
Conference of Chemical Engineering, ed. N. Mastorakis,
V. Mladenov, C.M. Travieso-Gonzalez and M. Kohler, 2011,
pp. 24–29.
16 G. W. Huber, P. O'Connor and A. Corma, Appl. Catal., A, 34 E. Laurent and B. Delmon, Appl. Catal., A, 1994, 109, 77.
2007, 329, 120.
35 A. Centeno, E. Laurent and B. Delmon, J. Catal., 1995, 154,
17 R. Kumar, B. S. Rana, R. Tiwari, D. Verma, R. Kumar,
288.
R. K. Joshi, M. O. Garg and A. K. Sinha, Green Chem., 2010, 36 O. V. Klimov, A. V. Pashigreva, M. A. Fedotov, D. I. Kochubey,
12, 2232.
Y. A. Chesalov, G. A. Bukhtiyarova and A. S. Noskov, J. Mol.
Catal. A: Chem., 2010, 322, 80.
37 H. Topsøe, B. S. Clausen and F. E. Massoth, Hydrotreating
Catalysts in Catalysis: Science and technology, ed. J. R.
Anderson and M. Boudart, Springer, Berlin, 1996, vol. 11.
18 I. Sebos, A. Matsoukas, V. Apostolopoulos and
N. Papayannakos, Fuel, 2009, 88, 145.
19 Ch. Templis, A. Vonortas, I. Sebos and N. Papayannakos,
Appl. Catal., B, 2011, 104(2011), 324.
20 C. Toth, P. Baladincz, S. Covacs and J. Hancsok, Clean 38 S. Eijsbouts, L. C. A. van den Oetelaar and R. R. Puijenbroek,
Technol. Environ. Policy, 2011, 13, 581. J. Catal., 2005, 229, 352.
21 A. Vonortas, Ch. Templis and N. Papayannakos, Energy Fuels, 39 S. Eijsbouts, L. C. A. van den Oetelaar, J. N. Louwen,
2012, 26, 3856.
R. R. Puijenbroek and G. C. van Leerdam, Ind. Eng. Chem.
Res., 2007, 46, 3945.
40 O. I. Senol, T.-R. Viljava and A. O. I. Krause, Catal. Today,
2005, 100, 331.
41 M. R. de Brimont, C. Dupont, A. Daudin, C. Geantet and
P. Raybaud, J. Catal., 2012, 286, 153.
42 M. Snare, I. Kubickova, P. Maki-Arvela, K. Eranen and
D. Y. Murzin, Ind. Eng. Chem. Res., 2006, 45, 5708.
43 E. Furimski and F. Massoth, Catal. Today, 1999, 52, 381.
22 J. Walendziewski, M. Stolarski, R. Luzny and B. Klimek, Fuel
Process. Technol., 2009, 90, 686.
23 B. Peng, C. Zhao, S. Kasakov, S. Foraita and J. A. Lercher,
Chem.–Eur. J., 2013, 19, 4732.
24 H. Zuo, Q. Liu, T. Wang, L. Ma, Q. Zhang and Q. Zhang,
Energy Fuels, 2012, 26, 3747.
25 E.-M. Ryymin, M. L. Honkela, T.-R. Viljava and
A. O. I. Krause, Appl. Catal., A, 2009, 358, 42.
2250 | RSC Adv., 2014, 4, 2242–2250
This journal is © The Royal Society of Chemistry 2014