Hydrodeoxygenation of fatty acids and triglycerides by Pt-loaded Nb2O5catalysts

Article abstract of DOI:10.1039/c4cy00757c

Platinum nanoparticles loaded onto various supports have been studied for the selective hydrogenation of lauric acid to n-dodecane. The activity depends on the support material and pre-reduction temperature. Pt/Nb₂O₅reduced at 300 °C gives the highest activity. Pt/Nb₂O₅shows higher activity than various Nb₂O₅-supported transition metals (Ir, Re, Ru, Pd, Cu, Ni). Under solvent-free conditions Pt/Nb₂O₅is effective for the hydrodeoxygenation of lauric, capric, palmitic, myristic, oleic, and stearic acids under 8 bar H₂at 180-250 °C, which gives high yields (88-100%) of linear alkanes with the same chain length as the starting compound. Tristearin is also converted to give 93% yield of n-octadecane. Pt/Nb₂O₅shows more than 60 times higher turnover number (TON) than the previously reported catalysts for the hydrogenation of stearic acid to n-octadecane. Mechanistic study shows a consecutive reaction pathway in which lauric acid is hydrogenated to 1-dodecanol, which undergoes esterification with lauric acid as well as hydrogenation to n-dodecane. The ester undergoes hydrogenolysis to give the alcohol, which is hydrogenated to the alkane. Infrared (IR) study of acetic acid adsorption on Nb₂O₅indicates that Lewis acid-base interaction of Nb cation and carbonyl oxygen, which suggests a possible role of Nb₂O₅as an activation site of carbonyl groups during hydrodeoxygenation. This journal is

Full text of DOI:10.1039/c4cy00757c

View Article Online
View Journal
Catalysis
Science &
Technology
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: K. Shimizu, K.
Kon, S. Takakusagi and W. Onodera, Catal. Sci. Technol., 2014, DOI: 10.1039/C4CY00757C.
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
Accepted Manuscripts are published online shortly after
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
Information for Authors.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the Ethical guidelines still
apply. In no event shall the Royal Society of Chemistry be held
responsible for any errors or omissions in this Accepted Manuscript
or any consequences arising from the use of any information it
contains.
www.rsc.org/catalysis

Page 1 of 17
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C4CY00757C
Hydrodeoxygenation of fatty acids and triglyceride by Pt-loaded Nb2O5
catalysts
Kenichi Kon a, Wataru Onodera,a Satoru Takakusagi,a Kenꢀichi Shimizu a,b,
*
a Catalysis Research Center, Hokkaido University, Nꢀ21, Wꢀ10, Sapporo 001ꢀ0021, Japan
b
Elements Strategy Initiative for Catalysts and Batteries, Kyoto University, Katsura, Kyoto
615ꢀ8520, Japan
*Corresponding author
Kenꢀichi Shimizu
Catalysis Research Center, Hokkaido University, Nꢀ21, Wꢀ10, Sapporo 001ꢀ0021, Japan
Eꢀmail: kshimizu@cat.hokudai.ac.jp, Fax: +81ꢀ11ꢀ706ꢀ9163
Abstract
Platinum nanoparticles loaded onto various supports have been studied for the selective
hydrogenation of lauric acid to nꢀdodecane. The activity depends on the support material and
preꢀreduction temperature. Pt/Nb2O5 reduced at 300 °C gives the highest activity. Pt/Nb2O5
shows higher activity than Nb2O5ꢀsupported various transition metals (Ir, Re, Ru, Pd, Cu, Ni).
Under solventꢀfree conditions Pt/Nb2O5 is effective for hydrodeoxygenation of lauric, capric,
palmitic, myristic, oleic, and stearic acids under 8 bar H2 at 180ꢀ250 °C, which gives high yield
(88ꢀ100%) of linear alkanes with the same chain length as the starting compound. Tristearin is
also converted to give 93% yield of nꢀoctadecane. Pt/Nb2O5 shows more than 60 times higher
turnover number (TON) than the previously reported catalysts for hydrogenation of stearic acid
to nꢀoctadecane. Mechanistic study shows a consecutive reaction pathway in which lauric acid
is hydrogenated to 1ꢀdodecanol, which undergoes esterification with lauric acid as well as
hydrogenation to nꢀdodecane. The ester undergoes hydrogenolysis to give the alcohol, which is
hydrogenated to the alkane. Infrared (IR) study of acetic acid adsorption on Nb2O5 indicates that
Lewis acidꢀbase interaction of Nb cation and carbonyl oxygen, suggesting a possible role of
Nb2O5 as activation site of carbonyl groups during the hydrodeoxygenation.
Introduction
Developing alternative fuels from nonꢀfood biomass has become a key factor to ensure the
sustainable growth through minimizing the carbon footprint. In this context, triglyceride and
fatty acids have been identified as promising resources of bioꢀdiesel,1ꢀ5 because they can be
easily and economically produced from microalgae which has rapid growth rate and high lipid
content. As comprehensively reviewed by Murzin et al.,2 three upgrading processes, cracking,
transesterification and deoxygenation, have been developed to produce bioꢀdiesel.
Deoxygenation of triglyceride/fatty acid feed under hydrogen is of particular importance,
because the reaction reduces highꢀoxygen content and related acidity of fatty acids, yielding
saturated hydrocarbons suitable for dropꢀin diesel fuel. As comprehensively described by
Lercher et al.1 and Bitter et al.3 catalytic deoxygenation of biomassꢀderived fatty acid
derivatives can be classified into three reactions: decarbonylation, decarboxylation and
hydrodeoxygenation. Decarbonylation and decarboxylation yield hydrocarbons with one carbon
1

Catalysis Science & Technology
Page 2 of 17
View Article Online
DOI: 10.1039/C4CY00757C
atom less than the fatty acids, while hydrodeoxygenation gives hydrocarbons with the same
chain length as the starting compound. Hydrodesulfurization catalysts (sulfided CoMo or NiMo
oxides), which are selective to hydrodeoxygenation,6,7 can suffer from sulfur leaching. In
8
contrast, most of the supported group 9 and 10 metal catalysts,8ꢀ15 such as Ni/ZrO2 and
supported Pd, Pt, and Rh catalysts,9ꢀ13 suffer from high selectivity to decarbonylation or
decarboxylation products even in the presence of H2. A few catalysts (Niꢀloaded zeolites,16
PtꢀRe/ZSMꢀ5,17 andtungsten and molybdenum carbides18ꢀ21) have been reported to be effective
for selective transformation of fatty acids into dieselꢀrange alkanes without a loss of the chain
length of the corresponding fatty acid at 260ꢀ350 °C under 10ꢀ50 bar H2. However, they have
still problems such as high catalyst loading (low turnover number, TON), formation of small
amount of decarbonylation and cracking byproducts and needs of high pressure H2 and high
temperature. In this paper we report selective and quantitative hydrodeoxygenation of stearic
acid to nꢀoctadecane under milder conditions (180 °C, 2 or 8 bar H2) than previous studies using
Pt/Nb2O5, which shows more than two orders of magnitude higher TON than Pt/SiO2 and
previous catalysts for the selective hydrodeoxygenation of fatty acids. Substrate scope, kinetic
and in situ IR studies are also carried out to discuss the catalytic performance, reaction pathways,
and a possible role of Nb2O5 in the present catalytic system.
Experimental
General
Commercially available organic compounds (from Tokyo Chemical Industry or Kanto
Chemical) were used without further purification. The GC (Shimadzu GCꢀ14B) and GCMS
(Shimadzu GCMSꢀQP2010) analyses were carried out with Ultra ALLOY capillary column
UA+ꢀ1 (Frontier Laboratories Ltd.) using nitrogen and He as the carrier gas.
Catalyst preparation
Nb2O5 (54 m2 gꢀ1) was prepared by calcination of niobic acid (Nb2O5ꢁnH2O, kindly supplied by
CBMMI) at 500 °C for 3 h. SiO2 (Qꢀ10, 300 m2 gꢀ1) was supplied from Fuji Silysia Chemical
Ltd. H+ꢀtype MFI zeolite (HMFI) with a SiO2/Al2O3 ratio of 22.3 was kindly supplied by Tosoh
Co. HBEA zeolite (JRCꢀZꢀHB25, SiO2/Al2O3= 25±5), silicaꢀalumina (JRCꢀSALꢀ2,
SiO2/Al2O3=5.6), MgO (JRCꢀMGOꢀ3), TiO2 (JRCꢀTIOꢀ4), CeO2 (JRCꢀCEOꢀ3) was supplied
from Catalysis Society of Japan. γꢀAl2O3 was prepared by calcination of γꢀAlOOH (Catapal B
Alumina purchased from Sasol) for 3 h at 900 °C. ZrO2 was prepared by calcination (500 °C for
3 h) of ZrO2ꢁnH2O prepared by hydrolysis of zirconium oxynitrate 2ꢀhydrate in water by
aqueous NH4OH solution, followed by filtration of precipitate, washing with water three times,
and drying at 100 °C.
Precursor of Pt/Nb2O5 was prepared by an impregnation method; a mixture of Nb2O5 and an
aqueous HNO3 solution of Pt(NH3)2(NO3)2 was evaporated at 50 °C, followed by drying at
90 °C for 12 h. Before each catalytic experiment, the Pt/Nb2O5 catalyst (with Pt loading of 5
wt%) was prepared by in situ preꢀreduction of the precursor in a pyrex tube under a flow of H2
(20 cm3 minꢀ1) at 300 °C for 0.5 h. Other supported Pt catalysts (Pt = 5 wt%) were prepared by
the same method. Nb2O5–supported metal catalysts, M/Nb2O5 (M = Ni, Cu, Ru, Pd, Ag, Re, Ir)
with metal loading of 5 wt% were prepared by impregnation method in the similar manner as
Pt/Nb2O5 using aqueous solution of metal nitrates (for Co, Ni, Cu, Ag), RuCl3, IrCl3ꢀnH2O,
NH4ReO4 or aqueous HNO3 solution of Pd(NO3)2. A commercial Ptꢀloaded carbon catalyst
2

Page 3 of 17
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C4CY00757C
(Pt/C, Pt = 5 wt%) was purchased from N.E. Chemcat, Corporation.
Characterization
The temperature programmed reduction under H2 (H2ꢀTPR) was carried out with BELCAT
(BELL Japan Inc.). Unreduced precursor of Pt/Nb2O5, Pt(NH3)2(NO3)2ꢀloaded Nb2O5 (20 mg),
was mounted in a quartz tube, and the sample was heated with a temperature rampꢀrate of 2 ˚C
minꢀ1 in a flow of 5% H2/Ar (20 cm3 minꢀ1). The effluent gas was passed through a trap
containing MS4Å to remove water, then through the thermal conductivity detector. The amount
of H2 consumed during the experiment was detected by a thermal conductivity detector.
The number of surface metal atoms in Pt/Nb2O5, preꢀreduced in H2 at 300 °C for 0.5 h, was
estimated from the CO uptake of the samples at room temperature using the pulseꢀadsorption of
CO in a flow of He by BELCAT. The average particle size was calculated from the CO uptake
assuming that CO was adsorbed on the surface of spherical Pt particles at CO/(surface Pt atom)
= 1/1 stoichiometry.
TEM measurements were carried out by using a JEOL JEMꢀ2100F TEM operated at 200 kV.
In situ IR spectra were recorded using a JASCO FT/IRꢀ4200 equipped with an MCT detector.
For the IR studies of CO and pyridine adsorption, we used a flowꢀtype IR cell connected to a
flow reaction system. For the IR study of acetic acid adsorption, we used a closed IR cell
connected to an evacuation system. The sample was pressed into a 40 mg of selfꢀsupporting
wafer (φ = 2 cm) and mounted into the IR cell with CaF2 windows. Spectra were measured
accumulating 15 scans at a resolution of 4 cmꢀ1. A reference spectrum of the catalyst wafer in
He taken at measurement temperature was subtracted from each spectrum. For the COꢀIR study,
a disk of Pt/Nb2O5 was heated in H2 flow (20 cm3 minꢀ1) at 300 °C for 0.5 h, followed by
exposure to a flow of CO(5%)/He (20 cm3 minꢀ1) for 180 s at 40 °C, and by purging with He for
600 s. Then, IR measurement was carried out. For the pyridineꢀIR study, the Pt/Nb2O5 disc,
preꢀreduced in H2 at 300 °C for 0.5 h, was exposed to 1ꢂL of liquid pyridine vaporized at
200 °C under He flow at 150 °C. Then, the disk was purged with He for 600 s, and IR
measurement was carried out. During the acetic acidꢀIR study, the IR cell was cooled by an ice
bath, and the thermocouple near the sample showed 8 ± 1 oC. The sample disc (Nb2O5 or SiO2)
in the closed IR cell, preꢀevacuated at 500 °C for 0.5 h, was exposed to 20 Pa of acetic acid at 8
oC for 600 s, then IR spectra were measured at 8 oC.
Typical procedures of catalytic reactions
Pt/Nb2O5 (0.07, 0.2 or 1 mol%ꢀPt with respect to carboxylic acid) preꢀreduced at 300 °C was
used as a standard catalyst. The other catalysts were also preꢀreduced at 300 °C before reaction.
After the preꢀreduction, the catalyst in the closed glass tube sealed with a septum inlet was
cooled to room temperature under H2 atmosphere. The mixture of carboxylic acid (1 mmol) and
nꢀdecane or nꢀdodecane (0.2 mmol) was injected to the preꢀreduced catalyst inside the glass
tube through the septum inlet. Then, the septum was removed under air, and a magnetic stirrer
was put in the tube, followed by inserting the tube inside stainless autoclave with a dead space
of 28 cm3. Soon after being sealed, the reactor was flushed with H2 from a high pressure gas
cylinder and charged with H2 (typically 8 bar) at room temperature. Then, the reactor was
heated typically at 180 ˚C under stirring (150 rpm). After the reaction, acetone (10 cm3) was
added to the mixture, and conversion and yields of products were determined by GC using
nꢀdecane or nꢀdodecane as an internal standard. The products were identified by GCꢀMS
3

Catalysis Science & Technology
Page 4 of 17
View Article Online
DOI: 10.1039/C4CY00757C
equipped with the same column as GC and by comparison with commercially available
products.
Results and discussion
Characterization of Pt/Nb2O5
We characterized the structure of the representative catalyst, Pt/Nb2O5 preꢀreduced at 300 °C.
H2ꢀTPR of the unreduced precursor, Pt(NH3)2(NO3)2ꢀloaded Nb2O5, is shown in Fig. 1. The H2
consumption due to the reduction of oxidic Pt to metallic Pt is observed mostly below 100 °C.
Fig. 2 shows the particle size distribution of Pt/Nb2O5 obtained by transmission electron
microscopy (TEM) analysis. The volumeꢀarea mean diameter of Pt particle was 4.7 nm. The
number of exposed Pt0 sites estimated by the CO adsorption amount (28.3 ꢂmol gꢀ1ꢀcat) is
smaller than the number of the surface Pt atoms (39.4 ꢂmol gꢀ1ꢀcat) estimated by using the
volumeꢀarea mean diameter of Pt by TEM and by assuming that CO is adsorbed on the surface
of spherical Pt particles at a stoichiometry of CO/(surface Pt atom) = 1/1. This result is
consistent with the previous result that increase in the H2ꢀtreatment temperature of Pt/Nb2O5
leads to decrease in hydrogen and CO uptake on the catalyst.22ꢀ24 This phenomenon has been
generally called strong metalꢀsupport interaction (SMSI),22ꢀ34 where a metal surface is partially
covered with suboxide (NbOx) species produced by partial reduction of the support (Nb2O5) at
high temperature (200ꢀ500 oC). The SMSI phenomenon was also observed in Nb2O5–supported
Pd, Rh and Ir catalysts.26ꢀ28 The IR study of CO adsorption on Pd/Nb2O5 showed that increase in
the reduction temperature of Pd/Nb2O5 from 100 oC to 300 oC resulted in the suppressions of the
band at 1938 cmꢀ1 due to the bridged CO on Pd.27 This observation was assigned to the blocking
of the Pd surface by migrated NbOx species. As shown in Fig. 3, our IR experiments of CO
adsorption on Pt/Nb2O5 gave similar results. The IR spectrum of CO on Pt/Nb2O5 reduced at
100 °C showed bands in a range 2000ꢀ2080 cmꢀ1 due to linearly coordinated CO on Pt0 and a
band at 1840 cmꢀ1 due to bridged CO on flat surface of Pt.24 The increase in the reduction
temperature of Pt/Nb2O5 from 100 oC to 300 oC resulted in the large suppression of the bridged
CO band and the linear CO bands (especially at low wavenumber region). The result supports
o
the SMSI model of Pt/Nb2O5 reduced at 300 C, where the Pt surface is partly covered with
NbOx species. Summarizing the structural results, it is concluded that the dominant Pt species
in Pt/Nb2O5 preꢀreduced at 300 oC are Pt metal particles whose surface is partially covered with
NbOx species.
Optimization of catalysts and conditions
We studied the influence of various catalyst parameters on the catalytic activity for
hydrogenation of lauric acid in solventꢀfree conditions under 8 bar H2 at 180 ˚C for 4 h using 1
mol% of the catalyst (Table 1). First, various metal (M)ꢀloaded Nb2O5 catalysts (M= Pt, Ir, Ru,
Re, Pd, Cu, Ni), preꢀreduced at 300°C, were studied (entries 2ꢀ8). It was found that Pt/Nb2O5
showed the highest yield of nꢀdodecane. 1ꢀDodecanol and dodecyl dodecanoate were also
produced as byproducts. Next, we screened a series of supported Pt catalysts preꢀreduced at
300°C (entries 2,10ꢀ20). The activity depended strongly on the support materials, and Pt/Nb2O5
(entry 2) showed more than 10 times higher yield of nꢀdodecane (60%) than the other Ptꢀloaded
metal oxides (entries 10ꢀ20). Pt/TiO2 (entry 13) and niobic acidꢀsupported Pt (entry 10) also
showed high conversion, but these catalysts gave partially reduced products (dodecyl
dodecanoate and 1ꢀdodecanol) as main products. Pt metal particles on an inert support (Pt/SiO2,
4

Page 5 of 17
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C4CY00757C
entry 12) and Nb2O5 itself (entry 1) showed no yield of dodecane, indicating that Pt metal
particles and the support itself could not catalyze the reaction under the present conditions. The
Pt/Nb2O5 catalyst reduced at 100 °C (entry 9) showed lower yield of nꢀdodecane (16%) than
that reduced at 300 °C (60%).
With the most effective catalyst, Pt/Nb2O5 reduced at 300 °C, we studied the optimization of
reaction conditions. Fig. 4 shows the effect of hydrogen pressure on conversion and product
yields for the reaction after 4 h. Note that the conversion and yield at H2 pressure of 0 bar
correspond to the reaction under 1 bar N2. The dodecane yield was highest at H2 pressure of 8
bar. Further increase in the H2 pressure decreased the yield of dodecane and increased those of
dodecanol and ester. This may suggests a competitive activation of hydrogen and lauric acid at
the same catalytic site under high H2 pressure .
Fig. 5 shows the time course of the reaction under solventꢀfree conditions at 180 °C in 8 bar H2.
Note that the conversion and yield at t = 0 h were measured after addition of the reaction
mixture to the preꢀreduced Pt/Nb2O5 under ambient conditions. The profile is characteristic to
consecutive reaction mechanism; 1ꢀdodecanol and dodecyl dodecanoate, formed at an initial
induction period, were consumed after 1 h to give the overꢀhydrogenated product, nꢀdodecane.
After 24 h, the yield of nꢀdodecane was 99%, and no formation of byproducts was observed by
GCꢀMS analysis. Next, we tested the reaction under 8 bar H2 for 24 h in different solvents
(Table 2). The reaction under solventꢀfree conditions gave the highest yield (99%) of
nꢀdodecane. The reaction in toluene also gave good yield (90%). Use of oꢀxylene, mesitylene,
diglyme and water resulted in moderate yield (31ꢀ43%) with coꢀproduction of 1ꢀdodecanol and
dodecyl dodecanoate as byproducts.
Performance of Pt/Nb2O5-catalyzed hydrogenation of fatty acids
From the above optimization experiments, we found the best conditions (8 bar H2, 24 h, 180 °C,
1 mol% Pt/Nb2O5, without solvent), which gave a quantitative dodecane yield without any
byproducts (Table 3, entry 1). ICPꢀAES analysis of the solution after the reaction confirmed that
the contents of Pt and Nb in the solution were below the detection limits. After the standard
reaction for 24 h, the catalyst was separated from reaction mixture by centrifugation and was
dried at 90 ºC at 12 h and then reduced in H2 at 300 ºC for 0.5 h. The recovered catalyst showed
quantitative yield of nꢀdodecane (Table 3, entries 2,3). These results indicate that Pt/Nb2O5 acts
as reusable heterogeneous catalysts for this reaction. It is important to note that this system is
effective even under low H2 pressure of 2 bar, giving a quantitative yield (equation 1).
O
5
Pt/Nb2O5 (1 mol%)
+ H2
(1)
+ 2H2O
o
5
OH
180 C, 24 h
> 99 %
1 mmol
neat
2 bar
As summarized in Table 3, the scope of the liquidꢀphase hydrogenation using Pt/Nb2O5 under
solventꢀfree conditions was further expanded to various fatty acids and tristearin. The reactions
of capric, palmitic, myristic, oleic, and stearic acids at low hydrogen pressure (8 bar) at
180ꢀ250 °C resulted in high yield (88ꢀ99%) of linear alkanes with the same chain length as the
starting compound, and yields of byproducts (lower or branched alkanes, alcohols and esters)
were below the detection limit of GC analyses. Tristearin was also converted to give 93% yield
5

Catalysis Science & Technology
Page 6 of 17
View Article Online
DOI: 10.1039/C4CY00757C
of nꢀoctadecane. Another important aspect of this catalytic system is high TON. As shown in
entry 8, hydrogenation of 2 mmol stearic acid in the presence of small amount of the catalyst
(0.07 mol%) at 250 °C under 8 bar H2 resulted in 96% yield of nꢀoctadecane, corresponding to
TON of 1371 with respect to total number of Pt atoms in the catalyst. As compared in Table 4,
this value is more than 60 times higher than those of previously reported heterogeneous
catalysts. Under the same conditions, Pt/SiO2 gave no yield of nꢀoctadecane but gave 7% yield
of a hydrocarbon with one carbon atom less than the fatty acids.
Reaction pathways
The kinetic profile for hydrogenation of lauric acid (Fig. 5) indicates the consecutive reaction
pathway; dodecanol and dodecyl dodecanoate, formed at an initial induction period, are
hydrogenated to yield dodecane as the final product. To give further evidences on the tentative
mechanism, we compared the initial rates of the individual steps. Table 5 lists the initial
formation rates of the products for hydrogenation of lauric acid, 1ꢀdodecanal, 1ꢀdodecanol or
dodecyl dodecanoate (ester), together with the ester formation rate by the reaction of lauric acid
with 1ꢀdedecanol. For the hydrogenation of lauric acid, the formation of nꢀdodecane (5 mmol
hꢀ1) is slower than the formation of 1ꢀdodecanol (17 mmol hꢀ1) and the ester (31 mmol hꢀ1). The
hydrogenation of 1ꢀdodecanal, 1ꢀdodecanol and dodecyl dodecanoate resulted in the formation
of nꢀdodecane as the main product. The order of the nꢀdodecane formation rate is as follows:
dodecanal (15 mmol hꢀ1) > dodecanol (13 mmol hꢀ1) > dodecyl dodecanoate (8 mmol hꢀ1) >
lauric acid (5 mmol hꢀ1). Considering the facts that the esterification of lauric acid with
1ꢀdodecanol (63 mmol hꢀ1) shows the highest rate in the reactions in Table 5, we propose a
possible reaction pathway in Scheme 1. The carboxylic acid is hydrogenated to an alcohol
possibly via an aldehyde. However, the fact that 1ꢀdodecanol is not produced by the
hydrogenation of 1ꢀdodecanal (entry 2 in Table 5) suggests a direct hydrogenation of carboxylic
acid to alcohol. The alcohol undergoes esterification with the carboxylic acid as well as
hydrogenation to an alkane (dodecane). The ester undergoes hydrogenolysis to give the alcohol,
which is hydrogenated to the alkane. During the model hydrogenation of dodecanol (Table 5,
entry 2), we did not observe dodecene as a byproduct, which exclude a possible formation of
dodecane via dehydration of the alcohol to alkene followed by its hydrogenation. The proposed
pathway in Scheme 1 is consistent with that proposed by Lercher for the catalytic hydrogenation
of stearic acid to octadecane over Ni/HBA catalyst.16 They proposed the production of
octadecane by the consecutive pathway: stearic acid hydrogenation to 1ꢀoctadecanol, followed
by 1ꢀoctadecanol hydrogenation to octadecane.
Possible role of Nb2O5
As discussed above, Pt/Nb2O5 showed significantly higher activity than the other Ptꢀloaded
catalysts and Nb2O5 itself, suggesting a synergy between Pt and the Nb2O5 support. On the basis
of our characterization results and the SMSI model of the Pt/Nb2O5 system in the literatures,22ꢀ24
we have concluded that that the dominant Pt species in the most effective catalyst, Pt/Nb2O5
reduced at 300 oC, are the Pt metal particles whose surface is partly covered with NbOx species.
As shown in Table 1, increase in the reduction temperature of Pt/Nb2O5 from 100 oC to 300 oC
increased its activity for the hydrodeoxygenation. The COꢀIR result in Fig. 3 suggests that
o
increase in the reduction temperature of Pt/Nb2O5 from 100 oC to 300 C resulted in nearly
complete blocking of the bridged site on a flat Pt surface by NbOx species. Summarizing these
6

Page 7 of 17
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C4CY00757C
o
results, the higher activity of the Pt/Nb2O5 catalyst reduced at 300 C is explained by a partial
decoration of Pt by NbOx in which relatively large number of Pt0 atoms can be located at
metalꢀNbOx perimeter sites. Pt/TiO2 is also well known as a SMSI catalyst.29ꢀ34 Table 1 shows
o
that Pt/TiO2 (reduced at 300 C) is a moderately effective catalyst; it shows higher yields of
dodecane and partially reduced products (dodecyl dodecanoate and dodecanol) than nonꢀSMSI
catalysts (such as Pt/C, Pt/SiO2 and Pt/Al2O3). These facts lead to a hypothesis that cooperation
between Pt metal and adjacent NbOx (or TiOx) species is responsible for the high catalytic
activity.
It is known that the partial blocking of Pt by suboxides (NbOx or TiOx) causes the formation
of new perimeter sites between Pt and the suboxide.22ꢀ24 Yoshitake and Iwasawa22 studied
o
hydrogenation of acrolein by the Pt/Nb2O5 catalyst in SMSI state (reduced at 500 C). They
propose that the higher selectivity in C=O bond reduction than C=C bond reduction is caused by
a cooperation between Pt and NbOx at the perimeter sites, in which H2 is dissociated by Pt and
acrolein is activated by NbOx. Vannice29 also showed a significant enhancement in the
selectivity in C=O bond reduction for hydrogenation of crotonaldehyde by Pt/TiO2 reduced at
o
500 C. Somorjai and Bell showed that activity of Rh foil for CO and CO2 hydrogenation was
increased by partially decorating the Rh surface with suboxides such as NbOx, and the activity
for the decorated Rh surface increased with increase in the Lewis acidity of the suboxide.37 This
rate enhancement is attributed to the activation of the C=O bond via Lewis acidꢀbase interaction
between the cationic sites in the suboxide and the oxygen atom of CO or CO2.37ꢀ39 Enhanced
selectivity or activity by metalꢀLewis acid cooperation at metalꢀoxide interface has been
reported for various catalytic systems for hydrogenation of CꢀO or C=O bonds.22,26ꢀ31,35ꢀ43
As evidenced by the IR band due to coordinated pyridine on Lewis acid site (1445 cmꢀ1) in the
pyridine adsorption IR experiment (Fig. 6), Pt/Nb2O5 has Lewis acid sites (exposed Nbn+
cations). Then, to evaluate the Lewis acidꢀbase interaction between Nb cations and carbonyl
oxygens of carboxylic acids in the present system, we carried out IR observation of acetic acid
on Nb2O5. The spectrum in Fig. 7 shows the C=O stretching band of adsorbed acetic acid30 at
lower wavenumber (1660 cmꢀ1) than that for non Lewis acidic oxide, SiO2 (1715 cmꢀ1). This
indicates charge transfer from the oxygen atom of carboxyl group (Lewis base) to the Nb5+
cation (Lewis acid) site of the Nb2O5 surface. Taking these discussions into consideration, we
propose that polarization of the C=O bond of carboxylic acid by Lewis acidꢀbase interaction of
the Nb cation and carboxyl group is the primary important role of the Nb2O5 support in the
present catalytic system (Scheme 2). The high activity of Pt/Nb2O5 can be caused by
cooperation between Pt and NbOx at the perimeter sites, in which H2 is dissociated by Pt and
carboxylic acid is activated by NbOx via the Lewis acidꢀbase interaction. The increase in the
o
activity with increase in the reduction temperature of Pt/Nb2O5 from 100 oC to 300 C could be
explained by an increase in the PtꢀNbOx perimeter length caused by a migration of NbOx
islands onto Pt metal surface.
Conclusions
Pt/Nb2O5 was found to be an effective catalyst for hydrodeoxygenation of lauric, capric,
palmitic, myristic, oleic, and stearic acids at low H2 pressure (8 bar) at 180ꢀ250 °C under
solventꢀfree conditions, which gives high yield of linear alkanes with the same chain length as
the starting compound. Tristearin is also converted to give 93% yield of nꢀoctadecane under 50
7

Catalysis Science & Technology
Page 8 of 17
View Article Online
DOI: 10.1039/C4CY00757C
bar H2 at 260 °C. Pt/Nb2O5 shows more than 60 times higher TON than previously reported
catalysts for hydrogenation of stearic acid to nꢀoctadecane. Mechanistic study showed the
following reaction pathways. Carboxylic acid is hydrogenated to alcohol, which undergoes
esterification with carboxylic acid as well as hydrogenation to alkane. The ester undergoes
hydrogenolysis to give alcohol, which is hydrogenated to alkane. Pt/Nb2O5 in a partial SMSI
state showed significantly higher activity than the nonꢀSMSI catalysts. Combined with the IR
evidence on the Lewis acidꢀbase interaction between the surface Nb cation and carboxyl oxygen
of an adsorbed carboxylic acid, we propose that the high activity of Pt/Nb2O5 is caused by
cooperation between Pt and NbOx at the perimeter sites, in which H2 is dissociated by Pt and
carboxylic acid is activated by NbOx via the Lewis acidꢀbase interaction.
Acknowledgement
This work was supported by a GrantꢀinꢀAid for Scientific Research on Innovative Areas "Nano
Informatics" (25106010) from JSPS and a MEXT program "Elements Strategy Initiative to
Form Core Research Center".
Notes and references
1. C. Zhao, T. Brück and J. A. Lercher, Green Chem., 2013, 15, 1720–1739.
2. S. Lestari, P. MäkiꢀArvela, J. Beltramini, G. Q. Max Lu and D. Yu. Murzin, ChemSusChem,
2009, 2, 1109–1119.
3. R. W. Gosselink, S. A. Hollak, S.ꢀW. Chang, J. van Haveren, K. P. de Jong, J. H. Bitter and D.
S. van Es, ChemSusChem, 2013, 6, 1576−1594.
4. F. Shi, P. Wang, Y. H. Duan, D. Link and B. Morreale, RSC Adv., 2012, 2, 9727– 9747.
5. J. C. SerranoꢀRuiz, E. V. RamosꢀFernandez and A. SepulvedaꢀEscribano, Energy Environ.
Sci., 2012, 5, 5638–5652.
6. O. I. Senol, E. M. Ryymin, T. R. Viljava and A. O. I. Krause, J. Mol. Catal. A, 2007, 268,
1–8.
7. Y. Liu, R. SoteloꢀBoyas, K. Murata, T. Minowa and K. Sakanishi, Energy Fuels, 2011, 25,
4675–4685.
8. B. Peng, C. Zhao, S. Kasakov, S. Fortaita and J. A. Lercher, Chem. Eur. J., 2013, 19,
4732–4741.
9. I. Kubičková, M. Snåre, K. Eränen, P. MäkiꢀArvela and D. Yu. Murzin, Catal. Today, 2005,
106, 197–200.
10. B. Rozmysłowicz, P. MäkiꢀArvela, A. Tokarev, A.ꢀR. Leino, K. Eränen and D. Y. Murzin,
Ind. Eng. Chem. Res., 2012, 51, 8922–8927.
11. J. Han, H. Sun, Y. Ding, H. Lou and X. Zheng, Green Chem., 2010, 12, 463−467.
12. J. Han, H. Sun, J. Duan, Y. Ding, H. Lou and X. Zheng, Adv. Synth. Catal., 2010, 352,
1805−1809.
13. C. Wang, Z. Tian, L. Wang, R. Xu, Q. Liu, W. Qu, H. Ma and B. Wang, ChemSusChem,
2012, 5, 1974–1983.
14. M. Snåre, I. Kubičková, K. Eränen, P. MäkiꢀArvela and D. Yu. Murzin, Ind. Eng. Chem.
Res., 2006, 45, 5708–5715.
15. P. MäkiꢀArvela, I. Kubičková, M. Snåre, K. Eränen and D. Yu. Murzin, Energy Fuels, 2007,
21, 30–41.
8

Page 9 of 17
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C4CY00757C
16. B. Peng, Y. Yao, C. Zhao and J. A. Lercher, Angew. Chem., Int. Ed., 2012, 51, 2072–2075.
17. K. Murata, Y. Liu, M. Inaba and I. Takahara, Energy Fuels, 2010, 24, 2404–2409.
18. R. W. Gosselink, D. R. Stellwagen and J. H. Bitter, Angew. Chem., Int. Ed., 2013, 52,
5089–5092.
19. S. A. W. Hollak, R. W. Gosselink, D. S. van Es and J. H. Bitter, ACS Catal., 2013, 3,
2837−2844.
20. J. Han, J. Duan, P. Chen, H. Lou, X. Zheng and H. Hong, Green Chem., 2011, 13,
2561−2568.
21. J. X. Han, J. Z. Duan, P. Chen, H. Lou and X. M. Zheng, Adv. Synth. Catal., 2011, 353,
2577–2583.
22. H. Yoshitake and Y. Iwasawa, J. Catal., 1990, 125, 227–242.
23. D. A. G. Aranda, A. L. D. Ramos, F.B. Passos and M. Schmal, Catal. Today, 1996, 28,
119–125.
24 D. A. G. Aranda and M. Schmal, J. Catal. 171, 1997, 398–405.
25. S. J. Tauster, S. C. Fung and R. L. Garten, J. Am. Chem. Soc., 1978, 100, 170–171.
26. H. Yoshitake, K. Asakura and Y. Iwasawa, J. Chem. Soc., Faraday Trans., 1, 1989, 85,
2021−2024.
27. A. Maeda, F. Yakamawa, K. Kunimori and T. Uchijima, Catal. Lett., 1990, 4, 107–112.
28. T. Uchijima, Catal. Today, 1996, 28, 105–117.
29. A. Dandekar and M. A. Vannice, J. Catal., 1999, 183, 344–354.
30. W. Rachmady and M. A. Vannice, J. Catal., 2002, 207, 317−330.
31. P. Reyes, M.C. Aguirre, I. MelianꢀCabrera, M. Lopez Granados and J. L. G. Fierro, J. Catal.,
2002, 208, 229–237
32. J. Sa, J. Bernardi and J. A. Anderson, Catal. Lett., 2007, 114, 91–95..
33. V. A. D. O’Shea, M. C. A. Galvan, A. E. P. Parts, J. M. CamposꢀMartin and J. L. G. Fierro,
Chem. Commun., 2011, 47, 7131–7133.
34. S. Bonanni, K. AïtꢀMansour, H. Brune and W. Harbich, ACS Catal., 2011, 1, 385–389.
35. T. Iizuka, Y. Tanaka and K. Tanabe, J. Mol. Catal., 1982, 17, 381–389.
36. M. A. Vannice, Catal. Today, 1992, 12, 255–267.
37. A. Boffa, C. Lin, A. T. Bell and G. A. Somorjai, J. Catal., 1994, 149, 149–158.
38. A. T. Bell, J. Mol. Catal. A, 1995, 100, 1–11.
39. Y. Borodko and G. A. Somorjai, Appl. Catal. A, 1999, 186, 355−362.
40. P. M. Maitlis and V. Zanottib, Chem. Commun., 2009, 45, 1619–1634.
41. Y. Nakagawa, M. Tamura and K. Tomishige, ACS Catal., 2013, 3, 2655–2668
42. D. A. Ruddy, J. A. Schaidle, J. R. Ferrell III, J. Wang, L. Moens and J. E. Hensley, Green
Chem., 2014, 16, 454–490.
43 S. Nishiyama, T. Hara, S. Tsuruya and M. Masai, J. Phys. Chem. B, 1999, 103, 4431–4439.
9

Catalysis Science & Technology
Page 10 of 17
View Article Online
DOI: 10.1039/C4CY00757C
Table 1. Hydrogenation of lauric acid by 5 wt% metalꢀloaded catalysts.
O
cat. (1 mol%)
OH
5
+ H2
5
O
180 oC, 4 h
neat
OH
5
C12H25
8 bar
C11H23
O
1 mmol
Entry
Catalyst
Conv. (%)
Yield (%)
dodecane
dodecanol
ester a
1
2
3
4
5
6
7
8
9
Nb2O5
29
100
86
73
23
21
24
23
93
98
2
0
60
19
4 (8) c
1
0
7
0
21
31
29
4
Pt/Nb2O5
Ir/Nb2O5
Ru/Nb2O5
Re/Nb2O5
Pd/Nb2O5
Cu/Nb2O5
Ni/Nb2O5
12
9
0
1
0
4
1
0
0
0.4
16
13
1
0
0
b
Pt/Nb2O5
11
19
0
36
24
0
10
11
12
13
14
15
16
17
18
19
20
Pt/niobic acid
Pt/C
Pt/SiO2
1
0
0
0
Pt/TiO2
82
42
8
6
13
0
41
1
Pt/SiO2ꢀAl2O3
Pt/BEA
3
2
0
2
Pt/MFI
61
14
14
2
1
0
1
Pt/ZrO2
2
0
9
Pt/Al2O3
Pt/CeO2
Pt/MgO
1
0
1
1
0
2
44
1
3
0
a dodecyl dodecanoate.
b Preꢀreduced at 100 °C for 0.5 h.
c undecane
10

Page 11 of 17
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C4CY00757C
Table 2. Hydrogenation of lauric acid by Pt/Nb2O5.a
Solvent
Conv.(%)
Yield (%)
nꢀdodecane
1ꢀdodecanol
esterb
0
Solventꢀfree
toluene
100
100
100
100
100
97
99
90
42
31
43
40
0
9
2
oꢀxylene
mesitylene
diglyme
H2O
12
17
17
49
39
45
40
10
a Conditions: lauric acid (1 mmol), catalyst (1 mol%), solvent (0.2 g), 180 °C, 8 bar H2, 24 h.
b dodecyl dodecanoate.
Table 3. Hydrogenation of fatty acids by Pt/Nb2O5.a
Entry
fatty acid
Pt
T
t
Conv.
(%)
Yield (%)
alcohol
b
(mol%)
(°C)
(h)
nꢀCnH2n+2
ester
1
2 c
3 d
4
Lauric acid
1
1
180
180
180
180
250
250
250
250
260
24
24
24
24
24
24
24
48
24
100
100
100
100
100
100
100
100
100
99
99
99
88
90
89
99
96
93
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
Capric acid
Palmitic acid
Myristic acid
Oleic acid
1
5
0.2
0.2
0.2
0.07
1
6
7
8e
9f
Stearic acid
Tristearin
a 1 mmol fatty acid, solventꢀfree, 8 bar H2.
b Linear alkanes with the same chain length as the starting compound.
c1st reuse.
d 2nd reuse.
e 2 mmol stearic acid, 8 bar H2.
f 1 mmol tristearin, 50 bar H2, 1 mol% Pt cat. with respect to carboxyl group.
11

Catalysis Science & Technology
Page 12 of 17
View Article Online
DOI: 10.1039/C4CY00757C
Table 4. Heterogeneous catalysts for hydrogenation of stearic acid to octadecane.a
T
t
(h)
48
48
8
PH2
(bar)
8
Yield
(%)
TOF
(hꢀ1)
Catalyst
mol%
TON
ref.
(°C)
250
250
260
280
350
this study a
this study a
Pt/Nb2O5
Pt/SiO2
0.07
0.07
9.65
22.3
2.9
96
0 (7)b
97.3
29
1371
ꢀ
8
ꢀ
Ni/HBEA
Mo2C/C
40
10
50
1.3
3.4
4.6
10
16
21
18
4
75
0.8
23
Wꢀbased cat.
5
67
a Conditions: 2 mmol stearic acid, solventꢀfree, 8 bar H2.
b Yield of nꢀheptadecane.
Table 5. Hydrogenation of model compounds by Pt/Nb2O5.a
Entry
Substrate
Initial rate / mmol hꢀ1
nꢀdodecane
1ꢀdodecanol
esterb
1
2
lauric acid
1ꢀdodecanal
1ꢀdodecanol
esterb
5
15
13
8
17
0
ꢀ
31
ꢀ
3
ꢀ
4
0
ꢀ
5 c
esterification
ꢀ
ꢀ
63
a Conditions: 1 mmol substrate, solventꢀfree, 1 mol% Pt catalyst, 180 °C, 8 bar H2.
b dodecyl dodecanoate.
c Esterification of lauric acid (1 mmol) with 1ꢀdedecanol (1 mmol) at 180 °C in 1 atm N2.
100
200
300
400
T (oC)
Fig. 1 H2ꢀTPR profile for unreduced precursor of Pt/Nb2O5.
12

Page 13 of 17
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C4CY00757C
40
30
20
10
0
TH2 = 300 oC
D = 4.7±1.1 nm
0
2
4
6
8
10
Particle size / nm
Fig. 2 A representative TEM image and Pt particle size distribution of Pt/Nb2O5 (preꢀreduced at
300 oC). The volumeꢀarea mean diameter of Pt particle was 4.7 ± 1.1 nm.
linear CO
0.1
bridged CO
100 oC
300 oC
2100 2000 1900 1800
Wavenumber /cm-1
Fig. 3 IR spectra (40 oC) of CO adsorbed on Pt/Nb2O5 preꢀreduced at 100 and 300 oC.
13

Catalysis Science & Technology
Page 14 of 17
View Article Online
DOI: 10.1039/C4CY00757C
100
80
60
40
20
0
conv.
dodecane
ester
dodecanol
0
10
20
PH2 / bar
Fig. 4 Effect of H2 pressure on conversion and products yields for lauric acid hydrogenation by
Pt/Nb2O5 (preꢀreduced at 300 oC).
100
lauric acid
80
60
40
20
dodecane
ester
dodecanol
0
10
t / h
20
Fig. 5 Timeꢀconversion profile for lauric acid hydrogenation by Pt/Nb2O5 (preꢀreduced at 300
oC).
14

Page 15 of 17
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C4CY00757C
1445 cm-1
(Py on Nb5+)
0.01
1538 cm-1
(PyH+)
1550
1500
1450
1400
Wavenumber/ cm-1
Fig. 6 IR spectra of pyridine adsorbed on Pt/Nb2O5 at 150 oC.
1715 cm-1
0.1
SiO2
1660 cm-1
Nb2O5
1800
1600
1400
Wavenumber/ cm-1
Fig. 7 IR spectra of acetic acid adsorbed on Nb2O5 and SiO2 at 8 oC.
Scheme 1 Proposed reaction pathway for hydrodeoxygenation of lauric acid by Pt/Nb2O5.
15

Catalysis Science & Technology
Page 16 of 17
View Article Online
DOI: 10.1039/C4CY00757C
Scheme 2 Proposed cooperative mechanism of hydrodeoxygenation of carboxylic acids by
Pt/Nb2O5 in a partially SMSI state.
16

Page 17 of 17
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C4CY00757C
Table of Contents
Pt/Nb2O5 shows more than 60 times
higher TON than non-SMSI Pt catalysts
O
R
R
+ H2O
OH
and
previous
catalysts
for
+ H2
R
hydrodeoxygenation of stearic acid to
n-octadecane at 180 °C in 8 bar H2. Nb2O5
can acts as activation sites of carbonyl
groups.
High yield of n-alkane
TON = 1371
Pt/Lewis acid cooperation
SMSI state
δ+
HO
-
δ
H
H
O
Nbn+
Pt
1