Catalytic conversion of Jatropha oil to alkanes under mild conditions with a Ru/La(OH)3 catalyst

Article abstract of DOI:10.1039/c4gc02406k

The long-chain alkanes obtained from the hydrodeoxygenation of plant oils are ideal substitutes for diesel. In this work, a new efficient catalytic system was established for the conversion of plant oil to long-chain alkanes under mild conditions with a bi-functional Ru/La(OH)₃ catalyst. The hydrodeoxygenation of stearic acid was performed in an autoclave with Ru-based catalysts with different supports (HZSM-5, ZSM-5, SiO₂-Al₂O₃, SiO₂, ZrO₂, Mg(OH)₂, La(OH)₃, and La₂O₃). Among these catalysts, Ru supported on basic La(OH)₃ showed a remarkable catalytic performance for the reaction. Over 98% of long-chain alkanes were obtained with 100% conversion of stearic acid at 200 °C and 4 MPa H₂. When crude Jatropha oil was hydrogenated, about 80.7 wt% of long chain alkanes were obtained under the optimized conditions (200 °C, 4 MPa H₂, 8 h). The high efficiency of the Ru/La(OH)₃ catalyst could be due to a co-effect of the high hydrogenation activity of Ru and the basic La(OH)₃ support which can attract the acidic raw material. Additionally, the Ru/La(OH)₃ catalyst was recycled four times and maintained a good activity and stability. The reaction pathway was also explored by using stearic acid as a model compound. Hydrogenation-decarbonylation could be the main pathway to produce n-heptadecane, which has one carbon atom less than stearic acid.

Full text of DOI:10.1039/c4gc02406k

Green Chemistry
View Article Online
View Journal | View Issue
PAPER
Catalytic conversion of Jatropha oil to alkanes
under mild conditions with a Ru/La(OH) catalyst†
3
Cite this: Green Chem., 2015, 17,
888
2
a
a
b
a
a
Jian-hua Guo, Guang-yue Xu, Fei Shen, Yao Fu, Ying Zhang* and
a
Qing-xiang Guo
The long-chain alkanes obtained from the hydrodeoxygenation of plant oils are ideal substitutes for
diesel. In this work, a new efficient catalytic system was established for the conversion of plant oil to
long-chain alkanes under mild conditions with a bi-functional Ru/La(OH)
tion of stearic acid was performed in an autoclave with Ru-based catalysts with different supports
HZSM-5, ZSM-5, SiO –Al , SiO , ZrO , Mg(OH) , La(OH) , and La ). Among these catalysts, Ru sup-
ported on basic La(OH) showed a remarkable catalytic performance for the reaction. Over 98% of long-
chain alkanes were obtained with 100% conversion of stearic acid at 200 °C and 4 MPa H . When crude
Jatropha oil was hydrogenated, about 80.7 wt% of long chain alkanes were obtained under the optimized
conditions (200 °C, 4 MPa H , 8 h). The high efficiency of the Ru/La(OH) catalyst could be due to a co-
effect of the high hydrogenation activity of Ru and the basic La(OH) support which can attract the acidic
3
catalyst. The hydrodeoxygena-
(
2
2
O
3
2
2
2
3
2 3
O
3
2
2
3
3
Received 7th December 2014,
Accepted 2nd March 2015
raw material. Additionally, the Ru/La(OH) catalyst was recycled four times and maintained a good activity
3
and stability. The reaction pathway was also explored by using stearic acid as a model compound. Hydro-
DOI: 10.1039/c4gc02406k
genation–decarbonylation could be the main pathway to produce n-heptadecane, which has one carbon
www.rsc.org/greenchem
atom less than stearic acid.
sis of recent research focuses on the conversion of oil to
diesel-range alkanes, the second generation of biodiesel,
Introduction
The rapid depletion of fossil fuels and the tight carbon dioxide instead of esters.
emission limits necessitate the production of liquid fuels from
Hydrotreatment has been widely studied and carried out
new resources. Biomass, which is renewable, CO -neutral and efficiently for the transformation of plant oils. Sulfided Ni–Mo
2
abundant, is a prospective substitute. The long chain (C12–C22
)
and Co–Mo are conventional hydrotreating catalysts. However,
fatty acid compounds in the triglyceride structures make it the harsh reaction conditions and sulfur leaching hinder their
9
–18
possible to produce high quality transportation fuels from oils, applications.
Supported Pt, Pd, and Ni catalysts also show
which cannot be directly used as fuels due to their high vis- high activity and selectivity for the conversion of plant oils to
1
,2
16–27
cosity, low volatility and instability. Substantial research has alkanes.
Murata et al. prepared a Pt/HZSM-5 catalyst,
been carried out to convert plant oils into fuels. Fatty acid which showed a good activity for the hydrotreatment of Jatro-
esters, the first generation of biodiesel, are produced by the pha oil and vegetable oils to long-chain alkanes at 270–300 °C.
3
–5
transesterification of plant oils with alcohols.
However, the When Pt/HZSM-5 was modified with rhenium, a high oil/cat.
poor flow properties of the obtained biodiesel at low tempera- ratio (up to 10) was also applicable for the hydrotreatment
6
–8
19
tures limit its application in engines.
Since long-chain process with a 67 wt% yield of long-chain alkanes. Lercher’s
hydrocarbons are the major components in diesel, the empha- group reported that Ni/ZrO and Ni/HBeta were efficient and
2
2
0,21
stable catalysts.
In the presence of Ni/ZrO , 75 wt% yield of
2
liquid alkanes and 70 wt% yield of n-heptadecane were
obtained from hydrotreating microalgae oil at 270 °C through
a
Collaborative Innovation Center of Chemistry for Energy Materials, Anhui Province
a
hydrogenolysis/hydrogenation–decarbonylation route. In
Key Laboratory of Biomass Clean Energy and Department of Chemistry, University of
Science and Technology of China, Hefei, 230026, China.
addition, in the presence of Ni/Hbeta, the dehydration of
alcohol intermediates was the main path and led to 60 wt%
yield of octadecane from hydrotreating microalgae oil at
E-mail: zhzhying@ustc.edu.cn; Fax: +86-551-63606689; Tel: +86-551-63603463
b
Post-doctoral Workstation, China HuanQiu Contracting & Engineering Corp.,
Beijing, 100012, China. E-mail: shenfei@ hqcec.com
260 °C. Recently, carbon supported tungsten- or molybdenum-
†
Electronic supplementary information (ESI) available: The characterization
CO-chemisorptions, ICP, XRD and XPS) of the Ru-catalyst. See DOI: 10.1039/
c4gc02406k
based materials were also used as catalysts for converting
(
2
8–30
biomass-derived fatty acids to hydrocarbons at 350 °C.
2888 | Green Chem., 2015, 17, 2888–2895
This journal is © The Royal Society of Chemistry 2015

View Article Online
Green Chemistry
Paper
Since it is necessary to consider the energy input during 280 °C for 3 h. The flow rate of hydrogen and nitrogen was 10
liquid fuel production, the milder the conversion conditions, and 100 mL min− , respectively.
the higher the energy efficiency. In this study, we use sup-
1
ported Ru catalysts for the hydrotreatment of Jatropha oil, a
kind of renewable but inedible plant oil, under mild con-
Characterization of the catalysts
FT-IR spectra were recorded on a Nicolet 8700 FT-IR spectro-
meter. The samples were prepared as follows: 0.2 g of the
support (HZSM-5, ZSM-5, SiO –Al O , SiO , ZrO , Mg(OH) ,
ditions to produce
different supports (HZSM-5, ZSM-5, SiO
OH) , La(OH) , and La O ) were prepared and characterized
C
15–C18 alkanes. The catalysts with
2
–Al , SiO , ZrO , Mg
2
O
3
2
2
2
2
3
2
2
2
(
2
3
2 3
La(OH) , or La O ) was mixed with 20 mL of a 0.05 M n-hexane
3
2 3
by different techniques. Stearic acid was used as a model com-
pound to investigate the effect of the support, temperature,
hydrogenation pressure and solvent. The stability of the
optimal catalyst was tested. The reaction pathways for fatty
acids and Jatropha oil were also investigated.
solution of stearic acid and this was stirred at room tempera-
ture for 2 h. Then the solid was separated from the mixture by
centrifugation. After washing with n-hexane 20 times, the
2
residue was dried at 120 °C for 12 h in a N atmosphere.
Nitrogen adsorption measurements were performed using a
Coulter SA 3100 adsorption analyzer, which reports the adsorp-
tion isotherm, specific surface area and pore volume automati-
cally. The Brunauer–Emmett–Teller (BET) equation was used
to calculate the surface area over a range of relative pressures
between 0.05 and 0.20. The pore size was calculated from the
adsorption branch of the isotherms using the thermodynamic
based Barrett–Joyner–Halenda (BJH) method.
A micromeritics ASAP 2020 analyzer (Tristar II 3020M) was
used to measure the CO chemisorption isotherms. The dis-
persion was calculated according to the CO chemisorption
results.
Experimental
3 3 2
La(NO ) ·6H O (99.99% metals basis), stearic acid (≥98.0%,
AR), stearic alcohol (>99.0%, AR) and n-eicosane (>99.0%, AR)
were purchased from Aladdin Reagents (Shanghai) Co., LTD.
MgCl
2
·6H
2
O (≥98%, AR), ammonium hydroxide (AR) and SiO
2
were purchased from Sinopharm Chemical Reagent Co., Ltd.
Crude Jatropha oil was provided by Yunnan Shenyu New
Energy Co., Ltd. SiO –Al O and ZrO were obtained from
2 2 3 2
Saint-Gobain NorPro. ZSM-5 and HZSM-5 were purchased
from the catalyst plant of Nankai University. All chemicals
were obtained commercially and used without further
treatment.
XRD analysis was conducted on an X-ray diffractometer
(
TTR-III, Rigaku Corp., Japan) using Cu Kα radiation (λ =
1.54056 Å). The data were recorded over a 2θ range of 10–70°.
XPS data was obtained with an X-ray photoelectron spectro-
meter (ESCALAB250, Thermo-VG Scientific, USA) using mono-
Preparation of the catalysts
chromatized Al Kα radiation (1486.92 eV).
The characterization of the Ru/La(OH) catalyst is given in
the ESI.†
La(OH)
3
was prepared by a precipitation method. First, 4.33 g
3
of La(NO ) ·6H O was dissolved in 200 mL of water and stirred
3
3
2
for 2 h. Then 2 M ammonia hydroxide was added into the
mixture to adjust the solution pH to 10 and this was then
stirred at room temperature for another 2 h. Then the solid
Experimental procedure
was filtered off, washed and dried overnight at 105 °C. After In a typical test, 1 mmol of stearic acid (or 0.2 g of Jatropha
calcination at 700 °C for 4 h, the solid was stirred for 2 h in oil), 0.2 g of Ru/La(OH) , and 20 mL of n-hexane were added
water. La(OH) was obtained after filtration and dried at 40 °C into a 50 mL Parr reactor with a quartz lining. After purging
3
3
for 12 h.
2 2
the reactor with H , the reaction was conducted with 4 MPa H
The catalysts were prepared by a deposition precipitation (room temperature) at 200 °C for 4 h with a stirring speed of
method. A calculated amount of the support (HZSM-5, ZSM-5, 1000 rpm. The reactor was heated from room temperature to
−
1
SiO
persed in water and then a water solution of RuCl
added and the mixture was stirred at room temperature for cool it down to room temperature. The liquid, gas and solid
h. Subsequently, 2 M ammonia hydroxide was added into were taken out for further analysis. The liquid products were
2
–Al
2
O
3
, SiO
2
, ZrO
2
, Mg(OH)
2
or La(OH)
3
) was first dis- the reaction temperature at a ramp rate of 10 °C min . After
3
·3H O was the reaction, the reactor was put into water straight away to
2
2
the mixture to adjust the solution pH to 10 and this was then analyzed by a gas chromatograph (GC, Kexiao 1690) with an
stirred at room temperature for another 2 h. After filtration, OV1701 capillary column. Both injection and detection tem-
washing and drying overnight at 40 °C, the catalysts were peratures were 320 °C. The column temperature was kept at
2 2
reduced in a H and N atmosphere at 280 °C for 3 h. The flow 250 °C. n-Eicosane was used as the internal standard to deter-
−
1
rate of hydrogen and nitrogen was 10 and 100 mL min
respectively.
,
mine the product amount. N was used as a carrier gas and
the column head pressure was 0.1 MPa. The products were
2
As for Ru/La
persed in acetone and then added an acetone solution of (Agilent 5975C with a Triple-Axis Detector) with a HP-5
RuCl ·3H O was added and the mixture was stirred at room column. The column temperature was increased from 40 to
temperature for 2 h. After filtration and drying overnight at 180 °C at a ramp rate of 20 °C min and then to 280 °C at a
2 3 2 3
O , a calculated amount of La O was first dis- identified by a GC (Agilent 7890A)–mass spectrometer detector
3
2
−
1
0 °C, the catalyst was reduced in a H and N atmosphere at ramp rate of 5 °C min− . Both the injection and detection
1
4
2 2
This journal is © The Royal Society of Chemistry 2015
Green Chem., 2015, 17, 2888–2895 | 2889

View Article Online
Paper
Green Chemistry
temperatures were 320 °C. Highly pure helium was used as a Table 2 Comparison of stearic acid conversions over different catalysts
−1
a
at 200 °C
carrier gas and the flow rate was 1 mL min
.
The conversion and selectivity were calculated by mol%
when stearic acid was used as the starting reactant:
Sel./%
b
b
b
ꢀ
ꢁ
Entry Catalyst
Conv./% n-C17 n-C18
C
A
E
molar amount of stearic acid after reaction
molar amount of stearic acid in the starting material
Conv:ð%Þ ¼ 1 ꢀ
1
2
3
4
5
6
7
8
9
Ru/HZSM-5
Ru/ZSM-5
39.8
38.5
47.3
67.2
82.8
91.5
100
71.5
58.4
5.5
1
1.4
0.8
1.6
1.9
1.8
2.3
2.1
3.1
—
1.4 50.7 0.7
0.8 53.6
14.8 41.1 2.9
5.5 32.4 0.4
ꢁ
100%
—
Ru/SiO
Ru/SiO
2
–Al
O
2 3
29.3
26.9
60.6
78.5
95.9
91
2
Ru/ZrO
Ru/Mg(OH)2
Ru/La(OH)
Ru/La
La(OH)
2
2.5 27.5
—
—
—
—
—
molar amount of each product
molar amount of the converted stearic acid
Sel:ð%Þ ¼
ꢁ 100%
4.6
0.3
1.5
1.9
—
1.5
—
3
2
O
3
The yield was calculated by wt% when Jatropha oil was
used as the starting reactant:
3
1.1
—
a
2
Reaction conditions: 1 mmol stearic acid, 0.2 g Ru-catalyst, 4 MPa H ,
h, and stirring at 1000 rpm. C: cracking products, A:1-octadecanol,
weight of each product
weight of starting material
b
4
Yieldðwt%Þ ¼
ꢁ 100%
E: stearyl stearate.
The separated catalyst was dried at 40 °C after filtration and
sequential washing with acetone. During the catalyst stability
test, the catalyst was then reused without any further n-heptadecane, n-octadecane, 1-octadecanol, and cracking pro-
treatment.
ducts (C9–16 alkanes). The activity of Ru/HZSM-5 and Ru/ZSM-5
were very low under 200 °C (entries 1 and 2). The conversion
of stearic acid was 39.8% and 38.5%, respectively, while the
selectivity for long-chain alkanes was only 8.3% and 2.6%,
respectively, and the major product was 1-octadecanol (50.7%
Results and discussion
Analysis of Jatropha oil
2 2 3 2
and 53.6% selectivity, respectively). Ru/SiO –Al O and Ru/SiO
The composition of Jatropha oil is shown in Table 1. The main showed moderate catalytic activity, and the selectivity for
elementary composition of Jatropha oil is C (76.99%), H alkanes was 45.7% and 34.3%, respectively (entries 3 and 4).
(
11.77%), O (10.58%), and N (0.66%). The predominant fatty The selectivity for the cracking products was up to 14.8% when
acids in Jatropha oil are oleic acid (36.7%), linoleic acid using Ru/SiO –Al O as the catalyst. That is, Ru/SiO –Al O
2 3
2
2
3
2
(39.1%), palmitic acid (15.3%) and stearic acid (7.1%).
tends to produce lighter alkanes by C–C cleavage. The conver-
sion of stearic acid was 82.8% with 64.9% of long-chain
Catalyst screening
alkanes over the Ru/ZrO
, it can be seen that the Ru catalysts supported on basic
hydroxide supports (Mg(OH) and La(OH) ) showed high
2
catalyst (entry 5). From entries 6 and
7
Because Jatropha oil mainly consists of triglycerides and fatty
acids, stearic acid was selected as a model compound to test
the hydrotreatment process with Ru supported on different
carriers (the characterization of the catalysts is shown in
Table S1†). As shown in Table 2, the main hydrogenated
products obtained from stearic acid (C18 carboxylic acid) were
2
3
activity for converting stearic acid into long chain alkanes.
When stearic acid was hydrotreated in the presence of the Ru/
Mg(OH) catalyst at 200 °C and 4 MPa H , 78.5% of n-heptade-
2 2
cane, 2.3% of n-octadecane and 4.6% of cracking products
were obtained with 91.5% stearic acid conversion. The Ru/La
(
3
OH) catalyst exhibited a more outstanding activity: 95.9% of
n-heptadecane, 2.1% of n-octadecane and 0.3% of cracking
Table 1 Composition of Jatropha oil
products were obtained with 100% conversion. When La O
2
3
was employed as the support, the conversion decreased to
1.5% (entry 8). Pure La(OH) was also tested as a catalyst,
and showed very low catalytic activity. The conversion of
stearic acid was 58.4% and the yield of n-heptadecane was
1.1% (entry 9).
Composition
Content/%
7
3
Elementary composition
C
H
O
N
76.99
11.77
10.58
0.66
The high efficiency of the Ru/La(OH) catalyst could be due
3
a
Fatty composition
to a co-effect of the metal and support. Ru is an efficient
hydrogenation metal. The acidic raw material would be easily
adsorbed on the surface of the basic support due to their
special acid–base properties, and then be efficiently hydrogen-
ated to alkanes over Ru. To explore the interactions between
the supports and substrates, the supports were treated with
the stearic acid solution. The treated samples were analyzed by
FT-IR to check whether there is adsorbed stearic acid (Fig. 1).
Myristic
Palmitic
Palmitoleic
Stearic
Oleic
Linoleic
Arachidic
14 : 0
16 : 0
16 : 1
18 : 0
18 : 1
18 : 2
20 : 0
1.1
15.31
0.52
7.1
36.72
39.05
0.20
a
The nomenclature shows the number of carbon atoms and the
degree of instauration.
2890 | Green Chem., 2015, 17, 2888–2895
This journal is © The Royal Society of Chemistry 2015

View Article Online
Green Chemistry
Paper
stretching vibration of carboxylates. In the research performed
by Lercher’s group, it was proposed that stearic acid was
2
1
adsorbed at oxygen vacancies in ZrO
The new peaks in the spectra of the treated Mg(OH)
and La O supports are also attributed to the CvO stretching
2
to form carboxylates.
2
, La(OH)
3
,
2
3
vibration of carboxylates, so stearic acid also existed as carboxy-
lates on the basic supports. The new peaks appeared in the
treated La(OH) (1540 and 1467 cm− ) and La O (1539 and
1
3
2 3
468 cm− ) supports were almost at the same positions, while
1
1
these peaks were different in the treated Mg(OH)
(
2
support
1567 and 1468 cm− ), indicating that the interactions between
stearic acid and the support were different for different metal
hydr)oxides. The different interactions may cause the big
1
(
difference in the selectivity for long-chain alkanes with the La
or Mg catalyst. As shown in Table 2, the product distribution
3
with Ru/La(OH) (95.9% n-heptadecane, 2.1% n-octadecane,
and 0.3% cracking products) and Ru/La O (91% n-heptade-
2
3
cane, 3.1% n-octadecane, 1.5% 1-octadecanol, and 1.5% crack-
ing products) were similar. However, the product distribution
was quite different when Mg(OH)2 was employed as the
support (78.5% n-heptadecane, 2.3% n-octadecane, 1.9%
1-octadecanol, and 4.6% cracking products). Compared with
La(OH) , the structure of La O changed greatly after stearic acid
3
2 3
−
1
treatment. The peaks at about 3607 and 648 cm , the charac-
teristic peaks for La(OH)
3
1
3
,
increased significantly. In other
words, La O is unstable and reacts with stearic acid to form
2
3
La(OH)
content in the La
La(OH) support (6.19%) (see Table S2†). Combining this with
3
during the treatment process. Moreover, the carbon
2 3
O
support (2.71%) was lower than in the
3
the fact that the Ru content and dispersion on Ru/La
lower than on Ru/La(OH) , this may cause the lower activity of
the La O supported catalyst. Based on all the above obser-
2 3
O were
3
2
3
vations, in the reaction system, the substrate should be
adsorbed on the basic support to form carboxylates and then
it can be easily catalyzed by Ru to produce alkanes.
Fig. 1 IR spectra of the supports. The dotted lines represent the
untreated supports and the solid lines represent the treated supports.
Effect of the reaction conditions
The effect of the reaction conditions was investigated. The
The peaks at 2961 cm−1 (the C–H stretching vibration of results are shown in Table 3. At a lower temperature (180 °C),
−
1
–
CH ), 2920 and 2877 cm (the C–H stretching vibration of only 50.2% of stearic acid was converted and the major
3
−1
–CH
2
), 1701 cm
(the CvO stretching vibration) and product was 1-octadecanol (40.1% selectivity). While at a
46 cm− (the OH wagging vibration of –COOH) are the charac- higher temperature (220 °C), the selectivity of n-heptadecane
1
9
teristic peaks of stearic acid. In the IR spectra of the treated decreased to 91.6% and more cracking products (4.2% selecti-
ZrO , Mg(OH) , La(OH) , and La supports, the peaks at vity) were obtained. A temperature around 200 °C was suitable
961, 2920, and 2877 cm were much stronger than those of for the conversion of stearic acid to long-chain alkanes, and
the treated HZSM-5, ZSM-5, SiO –Al O , and SiO supports, the conversion could be fulfilled in 4 hours. The hydrotreating
2
2
3
2 3
O
−
1
2
2
2
3
2
which indicates that more organic species were adsorbed on pressure also influenced the stearic acid conversion. At a lower
these supports. However, the peaks at 1701 and 946 cm− dis- pressure (2 MPa H ), the conversion was 70.6%, and the
1
2
appeared in the spectra of the treated supports. This indicates selectivity of n-heptadecane was 82.7%. While at a higher
that no free stearic acid exists on the supports. Some new pressure (6 MPa H ), more n-octadecane (3.8%) and cracking
2
peaks were found for the treated ZrO , Mg(OH) , La(OH) , and products (5.0%) were obtained. Both a higher temperature and
2
2
3
La O supports compared with the untreated samples, while higher pressure tend to produce cracking products.
2
3
new peaks were barely found for the treated HZSM-5, ZSM-5,
The effect of solvents was also investigated. When nonane
SiO –Al O , and SiO supports. Stearic acid could exist as a or dodecane was used as the solvent, the product distribution
2
2
3
2
new form on these supports. For the ZrO sample, the new was similar to when hexane was used as the solvent (Table 3,
2
−
1
peaks at 1536 and 1467 cm could be attributed to the CvO entries 1, 6 and 7). Stearic acid can also be hydrogenated in
This journal is © The Royal Society of Chemistry 2015
Green Chem., 2015, 17, 2888–2895 | 2891

View Article Online
Paper
Green Chemistry
a
3
Table 3 Stearic acid conversion over the Ru/La(OH) catalyst
Sel./%
b
b
b
Entry
Solvent
T/°C
2
H /MPa
Conv./%
n-C17
n-C18
C
A
E
1
2
3
4
5
6
7
8
Hexane
Hexane
Hexane
Hexane
Hexane
Nonane
Dodecane
200
180
220
200
200
200
200
200
4
4
4
2
6
4
4
4
100
50.2
100
70.6
100
100
100
93.4
95.9
25.9
91.6
82.7
92.6
96.6
97.9
83.6
2.1
1.4
2.8
1.7
3.8
1.7
1.9
4.9
0.3
0.4
4.2
1.3
5.0
0.6
0.2
1.2
—
40.1
—
4.4
—
—
—
0.1
—
0.1
—
—
—
0.8
—
0.9
2
H O
a
b
Reaction conditions: 1 mmol stearic acid, 0.2 g Ru/La(OH)
E = stearyl stearate.
3
, 4 MPa H
2
, 4 h, and stirring at 1000 rpm. C = cracking products, A = 1-octadecanol,
aqueous solution, although it is insoluble in water (Table 3,
entry 8). The conversion (93.4%) was lower than that in alkane
solvents, and the n-heptadecane selectivity decreased to 83.6%
and the n-octadecane selectivity increased to 4.9%. In aqueous
solution, the conversion and selectivity of n-heptadecane both
decreased.
Conversion of Jatropha oil
Scheme 1 Pathways for the deoxygenation of a fatty acid.
Crude Jatropha oil was hydrotreated in the presence of the Ru/
La(OH) catalyst at 200 °C and 4 MPa H . The time-course of
3
2
the product distribution for the transformation of Jatropha oil
is shown in Fig. 2. Fatty acids (hexadecanoic acid and stearic was converted to liquid alkanes. The total carbon balance was
acid) and the corresponding alcohols (1-hexadecanol and 97%. It can be seen that the Ru/La(OH) catalyst is effective for
3
1
-octadecanol) were detected in the first 4 h. The yield of fatty converting Jatropha oil to alkanes under mild conditions
acids and fatty alcohols first increased and then decreased to (Scheme 1).
. The yield of long-chain alkanes, mainly n-heptadecane and
Moreover, the recyclability of the Ru/La(OH)3 catalyst was
0
n-pentadecane, increased with time. After 8 h, 80.7 wt% examined. A batch of Ru/La(OH) was repeatedly used for the
3
alkanes were obtained and odd carbon number chain alkanes conversion of Jatropha oil. As shown in Fig. 3, the catalyst
were the main products (64.8 wt% yield of n-heptadecane and maintained a good activity after being used four times.
1
1.7 wt% yield of n-pentadecane). The collected gas contained Although the yield of long-chain alkanes decreased, it was still
th
5
.2 wt% yield of CH
4
, 2.7 wt% yield of C
3
H
8
and trace C
2
H
6
.
around 70 wt% in the 4 run. The yields of n-hexadecane and
Based on the carbon content, 89.1% of carbon in Jatropha oil n-octadecane slightly increased in runs 2–4.
The fresh and used catalysts were characterized. There was
no significant difference in the XRD patterns of the Ru/La
Fig. 2 Product distribution for the transformation of Jatropha oil over
Ru/La(OH) at 200 °C as a function of time. Reaction conditions: 0.2 g
Jatropha oil, 0.2 g Ru/La(OH) , 200 °C, and stirring at 1000 rpm.
3
3
Fig. 3 The performance of the catalyst in the recyclability experiments.
2892 | Green Chem., 2015, 17, 2888–2895
This journal is © The Royal Society of Chemistry 2015

View Article Online
Green Chemistry
Paper
Table 4 Characteristics of the Ru/La(OH)
3
catalyst before and after
being used four times
BET surface
area (m g
Pore volume
Pore size
(nm)
2
−1
3
−1
)
(cm g
)
Ru/wt%
C/%
Fresh
Used
15.2
6.0
0.077
0.040
19.1
26.7
1.56
1.43
0.65
4.69
3
(OH) catalyst (Fig. S4†). As shown in Table 4, after using the
catalyst four times, the surface area and pore volume of the
Fig. 4 The conversion of stearic acid. Reaction conditions: 1 mmol
Ru/La(OH) catalyst decreased, while the pore size increased.
3
n-octadecanol, 0.2 g Ru/La(OH)
000 rpm.
3 2
, 200 °C, 4 MPa H , and stirring at
The content of Ru decreased from 1.56% to 1.43%. The carbon
content in the catalyst increased from 0.65% to 4.69%. Accord-
ing to all of the characterization data, the activity decrease of
the catalyst could be mainly due to the pore structure change,
Ru leaching and the active site coking.
1
Octadecanal was not detected during the conversion
because it was unstable and converted to either n-heptadecane
or 1-octadecanol. 1-Octadecanol was tested in a separate exper-
iment on Ru/La(OH) under the same conditions to explore
3
Mechanism of the reaction
The deoxygenation of fatty acids can occur in three ways: de- the conversion process (Fig. 5). The yield of n-heptadecane
carbonylation [Scheme 1, eqn (1)], decarboxylation [Scheme 1, increased almost linearly with 1-octadecanol conversion, but
3
2,33
eqn (2)], and hydrodeoxygenation [Scheme 1, eqn (3)].
only trace n-octadecane (<0.7%) was formed via the hydro-
Taking stearic acid as an example, as the major products deoxygenation reaction. 1-Octadecanol mainly underwent the
2
1
were alkanes with one carbon atom less (n-heptadecane), de- dehydrogen–decarbonylation process,
carbonylation [eqn (1)]/decarboxylation [eqn (2)] could be the octadecanal, and then decarbonylating to n-heptadecane.
main reaction pathways when catalyzed by Ru/La(OH) . Very
According to the discussion on the deoxygenation process
dehydrogenating to
3
few stearic acid molecules were converted through hydrodeoxy- of stearic acid, the possible pathway for Jatropha oil conversion
genation [eqn (3)] since it was difficult for the basic catalyst to is shown in Scheme 2. The first step is the hydrogenation of
convert alcohols to alkanes through dehydration and hydro- unsaturated carbon bonds, and then the obtained saturated
genation. Because hydrogen is not necessary for the decarboxy- triglycerides were hydrogenolyzed to propane and saturated
lation step, one reaction was conducted in a N atmosphere. fatty acids. This step can be derived from the following obser-
2
However, only 0.11% n-heptadecane, coming from the decar- vations. From GC-MS analysis on the products obtained after
boxylation of stearic acid, was detected after the reaction. 2 h, a few saturated fatty acid isopropyl/propyl esters and
2
Trace CO was also detected in the gas phase, therefore the stearic acid were detected, which could come from the hydro-
decarboxylation step existed but was not the main deoxygena- genolysis of triglycerides. No unsaturated fatty acids or esters
tion process under these conditions. Moreover, from the data were detected. Then the saturated fatty acids were mainly
in Table 3, the hydrogen pressure influenced the reaction. hydrogenated to the corresponding aldehyde (RCHO).
Therefore, stearic acid was converted mainly through the de-
carbonylation process, indicating that Ru/La(OH)3 was more
conducive to catalyze the decarbonylation reaction rather than
hydrodeoxygenation or decarboxylation reactions. There was
no CO but CH , trace CO and trace C H in the gas phase col-
4
2
2
6
lected after the reaction. The CH
nation of CO.
4
could come from the metha-
Another potential source of organic gas
CH4 and C H ) could be from the C–C cleavage of alkanes
1
7–19,34
(
2
4
over the Ru-catalyst.
To further explain the decarbonylation process, the conver-
sion of stearic acid was explored with Ru/La(OH)3 (Fig. 4).
Trace n-octadecane and cracking products (<0.9%) were
detected. The yield of n-heptadecane continuously increased
over 4 h to 95.9% with increasing conversion, while the yield
of 1-octadecanol increased gradually to 36.4% and then
decreased at higher conversion. It can be inferred that 1-octa-
decanol was an intermediate product.
Fig. 5 The conversion of n-octadecanol. Reaction conditions: 1 mmol
n-octadecanol, 0.2 g Ru/La(OH) , 200 °C, 4 MPa H , and stirring at
3 2
1000 rpm.
This journal is © The Royal Society of Chemistry 2015
Green Chem., 2015, 17, 2888–2895 | 2893

View Article Online
Paper
Green Chemistry
Scheme 2 The reaction pathway for the hydrotreatment of Jatropha oil.
However, the aldehyde was not detected because it was easily
converted to more stable compounds – the linear alkane with
one carbon atom less (RH) via a decarbonylation process and
5 H. Fukuda, A. Kondo and H. Noda, J. Biosci. Bioeng., 2001,
92, 405.
6 G. W. Huber, P. O’Connor and A. Corma, Appl. Catal., A,
2007, 329, 120.
2
the corresponding fatty alcohol (RCH OH) via a hydrogenation
process. Most fatty alcohol molecules were converted to the
linear alkane (RH) via a dehydrogenation–decarbonylation
process, and very few of them were converted to the linear
7 M. Stumborg, A. Wong and E. Hogan, Bioresour. Technol.,
1996, 56, 13.
8 P. Priecel, D. Kubička, L. Čapek, Z. Bastl and P. Ryšánek,
Appl. Catal., A, 2011, 397, 127.
alkane with the same carbon number (RCH ) via hydrodeoxy-
3
genation. Meanwhile, the decarboxylation process also existed
but can almost be ignored.
9 D. Kubička, P. Šimáček and N. Žilková, Top. Catal., 2009,
52, 161.
1
1
0 D. Kubička, M. Bejblová and J. Vlk, Top. Catal., 2010, 53, 168.
1 D. Kubička and L. Kaluža, Appl. Catal., A, 2010, 372, 199.
Conclusion
12 D. Kubička and J. Horáček, Appl. Catal., A, 2011, 394, 9.
1
3 R. Sotelo-Boyás, Y. Liu and T. Minowa, Ind. Eng. Chem. Res.,
011, 50, 2791.
In summary, a new efficient catalytic system was established
for the hydrodeoxygenation of Jatropha oil to long-chain
alkanes under mild conditions with a bi-functional Ru/La
2
14 R. Tiwari, B. S. Rana, R. Kumar, D. Verma, R. Kumar,
R. K. Joshi, M. O. Garg and A. K. Sinha, Catal. Commun.,
3 2
(OH) catalyst. At 200 °C and 4 MPa H , a relatively high yield
2011, 12, 559.
(80.7 wt%) of long chain alkanes was achieved. After using
1
1
5 D. Kubička and J. Horáček, Appl. Catal., A, 2011, 394, 9.
6 P. Šimáček, D. Kubička, G. Šebor and M. Pospíšil, Fuel,
four times, about 70 wt% of alkanes was still obtained under
the same conditions. It was verified that hydrogenation–decar-
bonylation could be the main pathway to produce long-chain
alkanes from fatty acids. The La(OH) support could adsorb
3
the carboxyl group of the fatty acid, and thus promote the
hydrotreatment over the supported Ru catalyst.
2009, 88, 456.
1
1
1
2
2
2
2
2
7 B. Donnis, R. Egeberg, P. Blom and K. Knudsen, Top.
Catal., 2009, 52, 229.
8 S. Kovács, T. Kasza, A. Thernesz, I. W. Horváth and
J. Hancsók, Chem. Eng. J., 2011, 176–177, 237.
9 K. Murata, Y. Liu, M. Inaba and I. Takahara, Energy Fuels,
2
010, 24, 2404.
0 B. Peng, Y. Yao, C. Zhao and J. A. Lercher, Angew. Chem.,
012, 124, 2114, (Angew. Chem., Int. Ed., 2012, 51, 2072).
Acknowledgements
2
The authors are grateful to the National Basic Research
Program of China (2012CB215306), NSFC (21325208,
1 B. Peng, X. Yuan, C. Zhao and J. A. Lercher, J. Am. Chem.
Soc., 2012, 134, 9400.
2 C. Zhao, T. Brück and J. A. Lercher, Green Chem., 2013, 15,
2
1172209) and the Fundamental Research Funds for the
Central Universities for the financial support.
1720.
3 E. Santillan-Jimenez, T. Morgan, J. Lacny, S. Mohapatra
and M. Crocker, Fuel, 2013, 103, 1010.
4 A. T. Madsen, E. H. Ahmed, C. H. Christensen,
R. Fehrmann and A. Riisager, Fuel, 2011, 90, 3433.
Notes and references
1
S. Lestari, P. Mäki-Arvela, J. Beltramini, G. Q. M. Lu and
D. Y. Murzin, ChemSusChem, 2009, 2, 1109.
A. Srivastava and R. Prasad, Renewable Sustainable Energy
Rev., 2000, 4, 111.
25 B. Peng, C. Zhao, S. Kasakov, S. Foraita and J. A. Lercher,
Chem. – Eur. J., 2013, 19, 4732.
26 W. Song, C. Zhao and J. A. Lercher, Chem. – Eur. J., 2013,
19, 10060.
2
3
4
G. Knothe, Green Chem., 2011, 13, 3048.
Y. C. Sharma, B. Singh and J. Korstad, Green Chem., 2011, 27 K. Kon, W. Onodera, S. Takakusagi and K. Shimizu, Catal.
3, 2993. Sci. Technol., 2014, 4, 3705.
1
2894 | Green Chem., 2015, 17, 2888–2895
This journal is © The Royal Society of Chemistry 2015

View Article Online
Green Chemistry
Paper
2
2
3
3
8 J. Han, J. Duan, P. Chen, H. Lou and X. Zheng, Adv. Synth. 32 R. W. Gosselink, S. A. W. Hollak, S. W. Chang, J. van
Catal., 2011, 353, 2577.
9 R. W. Gosselink, D. R. Stellwagen and J. H. Bitter, Angew.
Chem., Int. Ed., 2013, 52, 5089.
0 S. A. W. Hollak, R. W. Gosselink, D. S. van Es and
J. H. Bitter, ACS Catal., 2013, 3, 2837.
Haveren, K. P. de Jong, J. H. Bitter and D. S. van Es, Chem-
SusChem, 2013, 6, 1576–1594.
33 M. Snåre, I. I. Kubičková, P. Mäki-Arvela, K. i. Eränen
and D. Yu. Murzin, Ind. Eng. Chem. Res., 2006, 45,
5708.
1 M. Aghazadeh, B. Arhami, A. M. Barmi, M. Hosseinifard, 34 S. Tada, R. Kikuchi, A. Takagaki, T. Sugawara, S. T. Oyama
D. Gharailou and F. Fathollahi, Mater. Lett., 2014, 115, 68–71. and S. Satokawa, Catal. Today, 2014, 232, 16.
This journal is © The Royal Society of Chemistry 2015
Green Chem., 2015, 17, 2888–2895 | 2895