Palladium-catalyzed transformation of renewable oils into diesel components

Article abstract of DOI:10.1002/adsc.201000195

A size-controlled palladium nanoparticle catalyst prepared by adsorption of colloidal palladium nanoparticles on barium sulfate is efficient and highly selective in transforming vegetable oils into diesel-like fuel. Preliminary kinetic investigations using model compounds indicated that decarboxylation of aliphatic esters on palladium in a hydrogenrich atmosphere showed a zero-order rate. Hydrogen temperature-programmed desorption measurements revealed that the high-temperature desorption of hydrogen species might be the rate-determining step.

Full text of DOI:10.1002/adsc.201000195

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
COMMUNICATIONS
DOI: 10.1002/adsc.201000195
Palladium-Catalyzed Transformation of Renewable Oils into
Diesel Components
Junxing Han,a Hui Sun,a Jinzhao Duan,a Yuqi Ding,a Hui Lou,a,*
and Xiaoming Zhenga
a
Key Lab of Applied Chemistry of Zhejiang Province, Department of Chemistry, Zhejiang University, Hangzhou 310028,
Peopleꢀs Republic of China
Fax : (+86)-571-8827-3593; phone: (+86)-571-8827-3593; e-mail: hx215@zju.edu.cn
Received: March 12, 2010; Revised: June 21, 2010; Published online: August 16, 2010
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/adsc.201000195.
Abstract: A size-controlled palladium nanoparticle
catalyst prepared by adsorption of colloidal palladi-
um nanoparticles on barium sulfate is efficient and
highly selective in transforming vegetable oils into
diesel-like fuel. Preliminary kinetic investigations
using model compounds indicated that decarboxyla-
tion of aliphatic esters on palladium in a hydrogen-
rich atmosphere showed a zero-order rate. Hydro-
gen temperature-programmed desorption measure-
ments revealed that the high-temperature desorp-
tion of hydrogen species might be the rate-deter-
mining step.
of decarboxylation and decarbonylation has been
studied by Murzin[10] and has been patented by sever-
al companies.[11–15] In our previous work the Pd/BaSO4
catalyst (conventional Rosenmund catalyst) showed a
selectivity of 98% to decarboxylates in converting
model methyl ester compounds into alkanes, but the
TOF was not satisfactory. Herein, we describe that
supported palladium nanoparticles prepared by ad-
sorption of colloidal Pd nanoparticles on BaSO4 are
efficient catalysts for the conversion of vegetable oils
into diesel-like paraffins.
Colloidal Pd nanoparticles were prepared by reduc-
ing the precursor Pd2+ ions with ethanol in the mi-
celles of hydrophilic block copolymers poly(N-vinyl-
2-pyrrolidone) (PVP). The as-synthesized Pd nanopar-
ticles were then supported onto BaSO4, and the cata-
lyst obtained was denoted as PdNP/BaSO4. Prepared
Keywords: esters; green chemistry; heterogeneous
catalysis; hydrogenation; nanoparticles; palladium
Pd
ACHTUNGTRENN(UGN PVP) colloids and PdNP/BaSO4 catalyst were
characterized by HR-TEM (see Figure 1), EDX and
Diminishing fossil fuel reserves and growing concerns UV-Vis (see Supporting Information, Figure S1). The
about global warming make the production of trans- TEM micrographs and corresponding histogram
portation fuels from renewable resources a hot showed that PdACTHNUTRGENUG(N PVP) colloids were monodisperse
domain all over the world.[1–4] Vegetable oils com- with a narrow size distribution and a maximum in the
posed primarily of triglycerides exist widespread in histogram at 6.1 nm (see Figure 1 a, inset, top right).
nature. Transforming vegetable oils into alkanes with- The HR-TEM image (see Figure 1 a, inset, bottom
out cracking may be a very important way to provide right) confirmed the crystalline nature of the Pd
high-grade renewable diesel-like fuels such as aviation nanoparticles, and the d spacing of 0.22 nm obtained
fuel and jet fuel in the near future. The key problem from the fringes of the lattice corresponded with the
at present is still to remove oxygen atoms of renewa- (111) lattice spacing of face-centered cubic (fcc) Pd.
ble oils present in the form of esters.
The resulting PdACTHNGUTERN(UNG PVP) colloids could be uniformly
According to the results of previous investiga- dispersed on BaSO4, and particle size and morpholo-
tions,[5–9] hydrodeoxygenation, decarboxylation and gy were identical to those of the unsupported colloids
decarbonylation were all participating in the deoxyge- (see Figure 1, b). No diffraction peaks corresponding
nation process of esters. In comparison with hydro- to Pd were observed in the XRD pattern (see Sup-
deoxygenation, decarboxylation and decarbonylation porting Information, Figure S2) indicating the high
require less hydrogen consumption and will be more dispersion of Pd nanoparticles and the low content of
promising from the industrial point of view. Therefore palladium compared with that of the support. The
the deoxygenation route of vegetable oils in the form
Adv. Synth. Catal. 2010, 352, 1805 – 1809
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1805

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Junxing Han et al.
COMMUNICATIONS
Figure 1. (a) TEM image of Pd nanoparticles with histogram showing the size distribution for the Pd colloids (inset, top
right), and HR-TEM image of Pd nanoparticles (inset, bottom right). (b) TEM image of Pd nanoparticles supported on
BaSO4.
specific area of PdNP/BaSO4 catalyst was 22.5 m2 gÀ1 hexane as solvent, and 0.797 g energy-rich paraffins
as determined by N2 adsorption.
per gram of sunflower oil were produced. The theo-
Initially the catalytic performance of the PdNP/ retical output was 0.81 g paraffins per gram of sun-
BaSO4 catalyst was examined using sunflower oil as flower oil based on triglycerides. Thus, the yield of
substrate, whereby, according to Peterson,[4] sunflower diesel components was over 95%, which is far better
oil is one of the greatest potential feedstocks for fuel than that achieved with the conventional Rosenmund
production, and the results are shown in Table 1. The catalyst (see Table 1, entries 2, 10 and 11). The TOF
optimal yield was obtained in the presence of 10 mL for PdNP/BaSO4 catalyst was 390 hÀ1 versus 170 hÀ1
Table 1. Performance of the PdNP/BaSO4 catalyst under different reaction conditions.
Entry
Temp. [8C]
Solvent amount [mL]
Product distribution[a]
Yield [ggÀ1][b]
C15
C16
C17
C18
C19
C21
1
280
280
280
280
280
280
280
270
300
270
290
0
Solid yellow wax
trace
2
10
10
20
10
10
10
10
10
10
10
0.0448
0.0238
0.0318
0.0522
0.0377
0.0674
0.0321
0.0462
0.0402
0.0457
0.0046
0.0060
0.0026
0.0033
0.0053
0.0152
0.0032
0.0033
0.0059
0.0026
0.6892
0.3670
0.5066
0.7023
0.5247
0.5304
0.5118
0.7075
0.5245
0.5894
0.0486
0.0785
0.0229
0.0258
0.0610
0.1683
0.0382
0.0237
0.0441
0.0165
0.0048
0.0041
0.0042
0.0048
0.0047
0.0108
0.0043
0.0051
0.0039
0.0040
0.0054
0.0039
0.0048
0.0057
0.0042
0.0062
0.0044
0.0057
0.0029
0.0046
0.7974
0.4833
0.5729
0.7941
0.6376
0.7983
0.5940
0.7915
0.6215
0.6628
3[c]
4
5[d]
6[c,d]
7[e]
8
9
10[f]
11[f]
[a]
Cn represents the number of carbon atoms of the alkanes.
Sum amount of paraffins per gram of sunflower oil.
PdNP/BaSO4 catalyst was used successively for the second cycle without any treatment.
PdNP/BaSO4 catalyst was calcined and then reduced in hydrogen.
[b]
[c]
[d]
[e]
The deactivated PdNP/BaSO4 catalyst which was successively used for two cycles and then calcined in air and reduced in
hydrogen.
Conventional Rosenmund catalyst.
[f]
1806
asc.wiley-vch.de
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2010, 352, 1805 – 1809

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Palladium-Catalyzed Transformation of Renewable Oils into Diesel Components
for the conventional Rosenmund catalyst. The conver-
To confirm the possibility as to whether the leach-
sion of sunflower oil measured by HPLC and GC-MS ing Pd atoms was responsible for the homogeneous
was approximately 100% (see Supporting Informa- catalytic activity, the solid catalyst was filtered from
tion, Figure S3). When the catalytic reaction was per- the liquid phase after 3 h of reaction, the filtrate was
formed without solvents, only a solid yellow wax, in- returned to a clean autoclave and allowed to react
soluble in hexane, DMF, methanol and water, was under the same reaction conditions for another 2 h,
produced.[16] Sunflower oil mainly consists of unsatu- but no further transformation of esters was observed.
rated oleic acid and linoleic acid and their triglycer- The result indicates that the decarboxylation of esters
ides. Hydrogenation of the C=C double bonds in the is a heterogeneous catalytic reaction on the surface of
molecules of sunflower oil took place at 180–2008C, the Pd nanoparticles.
which made the liquid oils solidify and inhibited con-
X-ray photoelectron spectra (XPS) of the support-
tact of hydrogen with the catalyst. Therefore, the sub- ed Pd catalysts were recorded to determine the Pd ox-
sequent decarboxylation reaction could not take place idation states. The spectra obtained for uncalcined,
effectively and only few paraffins was produced, and calcined and then reduced in hydrogen, and calcined
the result agreed well with that of our reported without further reduction palladium catalysts are col-
work.[9] Increasing the solvent amount to 20 mL re- lected in Figure 2. The XPS spectrum of freshly pre-
strained the conversion of sunflower oil (see Table 1, pared Pd nanoparticles displayed signals for palladi-
entry 4). It was speculated that the high diffusivity of um(0) 3d5/2 (Eb =335.0 and 340.3 eV). After calcina-
supercritical hexane could reduce mass-transfer resist- tion at 3008C in air, only the PdO peaks (Eb =336.3
ance and the strong solvent power of the solvent en- and 341.6 eV) were detected. Upon reduction in hy-
hanced the desorption rate of alkanes generated on drogen, the Pd 3d5/2 shifted to 335.0 eV designated to
the catalyst surface, while excess hexane inhibited the palladium(0). The XPS experiments confirmed the
conversion of the volume-expansion decarboxylation conclusion that palladium(0) provided active sites for
reaction of esters.[7] And so the yield dropped down the decarboxylation of renewable oils, although PdCl2
when 20 mL of hexane were used. Decarboxylation of showed a high catalytic activity for decarbonylation of
aliphatic acid methyl and ethyl esters could be acyl chlorides.[18]
achieved with the conventional Rosenmund catalyst
With methyl palmitate as a model compound, we
in the presence of hydrogen at 2708C.[9] But the de- found that the selectivity of decarboxylates was a
carboxylation temperature should be increased to time-independent but temperature-dependent param-
2808C on account of the steric situation of triglycer- eter (see Supporting Information, Figure S6). Selec-
ides. Further raising the reaction temperature to tivity of decarboxylates was maintained above 98% as
3008C increased the conversion of intermediates, but the reaction proceeded, while on an increase of the
the yield of diesel components varied little. Hence, reaction temperature, the proportion of decarboxy-
2808C was selected for the following study. Li lates raised.[19] Preliminary decarboxylation kinetics
et al.[16,17] examined the effect of the capping agents in were investigated with two kinds of different initial
colloidal solution and found that strong adsorption of reactant concentration. Both of them manifested the
the stabilizers on the active sites of the nanoparticles same rate constant, which indicated that the decar-
resulted in a loss of catalytic efficiency. Our results
manifested that washing the supported PdACHTNUTRGNE(UNG PVP) col-
loids with water and ethanol could remove most of
the organic stabilizers capping the surface of Pd cores,
and this conclusion was confirmed by FT-IR and
TGA (see Supporting Information, Figures S4 and
S5) as well as by comparing the catalytic performance
of the uncalcined PdNP/BaSO4 catalyst with that of
the calcined catalyst (see Table 1, entries 2 and 5).
Both uncalcined and calcined PdNP/BaSO4 catalysts
exhibited an obvious deactivation during the succes-
sively recyclable tests (see Table 1, entries 3 and 6).
However, the deactivated catalyst could regain its ac-
tivity after simple calcination in air at 3008C (see
Table 1, entry 7). TGA measurements manifested that
it was not the carbon deposit that resulted in the de-
activation of the PdNP/BaSO4 catalyst. We speculate
that low-boiling-point compounds may be adsorbed
on the outer surface of the PdNP/BaSO4 catalyst,
which inhibits the decarboxylation activity.
Figure 2. XPS of the Pd 3d region of the PdNP/BaSO4 cata-
lyst: (a) uncalcined, (b) calcined and then reduced in hydro-
gen, and (c) calcined but not reduced.
Adv. Synth. Catal. 2010, 352, 1805 – 1809
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1807

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Junxing Han et al.
COMMUNICATIONS
tivity of 99% (see Supporting Information, Fig-
ure S8).
In conclusion, the PdNP/BaSO4 catalyst prepared
by an adsorption method is efficient and highly selec-
tive in transforming vegetable oils into diesel-like par-
affins. Preliminary kinetic investigations using methyl
palmitate as a model compound indicated that the de-
carboxylation of aliphatic esters on palladium in a hy-
drogen-rich atmosphere showed a zero-order rate. H2-
TPD measurements revealed that the high-tempera-
ture desorption of hydrogen species might be the
rate-determining step.
Experimental Section
Figure 3. H2-TPD profiles of typical samples: (a) BaSO4 sup-
port, (b) Pd/BaSO4 catalyst, and (c) PdNP/BaSO4 catalyst.
Commercial edible grade sunflower oil (palmitic acid 6.3%,
stearic acid 5.6%, oleic acid 24.3%, linoleic acid 62.7% and
others 1.1%) was purchased from local supermarket and
used without any treatment. Poly(N-vinyl-2-pyrrolidone)
(PVP K-30) with an average molecular mass of 40000
served as a stabilizer to prevent the aggregation of Pd nano-
particles.[20]
boxylation of aliphatic esters exhibited a zero-order
rate with a rate constant of 0.023 molLÀ1 hÀ1 at 2808C
(see Supporting Information, Figure S7).
Hydrogen temperature-programmed desorption
(H2-TPD) measurements of the typical samples are
collected in Figure 3. As shown in Figure 3, Pd/BaSO4
only exhibited a broad hydrogen desorption peak at
200–6008C with the maximum hydrogen desorption
amount at 4608C and PdNP/BaSO4 showed a temper-
ature range from 1708C to more than 6008C for hy-
drogen desorption with two main peaks at 3508C and
4508C, respectively. No obvious H2 desorption peaks
were detected on the BaSO4 support. We speculate
that the high-temperature desorbed hydrogen species
might exclusively participate in the rate-determining
step in a hydrogen-rich atmosphere, which made the
decarboxylation of esters an apparent zero-order re-
action. When active hydrogen species on the surface
of the catalyst participated in the reaction and were
consumed, they could be complemented simultane-
ously in the hydrogen-rich atmosphere, which resulted
in the concentration of the hydrogen species on the
surface of the catalyst being constant. Therefore the
reaction rate had nothing to do with the concentra-
tion of the hydrogen species on the surface of the cat-
alyst, so that the decarboxylation of esters manifested
an apparent zero-order reaction. PdNP/BaSO4 exhib-
ited a better low-temperature hydrogen desorption
TEM images were taken on a FEI Tecnai F30 microscope
operated at an accelerating voltage of 300 kV. XPS experi-
ments were performed using a VG ESCLAB MARK II
spectrometer with an Mg-Ka X-ray source. TGA profiles
were recorded on a Perkin–Elmer TGA7 thermo gravimet-
ric analyzer. Infrared (IR) spectra were measured on a
Nicolet NEXUS 470 spectrophotometer using KBr pellets.
H2-TPD profiles were collected on a Dycor 200 quadrupole
mass spectrometer with m/e (2, 16, 18, 28, 32 and 44) traced
simultaneously.
Procedure for the Barium Sulfate-Supported
Colloidal PdACTHNUTRGNE(NUG PVP) Nanoparticles Catalysts
A solution containing H2PdCl4, PVP (the molar ratio of N
to Pd was 10), HCl and deionized water was heated. When
the solution began to reflux, ethanol was added. The solu-
tion was then refluxed for 3 h, and a dark-brown colloidal
Pd solution was formed. After being cooled to room temper-
ature, the solution was mixed with barium sulfate and stirred
vigorously for 12 h. Then, the suspension was filtered and
washed with deionized water and ethanol. Finally, the ob-
tained powder was allowed to dry at 708C.
For calcined PdNP/BaSO4 catalysts, the dry sample was
calcined in air at 3008C for 1 h. The residue was reduced at
2508C in hydrogen and cooled to room temperature before
property than that of Pd/BaSO4, which might be the the catalytic reactions.
reason that the TOF of PdNP/BaSO4 was twice more
than that of Pd/BaSO4.
PdNP/BaSO4 catalyst has overcome the deactiva-
tion shortcoming of Rosenmund catalyst in catalytic
recycling for decarboxylation of biodiesel into diesel.
No deactivation was detected after six consecutive
run without any treatment of the catalyst. The yield
of decarboxylates was above 98% with a steady selec-
Typical Procedure for Decarboxylation
Sunflower oil, freshly prepared PdNP/BaSO4 catalyst and
hexane were added into a 100-mL stainless steel autoclave.
The reaction system was pressurized with hydrogen and
heated to react at 2808C for 6 h. After the reaction, gas and
liquid products were collected and then detected by gas
chromatography (GC), HPLC and GC-MS, respectively.
1808
asc.wiley-vch.de
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2010, 352, 1805 – 1809