See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/327392236

Catalytic hydrogenation of soybean oil-derived fatty acid methyl esters over

Pd supported on Zr-SBA-15 with various Zr loading levels for enhanced
oxidative stability

Article · September 2018

CITATIONS READS

0 183

5 authors, including:

Chachchaya Thunyaratchatanon Apanee Luengnaruemitchai

Chulalongkorn University Chulalongkorn University
7 PUBLICATIONS   24 CITATIONS    95 PUBLICATIONS   2,227 CITATIONS   

SEE PROFILE SEE PROFILE

Nuwong Chollacoop Shih-Yuan Chen

National Metal and Materials Technology Center National Institute of Advanced Industrial Science and Technology
104 PUBLICATIONS   2,645 CITATIONS    55 PUBLICATIONS   997 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Production of Biobutanol From Acid Pretreated Corncob Using Clostridium beijerinckii TISTR 1461: Process Optimization Studies View project

Effect of Phosphoric Acid Pretreatment of Corncobs on the Fermentability of Clostridium beijerinckii TISTR 1461 for Biobutanol Production View project

All content following this page was uploaded by Apanee Luengnaruemitchai on 06 September 2018.

The user has requested enhancement of the downloaded file.

 Fuel Processing Technology 179 (2018) 422–435

Contents lists available at ScienceDirect

Fuel Processing Technology

journal homepage: www.elsevier.com/locate/fuproc

Research article

Catalytic hydrogenation of soybean oil-derived fatty acid methyl esters over T

Pd supported on Zr-SBA-15 with various Zr loading levels for enhanced
oxidative stability
⁎
Chachchaya Thunyaratchatanona,b, Apanee Luengnaruemitchaia,b,c, , Nuwong Chollacoopd,
Shih-Yuan Chene, Yuji Yoshimurad,e
a
The Petroleum and Petrochemical College, Chulalongkorn University, Soi Chula 12, Phayathai Rd., Pathumwan, Bangkok 10330, Thailand
b
Center of Excellence on Petrochemical and Materials Technology, Chulalongkorn University Research Building, Soi Chula 12, Phayathai Rd., Bangkok 10330, Thailand
c
Center of Excellence on Catalysis for Bioenergy and Renewable Chemicals (CBRC), Chulalongkorn University, Phayathai Rd., Pathumwan, Bangkok 10330, Thailand
d
Renewable Energy Laboratory, National Metal and Materials Technology Center (MTEC), 114 Thailand Science Park, Paholyothin Rd., Klong 1, Klong Luang,
Pathumthani 12120, Thailand
e
Energy Catalyst Technology Group, Research Institute of Energy Frontier (RIEF), National Institute of Advanced Industrial Science and Technology (AIST), 16-1
Onogawa, Tsukuba 305–8569, Japan

A R T I C LE I N FO A B S T R A C T

Keywords: Partial hydrogenation of soybean oil (SO)-derived biodiesel, as fatty acid methyl esters (FAMEs), over meso-
Biodiesel porous palladium (Pd)/SBA-15 materials was examined in a semi-batch reactor. The amount of Pd/SBA-15
Partial hydrogenation catalyst was varied from 0.5–1% by weight (wt%) of the SO feedstock. The selectivity of saturated FAMEs
Oxidative stability (C18:0), which contribute to an excellent oxidative stability but undesirable cold flow properties, over the Pd/
Pd catalyst
SBA-15 catalyst at different wt% was poor after a 4-h reaction period. To improve the catalytic performance,
Zr-SBA-15
zirconium (Zr) species were incorporated in the SBA-15, resulting in Zr-SBA-15 with tunable acidity and pore
XPS
properties. Partial hydrogenation of SO-FAME over 0.75 wt% Pd/xZr-SBA-15 catalysts (where x is the Zr/silicon
(Si) molar ratio and ranged from 0.01 to 0.11) was performed for 2 h. The catalytic performance and the se-
lectivity towards trans-monounsaturated FAMEs (trans-C18:1), in terms of the turnover frequency (TOF) and
trans-C18:1/cis-C18:1 ratio, respectively, were considered at 10% and 80% conversion levels of diunsaturated
FAMEs (C18:2). The incorporation of Zr into SBA-15 resulted in an enhanced TOF due to the greater adsorption
of both H2 and the basic double bonds of FAME molecules on the electron-deficient catalyst surfaces. The TOF for
the Pd/xZr-SBA-15 catalysts increased as the Zr/Si molar ratio increased up to 0.07, and then slightly dropped at
higher Zr/Si molar ratios, presumably due to the hindrance of adsorbed-reacted FAME molecules with strong
adsorption and the worsening of physical properties, i.e. pore width (DP). The selectivity towards trans-C18:1
was directly correlated with the Zr/Si molar ratio. Since the strong adsorption of FAME molecules took place on
the catalyst surface, the high level of trans-C18:1 was consequently generated. The oxidative stability of the
partially hydrogenated biodiesels was improved by more than four-fold compared to that of the starting SO-
FAMEs.

1. Introduction Thailand, the Royal Thai government recently proposed an Alternative

Energy Development Plan (AEDP2015–2036) to focus on the develop-
The rapidly increasing global population and economic growth over ment of a low-carbon society by promoting the use of renewable energy
the past few decades have resulted in a corresponding increased de- and resources, such as solar, wind, hydro, bio-energy and bio-fuels [1].
mand for electricity, transport and agricultural products based on fossil- Among these, bio-fuels and especially biodiesel are potential candidates
fuels. The increased emission of greenhouse gases, mainly as carbon for widespread use as an alternative transport fuel owing to its biode-
dioxide, derived from fossil-based industries has caused severe pro- gradable nature, sustainable renewability and high cetane number.
blems in environmental pollution and global warming. Within Specifically, biodiesel has lower exhaust emissions (unburned

⁎
Corresponding author at: The Petroleum and Petrochemical College, Chulalongkorn University, Soi Chula 12, Phayathai Rd., Pathumwan, Bangkok 10330,
Thailand.
E-mail address: [email protected] (A. Luengnaruemitchai).

https://doi.org/10.1016/j.fuproc.2018.07.014
Received 29 May 2018; Received in revised form 6 July 2018; Accepted 15 July 2018
0378-3820/ © 2018 Elsevier B.V. All rights reserved.
C. Thunyaratchatanon et al. Fuel Processing Technology 179 (2018) 422–435

hydrocarbons and particulate matters) than petroleum-based fuels, performance of Pd catalysts in the partial hydrogenation of soybean oil
which is in accordance with the more earnest environmental emission (SO)-derived FAMEs (SO-FAMEs).
regulations [2,3]. In this study, the motivation was to improve the characteristics and
The commercial production of biodiesel is usually as fatty acid acid properties of Zr-SBA-15 by Zr incorporation into the silica frame-
methyl esters (FAMEs), formed via the transesterification of vegetable work of SBA-15. The Zr-SBA-15 materials were prepared by direct hy-
oils with an excess amount of methanol and catalyzed by homogeneous drothermal synthesis using a Zr/Si molar ratio range from 0.01–0.11.
alkali bases under a mild reaction condition (60–80 °C at atmospheric The reactions were performed in a semi-batch reactor at 100 °C and a
pressure) [4,5]. However, the resulting biodiesel usually has a poor 4 bar H2 atmosphere. The hydrogenation activity was examined in
oxidative stability and cold flow properties, limiting its application to terms of the turnover frequency (TOF), while the selectivity was ob-
being blended with petro-diesel. In general, fuel properties are asso- served in terms of the ratio of trans- to cis-monounsaturated FAMEs
ciated with the structural features of the starting oil or fat, such as the (trans-C18:1/cis-C18:1 ratio) at low and high conversion levels of
chain length, degree of chain branching and the nature of individual diunsaturated FAMEs (C18:2). X-ray diffractometry (XRD), nitrogen
FAME with various numbers of carbon double bonds (C]C). Those (N2) adsorption-desorption, carbon monoxide (CO) pulse chemisorp-
biodiesels with a high content of polyunsaturated FAMEs have a low tion, transmission electron microscopy (TEM), temperature pro-
oxidative stability, whereas those containing a high content of saturated grammed desorption of ammonia (NH3-TPD) and X-ray photoelectron
FAMEs have poor cold flow properties. To obtain a high-quality bio- spectroscopy (XPS) analyses were employed to characterize the crys-
diesel with a balanced oxidation stability and cold flow property, up- tallinity phases, textural and structural properties, metal dispersion and
graded biodiesel enriched in monounsaturated FAMEs is expected to be size, acid properties and chemical states of the catalysts. The chemical
suitable for most practical applications, especially high blend fuels functional groups and FAME compositions of the feed SO-FAMEs and
[5–7]. the partially hydrogenated (PH)-biodiesels were analyzed using atte-
Partial hydrogenation of FAMEs in the presence of transition metal nuated total reflectance-Fourier transform spectroscopy (ATR-FTIR)
catalysts has been studied as a means to decrease the level of C]C and gas chromatography with flame ionization detection (GC-FID),
bonds within the fatty acid chains to form a limited level of saturated respectively. In particular, the fuel properties, including the oxidative
FAMEs [8,9] In addition to the degree of hydrogenation, the iso- stability and cold flow properties, were determined according to the
merization of cis-monounsaturated FAME to its thermodynamically standard test methods [22, 23].
stable trans-configuration during the hydrogenation has to be averted
due to its higher crystallization point, which results in worse cold flow 2. Experimental
properties [9,10].
Recent studies have demonstrated that the supported palladium 2.1. Materials and chemicals
(Pd) catalysts are able to perform hydrogenation in the oil phase due to
their high hydrogen adsorption capacity, and to provide a higher cat- A nonionic triblock copolymer Pluronic P123 (EO20PO70EO20,
alytic performance than other active metal catalysts, such as platinum, Mw = 5800), which was utilized as an organic structure directing agent,
rhodium, rubidium and nickel [9,11–13]. Another considerable and was obtained from Aldrich. Tetraethyl orthosilicate (TEOS, 98%) and
widely studied strategy for improving the catalyst performance in the ZrOCl2·8H2O, which were used as the silica source and the zirconium
hydrogenation process is to develop outstanding catalyst supports. The precursor, were purchased from Fluka and Acros, respectively.
preparation of a mesoporous supporting material with different acidity Hydrochloric acid (HCl) was bought from Quality Reagent Chemical
and pore structures has been investigated recently to further tune the (QRëC). Palladium ammonium chloride (Pd(NH3)4Cl2·xH2O), used as
size and surface electronic property of the supported metal catalyst. the Pd precursor, was acquired from N.E. Chemcat Corporation, Japan.
Mesoporous silica materials, especially SBA-15, are extremely attractive The SO used as the starting material for transesterification, was bought
because of their excellent characteristics of a high surface area, high from the Thai Vegetable Oil Co. Ltd. Methanol (analytical reagent
surface-to-volume ratio, highly-ordered porous framework, tunable and grade) and potassium hydroxide (KOH; 85%) were purchased from RCI
uniform mesoporous size, super hydrothermal stability and high capa- Labscan Limited, whilst n-heptane (analytical reagent grade) was ob-
city to functionalize its surface [14–16]. For example, SBA-15 with its tained from Fisher Scientific. Other chemical reagents, including so-
high surface area and narrow pore size distribution contributed greatly dium sulphate anhydrous (Na2SO4; 99%) and sodium chloride (NaCl)
to a high Pd dispersion [17], while the well dispersed metal nano- were bought from Ajax Finechem Pty. Ltd.
particles (NPs) on the SBA-15 support promoted a higher affinity to-
wards hydrogen (H2) molecules [16]. 2.2. Synthesis of the mesoporous support
Nevertheless, the lack of an acidic environment on the pure siliceous
SBA-15 limits its utilization in a variety of chemical applications. To 2.2.1. SBA-15
overcome this problem, the incorporation of zirconium (Zr) into the The parent SBA-15 sample was synthesized by a hydrothermal
SBA-15 framework (Zr-SBA-15) to create potential Lewis acid proper- process as previously reported [24]. In a typical synthesis procedure, a
ties has been extensively applied in many acid-catalyzed reactions, such solution of P123 (1 g), deionized water (40.6 g) and 2 M HCl solution
as oxidation, hydroxylation, transesterification and Fischer–Tropsch (40 g) were prepared in a sealed bottle at 35 °C. After that, TEOS (2.1 g)
synthesis [18–21]. However, the Zr-SBA-15 material has not been was added dropwise into the prepared solution and maintained at 35 °C
broadly applied in the partial hydrogenation of biodiesel. Chen et al. for 24 h to give a milky solution mixture with a molar ratio of 0.017
[8] researched the performance of Pd catalysts with various meso- P123: 1.0 TEOS: 7.9 HCl: 221 H2O. This was then transferred into a
porous supports, SBA-15, Zr-SBA-15 (Zr/Si molar ratio = 0.05) and 125-mL Teflon-lined stainless-steel autoclave and aged at 100 °C for
commercial γ-Al2O3, for upgrading the oxidation stability and cold flow another 24 h. The dispersed particles were harvested by filtration,
properties of palm oil-derived biodiesel using a continuous fixed-bed washed with a large amount of deionized water, dried at 100 °C over-
reactor. The hydrogenation activity of Pd/SBA-15 and Pd/Zr-SBA-15 night and then calcined in air at 500 °C for 12 h at a 1 °C/min heating
were quite high and analogous due to the ultra-small Pd NPs with rate.
electron-deficient surfaces. With the higher acidic environment of Pd/
Zr-SBA-15, the level of trans-monounsaturated FAME produced was 2.2.2. xZr-SBA-15
slightly greater than that with Pd/SBA-15. Accordingly, the outstanding The Zr was incorporated into the synthesized SBA-15 by direct hy-
features of a Zr-SBA-15 support together with its optimum level of acid drothermal treatment as previously reported [25] and described in
properties makes it an interesting candidate for enhancing the detail [26], using various molar ratios of Zr/Si (0.01 to 0.11). The Zr-

423
C. Thunyaratchatanon et al. Fuel Processing Technology 179 (2018) 422–435

incorporated SBA-15 supports were denoted as xZr-SBA-15, where x Surface acidic sites of catalysts were determined by NH3-TPD using
refers to the Zr/Si ratio. In this synthesis method, P123 (1 g), deionized a Thermo Finnigan TPDRO 1100 series instrument with a thermal
water (40 g), NaCl (0.59 g) and a different amount of Zr precursor were conductivity detector (TCD). The catalyst samples were first pretreated
firstly homogeneously mixed in a sealed bottle at 35 °C. TEOS (2.1 g) with N2 at 300 °C for 1 h, cooled to room temperature and then purged
was then gradually dropped into the prepared solution mixture and with NH3 for another 30 min. The NH3-TPD patterns were collected
stirred altogether at 35 °C for 24 h, forming a milky mixture with a under He at a flow rate of 30 mL/min by heating from 40 to 900 °C with
molar ratio of 0.017 P123: 1.0 TEOS: 0.01–0.11 ZrOCl2.8H2O: 221 H2O: a heating rate of 10 °C/min.
1.0 NaCl. The mixture was transferred into the Teflon-lined stainless Surface elements and chemical states of catalysts were identified by
steel autoclave and aged at 90 °C for another 24 h. The white-colored XPS, recorded on a Kratos Axis Ultra DLD photoelectron spectrometer
particles was recovered by filtration, washed with a large amount of equipped with a magnetic immersion lens and charge neutralizer. The
deionized water and subsequently dried at 50 °C for 16 h. Finally, the analyses were performed under a high vacuum condition
white-fine powder of xZr-SBA-15 material was acquired after calcina- (≤5 × 10−7 Torr). A monochromatic Al Kα (hv = 1486.6 eV) source
tion using the same heating condition as the Section 2.2.1. operated at 15 kV was employed as the X-ray radiation source. The
recorded spectra were calibrated using the binding energy peak of C 1s
2.3. Catalyst preparation at 284.6 eV.

The Pd/SBA-15 and Pd/xZr-SBA-15 catalysts were prepared by in- 2.5. Transesterification of SO with methanol to SO-FAMEs
cipient wetness impregnation using a Pd loading level of 1% by weight
(wt%). The synthesized supports were first dried in an oven at 110 °C The SO (180 g) was added to a 500-mL three-necked round-bottom
overnight to remove any residual adsorbed water. The aqueous solution flask and heated to 60 °C. Methanol (61 g) and 1.8 g KOH (1 wt%
containing the Pd precursor was then impregnated into the dried sup- compared to SO) were prepared by homogeneous mixing and then
port under vacuum for 24 h, and the catalyst was then dehydrated using added into the heated SO (9:1 methanol: SO molar ratio) and main-
a vacuum oven at 110 °C for 6 h. To eliminate any adsorbed organic tained at this temperature for 1 h. The solution was then transferred
compounds, the catalyst was calcined at 300 °C for 3 h under an oxygen into a separation funnel to allow phase separation of the glycerine and
flow of 500 mL/min and afterwards for 2 h under a H2 gas flow of biodiesel overnight. The lower phase (glycerine) was removed and the
100 mL/min. upper phase (biodiesel) was extracted with distilled water at 60 °C
several times in order to remove the excess methanol, KOH and possible
2.4. Catalyst characterization soap. Finally, the obtained biodiesel (SO-FAMEs) was kept in a sealed-
glass bottle containing anhydrous Na2SO4 to adsorb the residual water.
Crystallinity and composition of the crystalline phases of the cata-
lysts were examined by XRD analysis. The small-angle X-ray diffraction 2.6. Partial hydrogenation of the SO-FAMEs
patterns were recorded on a Rigaku TTRAX III with an angular 2θ-
diffraction in the 0.5–3.0° range, whereas the wide-angle X-ray dif- Partial hydrogenation of the SO-FAMEs (from Section 2.5) to yield
fraction patterns were obtained using a Rigaku DMAX 2200 HV at a the PH-biodiesel was performed in a 300-mL stainless-steel semi-batch
scan speed of 1°/min over a 2θ range of 10–80°. The Cu Kα radiation reactor at 100 °C and a H2 pressure of 4 bar. Firstly, the desired amount
with a wavelength of 1.541 Å was utilized as the X-ray source, and was of reduced catalyst (0.5, 0.75 and 1 wt% compared to biodiesel) was
generated by a ceramic X-ray tube at 50 kV and 300 mA for the small- put into the reactor containing 130.4 g of the SO-FAMEs. The reactor
angle region, and at 40 kV and 30 mA for the wide-angle measurement. was then purged with N2 to eliminate any residual air, followed by H2
The analysis was performed by placing the finely-powdered samples at a flow rate of 150 mL/min, maintaining the H2 pressure at 4 bar. To
onto glass holders, which were then set at the center of the dif- obtain a perfect mix between the feed SO-FAMEs, catalyst and H2, a
fractometer. stirrer was applied at around 500 rpm. The temperature was gradually
Textural properties of the catalysts were examined using a increased to the set point, and subsequently the first hydrogenated SO-
Quantachrome Autosorb–1 MP surface area analyzer. Each catalyst FAMEs was instantly collected. Subsequent biodiesel sampling was
sample was degassed at 250 °C for 12–16 h under a high vacuum con- performed every 30 min during the 4-h reaction time.
dition (≤10 μm Hg/min) to remove any volatile species adsorbed on
the sample surface. The specific surface area (SBET) was calculated by 2.7. Analysis of feed SO-FAMEs and the PH-biodiesel
the multi-point Brunauer–Emmett–Teller (BET) method. The total pore
volume (VTotal) was obtained from the adsorbed N2 after complete The chemical functional groups of the SO, the feed SO-FAMEs and
condensation at a relative pressure of about 0.995. The pore size dis- PH-biodiesel products were analyzed by ATR-FTIR using a Thermo
tribution (DP) was calculated on the basis of desorption branches of Scientific iD7 ATR accessory for the Thermo Scientific Nicolet iS5
nitrogen isotherms using the Barrett–Joyner–Halenda (BJH) method. Fourier transform infrared spectrometer. A drop of liquid samples was
Metal dispersion and metal particle size were analyzed by CO pulse placed onto the monolithic diamond crystal plate. The IR spectra were
chemisorption using an Ohkura Riken R6015 instrument. The catalysts recorded in the range of 3500–650 cm−1 at a 4 cm−1 resolution with 64
were placed in quartz tube and reduced under a H2 atmosphere at scans.
300 °C for 4 h, then purged with helium (He) and cooled down to room The fatty acid composition of the feed SO-FAMEs and the PH-bio-
temperature. Afterwards, the pretreated catalysts were pulsed by a 10% diesel were determined using a Hewlett Packard 5890 Series II GC-FID
(v/v) CO in a He flow until no more CO was adsorbed. The metal dis- equipped with an HP–88 (100 m × 0.25 mm × 0.2 μm) fused-silica ca-
persions were calculated using a CO/Pd stoichiometric ratio of 1. pillary column with He as the carrier gas at a flow rate of 70 mL/min.
The micro-structure and micro-texture of catalyst samples were Each sample (2 μL) was injected into the oven (155 °C) with an injector
characterized by TEM using a JEOL JEM-2100 model instrument. and detector temperature of 200 °C and 230 °C, respectively, with a split
Moreover, energy-dispersive X-ray spectroscopy (EDS) was applied for ratio of 75:1. After a 20-min isothermal period, the temperature was
elemental identification. Typically, tiny amounts of powder samples increased to 220 °C at 2 °C/min and held at 220 °C for 10 min.
were dispersed in ethanol and then placed in an ultrasonic sonicator for Oxidative stability of the feed SO-FAMEs and the PH-biodiesel
30 min. A drop of the sample solution was added onto a carbon-covered products, corresponding to the induction period (IP; h), was measured
grid, dried at room temperature overnight and finally kept in a de- by the Rancimat method according to EN 14214 [22] using a Metrohm
siccator in order to evade any moisture. 743 instrument. The feed SO-FAMEs and the PH-biodiesels (after the

424
C. Thunyaratchatanon et al. Fuel Processing Technology 179 (2018) 422–435

complete reaction) were examined in heating blocks at 110 °C under an standards, PDF-2/Release 2012 RDB. All diffraction profiles demon-
air flow rate of 10 L/h. During this time, the volatile compounds, which strated a broad peak located at around 2θ of 22°, attributed to the ty-
were the byproducts of FAME degradation, were released and recorded pical peak of amorphous silicate materials (DB 00-001-0649) [8,18].
by the conductivity cells. The infection point of the conductivity vs. Meanwhile, no peak that correlated with Pd and/or Zr was detected,
time curve designates as the IP (h). which was due to the high metal dispersion (low density) and/or small
The cold flow properties, used for determining the operating cap- Pd or Zr NP size on the prepared supports [8,28,29]. To affirm the high
ability of fuels, are normally evaluated as the cloud point (CP) and pour Pd dispersion and the nano-sized Pd particles, CO pulse chemisorption
point (PP). The CP, which is the temperature at which wax crystals start analysis was utilized (see Section 3.1.3).
to cloud the fuel, was assessed following the ASTM D2500 protocol
[23]. The PP, which is the lowest temperature at which a fuel loses its
flow characteristics, was investigated according to ASTM D97 [23]. The 3.1.2. N2 adsorption-desorption
tests were performed using a CPP 5Gs automated CP and PP analyzer. The N2 adsorption-desorption isotherms and pore size distribution
The SO-FAMEs and the PH-biodiesel samples were put into the test jar curves for the Pd/SBA-15 and Pd/xZr-SBA-15 catalysts are presented in
and the temperature was continuously decreased. These characteristics Fig. 2. The isotherms of the SBA-15 samples featured hysteresis loops
were reported on screen to one decimal place. with sharp adsorption and desorption branches, indicating a narrow
mesopore size distribution. All samples exhibited representative type IV
isotherms, according to the classification of IUPAC, with a well-defined
3. Results and discussion
hysteresis loop in a relative pressure (P/P0) range of around 0.4–0.8,
suggesting the presence of ordered mesopores in the frameworks with
3.1. Characterization and properties of each catalyst
narrow pore size distribution [20,27,30]. These results are in accord
with the XRD profiles. The alternation of the hysteresis loop shapes was
3.1.1. XRD analysis
distinctly noticed for the Pd supported on xZr-SBA-15 materials, which
The small-angle and wide-angle XRD patterns of the Pd/SBA-15 and
was due to the influence of the Zr incorporation rather than that of the
Pd/xZr-SBA-15 catalysts are illustrated in Fig. 1. In the small-angle
Pd loading. The addition of Zr aided in the formation of P123 micelles
region, three conspicuous peaks centered at 2θ = 0.90°, 1.56° and
in a self-generated acidic environment, resulting in the well-ordered Zr-
1.80°, corresponding to the (100), (110) and (200) reflections, were
SBA-15 materials. However, the increased Zr/Si molar ratio (> 0.07)
observed in the XRD pattern of Pd/SBA-15, indicating the existence of
caused a distortion in the structure as well as a partial blocking of
well-ordered hexagonal arrays and the two-dimensional channel
mesopore channels, leading to the decreased pore diameter distribu-
structures of p6mm symmetry [8,16,18,27]. For Pd/xZr-SBA-15, how-
tion, which has been described previously [26].
ever, the intensity of all the diffraction peaks decreased and slightly
The textural properties of the catalysts are summarized in Table 1.
shifted to a higher 2θ value, indicating a decrease in the ordered level
The impregnation of Pd on the synthesized supports resulted in a slight
of structure as well as in the mesoporous diameter. Consequently, the
decrease in the SBET and VTotal in comparison with the pure supports
regular interspace distance between the walls of the mesoporous
(not shown but reported in [26]). Surprisingly, the DP of the catalysts
channels (d100) and unit cell parameter (a0) were also decreased
was hardly altered from that of the SBA-15 and xZr-SBA-15 supports,
(Table 1). As previously reported [26], the diminishing intensity of
accounting for the high dispersion level and diminutive size of the Pd
these peaks in all the synthesized xZr-SBA-15 supports was due to the
NPs. As a result, it could be epitomized that the blockage of the pore
increasing Zr/Si ratios, and so the existence of Pd did not destroy the
mouths by the Pd particles was negligible.
structure of the support materials. For the wide-angle region, the
characteristic peaks were determined according to ICDD reference

Fig. 1. Representative XRD patterns of the Pd/SBA-15 and Pd/xZr-SBA-15 catalysts by (A) small-angle and (B) wide-angle.

425
C. Thunyaratchatanon et al. Fuel Processing Technology 179 (2018) 422–435

Table 1
Characteristics of the Pd/xZr-SBA-15 catalysts.
Catalyst SBETa (m2/g) VTotalb (mL/g) DPc (nm) d100 (nm) a0d (nm) δe (nm) Pd NPsf Acid sitesg (mmol/g)

(%) (nm)

Pd/SBA-15 626 0.94 6.1 9.8 11.3 5.2 40 1.22 N/A

Pd/0.01Zr-SBA-15 617 0.86 6.1 9.3 10.7 4.6 26 1.88 0.04
Pd/0.03Zr-SBA-15 620 0.78 6.1 9.6 11.1 5.0 30 1.62 0.29
Pd/0.05Zr-SBA-15 617 0.76 4.6 9.6 11.1 6.5 30 1.61 0.47
Pd/0.07Zr-SBA-15 599 0.73 4.6 9.4 10.8 6.2 27 1.80 0.46
Pd/0.09Zr-SBA-15 561 0.66 3.6 9.2 10.6 7.0 31 1.58 0.47
Pd/0.11Zr-SBA-15 537 0.55 3.6 8.5 9.8 6.2 27 1.83 0.60

a
Determined by the BET method.
b
Determined by N2 adsorption at a relative pressure (P/P0) = 0.995.
c
Obtained from the maximum of BJH method desorption pore size distribution.
d
Calculated from the position of the (1 0 0) diffraction as a0 = 2d100/ 3 .
e
Pore wall thickness = a0 − DP.
f
Measured by CO pulse chemisorption.
g
Calculated from NH3-TPD profiles in the range of 160–300 °C.

Fig. 2. Representative N2 adsorption-desorption results of the Pd/SBA-15 and Pd/xZr-SBA-15 catalysts, shown as the (A) isotherms and (B) pore size distribution.

3.1.3. CO pulse chemisorption supports was destroyed with a high Zr/Si molar ratio, especially for the
The Pd dispersion level (%) and particle sizes for all the catalysts are Pd/0.11Zr-SBA-15 catalyst. As clearly seen in these images, the pore
tabulated in Table 1. The Pd/SBA-15 catalyst clearly had the highest Pd walls (dark region) were broader as the Zr/Si molar ratio increased,
dispersion level and smallest Pd NPs at 40% and 1.22 nm, respectively, while the pore channels (clear region) narrowed, indicating the de-
due to having the largest support surface area [17,31,32]. When Zr was creased DP. In the same way as the XRD profiles and N2 adsorption-
introduced into the synthesized materials, the Pd dispersion level was desorption results, the deterioration of the highly-ordered structure of
reduced 1.29- to 1.54-fold to around 26–31%, with correspondingly the catalysts were clearly associated with a large Zr incorporation and
larger Pd NPs (1.58–1.88 nm) than in Pd/SBA-15. This was due to the not the Pd impregnation.
lower surface area of the xZr-SBA-15 supports and/or metal-metal at- Apart from the ordered channels, the appearance of tiny dark spots
traction [8,17,33,34]. These CO pulse chemisorption results were sup- that were well-dispersed on the supports, which were the Pd NPs (active
ported by the absence of the metal crystalline phases in the XRD pat- sites), was evidently noticed. To confirm the presence of Pd NPs, TEM-
terns (Section 3.1.1) as well as the high surface area of catalysts EDS analysis was applied (SI, Fig. 2S). The EDS spectra of all the cat-
obtained from N2 adsorption-desorption analysis (Section 3.1.2). alysts except for Pd/SBA-15 showed the characteristic signals of Si, O,
Pd, Zr and Cu, which were attributed to the silicate framework of the
3.1.4. TEM analysis support, metal active sites, incorporated heteroatoms and the copper
The TEM photographs of the Pd/SBA-15 and Pd/xZr-SBA-15 cata- grid, respectively. Thus, the presence of Pd and Zr NPs was confirmed
lysts are displayed in Fig. S1 in the Supporting Information (SI). It is even though no diffraction peak was revealed in the XRD patterns.
important to note that the well-ordered mesoporous structures with
long cylindrical channels of catalysts were well preserved after the 3.1.5. NH3-TPD analysis
metal impregnation [15,20], albeit part of the pore channels of the The NH3-TPD profiles of all the support materials and catalysts are

426
C. Thunyaratchatanon et al. Fuel Processing Technology 179 (2018) 422–435

Fig. 3. Representative NH3-TPD profiles of the (A) SBA-15 and xZr-SBA-15 supports and (B) Pd/SBA-15 and Pd/xZr-SBA-15 catalysts.

displayed in Fig. 3. As noticed, the distinct peak below 160 °C, which investigated by XPS. To examine the surface elements, the survey scan
was assigned to the NH3 physisorption on the materials' surfaces, was (wide-scan) of the XPS was applied. The interesting peaks of O 1s, Zr 3p
observed for all the samples due to the NH3 adsorption at room tem- and Pd 3d were visualized in most of the Pd catalysts (SI, Fig. S3),
perature (~30 °C; Section 2.4) [25]. As increased the Zr/Si molar ratio, supporting the successful Zr incorporation and Pd impregnation. As for
the tail of this peak was gradually increased in the desorption range of the Pd/SBA-15 and Pd/0.01Zr-SBA-15, the characteristic peak corre-
160–300 °C, which was due to the formation of new peak correlated wih sponding to Pd 3d and Zr 3p was hard to discover in the wide-scan
ZreOeSi and/or ZreOeH groups with weak to medium strength. This spectra, presumably due to the well-dispersed and diminutive-sized
was also substantial evidence that the acidity of the xZr-SBA-15 was metal NPs on the catalyst surface. However, the presence of these
higher than that of the parent SBA-15. At temperature above 300 °C, a metals was justified by CO pulse chemisorption, TEM-EDS and high-
broad peak ascribed to the dehydration condensation of silanol func- resolution XPS (narrow-scan) analyses. For the other catalysts, it was
tional group (SieOeH) was evident in the SBA-15 support because of noteworthy that the intensity of the Zr 3p signals increased as the Zr/Si
abundance of surface silanol. Nevertheless, this peak gradually became molar ratio of the synthesized supports was increased.
smaller as the Zr/Si molar ratio increased, due to its better condensa- To more thoroughly examine the chemical states of the catalysts,
tion resulting from the salt effect of Zr4+ species on P123 micelles. In narrow-scan XPS analysis was performed. The XPS spectra with de-
the case of Pd catalysts, the desorption profiles were analogous to that convolution for the O 1s, Zr 3p and Pd 3d are illustrated in Figs. 4 and 5,
of their corresponding supports, especially for the presence of the peak respectively. For the chemical states of O 1s, the XPS spectrum of Pd/
at around 160–300 °C, indicating the existence of weak to medium acid SBA-15 (Fig. 4A) was curve-fitted into two identical peaks centered at a
sites. In accordance with the NH3-TPD results, it was evidently epito- binding energy (BE) of 533.6 eV and 531.2 eV, which were attributed to
mized that the Pd/xZr-SBA-15 catalysts were more electron-deficient the SieOeSi (amorphous SiO2 support) and M–O (in this case PdeO)
than the Pd/SBA-15. bonds, respectively [39,40]. The latter peak arose from the charge-
To measure the actual acid sites, the NH3-TPD profiles only in the transfer process of the Pd metal and oxygen, contributing to the ex-
range of 160–300 °C were determined. The density of acid sites for all istence of an identical peak at a lower BE. As for the Pd/xZr-SBA-15
the supports and catalysts are summarized in Table S1 (SI) and Table 1, catalysts (Fig. 4B–G), another characteristic peak was clearly observed
respectively. In the case of the SBA-15 support, no acidity was detected at a BE of around 531.3–532.7 eV, which was ascribed to the bridging
due to the absence of acidic species. The density of acid sites was close oxygen atoms in the silica framework as SieOeZr bonds [28,41,42],
to 0.04 mmol/g for 0.01Zr-SBA-15, and then increased to a maximum of indicating the Zr incorporation into the silica frameworks. In addition,
0.60 mmol/g as the Zr/Si molar ratio increased to 0.11. The substitu- the identical M–O peak, could be assigned to both PdeO and ZreO due
tion of the larger atomic size Zr4+ ions (0.084 nm) into the smaller to the presence of Zr species.
atomic size of Si4+ ions (0.026 nm) in the framework caused the dif- The chemical states of Zr 3p and Pd 3d were also examined (Fig. 5),
ference between the bond length of ZreOeSi and SieOeSi, resulting in where the XPS spectrum of Pd/SBA-15 (Fig. 5A) obviously showed the
the higher acidity [35]. When Pd was impregnated onto the synthesized two doublets of Pd 3d. As expected, the main doublets at a BE of around
SBA-15 and xZr-SBA-15 materials, the number of total acid sites of all 335.6 eV and 340.9 eV were identified to Pd0 3d5/2 and Pd0 3d3/2, re-
the catalysts was slightly diminished compared to the pure synthesized spectively, due to the fact that the catalysts were reduced prior to ap-
supports, which was due to the location of Pd that covered the disclosed plication in the partial hydrogenation. Meanwhile, the other doublets
Zr species on the supports [36–38]. centered at 337.8 eV and 344.8 eV were assigned to PdO 3d5/2 and PdO
3d3/2, respectively, suggesting that the Pd NPs might be exposed to
atmospheric oxygen during catalyst storage and/or preparation prior to
3.1.6. XPS analysis characterization [43–45]. Noticeably, two other doublets of the
The surface elements and chemical states of the Pd catalysts were

427
C. Thunyaratchatanon et al. Fuel Processing Technology 179 (2018) 422–435

(caption on next page)

428
C. Thunyaratchatanon et al. Fuel Processing Technology 179 (2018) 422–435

Fig. 4. Representative XPS spectra of O 1s of the (A) Pd/SBA-15, (B) Pd/0.01Zr-SBA-15, (C) Pd/0.03Zr-SBA-15, (D) Pd/0.05Zr-SBA-15, (E) Pd/0.07Zr-SBA-15, (F)
Pd/0.09Zr-SBA-15 and (G) Pd/0.11Zr-SBA-15.

distinctive Zr 3p peaks were also detected in the XPS spectra of all Pd/ 3.2.2. GC-FID analysis
xZr-SBA-15 catalysts (Fig. 5B–G). The main doublets appeared at The synthesized SO-FAMEs had a proximate analysis of 99.1%
around 333.1–333.9 eV and 347.4–348.3 eV and were specified to Zr4+ FAME components, including methyl linoleate (C18:2, 50.23%), cis-
ions in both the framework and as excess ZrO2 particles [46,47]. Ac- methyl oleate (cis-C18:1, 25.36%), methyl palmitate (C16:0, 11.60%),
cording to the literature [26], it could be epitomized that most of the methyl linolenate (C18:3, 5.60%) and methyl stearate (C18:0, 3.91%).
Zr4+ ions were located in the silica framework rather than on the Most of the geometric isomers were in the cis-configuration and no
catalyst surfaces, albeit the amounts of ZrO2 clusters, which were trans-isomers were observed in the SO-FAMEs, consistent with the ATR-
placed nearby the superficial areas of xZr-SBA-15 materials, increased FTIR spectra results. A high degree of polyunsaturated FAMEs (C18:3
with increasing Zr/Si molar ratios. The other small doublets at a BE of and C18:2) causes a substandard oxidative stability in the biodiesel
about 329.0–329.6 eV and 344.3–345.4 eV were assigned to sub-oxide owing to their higher autoxidation rate, which for C18:3 and C18:2 are
of Zr species, resulting from the reduction stage in the catalyst pre- 98-fold and 41-fold, respectively, higher than that for C18:1 [53]. Ac-
paration (Section 2.3) [48]. cordingly, it is required to decrease the level of unsaturated fatty acid
chains without increasing the saturation level by partial hydrogenation
of the FAMEs.
3.2. Analysis of the synthesized feed SO-FAMEs

3.2.1. ATR-FTIR 3.3. Partial hydrogenation of the SO-FAMEs

The ATR-FTIR spectra of the commercial SO and the synthesized SO-
FAMEs (from Section 2.5) in the range of 3500–650 cm−1 are depicted 3.3.1. Effect of the Pd/SBA-15 catalyst loading level
in Fig. 6a and b. The characteristic peaks of the ATR-FTIR spectra were The amount of catalyst was previously specified at only 1 wt% of the
identified with reference to previous reports [49–51]. For both SO and feed biodiesel, a level at which it could transform most polyunsaturated
SO-FAMEs, no adsorption band was observed in the 3500–3000 cm−1 FAMEs to C18:1 in a limited reaction time of 4 h. In this present study,
region associated with OeH bonds, suggesting the absence of water where a new type of support (SBA-15) was established that was ex-
molecules or unreacted methanol in these oil samples. Also, the band pected to have a higher catalytic performance, the catalyst loading level
located at a wavenumber of 3008 cm−1 was ascribed to the ]CeH was studied in the range of 0.5–1 wt% of the feed SO-FAMEs. The
stretching vibration of the cis-C]C bond (alkene groups), indicating partial hydrogenation of SO-FAMEs was performed with stirring at
that the samples contained unsaturated FAMEs with a cis-configuration. 500 rpm for 4 h at 100 °C and 4 bar of H2 at a 150 mL/min flow rate.
The bands at 2922 cm−1 and 2853 cm−1 were detected in both the SO The alteration in the FAME compositions of the obtained PH-bio-
and SO-FAMEs and were ascribed to the asymmetric and symmetric diesels, in terms of the C18:0, cis-C18:1, trans-C18:1, C18:2 and C18:3
stretching vibrations of the CeH alkane groups, respectively. More levels, with different catalyst loading levels during a 4-h hydrogenation
specifically, these two bands suggested the presence of methyl (CH3) or reaction are illustrated in Fig. 7. Almost all of the C18:3 and C18:2
methylene groups in the ester chains of the SO-FAMEs. The strongest contents were rapidly diminished, with the cis-C18:1, trans-C18:1 and
peak absorbed at 1741 cm−1 was attributed to C]O groups with a C18:0 levels being substantially increased in the 4-h reaction. That the
stretching mode of vibration, indicating the appearance of esters in level of C18:0 was minimal reflects the principle of the partial hydro-
both the SO (glyceryl ester) and SO-FAMEs (methyl ester). Thus, con- genation of unsaturation. The maximum cis-C18:1 level at all catalyst
version of triglycerides in the SO to FAMEs was achieved [49,51]. loading levels was produced at 2 h of the reaction and then decreased
Asides those absorbance regions, the ATR-FTIR spectrum in the continuously thereafter leading to an accelerated rise in the C18:0
range of 1500–900 cm−1 is well known as the fingerprint region that content. Likewise, the conversion of polyunsaturated FAMEs to the
corresponded to the actual chemical discrimination between SO and its highest level of trans-C18:1 was evident at a reaction time of around
biodiesel [49]. In this region, the vibration bands of the bending of CH3 2.5–3 h, but afterwards deep hydrogenation was performed. Therefore,
and/or CH2 aliphatic groups were exhibited at 1460 cm−1 and a 2-h reaction was appropriate as a compromise between a high cata-
1377 cm−1 for SO as well as at 1460 cm−1, 1435 cm−1 and 1361 cm−1 lytic conversion without increasing the C18:0 content but keeping the
for the SO-FAMEs. In particular, the peaks at 1377 cm−1 and highest cis-C18:1 formation. These results were in good agreement with
1435 cm−1 were assigned to the asymmetric stretching of OeCH2 for the ATR-FTIR spectra (Fig. 6), in which the absence and presence of
the moiety of mono-, di- and triglycerides and asymmetric stretching of vibration bands at 3008 cm−1 and 966 cm−1, corresponding to the al-
OeCH3 for methyl ester, respectively. In addition, the absorbance kene groups and trans-configuration of –CH3, respectively, were clearly
bands related to the stretching vibrations of CeO and CeOeC for SO observed after partial hydrogenation with different catalyst loading
were revealed at 1236 cm−1 and 1098 cm−1, while they were located at levels. Meanwhile, the reduction of the absorbance band at 722 cm−1,
1244 cm−1, 1195 cm−1, 1169 cm−1, 1120 cm−1 and 1016 cm−1 in the which was attributed to the cis-configuration, was perceived as well.
SO-FAMEs. The characteristic peaks that appeared at 1169 cm−1 and To determine the suitable amount of Pd/SBA-15 catalyst in the re-
1016 cm−1 corresponded to the CeO symmetric and anti-symmetric action, a comparison of both the trans-C18:1 selectivity, estimated from
stretching vibrations, respectively. Importantly, the band located at the trans-C18:1/cis-C18:1 ratio, and the C18:2 conversion level in a 2-h
1196 cm−1 in the spectrum of the SO-FAMEs, attributed to O–CH3 hydrogenation reaction was evaluated (Fig. 8). As expected, at higher
stretching, confirmed the presence of FAMEs (biodiesel). The other two catalyst loading levels, a larger selectivity towards trans-C18:1 and a
peaks at a wavelength of 966 cm−1 and 722 cm−1 were ascribed to the higher C18:2 conversion level were obtained [54]. Hence, the 0.75 wt%
trans- and cis-configuration of eCH, respectively. Noticeably, the peak Pd/xZr-SBA-15 catalyst was selected for further application in this
at 966 cm−1 was very weak in the ATR-FTIR spectra of both SO and SO- hydrogenation process, based upon a compromise between the catalyst
FAMEs, which conformed well to the fact that the natural occurrence of selectivity and activity. From this result, it was epitomized that the
the cis-isomer in oil is more favorable than that of the trans-isomer [52]. addition of Pd onto the outstanding SBA-15 support enhanced the
To determine the yield and type of FAMEs in the biodiesel derived from catalytic activity and cis-C18:1 formation within a 2-h reaction, al-
soybean oil, GC-FID analysis was then performed. lowing a smaller catalyst loading level and shorter reaction time than
previously reported [5,55].

429
C. Thunyaratchatanon et al. Fuel Processing Technology 179 (2018) 422–435

(caption on next page)

430
C. Thunyaratchatanon et al. Fuel Processing Technology 179 (2018) 422–435

Fig. 5. Representative XPS spectra of Zr 3p and Pd 3d of the (A) Pd/SBA-15, (B) Pd/0.01Zr-SBA-15, (C) Pd/0.03Zr-SBA-15, (D) Pd/0.05Zr-SBA-15, (E) Pd/0.07Zr-
SBA-15, (F) Pd/0.09Zr-SBA-15 and (G) Pd/0.11Zr-SBA-15.

the incorporation of Zr atoms, leading to electron-deficient surfaces

of both supports and Pd0 and so, as a result, the adsorption of H2 and
unsaturated bonds (electron-rich) of FAME molecules on the cata-
lysts was augmented leading to an enhanced hydrogenation of the
adsorbed FAME molecules, as described by Scheme 1 [44,56,57].
However, the decreased TOF levels detected in catalysts with higher
Zr levels (Pd/0.09Zr-SBA-15 and Pd/0.11Zr-SBA-15) may have re-
sulted from the steric hindrance of the strongly adsorbed-reacted
FAME molecules on the catalyst surface. Another possibility was the
reduced textural properties (Table 1), especially the decreased DP,
compared to the other catalysts, which would lead to a more difficult
diffusion for the FAME molecules through the pore channels of the
supports, in which the Pd0 active sites were located, and hence a
longer reaction time than that for the Pd/0.07Zr-SBA-15 catalyst.
Analogously, the reduction in the D P during the synthesis of the SBA-
15 materials with high Zr loadings, which contributed to their lower
catalytic conversion, was previously proposed by Yusoff and Ab-
dullah [58].
At an 80% C18:2 conversion level, the levels of TOF for almost all
the catalysts were lower than those in the corresponding 10% C18:2
conversion level, presumably due to the lower amount of both basic
C]C bonds in the FAME molecules as well as unoccupied Pd0 active
sites of the catalyst. This could suggest that the influence of the elec-
tron-deficient catalyst surfaces might abate, and so the time required
for obtaining a high C18:2 conversion level was quite long. As evident
in Fig. 9, the trends in the TOF changes in most catalysts at an 80%
C18:2 conversion level were similar to those at a 10% C18:2 conversion
level. More specifically, a slight increase in TOF for the Pd/0.01Zr-SBA-
15 and Pd/0.11Zr-SBA-15 catalysts was evident and the highest TOF
level was acquired from the Pd/0.07Zr-SBA-15 catalyst. This phenom-
Fig. 6. Representative ATR-FTIR spectra of the (a) SO, (b) SO-FAMEs and the enon could be explained separately for each catalyst.
PH-biodiesel using (c) 0.5 wt% Pd/SBA-15, (d) 0.75 wt% Pd/SBA-15 and (e) For Pd/0.01Zr-SBA-15, the increase in TOF might arise from the
1 wt% Pd/SBA-15 catalysts. gradual but continuous adsorption and desorption of FAME mole-
cules on the catalyst surface, allowing access to other FAME mole-
cules without any hindrance from the reacted FAME molecules. In
contrast, the insignificant increase in TOF for Pd/0.11Zr-SBA-15,
3.3.2. Effect of the Pd/xZr-SBA-15 catalysts
compared to the other catalysts at an 80% C18:2 conversion level,
The Pd/xZr-SBA-15 catalysts were examined for their catalytic be-
might result from the abundance of Zr sites on the catalyst that en-
haviors in the partial hydrogenation of the SO-FAMEs under the same
hanced the adsorption possibility of unreacted unsaturated FAMEs.
conditions as that used for the Pd/SBA-15 catalyst (Section 3.3.1), ex-
In spite of having the highest Zr loading level, its TOF was a little
cept at a catalyst loading level and total reaction time of 0.75 wt% and
lower than that of Pd/0.07Zr-SBA-15, as previously described, sug-
2 h, respectively.
gesting that the approach of other FAME molecules was blocked by
The hydrogenation activity of all the catalysts was considered in
the adsorbed-reacted molecules located at the high electron-deficient
terms of the TOF, derived using Eq. (1), which when compared at the
surface of the Pd/0.11Zr-SBA-15 catalyst, together with the worse
conversion levels of C18:2 were around 10% and 80%.
physical properties. In brief, the optimal physicochemical properties,
TOF (h−1) = ⎡
Converted C18: 2 × Amount of biodiesel (g) × (MW of C18: 2)−1 ⎤ i.e. the amount of Zr sites and fine textural properties, were the
⎢ metal loading× metal dispersion × Wcatalyst (g) × (MW of metal)−1 ⎥ ÷ time
⎣ ⎦ important factors to achieve a high TOF in this hydrogenation re-
(1) action.
The trans-C18:1 selectivity at a 10% C18:2 conversion level was
The alterations in the FAME composition at a low (10%) and high
very small and equivalent (~0.10–0.12) for all the catalysts, which
(80%) C18:2 conversion level are exhibited in Table 2, while the TOF
resulted from the low isomerization from cis-C18:1 to trans-C18:1 at
and trans-C18:1 selectivity at both conversion levels for all catalysts are
the beginning of the reaction. In addition, it was specifically ob-
shown in Fig. 9.
served that the lowest trans-C18:1 selectivity (around 0.10) was
At a C18:2 conversion level of around 10%, the FAME composition
found with the Pd/0.07Zr-SBA-15 catalyst. However, at an 80%
in the PH-biodiesel was only slightly altered from that in the feed SO-
C18:2 conversion level the trans-C18:1 selectivity was significantly
FAMEs, where the C18:3 and C18:2 contents were diminished, and the
larger than that at a 10% C18:2 conversion level, which was due to
amount of C18:1 was increased with only a limited level of C18:0 for-
the fact that the trans-configuration is produced together with the
mation (Table 2). As noticed in Fig. 9, the lowest TOF (around
hydrogenation reaction. The selectivity towards trans-C18:1 was in-
4279 h−1) was presented by Pd/SBA-15. The TOF values increased with
significantly increased with increasing Zr/Si molar ratios of the
increasing Zr/Si molar ratios of the synthesized supports up to 0.07, and
support materials. The stronger adsorption of the basic C]C bonds of
then slightly decreased thereafter at higher Zr/Si molar ratios.
FAME molecules onto the acid catalyst surfaces would result in a
With respect to the increased TOF, the major reason for this
higher level of the trans-isomer [59].
phenomenon was the change in the acidity of the supports through

431
C. Thunyaratchatanon et al. Fuel Processing Technology 179 (2018) 422–435

Fig. 7. FAME composition of the PH-biodiesel during the 4-h reaction using (A) 0.5 wt%, (B) 0.75 wt% and (C) 1 wt% Pd/SBA-15 catalysts compared to the feed SO-
FAMEs.

3.4. Properties of the feed SO-FAMEs and PH-biodiesels described previously, the incorporation of Zr into the parent SBA-15
together with suitable physicochemical properties, which were pro-
The fuel properties (oxidative stability, CP and PP) of the feed SO- vided by the 0.05Zr-SBA-15 and 0.07Zr-SBA-15 materials, promoted the
FAMEs and the PH-biodiesels are tabulated in Table 3. The poly- acceleration of the basic FAME molecule adsorption on the Pd catalyst
unsaturated FAMEs contents are important determinants in the oxida- surface. As a result, the hydrogenation activity was improved and
tive stability of biodiesel. The feed SO-FAMEs had an abundant poly- consequently enhanced the oxidative stability above that of Pd/SBA-15.
unsaturated composition and so an oxidative stability of only 2 h, which In the case of the 2-h hydrogenation, the oxidative stability of the PH-
would be considered as a substandard biodiesel according to the Eur- biodiesels was decreased compared with the 4-h reaction, being in the
opean Standard (EN14214:2012) and Thai standard, by which liquid range of 32–60 h (4- to 7-fold higher than the European standard),
fuel has to resist oxidation for at least 8 h [60] and 10 h [52], respec- whilst the CP and PP were improved to around 6–10 °C and 3–6 °C,
tively. The polyunsaturated FAME contents were reduced by the partial respectively. To apply these PH-biodiesels in cold climate countries,
hydrogenation, leading to a better oxidative stability. With respect to other techniques, such as the addition of chemical additives, blending
the catalyst loading level compositions in a 4-h reaction, the amount of with other diesel fuels and winterization, would have to be used to
C]C bonds in the FAME molecules was markedly hydrogenated to improve their cold flow properties [61,62].
saturated, where higher catalyst loading levels gave larger proportions
of C18:0 in the obtained PH-biodiesel, and consequently a better oxi- 4. Conclusions
dative stability. Nevertheless, the cold flow properties of the PH-bio-
diesels were considerably worsened (CP of ~25–27 °C and PP of The influence of the Pd/SBA-15 catalyst loading level (0.5, 0.75 and
~24–26 °C). With respect to the various xZr-SBA-15 supports, the re- 1 wt%) was investigated in the partial hydrogenation of SO-FAMEs to
action time was reduced to 2 h in order to decrease the level of C18:0 improve the oxidative stability of the biodiesel. The highest saturated
formation, in order to obtain a compromise between the oxidative FAME content was produced with 1 wt% catalyst in a 4-h hydrogena-
stability and cold flow properties of the resultant PH-biodiesel. As tion reaction, giving an excellent oxidative stability but a substantially

432
C. Thunyaratchatanon et al. Fuel Processing Technology 179 (2018) 422–435

Fig. 9. Ratio of trans-C18:1/cis-C18:1 in the PH-biodiesels and TOF at the C18:2

conversion levels of around (A) 10% and (B) 80%.

performance. The decrease level of polyunsaturated FAMEs through

Fig. 8. Ratio of trans-C18:1/cis-C18:1 and the C18:2 conversion level in the PH- their partial catalytic hydrogenation could markedly improve the oxi-
biodiesel at a 2-h reaction using various amounts of Pd/SBA-15 catalyst. dative stability of the PH-biodiesels, even though the CP and PP were
inevitably worsened.
worsened CP and PP. To compromise between these fuel properties,
partial hydrogenation was then performed for 2 h with 0.75 wt% Pd/
xZr-SBA-15 catalyst. The modification of the SBA-15 support by the Acknowledgements
incorporation of Zr atoms (Zr/Si molar ratios of 0.01–0.11) was also
examined for their catalytic performance in this reaction. All the Pd/ The authors acknowledge the contributions and financial support of
xZr-SBA-15 catalysts had a higher TOF at 10 and 80% C18:2 conversion the following organizations: the 24th Year PPC Fund for Ph.D. Program;
levels compared to the Pd/SBA-15 catalyst. This was due to the phy- the Ratchadapisake Sompote Endowment Fund; and Center of Excellence
sicochemical properties of the catalysts. The electron-deficient catalyst for Petrochemical and Materials Technology, Chulalongkorn University for
surfaces, ascribed to the existence of Zr species, caused an enhanced fellowship support to Ms. Chachchaya Thunyaratchatanon. This research
adsorption of H2 and basic unsaturated FAMEs, and consequently an was supported by the Government Budget (GRB_APS_49_59_63_08) and by
improved hydrogenation activity. Nevertheless, increasing the Zr/Si Chulalongkorn University (CU-GES-60-04-63-03), Thailand. We are also
molar ratio up to 0.07 resulted in a much stronger adsorption of reacted much obliged to the Bangchak Corporation P.C.L., and Renewable Energy
FAME molecules on the catalyst surface, leading to stearic hindrance of Laboratory, National Metal and Materials Technology Center (MTEC),
other FAME accessibility. Also, the fine textural characteristic, i.e. pore Thailand for instrumental analysis support.
width, was the necessary determinant for providing the high catalytic

Table 2
FAME composition, reaction time and TOF of the feed SO-FAMEs and the PH-biodiesels using Pd/xZr-SBA-15 catalysts at a 10% and 80% C18:2 conversion level.
Catalyst FAME composition (%)a Reaction time (h) TOF (h−1)

C18:3 C18:2 C18:1 C18:0

Feed SO-FAMEs 5.6 50.2 25.5 3.9 – –

10% C18:2 conversion level

Pd/SBA-15 2.8 45.1 34.2 4.0 0.3 4279.6
Pd/0.01Zr-SBA-15 3.1 45.1 33.5 4.0 0.3 6227.9
Pd/0.03Zr-SBA-15 3.7 45.1 33.5 4.0 0.3 6363.0
Pd/0.05Zr-SBA-15 3.8 45.1 33.2 4.2 0.2 9146.4
Pd/0.07Zr-SBA-15 3.9 45.1 33.1 4.2 0.2 11,740.5
Pd/0.09Zr-SBA-15 3.8 45.1 33.4 4.2 0.2 9078.2
Pd/0.11Zr-SBA-15 3.7 45.0 33.4 4.2 0.2 8927.0

80% C18:2 conversion level

Pd/SBA-15 – 10.2 71.2 5.0 1.7 5590.0
Pd/0.01Zr-SBA-15 – 10.2 70.9 4.9 1.9 7728.1
Pd/0.03Zr-SBA-15 – 10.1 71.3 4.9 2.0 6527.9
Pd/0.05Zr-SBA-15 – 10.2 71.1 5.2 1.8 7004.1
Pd/0.07Zr-SBA-15 – 10.1 71.3 5.4 1.8 7812.5
Pd/0.09Zr-SBA-15 – 10.1 71.0 5.4 1.9 6726.0
Pd/0.11Zr-SBA-15 – 10.2 71.1 5.2 1.9 7609.7

a
When the C18:2 conversion level was approximately 10% or 80%.

433
C. Thunyaratchatanon et al. Fuel Processing Technology 179 (2018) 422–435

Scheme 1. Schematic representation of the hydrogenation mechanism on the Pd/SBA-15 and Pd/Zr-SBA-15 catalyst surfaces.

Table 3 References
The properties of the feed SO-FAMEs and the PH-biodiesels after a 4-h and a 2-h
reaction. [1] Thailand Department of Renewable Energy Development and Energy Efficiency,
The Alternative Energy Development Plan 2015–2036 (AEDP2015–2036), See the
Fuel properties of the biodiesel
following, 2015. http://www.eppo.go.th/images/POLICY/ENG/AEDP2015ENG.
pdf.
Oxidative stability (h)a CP (°C)b PP (°C)c [2] P. Verma, M.P. Sharma, G. Dwivedi, Impact of alcohol on biodiesel production and
properties, Renew. Sust. Energ. Rev. 56 (2016) 319–333.
Feed SO-FAMEs 2 2 0 [3] M. Faried, M. Samer, E. Abdelsalam, R.S. Yousef, Y.A. Attia, A.S. Ali, Biodiesel
PH-biodiesel after a 4-h reaction using various catalyst loading levels production from microalgae: processes, technologies and recent advancements,
0.5 wt%Pd/SBA-15 46 25 24 Renew. Sust. Energ. Rev. 79 (2017) 893–913.
0.75 wt%Pd/SBA-15 113 25 24 [4] B. Sajjadi, A.A.A. Raman, H. Arandiyan, A comprehensive review on properties of
1 wt%Pd/SBA-15 145 27 26 edible and non-edible vegetable oil-based biodiesel: composition, specifications and
prediction models, Renew. Sust. Energ. Rev. 63 (2016) 62–92.
PH-biodiesel after a 2-h reaction using various xZr-SBA-15 supportsd [5] P. Hongmanorom, A. Luengnaruemitchai, N. Chollacoop, Y. Yoshimura, Effect of
Pd/SBA-15 43 10 6 the Pd/MCM-41 pore size on the catalytic activity and cis–trans selectivity for
Pd/0.01Zr-SBA-15 37 9 3 partial hydrogenation of canola biodiesel, Energy Fuel 31 (2017) 8202–8209.
Pd/0.03Zr-SBA-15 32 6 3 [6] N. Numwong, A. Luengnaruemitchai, N. Chollacoop, Y. Yoshimura, Effect of SiO2
Pd/0.05Zr-SBA-15 60 9 6 pore size on partial hydrogenation of rapeseed oil-derived FAMEs, Appl. Catal. A-
Pd/0.07Zr-SBA-15 53 7 3 Gen. 441–442 (2012) 72–78.
Pd/0.09Zr-SBA-15 35 6 3 [7] B.S. Souza, D.M.M. Pinho, E.C. Leopoldino, P.A.Z. Suarez, F. Nome, Selective partial
Pd/0.11Zr-SBA-15 37 6 3 biodiesel hydrogenation using highly active supported palladium nanoparticles in
imidazolium-based ionic liquid, Appl. Catal. A-Gen. 433–434 (2012) 109–114.
a [8] S.-Y. Chen, L. Attanatho, T. Mochizuki, Y. Abe, M. Toba, Y. Yoshimura,
Determined by the Rancimat method according to EN 14214.
b C. Kumpidet, P. Somwonhsa, S. Lao-ubol, Upgrading of palm biodiesel fuel over
Evaluated according to ASTM D2500. supported palladium catalysts, C. R. Chim. 19 (2016) 1166–1173.
c
Measured according to ASTM D97. [9] Y. Zhao, Y. Ren, R. Zhang, L. Zhang, D. Yu, L. Jiang, W. Elfalleh, Preparation of
d
Note: Amount of utilized catalyst was 0.75 wt% compared to the SO- hydrogenated soybean oil of high oleic oil with supported catalysts, Food Biosci. 22
FAMEs. (2018) 91–98.
[10] K. Wadumesthrige, S.O. Salley, K.Y.S. Ng, Effects of partial hydrogenation, epox-
idation, and hydroxylation on the fuel properties of fatty acid methyl esters, Fuel
Appendix A. Supplementary data Process. Technol. 90 (2009) 1292–1299.
[11] P. Stathis, D. Stavroulaki, N. Kaika, K. Krommyda, G. Papadogianakis, Low trans-
isomers formation in the aqueous-phase Pt/TPPTS-catalyzed partial hydrogenation
TEM micrographs and EDS spectra of all catalysts, total acid sites of of methyl esters of linseed oil, Appl. Catal. B-Environ. 209 (2017) 579–590.
pure SBA-15 and xZr-SBA-15 supports and wide-scan XPS spectra of [12] Y. Song, U. Sanyal, D. Pangotra, J.D. Holladay, D.M. Camaioni, O.Y. Gutiérrez,
catalysts. Supplementary data to this article can be found online at J.A. Lercher, Hydrogenation of benzaldehyde via electrocatalysis and thermal cat-
alysis on carbon-supported metals, J. Catal. 359 (2018) 68–75.
https://doi.org/10.1016/j.fuproc.2017.02.007.
[13] A.B. Bindwal, A.H. Bari, P.D. Vaidya, Kinetics of low temperature aqueous-phase
hydrogenation of model bio-oil compounds, Chem. Eng. J. 207–208 (2012)
725–733.
[14] S. Li, Y. Wang, L. Gao, Y. Wu, X. Yang, P. Sheng, G. Xiao, Short channeled Ni-co/
SBA-15 catalysts for highly selective hydrogenation of biomass-derived furfural to

434
C. Thunyaratchatanon et al. Fuel Processing Technology 179 (2018) 422–435

tetrahydrofurfuryl alcohol, Microporous Mesoporous Mater. 262 (2018) 154–165. [39] S.K. Martha, J. Nanda, G.M. Veith, N.J. Dudney, Surface studies of high voltage
[15] P.M. Carraro, A.A.G. Blanco, C. Chanquía, K. Sapag, M.I. Oliva, G.A. Eimer, lithium rich composition: Li1.2Mn0.525Ni0.175Co0.1O2, J. Power Sources 216 (2012)
Hydrogen adsorption of nickel-silica materials: role of the SBA-15 porosity, 179–186.
Microporous Mesoporous Mater. 248 (2017) 62–71. [40] M.M. Sabri, J. Jung, D.H. Yoon, S. Yoon, Y.J. Tak, H.J. Kim, Hydroxyl radical-
[16] N. Bouazizi, R. Ouargli, S. Nousir, R.B. Slama, A. Azzouz, Properties of SBA-15 assisted decomposition and oxidation in solution-processed indium oxide thin-film
modified by iron nanoparticles as potential hydrogen adsorbents and sensors, J. transistors, J. Mater. Chem. C 3 (2015) 7499–7505.
Phys. Chem. Solids 77 (2015) 172–177. [41] A. Sabbaghi, F.L.Y. Lam, X. Hu, High Zr-loaded SBA-15 cobalt catalyst for efficient
[17] I. Yuranov, L. Kiwi-Minsker, A. Renken, Structured combustion catalysts based on NOx reduction in lean-burn exhaust, Appl. Catal. A Gen. 508 (2015) 25–36.
sintered metal fibre filters, Appl. Catal. B Environ. 43 (2003) 217–227. [42] H. Ahn, H.-W. Chen, D. Landheer, X. Wu, L.J. Chou, T.-S. Chao, Characterization of
[18] L. Yang, X. Yang, E. Tian, V. Vattipalli, W. Fan, H. Lin, Mechanistic insights into the interfacial layer of ultrathin Zr silicate on Si (100) using spectroscopic ellipsometry
production of methyl lactate by catalytic conversion of carbohydrates on meso- and HRTEM, Thin Solid Films 455–456 (2004) 318–322.
porous Zr-SBA-15, J. Catal. 333 (2016) 207–216. [43] S. Ogawa, T. Fujimoto, T. Mizutani, M. Ogawa, N. Uchiyama, K. Kato, T. Ohta,
[19] L. Fuxiang, Y. Feng, L. Yongli, L. Ruifeng, X. Kechang, Direct synthesis of Zr-SBA-15 T. Yoshida, S. Yagi, Hydrogen storage of binary nanoparticles composed of Mg and
mesoporous molecular sieves with high zirconium loading: characterization and Pd, Int. J. Hydrog. Energy 40 (2015) 11895–11901.
catalytic performance after sulphated, Microporous Mesoporous Mater. 101 (2007) [44] Y. He, J. Fan, J. Feng, C. Luo, P. Yang, D. Li, Pd nanoparticles on hydrotalcite as an
250–255. efficient catalyst for partial hydrogenation of acetylene: effect of support acidic and
[20] Z. Li, J. Wu, J. Yu, D. Han, L. Wu, J. Li, Effect of incorporation manner of Zr on the basic properties, J. Catal. 331 (2015) 118–127.
Co/SBA-15 catalyst for the Fischer–Tropsch synthesis, J. Mol. Catal. A Chem. 424 [45] K. Zhang, C. Wang, D. Bin, J. Wang, B. Yan, Y. Shiraishi, Y. Du, Fabrication of Pd/P
(2016) 384–392. nanoparticle networks with high activity for methanol oxidation, Catal. Sci.
[21] W.-K. Chen, H.-H. Tseng, M.-C. Wei, E.-C. Su, I.-C. Chiu, Transesterification of ca- Technol. 6 (2016) 6441–6447.
nola oil as biodiesel over Na/Zr-SBA-15 catalysts: effect of zirconium content, Int. J. [46] S. Cheng, N. Shang, X. Zhou, C. Feng, S. Gao, C. Wang, Z. Wang, High catalytic
Hydrog. Energy 39 (2014) 19555–19562. activity of bimetallic AgPd alloy supported on UiO-66 derived porous carbon for
[22] A.L. de Carvalho, E.A. Cardoso, G.O. da Rocha, L.S.G. Teixeira, I.M. Pepe, transfer hydrogenation of nitroarenes using formic acid-formate as hydrogen
D.M. Grosjean, Carboxylic acid emissions from soybean biodiesel oxidation in the source, New J. Chem. 41 (2017) 9857–9865.
EN14112 (Rancimat) stability test, Fuel 173 (2016) 29–36. [47] L. Yang, X. Yang, S. Lin, R. Zhou, Insights into the role of a structural promoter (Ba)
[23] M. Tesfaye, V. Katiyar, Microwave assisted synthesis of biodiesel from soybean oil: in three-way catalyst Pd/CeO2–ZrO2 using in situ DRIFTS, Catal. Sci. Technol. 5
effect of poly (lactic acid)-oligomer on cold flow properties, IC engine performance (2015) 2688–2695.
and emission characteristics, Fuel 170 (2016) 107–114. [48] K. Idczak, P. Mazur, S. Zuber, L. Markowski, M. Skiścim, S. Bilińska, Growth of thin
[24] D. Zhao, J. Feng, Q. Huo, N. Melosh, G.H. Fredrickson, B.F. Chmelka, G.D. Stucky, zirconium and zirconium oxides films on the n-GaN (0001) surface studied by XPS
Triblock copolymer syntheses of mesoporous silica with periodic 50 to 300 ang- and LEED, Appl. Surf. Sci. 304 (2014) 29–34.
strom pores, Science 279 (1998) 548–552. [49] T.G. Kudre, N. Bhaskar, P.Z. Sakhare, Optimization and characterization of bio-
[25] S.-Y. Chen, J.-F. Lee, S. Cheng, Pinacol-type rearrangement catalyzed by Zr-in- diesel production from rohu (Labeo rohit) processing waste, Renew. Energy 113
corporated SBA-15, J. Catal. 270 (2010) 196–205. (2017) 1408–1418.
[26] C. Thunyaratchatanon, A. Luengnaruemitchai, T. Chaisuwan, N. Chollacoop, S.- [50] N. Santiago-Torres, I.C. Romero-Ibarra, H. Pfeiffer, Sodium zirconate (Na2ZrO3) as a
Y. Chen, Y. Yoshimura, Synthesis and characterization of Zr incorporation into catalyst in a soybean oil transesterification reaction for biodiesel production, Fuel
highly ordered mesostructured SBA-15 material and its performance for CO2 ad- Process. Technol. 120 (2014) 34–39.
sorption, Microporous Mesoporous Mater. 253 (2017) 18–28. [51] D.A. Torres-Rodríguez, I.C. Romero-Ibarra, I.A. Ibarra, H. Pfeiffer, Biodiesel pro-
[27] Y. Wei, Y. Li, Y. Tan, Z. Wu, L. Pan, Y. Liu, Short-channeled mesoporous Zr–Al-SBA- duction from soybean and Jatropha oils using cesium impregnated sodium zirco-
15 as a highly efficient catalyst for hydroxyalkylation of phenol with formaldehyde nate as a heterogeneous base catalyst, Renew. Energy 93 (2016) 323–331.
to bisphenol F, Chem. Eng. J. 298 (2016) 271–280. [52] N. Numwong, A. Luengnaruemitchai, N. Chollacoop, Y. Yoshimura, Partial hydro-
[28] A. Sabbaghi, F.L.Y. Lam, X. Hu, Zr-SBA-15 supported Ni catalysts for lean NOx genation of polyunsaturated fatty acid methyl esters over Pd/activated carbon:
reduction, J. Mol. Catal. A Chem. 409 (2015) 69–78. effect of type of reactor, Chem. Eng. J. 210 (2012) 173–181.
[29] A. Rodríguez-Gómez, F. Platero, A. Caballero, G. Colón, Improving the direct [53] G. Knothe, Some aspects of biodiesel oxidative stability, Fuel Process. Technol. 88
synthesis of hydrogen peroxide from hydrogen and oxygen over Au-Pd/SBA-15 (2007) 669–677.
catalysts by selective functionalization, Mol. Catal. 445 (2018) 142–151. [54] Y. Hou, L. Xu, Z. Wei, Y. Liu, X. Li, S. Deng, Reaction process and kinetics of the
[30] O.Y. Gutiérrez, F. Pérez, G.A. Fuentes, X. Bokhimi, T. Klimova, Deep HDS over selective hydrogenation of resorcinol into 1,3-cyclohexanedione, J. Taiwan Inst.
NiMo/Zr-SBA-15 catalysts with varying MoO3 loading, Catal. Today 130 (2008) Chem. Eng. 45 (2014) 1428–1434.
292–301. [55] C. Thunyaratchatanon, J. Jitjamnong, A. Luengnaruemitchai, N. Numwong,
[31] N.K. Nag, A study on the dispersion and catalytic activity of gamma alumina-sup- N. Chollacoop, Y. Yoshimura, Influence of mg modifier on cis-trans selectivity in
ported palladium catalysts, Catal. Lett. 24 (1994) 37–46. partial hydrogenation of biodiesel using different metal types, Appl. Catal. A-Gen.
[32] C.D. Evans, S.A. Kondrat, P.J. Smith, T.D. Manning, P.J. Miedziak, G.L. Brett, 520 (2016) 170–177.
R.D. Armstrong, J.K. Bartley, S.H. Taylor, M.J. Rosseinsky, G.J. Hutchings, The [56] S.D. Lin, M.A. Vannice, Hydrogenation of aromatic hydrocarbons over supported Pt
preparation of large surface area lanthanum based perovskite supports for AuPt catalysts. III. Reaction models for metal surfaces and acidic sites on oxide supports,
nanoparticles: tuning the glycerol oxidation reaction pathway by switching the J. Catal. 143 (1993) 563–572.
perovskite B site, Faraday Discuss. 188 (2016) 427–450. [57] J. Zhao, H. Chen, J. Xu, J. Shen, Effect of surface acidic and basic properties of the
[33] S. Grimme, J.-P. Djukic, Cation–Cation “Attraction”: when London dispersion at- supported nickel catalysts on the hydrogenation of pyridine to piperidine, J. Phys.
traction wins over coulomb repulsion, Inorg. Chem. 50 (2011) 2619–2628. Chem. C 117 (2013) 10573–10580.
[34] P. Pyykko, N. Runeberg, F. Mendizabal, Theory of the d10–d10 closed-Shell attrac- [58] M.H.M. Yusoff, A.Z. Abdullah, Effects of zirconia loading in sulfated zirconia/SBA-
tion: 1. Dimers near equilibrium, Chem. Eur. J. 3 (1997) 1451–1457. 15 on esterification of palmitic acid with glycerol, Korean J. Chem. Eng. 35 (2018)
[35] C. Autthanit, B. Jongsomjit, Differences in characteristics of Zr/SBA-15 and bime- 383–393.
tallic Zr-La/SBA-15 prepared by sol–gel and hydrothermal methods, J. Porous. [59] H. Iida, D. Itoh, S. Minowa, A. Yanagisawa, A. Igarashi, Hydrogenation of soybean
Mater. 24 (2017) 1383–1394. oil over various platinum catalysts: effects of support materials on trans fatty acid
[36] K.K. Cheralathan, H. Hirata, M. Ogura, Assembling mode of alumina and zirconia levels, Catal. Commun. 62 (2015) 1–5.
particles inside the mesopores of SBA-15 under high loading, Stud. Surf. Sci. Catal. [60] M. Serrano, R. Oliveros, M. Sánchez, A. Moraschini, M. Martínez, J. Aracil,
174 (2008) 161–166. Influence of blending vegetable oil methyl esters on biodiesel fuel properties: oxi-
[37] G. Morales, A. Osatiashtiani, B. Hernández, J. Iglesias, J.A. Melero, M. Paniagua, dative stability and cold flow properties, Energy 65 (2014) 109–115.
D.R. Brown, M. Granollers, A.F. Lee, K. Wilson, Conformal sulfated zirconia [61] S. Nainwal, N. Sharma, A.S. Sharma, S. Jain, S. Jain, Cold flow properties im-
monolayer catalysts for the one-pot synthesis of ethyl levulinate from glucose, provement of Jatropha curcas biodiesel and waste cooking oil biodiesel using
Chem. Commun. 50 (2014) 11742–11745. winterization and blending, Energy 89 (2015) 702–707.
[38] A.I.M. Rabee, G.A.H. Mekhemer, A. Osatiashtiani, M.A. Isaacs, A.F. Lee, K. Wilson, [62] V. Makarevičienė, K. Kazancev, I. Kazanceva, Possibilities for improving the cold
M.I. Zaki, Acidity-reactivity relationships in catalytic esterification over ammonium flow properties of biodiesel fuel by blending with butanol, Renew. Energy 75
sulfate-derived sulfated zirconia, Catalysts 7 (7) (2017) 204–220. (2015) 805–807.

435

View publication stats