Bioresource Technology 101 (2010) S59–S61

Contents lists available at ScienceDirect

Bioresource Technology
journal homepage: www.elsevier.com/locate/biortech

Esterification of used vegetable oils using the heterogeneous WO3/ZrO2 catalyst

for production of biodiesel
Young-Moo Park a, Joon Yeob Lee a, Sang-Ho Chung a, In Seon Park a, Seung-Yeon Lee a,
Deog-Keun Kim b, Jin-Suk Lee b, Kwan-Young Lee a,*
a
Department of Chemical and Biological Engineering, Korea University, 5-1, Anam-dong, Sungbuk-ku, Seoul 136-701, Republic of Korea
b
Biomass Research Center, Korea Institute of Energy Research, Yuseong 305-600, Taejeon, Republic of Korea

a r t i c l e i n f o a b s t r a c t

Article history: Tungsten oxide zirconia, sulfated zirconia and Amberlyst-15 were examined as a catalyst for a conversion
Received 17 December 2008 of used vegetable oils (VOs) to fatty acid methyl esters (FAMEs). Among them, tungsten oxide zirconia
Received in revised form 4 April 2009 was a promising heterogeneous catalyst for the production of biodiesel fuels from used VOs because of
Accepted 10 April 2009
high activity in the conversion over 93% and no leaching WO3 in the esterification reaction. The reaction
Available online 9 May 2009
conditions were optimized. A study for optimizing the reaction parameters such as the reaction temper-
ature, stirring speed, WO3 loading over ZrO2 and reaction time, was carried out. The catalyst was charac-
Keywords:
terized by BET, XRD, FT-IR, and NH3-TPD. With increasing WO3 loading over ZrO2, the triclinic phase of
Biodiesel
Esterification
WO3 increased and the tetragonal phase of zirconia was clearly generated. The highest acid strength of
Tungsten oxide zirconia 20 wt% tungsten oxide zirconia catalyst was confirmed by NH3-TPD analysis and the result was correlated
Heterogeneous catalyst to the highest catalytic activity of the esterification reaction.
Free fatty acid Ó 2009 Elsevier Ltd. All rights reserved.

1. Introduction
Recently, many countries have reported rapid progress in
renewable fuels, with biodiesel being the most well known. Biodie-
sel fuel obtained from vegetable oils (VOs) has attracted consider-
able attention over recent decades as a renewable, biodegradable,
eco-friendly and non-toxic fuel, and several processes have been
developed for its production. Biodiesel is gaining prominence as
a substitute for petroleum-based diesel due to environmental con-
siderations and the depletion of vital resources, such as petroleum
and coal. The possible uses of renewable resources are fuels and
major feedstocks for the chemical industry. Petroleum is a rapidly
depleting natural resource, making an alternative renewable route
to petroleum a necessity. Furthermore, biodiesel is a clean burning
fuel with no sulfur emissions and no efficiency loss despite its heat
of combustion being slightly lower than that of petro-diesel (Sri-
vastava and Prasad, 2000), and no engine adjustments are neces-
sary. In addition, it is non-corrosive and can be produced under
low pressure and temperature conditions.
Biodiesel is commonly derived from triglycerides (VOs) and
methanol through transesterification using homogeneous cata-
lysts, such as alkali hydroxides (NaOH and KOH), carbonates, and
sodium and potassium alkoxides (Canakci and Van Gerpen, 2001,
2003; Dorado et al., 2002; Ghadge and Raheman, 2005). A gener-
* Corresponding author. Tel.: +82 2 3290 3299; fax: +82 2 926 6102.
E-mail address: kylee@korea.ac.kr (K.-Y. Lee).
ally mild process for production of biodiesel from VOs using heter-
ogeneous base catalysts was developed in our previous study (Kim
et al., 2004).
Despite the environmental advantage of the biodiesel, produc-
tion cost of biodiesel is pretty high compared to conventional die-
sel fuel. In order to reduce the production cost, used VOs has been
suggested as feedstock. However, the free fatty acids (FFAs) in used
VOs are known to cause severe problems for the transesterification
catalyzed by base catalyst. FFA can react with the base catalyst
(neutralization reaction), which brings a loss of catalyst and a pro-
duction of soap as by-product which accelerates the deactivation
of the base catalyst (Park et al., 2008). Consequently, FFA contained
in used VOs should be removed or converted to inert material. In
this work, we tried to convert FFA to fatty acid methyl esters
(FAMEs) using heterogeneous acid catalysts (Eq. (1)), which are
the components of biodiesel themselves. The catalytic esterifica-
tion activity of SO2
4 /ZrO2, WO3/ZrO2, and Amberlyst-15 was con-
firmed and compared. WO3/ZrO2 was finally chosen on account
of its high activity and stability. A series of experiments were per-
formed using WO3/ZrO2 with temperature, WO3 loading, amount
of catalyst (g/ml (oil)) and reaction time as parameters to deter-
mine the optimal reaction conditions. The WO3/ZrO2 catalyst was
characterized by BET, X-ray diffraction (XRD), Fourier transform
infrared (FT-IR) spectroscopy, NH3-temperature programmed
desorption (TPD).
Free fatty acids ðFFAsÞ þ methanol
! Fatty acid methyl esters ðFAMEsÞ þ water ð1Þ

0960-8524/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.biortech.2009.04.025
S60 Y.-M. Park et al. / Bioresource Technology 101 (2010) S59–S61
2. Methods
The SO2
4 /ZrO2 catalyst was prepared by the dehydration of
H2SO4 and Zr(OH)4. WO3/ZrO2 was synthesized by impregnating
on Zr(OH)4 with an ammonium metatungstate (Aldrich) solution.
A series of catalysts with different WO3 loadings ranging from 10
to 30 wt% were prepared and calcined at 800 °C. The Amberlyst-
15 was purchased from Rohm and Haas. The esterification reaction
of soybean oil contained 4 wt% oleic acid to simulate the used VOs
was carried out in a 100 ml autoclave batch reactor. Generally used
oil contains FFA of 1–3.7 wt% depending on the sources. The load-
ing ratio of used oil and methanol added in excess was 1:9. The
reactions were carried out with different stirring speeds, WO3
loadings, temperature, amount of catalyst and reaction time. The
conversion of oleic acid (FFA) to the methyl ester was confirmed
by a titration method reported elsewhere (Park et al., 2008). The
XRD patterns of the catalysts were obtained using a Rigaku
D/max 2000 ultima plus diffractometer (monochromatic nickel
filter, Cu Ka radiation). An FT-IR spectrometer (Perkin–Elmer
Spectrum GX1) was used to acquire the transmission spectra. The
BET surface area and pore volume of the prepared catalysts were
measured using an ASAP 2100 BET surface analyzer. The acidity
of the catalysts was measured using the temperature programmed
desorption of NH3 (NH3-TPD) in a 3/8 in. stainless steel. reactor
equipped with an on-line gas chromatograph (Agilent 6890 N)
and a mass spectrometer detector (Agilent 5975).
3. Results
3.1. Reaction of heterogeneous acid catalyst for biodiesel
The WO3/ZrO2, SO2 4 /ZrO2 and Amberlyst-15 catalysts were
used for the esterification of 4 wt% oleic acid in soybean oil with
methanol. The experiments were carried out under the following
conditions: temperature of 75 °C, reaction time of 2 h, vegetable
oil:methanol molar ratio of 1:9, 0.29 g of catalyst per ml of oil
and a stirring speed of 800 rpm. All catalysts were confirmed to
show approximately 93% FFA conversion. Amberlyst-15 is a poly-
meric catalyst and should not be used at temperature over
180 °C because heat can cause collapse of the catalyst structure.
The concern with the SO2 4 /ZrO2 catalyst is the possible leaching
of SO2 4 from the catalyst into the liquid phase reaction (Park
et al., 2008). Therefore, it makes sulfur content elevate in biodiesel
fuel. The WO3/ZrO2 catalyst reveals as high activity as others and
leaching or loss of WO3 has not been reported but if so, it is not
harmful. WO3/ZrO2 was chosen as the catalyst in this paper after
considering its relative stability.
3.2. Effect of reaction conditions
The effect of the WO3 loading in the catalyst on the esterifica-
tion of FFA was examined using WO3 loadings of 10–30 wt%.
A 10 wt% WO3/ZrO2 showed the lowest FFA conversion of 78%.
The conversion increased gradually with increasing WO3 loading:
20 wt% WO3/ZrO2 catalyst was confirmed to have the highest FFA
conversion of 93%. But 30 wt% WO3/ZrO2 showed lower FFA con-
version of 89%. The increase of FFA conversion correlated to in-
crease of WO3 loading until 20 wt% WO3/ZrO2. On the other hand
this correlation was not maintained over the 20 wt% WO3 loading.
Therefore, the optimal WO3 loading was concluded to be in 20 wt%.
The effect of the reaction temperature was examined from 75 to
200 °C using the 20 wt% WO3/ZrO2 catalyst. The FFA conversion in-
creased with increasing temperature. At 75 °C, FFA conversion was
93%, which increased to 98% at 200 °C. The increased conversion
might be due to not only the increase of reaction rate by high tem-
Fig. 1. TPD results of the WO3/ZrO2 catalyst.
perature but also some improvement of the mass transfer limita-
tions between the reactant and catalyst. Generally soybean oil
(contained 4 wt% oleic acid) was known as viscous material, and
it showed a tendency to be attached on the catalyst during the
reaction. As the reaction temperature increased, the viscosity de-
creased and the miscibility of the oil and methanol also increased,
resulting in an increase in FFA conversion.
In order to investigate how pure FFA affects the catalytic activ-
ity, the reactions were carried out with oleic acid and methanol
only. The test was carried out at temperatures ranging from 75
to 200 °C using 100 wt% FFA (oleic acid) under the same reaction
conditions. The FFA conversion increased from 80% to 94% with
increasing reaction temperature from 75 to 200 °C. The catalytic
activity decreased in high FFA content, but the activity was re-
stored as elevating reaction temperature. As the result, WO3/ZrO2
catalyst was found that it was possible to apply to the high temper-
ature reaction.
3.3. Catalyst characterization
XRD analysis of the catalysts with different WO3 loading was
performed to certify structural property. Crystalline WO3 and
ZrO2 peaks were confirmed in the XRD patterns. The formation of
a triclinic phase WO3 with increasing WO3 loading was identified
by three peaks in the range of 2h = 23–25°. A tetragonal phase of
zirconia which could not be present in pure ZrO2 appeared in the
WO3/ZrO2 catalyst. With increasing WO3 loading over ZrO2, the tri-
clinic phase of WO3 increased and the tetragonal phase of zirconia
was clearly generated. Thus, the structures mentioned here are
thought to affect an enhancement in catalytic activity.
BET analysis was carried out to determine the surface area and
pore volume of the catalyst. The pure zirconium hydroxide de-
gassed at 200 °C showed a surface area and pore volume of
137.2 m2/g and 0.07 cm3/g, respectively. After calcination at
800 °C, the surface area and pore volume of ZrO2 decreased to
18.1 m2/g and 0.019 cm3/g, respectively. However, the addition of
WO3 to ZrO2 resulted in an increase in surface area to 55.1 m2/g
at a 30 wt% WO3 loading, which means that the added WO3 could
reduce the sintering of ZrO2 (Iglesia et al., 1993). With increasing
crystallization in the XRD, the pores became narrow or closed,
and the specific surface area decreased. Hence, the surface area is
dependent on the WO3 loading.
FT-IR is an important tool for analyzing the nature of the surface
tungsten oxide species on zirconia. The FT-IR spectrum of the WO3/
ZrO2 catalyst shows one distinct absorption band centered at
approximately 791 cm1, which was assigned to a W–O vibration
band from crystalline and non-crystalline species as reported pre-
viously (Boyse and Ko, 1997). Consequently, the enhanced crystal-
 Y.-M. Park et al. / Bioresource Technology 101 (2010) S59–S61
linity of WO3 with increase of WO3 loading over ZrO2 was con-
firmed by the high intensity of the band for W–O vibration.
The acidic properties of the WO3/ZrO2 catalyst were examined
by NH3-TPD analysis. Fig. 1 shows the NH3-TPD profiles of the cat-
alysts. The TPD profiles showed that in all samples, the acid
strength was broadly distributed over the temperature range with
the highest peaks of the 25 and 30 wt% WO3/ZrO2 catalysts ob-
served at approximately 250 °C. Over all temperature regions, the
TPD profiles of 10, 15, 25 and 30 wt% WO3/ZrO2 catalysts were sim-
ilar with a single desorption peak. However, the profile of the
20 wt% WO3/ZrO2 catalyst was distinguished from the others,
showing two desorption peaks; the first at 250 °C and the second
at 750 °C. This second desorption shoulder showed that the acid
site had a stronger interaction with the adsorbed NH3, indicating
that 20 wt% WO3/ZrO2 catalyst has strong acid site compared to
the other catalysts. The strong acidity of the catalyst is thought
to be one of the reason for the highest catalytic activity of the
catalyst.
4. Conclusions
The effect of the catalytic properties on the esterification of FFA
to FAME was examined using WO3/ZrO2 catalysts in a batch reac-
tor. The FFA conversion was compared with those of the SO2
ZrO2 and Amberlyst-15 catalysts under optimized reaction condi-
tions. All catalysts showed approximately 93% FFA conversion.
However, SO2 4 /ZrO2 and Amberlyst-15 exhibited some structural
problems. The main concern with the SO2 4 /ZrO2 catalyst was the
potential loss of SO2 4 . On the other hand, the structure of the
Amberlyst-15 catalyst might be decomposed at reaction tempera-
tures over 180 °C. Increasing the WO3 loading favored the crystal-
lization of triclinic phase WO3 and generation of the tetragonal
phase of zirconia. The IR band corresponding to W–O vibration
was clearly identified at high WO3 loading. Among the WO3/ZrO2
S61
catalysts with WO3 loading amount from 10 to 30 wt%, 20 wt%
WO3/ZrO2 showed the highest catalytic activity. TPD analysis indi-
cated that the 20 wt% WO3/ZrO2 catalyst had higher acidity than
the other samples, as evidenced by the existence of a high temper-
ature desorption peak at 750 °C. Therefore, it can be concluded that
the strong acidity of the catalyst is thought to be one of the reason
for the highest catalytic activity of the 20 wt% WO3/ZrO2 catalyst.
Acknowledgement
The authors wish to acknowledge the support by the Resource
Recycling R&D Center, 21C Frontier R&D Program.
References
Boyse, R.A., Ko, E.I., 1997. Crystallization behavior of tungstate on zirconia and its
relationship to acidic properties. Journal of Catalysis 171, 191–207.
Canakci, M., Van Gerpen, J., 2003. A pilot plant to produce biodiesel from high free
fatty acid feedstocks. Transactions of the American Society of Agricultural
Engineers 46, 945–954.
Canakci, M., Van Gerpen, J., 2001. Biodiesel production from oils and fats with high
free fatty acids. Transactions of the American Society of Agricultural Engineers
44, 1429–1436.
Dorado, M.P., Ballesteros, E., de Almeida, J.A., Schellert, C., Lohrlein, H.P., Krause, R.,
2002. An alkali-catalyzed transesterification process for high free fatty acid
waste oils. Transactions of the American Society of Agricultural Engineers 45,
4 /
525–529.
Ghadge, S.V., Raheman, H., 2005. Biodiesel production from mahua (Madhuca indica)
oil having high free fatty acids. Biomass & Bioenergy 28, 601–605.
Iglesia, E., Soled, S.L., Kramer, G.M., 1993. Isomerization of alkanes on sulfated
zirconia: promotion by Pt and by adamantyl hydride transfer species. Journal of
Catalysis 144, 238–253.
Kim, H.J., Kang, B.S., Kim, M.J., Park, Y.M., Kim, D.K., Lee, J.S., Lee, K.Y., 2004.
Transesterification of vegetable oil to biodiesel using heterogeneous base
catalyst. Catalysis Today 93–95, 315–320.
Park, Y.-M., Lee, D.W., Kim, D.K., Lee, J.S., Lee, K.Y., 2008. The heterogeneous catalyst
system for the continuous conversion of free fatty acids in used vegetable oils
for the production of biodiesel. Catalysis Today 131, 238–243.
Srivastava, A., Prasad, R., 2000. Triglycerides-based diesel fuels. Renewable and
Sustainable Energy Reviews 4, 111–133.