Catalysis Communications 9 (2008) 2140–2143

Contents lists available at ScienceDirect

Catalysis Communications
journal homepage: www.elsevier.com/locate/catcom

A new zinc hydroxide nitrate heterogeneous catalyst for the esterification

of free fatty acids and the transesterification of vegetable oils
Claudiney Soares Cordeiro, Gregorio Guadalupe Carbajal Arizaga, Luiz Pereira Ramos, Fernando Wypych *
CEPESQ – Research Center in Applied Chemistry, Department of Chemistry, Federal University of Paraná – P.O. Box 19081, 81531-990 – Curitiba, PR, Brazil

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

Article history: A new heterogeneous catalyst for the esterification of free fatty acids and the transesterification of veg-
Received 11 February 2008 etable oils is reported. The layered compound zinc hydroxide nitrate (Zn5(OH)8(NO3)2  2H2O) was very
Received in revised form 18 April 2008 effective in the alcoholysis of palm oil and the esterification of lauric acid with m(ethanol), even when
Accepted 23 April 2008
hydrated ethanol was used. Over the range of 100–140 °C, the ester yield was the highest at 140 °C, while
Available online 29 April 2008
the catalyst concentration had a much greater effect on ester yields than the molar ratio of alcohol to acid
did. Total ester contents above 95 wt% were obtained in both reactions and 93.2 wt% glycerin streams
Keywords:
were recovered as a result of methanolysis.
Layered zinc hydroxide nitrate
Biodiesel
Ó 2008 Elsevier B.V. All rights reserved.
Esterification
Transesterification

1. Introduction been proposed in a series of recent articles [3–10]. In these stud-

The current efforts to reduce diesel exhaust emissions have led
to an increased demand for biodiesel. In Brazil, the current regu-
lations require the addition of 2 wt% of biodiesel to conventional
diesel (B2) since January 2008, aiming to extend this requirement
to B5 in 2013 [1]. This represents an annual demand of 800 mil-
lion liters in 2008, and an approximate increase in demand to
about 2 billion liters in 2013. Most of the biodiesel produced
worldwide is based on alkaline transesterification in homoge-
neous media with sodium methoxide as the reaction catalyst.
However, this technology requires expensive and complex down-
stream operations for biodiesel purification. A suitable alternative
to reduce costs and simplify operations is the use of a solid cata-
lyst. Since water washing would no longer be required for cata-
lyst removal, waste water streams would be considerably
reduced, and glycerin would be recovered in higher concentra-
tions and purity [2]. Although several solid materials have been
reported as catalytically active for the esterification of free fatty
acids and the transesterification of oils and fats, the development
of new heterogeneous catalysts capable of carrying out both reac-
tions simultaneously is still a fundamental need to improve pro-
duction cost through the use of cheap raw materials such as
rendered fats, used frying oil, tap greases, and other high acid
number lipids.
Esterification and transesterification reactions using heteroge-
neous catalysts based on layered double hydroxides (LDHs) has
* Corresponding author. Tel.: +55 41 3361 3473; fax: +55 41 3361 3186.
E-mail address: wypych@quimica.ufpr.br (F. Wypych).
ies, LDHs have been used either as the actual active phase or as
supports for the immobilization of exogenous catalytic species.
In the former case, catalysis is due to the alkaline characteristics
of the material: a recent study of one LDH, namely, Li/Al–OH
crystals, showed that the entire outer surface (i.e. both the basal
face and the edges) is catalytically active in transesterification
reactions [4]. In the latter case, the catalytic species can be immo-
bilized either at their surface or between the LDH structure
layers.
In search for new catalysts and with the stimulating example
of LDH, our research group is currently investigating a class of
layered compounds referred to as layered hydroxide salts (LHSs).
LHS compounds have anionic exchange capacity and their struc-
ture is based on only one divalent metal. They are represented
by the general formula Mþ2 ðOHÞ2x ðAn Þx=n  yH2 O where the diva-
lent metal (M+2) and the anion (An) have many possible varia-
tions [11]. Several compounds are now under evaluation,
however, the objective of this short communication is to demon-
strate the potential of zinc hydroxide nitrate (Zn5(OH)8-
(NO3)2  2H2O), hereafter referred to as ZHN, as an alternative
heterogeneous catalyst for esterification and transesterification
reactions. The monoclinic structure of ZHN consists of brucite-like
layers oriented parallel to the 0 0 1 plane. One quarter of the octa-
hedrally coordinated zinc cations in the ZHN structure are dis-
placed from the main layer to tetrahedral sites located above
and below each empty octahedron. The exchangeable nitrate ions
are surrounded by water molecules and located between the lay-
ers, where they are not directly coordinated to the zinc atoms
[11–13].

1566-7367/$ - see front matter Ó 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.catcom.2008.04.015
 C.S. Cordeiro et al. / Catalysis Communications 9 (2008) 2140–2143
2. Materials and methods
The refined palm oil (RPO) used in this study was donated by
Agropalma (Belém, PA, Brazil). Other reagents (lauric acid, metha-
nol, ethanol, solvents) were analytical grade. Zinc laurate was syn-
thesized as described previously [14].
ZHN was synthesized by chemical precipitation at pH 9.0 and
characterized by X-ray powder diffraction (XRPD), Fourier-trans-
formed infrared spectroscopy (FTIR) and thermal analysis. How-
ever, the latter results were not included in the current article.
The solid catalysts were applied in (trans)esterification reac-
tions without any further treatment. Experiments were carried
out with 2–6 wt% of ZHN catalyst in relation to the amount of fatty
material used for alcoholysis. The reactions were carried out in a
Büchiglass Miniclave Drive pressure vessel at different tempera-
tures and alcohol:lipid (lauric acid and palm oil) molar ratios.
The reaction vessel was sealed, but not pressurized, so that its
internal pressure corresponded to the vapor pressure of the most
volatile component of the reaction mixture. In all experiments,
the system was agitated at 500 rpm for 2 h, the ZHN catalyst was
then recovered from the reaction mixture by filtration and the ex-
cess alcohol was recovered by evaporation at low pressure. In the
esterification reactions, the acid number of each untreated alco-
hol-free reaction aliquot was determined using the American Oil
Chemist’s Society (AOCS method Ca-5a-40) official method. The
alkyl esters were quantified by high-performance liquid chroma-
tography (HPLC). The transesterification phase separation in alco-
hol-free reaction aliquots was induced by centrifugation and the
top and bottom layers were analyzed by HPLC for total alkyl esters
and total glycerin content, respectively.
HPLC analysis was carried out in a Shimadzu LC10AD worksta-
tion (Shimadzu Scientifics Inc., Kyoto, Japan) fit with a SIL10A auto-
injector and a RID10A refractive index detector. Quantification was
always performed by external calibration and 20 lL was routinely
used as the sample size for injection. Reverse-phase high perfor-
mance liquid chromatography (RP-HPLC) was used to determine
the total amount of alkyl esters in the final product (biodiesel).
To this end, a Waters Spherisorb C18 column (4.6  250 mm,
5 lm) was used at 35 °C in isocratic elution with 9:1 (v/v) acetoni-
Intensity (a.u.)
29.4Å
29.4Å
10.2Å
0 10 20
2*theta (°)
Fig. 1. XRPD patterns and FTIR spectra of ZHN before (A,B) and after (C,D) the first reaction step, compared to a freshly synthesized sample of zinc laurate (E,F).
2141
trile:acetone at 0.9 mL/min. Quantitative analysis was based on
standard solutions of alkyl esters and free fatty acids (lauric acid)
in 0.05–2 mg/mL concentration range. The RP-HPLC method could
not quantify final product (biodiesel) components in concentra-
tions below 1 wt%. Glycerin analysis was performed by ion
exchange chromatography in the same HPLC workstation. A Bio-
Rad Aminex HPX-87 H column (4.6  250 mm, 5 lm) was used in
isocratic elution with 8 mM H2SO4 at 0.6 mL/min at 65 °C. Quanti-
tative analysis was based on standard solutions of pure glycerol in
0.25–3.15 mg/mL concentration range.
3. Results and discussion
ZHN was characterized by XRPD and FTIR (Fig. 1). The solid cat-
alyst consisted of a pure phase with a basal distance of 10.2 Å (Fig.
1A) and its FTIR spectra was typical of ZHN as reported previously
(Fig. 1B) [12,13]. The thermal stability of ZHN was also investigated
by thermal analysis (data not shown) and the solid catalyst was
shown to be stable until 170 °C [13].
Table 1 gives the results of ZHN as a heterogeneous catalyst in
the esterification of lauric acid. The performance of ZHN was
strongly influenced by the reaction temperature, as indicated by
experiments 1–3, which were carried out under otherwise identi-
cal conditions. The ester content of the final ester layer, measured
after removing the excess alcohol by evaporation, increased from
39.4 wt% at 100 °C to 87.1 wt% at 140 °C.
Experiments 3–6 (Table 1) were performed at the same temper-
ature (140 °C) under various methanol:lauric acid molar ratios
(MRL) and catalyst concentrations (wt% in relation to the oil mass).
When the catalyst concentration was maintained at 2 wt% and MRL
varied from 6:1 to 2:1 (experiments 3 and 4), the total ester con-
tent decreased from 87.1 to 84.9 wt%, indicating that the catalytic
response of ZHN was not very sensitive to a threefold decrease in
MRL. Likewise, when the catalyst concentration was kept at
6 wt% and the same variation in MRL was imposed, the increase
in the final ester content was negligible (see experiments 5 and 6).
The effect of catalyst concentration on the total ester content
was more significant than changes in MRL. Experiments 3 and 5
were done at the same MRL of 6:1, but experiment 5 had a 3-fold
XRPD FTIR
2919
(F)
1542
2952
2848
(D)
1393
1457
(B)
(E)
1620
(C)
1443
1344
3590
3475
(A)
30 40 50 4000 3500 3000 1500 1000 500
-1
Wavenumber (cm )
2142 C.S. Cordeiro et al. / Catalysis Communications 9 (2008) 2140–2143

Table 1 water (Table 1). The final ester yields obtained at 100 °C were
Esterification of lauric acid with methanol and ethanol 23.1, 21.8 and 20.6%, respectively. Hence, there was a linear corre-
Experiment Alcohola MRbL Catalyst Temperature Estersc lation (y = 0.2525x + 23.104, R2 = 0.9992) between reaction effi-
(wt%) (°C) (wt%) ciency (variable y, measured as total ethyl ester content) and
1 Methanol 6:1 2 100 39.4 ethanol moisture content (variable x), which demonstrates that
2 Methanol 6:1 2 120 48.1 ZHN is partially deactivated by water. However, 89 wt% of the reac-
3 Methanol 6:1 2 140 87.1 tion yield still remained in the esterification of lauric acid with
4 Methanol 2:1 2 140 84.9
5 Methanol 6:1 6 140 95.7
90 wt% ethanol. Similar trends were observed at 140 °C. The ethan-
6 Methanol 2:1 6 140 95.6 olysis yield decreased from 77.2 (12) to 67.9 (13) and 53.6 wt% (14)
7 Methanol 4:1 4 140 96.2 when anhydrous ethanol was replaced by 5 and 10 wt% hydrated
8 Methanol 4:1 4 140 97.4 ethanol, respectively (Table 1).
9 Ethanol 6:1 2 100 23.1
Both XRPD and FTIR characterization of the ZHN catalyst after
10 Ethanol 6:1 2 100 21.8
95 wt% one reaction cycle (Fig. 1C and D) revealed that the basal distance
11 Ethanol 6:1 2 100 20.6 of the solid catalyst changed from 10.2 to 29.4 Å, which is consis-
90 wt% tent with the layered structure of zinc laurate (Fig. 1E and F)
12 Ethanol 6:1 2 140 77.2 [14]. Although the structure of the used catalyst matches very well
13 Ethanol 6:1 2 140 67.9
with that of zinc laurate, intercalation compounds obtained by the
95 wt%
14 Ethanol 6:1 2 140 53.6 exchange of carboxylate anions within LDH have very similar XRPD
90 wt% and FTIR spectra [16,17]. A series of new experiments are under
15d Methanol 6:1 2 140 96.5 way to better characterize the active phase after recycling.
a
Anhydrous, unless indicated otherwise. To account for the possibility of having zinc laurate after the
b
MRL, alcohol:lauric acid molar ratio. first reaction step, the catalytic activity of freshly synthesized zinc
c
Total esters content in the final product (biodiesel), as determined by HPLC. laurate was determined and compared with that of ZHN under the
d
Experiment carried out with fresh synthesized zinc laurate. same reaction conditions. Zinc laurate produced a methyl ester
yield of 96.5 wt%, whereas ZHN reached conversions not beyond
larger catalyst concentration. This resulted in an increase in final 87.1 wt% (experiments 3 and 15 in Table 1). Since laurate ions were
ester content from 87.1 to 95.7 wt%. Likewise, for experiments car- the main species present in the lamellar structure after the first
ried out at a MRL of 2:1 (experiments 4 and 6), the same threefold reaction step, the catalyst loading for the second reaction step
increase in catalyst concentration improved the corresponding to- was adjusted to represent the same molar Zn content in relation
tal ester content from 84.9 to 95.6 wt%. to the oil mass, based on the ash obtained from the TG analysis
The statistical analysis of the reaction procedures used in this (data not shown).
study was evaluated from three replicates carried out at the center Esterification of lauric acid with the reused solid catalyst was
point of experiments 3–6 (see experiments 7 and 8 in Table 1). The carried out under the same conditions showed in experiment 1
total ester content in these replicates had a relative standard devi- (Table 1). The first use gave an acidity of 51.9% (39.4% of ester yield
ation of only 0.93% for an average ester content of 96.8 ± 0.9 wt%, based on HPLC analysis (Table 1)), whereas the second, third and
indicating that the system gives reproducible results. The high es- fourth use of the same batch of catalyst resulted in an acidity of
ter contents achieved in these replicates are an indication of binary 39.5, 54.3 and 58.3%, respectively. These preliminary data indicate
interaction between MRL and the catalyst concentration, because that the solid catalyst can be reused and may be applied for the
the center point yielded larger ester contents than those of exper- development of a future continuous process.
iments 3–6 in Table 1. Based on both XRPD and FTIR data (Fig. 1), the interlayer nitrate
Experiments were also carried out to evaluate whether ethanol ions present in the ZHN structure were exchanged by laurate ions
may replace methanol (both in anhydrous conditions) as the acyl- during the first reaction course and this exchange swelled the crys-
ating agent of free fatty acids. Experiment 9 was done under the tallites, facilitating the access of the reagents to the catalytic sites
same reaction conditions as experiment 1 (i.e. MRL of 6:1, 2 wt% present at the surface and edges of the layered crystal structure [4].
catalyst and 100 °C), except that ethanol was used instead of meth- Another effect was the apparent delamination of the crystals,
anol. At 100 °C, methanol proved to be a much better acylating which contributed to a further increase in the availability and
agent than ethanol, resulting in a final ester content significantly accessibility of the catalyst active sites.
larger than that obtained with ethanol (39.4 wt% in comparison The catalytic activity of ZHN was also tested in transesterifica-
with 23.1 wt%). This can be readily explained by the lower reactiv- tion reactions using RPO. For reactions carried out at 150 °C, the es-
ity of ethanol in relation to that of methanol, but other factors may ter layer obtained from RPO contained 95.7 wt% methyl esters,
also have influenced the reaction efficiency, such as the higher revealing that ZHN can also be used in the alcoholysis of vegetable
hygroscopic character of the former and the lower density of the oils. However, a high methanol:palm oil molar ratio of 48:1 (i.e. 16
latter. Such reactivity differences in esterification reactions based times the reaction stoichiometry) was required to achieve good
on homogeneous catalysts were also reported by other authors reaction efficiency. The purity of the glycerin layer derived from
[15]. Interestingly, when the same comparison was repeated at RPO methanolysis was also characterized by ion-exchange chro-
140 °C (experiments 3 and 12), the difference in ester yields was matography. Since no water was used for phase separation, the
smaller (87.1 wt% with methanol compared to 77.2 wt% with eth- glycerol concentration in this layer was as high as 93 wt% with
anol). We are currently investigating whether this gap can be fur- no detectable contamination of solid catalyst. This glycerol content
ther reduced. is much higher than the contents usually achieved by classical bio-
In many cases, high temperature treatments are required to diesel production technologies such as the alkaline transesterifica-
eliminate any water moiety from the catalyst surface, a process tion with homogeneous catalysts. The ‘‘crude” glycerin obtained by
usually referred to as activation. The ZHN catalyst was relatively ZHN is suitable for many chemical processes without any further
tolerant to the presence of water in the reaction media, which treatment and would require rather simple and energy-efficient
may be one of its main advantages. A series of experiments were downstream processing for its pharmaceutical grade upgrading.
carried out under the same esterification conditions with anhy- ZHN may have a high impact on the economy of mid- and large-
drous ethanol (9) and ethanol with 5 wt% (10) and 10 wt% (11) scale biodiesel operations, since not only will the process be
 C.S. Cordeiro et al. / Catalysis Communications 9 (2008) 2140–2143
simpler, but also will biodiesel purification and glycerin upgrading
be facilitated, thereby improving process sustainability greatly.
Furthermore, ZHN offers several additional advantages, such as
easy synthesis, totally safe handling, and suitability to low cost
processing of high acid number feedstocks. Its chemical composi-
tion is based on nontoxic zinc salts recoverable by simple filtration
and recyclable by mineral acid treatment and reprecipitation. In
addition, the ZHN catalyst may be synthesized within porous
matrices with high mechanical properties for application as reac-
tive fillers in continuous reactors.
Based on these preliminary findings, we are currently optimiz-
ing the use of ZHN for ethanol/methanol biodiesel production and
studying the catalytic activity of other types of layered hydroxide
salts.
Acknowledgments
The authors are grateful to the following Brazilian funding
agencies for financial support: CNPq, CAPES and FINEP. Thanks
are due to David A. Mitchell for help with the English expression.
2143
References
[1] L.P. Ramos, H.M. Wilhelm, Appl. Biochem. Biotechnol. 121–4 (2005) 807.
[2] A.C. Pinto, L.N.L. Guarieiro, M.J.C. Rezende, N.M. Ribeiro, E.A. Torres, W.A.
Lopes, P.A.P. Pereira, J.B. Andrade, J. Braz. Chem. Soc. 16 (2005) 1313.
[3] B.M. Choudary, M.L. Kantam, C.V. Reddy, S. Aranganathan, P.L. Santhi, F.
Figueras, J. Mol. Catal. A 159 (2000) 411.
[4] M.B.J. Roeffaers, B.F. Sels, H. Uji-I, F.C. De Schryver, P.A. Jacobs, D.E. De Vos, J.
Hofkens, Nature 439 (2006) 572.
[5] J.L. Shumaker, C. Crofcheck, S.A. Tackett, E. Santillan-Jimenez, M. Crocker, Catal.
Lett. 115 (2007) 56.
[6] Y. Liu, E. Lotero, J.G. Goodwin, X. Mo, Appl. Catal., A. 331 (2007) 138.
[7] W.L. Xie, H. Peng, L.G. Chen, J. Mol. Catal. A 246 (2006) 24.
[8] A. Corma, S.B. Abd Hamid, S. Iborra, A. Velty, J. Catal. 234 (2005) 340.
[9] D.G. Cantrell, L.J. Gillie, A.F. Lee, K. Wilson, Appl. Catal., A 287 (2005) 183.
[10] W.M. Antunes, C.O. Veloso, C.A. Henriques, Catal. Today 133–135 (2008) 548.
[11] G.G.C. Arizaga, K.G. Satyanarayana, F. Wypych, Solid State Ionics 178 (2007)
1143.
[12] W. Stahlin, H.R. Oswald, Acta Crystallogr., B 26 (1970) 860.
[13] G.G.C. Arizaga, A.S. Mangrich, F. Wypych, J. Colloid Interf. Sci. 320 (2008) 238.
[14] S. Barman, S. Vasudevan, J. Phys. Chem. B 110 (2006) 22407.
[15] M. Canakci, J. Van Gerpen, Trans. Am. Soc. Agric. Eng. 42 (1999) 1203.
[16] T. Itoh, N. Ohta, T. Shichi, T. Yui, K. Takagi, Langmuir 19 (2003) 9120.
[17] N. Nhlapo, T. Motumi, E. Landman, S.M.C. Verryn, W.W. Focke, J. Mater. Sci. 43
(2008) 1033.