Characterization of an alkaline earth metal-doped solid superacid and its activity for the esterification of oleic acid with methanol

Article abstract of DOI:10.1039/c5gc00188a

The leaching of grafted sulfate groups is one of the major issues for the solid acid catalysts with sulfate modification. A detailed study of sulfated alkaline earth metal-ferric composite oxides was carried out to suppress the leaching of sulfate groups as well as maintain high reactivity during the esterification of oleic acid. The acid properties and quantities of the active sites present on the catalysts were studied with pyridine-adsorption, FT-IR, TGA, and titration methods. The following order of acidic strength was observed: SO₄^2-/Sr-Fe oxide > SO₄^2-/Ca-Fe oxide > SO₄^2-/Mg-Fe oxide. After sulfate modification, sulfate ions are linked to the composite oxides in a bridged bidentate form. SO₄^2-/Sr-Fe oxide exhibits superacidic nature due to the high induction effect of the sulfated ion grafted on the Sr cation. TGA and FT-IR spectroscopic analyses provide new insights into the function of the iron cations implanted on the catalyst surface for enhancing the turnover frequency (TOF) of the active sites on SO₄^2-/Sr-Fe oxide in the esterification. It was verified that the active energy of SO₄^2-/Sr-Fe oxide for the esterification of oleic acid with methanol was as low as 28.53 ± 0.72 kJ mol^-1. In the reusability study, SO₄^2-/Sr-Fe oxide exhibited high reusability in the esterification, due to the high stability of the sulfate ion grafted on SO₄^2-/Sr-Fe oxide.

Full text of DOI:10.1039/c5gc00188a

View Article Online
View Journal
Green
Chemistry
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: C. Huang, C.
Yang, P. Gao, N. Wang, C. Chen and J. Chang, Green Chem., 2015, DOI: 10.1039/C5GC00188A.
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
Accepted Manuscripts are published online shortly after
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
Information for Authors.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the Ethical guidelines still
apply. In no event shall the Royal Society of Chemistry be held
responsible for any errors or omissions in this Accepted Manuscript
or any consequences arising from the use of any information it
contains.
www.rsc.org/greenchem

Page 1 of 27
Green Chemistry
View Article Online
DOI: 10.1039/C5GC00188A
Characterization of Alkaline Earth Metal-doped Solid
Superacid and Activity for the Esterification of Oleic Acid
with Methanol
a
a
a
a
Chien-Chang Huang* , Chieh-Ju Yang , Pei-Jyuan Gao , Nai-Ci Wang , Ching-Lung
b
b
Chen and Jo-Shu Chang
Abstract
The leaching of grafted sulfate groups is one of the major issues for the solid
acid catalysts with sulfate modification. A detailed study of sulfated alkaline earth
metalꢀferric composite oxides is investigated to suppress the leaching of sulfate
groups as well as maintain high reactivity during the esterification of oleic acid. The
acid properties and quantities of the active sites present on the catalysts were studied
with pyridineꢀadsorption FTꢀIR, TGA, and titration methods. The following order of
2
ꢀ
2ꢀ
2ꢀ
acidic strength was observed: SO /SrꢀFe oxide > SO /CaꢀFe oxide > SO /MgꢀFe
4
4
4
oxide. After sulfate modification, sulfate ions are linked to the composite oxides in a
2
ꢀ
bridged bidentate form. SO /SrꢀFe oxide exhibits superacidic nature due to the high
4
induction effect of the sulfated ion grafted on Sr cation. TGA and FTꢀIR spectroscopic
analyses provide new insights into the function of the iron cations implanted on the
catalyst surface for enhancing the turnover frequency (TOF) of the active sites on
2
ꢀ
SO /SrꢀFe oxide in the esterification. It was verified that the active energy of
4
2
ꢀ
SO /SrꢀFe oxide for the esterification of oleic acid with methanol was as low as
4
ꢀ1
2ꢀ
2
8.53±0.72 kJmol . In the reusability study, SO /SrꢀFe oxide exhibited high
4
reusability in the esterification, due to the high stability of the sulfate ion grafted on
2
ꢀ
SO /SrꢀFe oxide.
4
Keywords: Biodiesel, Alkaline earth metal ion, Oleic acid esterification, Sulfated
composite oxide.
1
. Introduction
Esterification is an important reaction for the preparation of flavor chemicals,
drugs, perfumes, cosmetic ingredients, solvents and biofuels. For biodiesel production
a
Department of Cosmetic Science, Providence University, 200 Sec. 7, Taiwan Boulevard, Shalu Dist., Taichung 433, Taiwan.
Fax: +886 42631 1167; Tel:+886 42632 8001 ext. 15411; E-mail: cchuang9@pu.edu.tw
Department of Chemical Engineering, National Cheng Kung University, 1 University Road, Tainan 701, Taiwan. Tex: +886
b
6894 4496; Tel: +886 6275 7575; E-mail: changjs@mail.ncku.edu.tw
1

Green Chemistry
Page 2 of 27
View Article Online
DOI: 10.1039/C5GC00188A
in particular, the acidꢀcatalyzed preꢀesterification of the free fatty acids (FFA) present
in low cost plant oils or waste cooking oils is a key step to decrease the level of FFA
and prevent the saponification of the free fatty acids before the oil can be used in the
1
,2,3,4,5
baseꢀcatalyzed transesterification as a feedstock
. At the industrial level, the
esterification reactions of fatty acids are performed in a homogeneously catalyzed
reaction. Mineral acids, such as H SO and H PO , are the conventional homogeneous
2
4
3
4
catalysts, due to their high reactivity and low prices. However, their usage is limited
because mineral acids accelerate the corrosion of the production equipments, and the
residual catalysts are difficult to remove from the product stream. An extra
purification step is thus required to remove the remaining strong acid in the products
derived from homogeneously acidꢀcatalyzed reactions.
Solid acid catalysts are environmentally benign substitutes for mineral acids to
meet the needs of green chemistry technologies. Since the reaction is catalyzed on a
solidꢀliquid interface, the used solid catalyst can be easily extracted from the reaction
medium. The used catalysts can then be recovered by washing with solvents or
6
,7
calcining at an appropriate temperature. In these respects, ionꢀexchange resins ,
8
,9,10,11,12,13
zeolites, as well as sulfated metal oxides
are the promising alternatives for
mineral acids. Among the several solid acid catalysts that have been reported to date,
sulfated metallic oxides are examples of highly acidic catalysts, and some of these
were classified as superacids.
8,9
In recent years, much attention has been paid to sulfated metal oxides. Zirconia ,
10,11
12,13
titanium , tin oxide
and related oxides modified by sulfate groups are the most
commonly examined solid acid catalysts in previous studies. It has been reported that
these catalysts provide significant catalytic reactivity in the esterification of
1
4
15
carboxylic acid with alcohols , transesterification of triglycerides with alcohols ,
13
FriedelꢀCrafts alkylation and acylation , oxidation of alkenes with hydrogen
1
6,17
18
peroxide
and the skeletal isomerization of alkanes . Sulfated zirconia is known
as one of the strongest solid acid catalysts, and has been well studied due to its
superacidic nature and consequent potential as a solid acid catalyst for numerous
8
,9, 19 , 20
acidꢀcatalyzing reactions
. In particular, sulfated zirconia shows superior
reactivity for the esterification of fatty acids, and its reactivity is much higher than
2
1
that of conventional mineral acids . Sulfated zirconia consequently seems to be a
good replacement for the mineral acids to carry out the esterification of FFA existing
in low cost plant oils.
Although much effort has been devoted to investigating the catalytic properties
of sulfated zirconia in the esterification of fatty acids with alcohols, sulfated zirconia
has several drawbacks that make its utilization in biodiesel production difficult at the
industrial scale. The major drawback is its short life time and easy deactivation.
2

Page 3 of 27
Green Chemistry
View Article Online
DOI: 10.1039/C5GC00188A
Ardizzone et al. noted that the sulfate groups on zirconia are thermally and chemically
22
unstable. They stated that sulfated zirconia has poor resistance to hydrolysis and
alcoholysis of the sulfate groups grafted on the zirconia surface. If the reaction
medium contains a polar substance or water, then the sulfate groups quickly leach out
ꢀ
2,23
as H SO and HSO
. The deactivation of sulfated zirconia in the esterification of
2
4
4
fatty acid is thus irreversible, and the leached sulfate groups also contaminated the
2
4
product stream of esterification in the biodiesel production process .
Several developments are thus needed to aid the commercial introduction of
sulfated metal oxide to the biodiesel production process. These improvements include
aspects such as: enhancing the stability of the grafted sulfate groups to avoid their
leaching in the reaction medium and in the recovery process; simplifying the
separation of used solid acid catalyst from the production stream to reduce operational
cost; lowering operation temperature by means of improving the reactivity; as well as
extending the lifetime to meet the demand of green technologies, in order to maintain
the reactivity during catalytic reaction. Therefore, an appropriate support for the
sulfate group is thus needed in order to meet the demands for an ideal solid acid
catalyst. Alkaline earth metal oxides have been reported as a base catalyst in the
literatures. The strong basic character of strontium oxide is one of its bestꢀknown
chemical properties. In our previous study, it was found that the composite oxide
prepared by incorporating strontium in ferric oxide matrix (SrꢀFe oxide) exhibited
2
5
high baseꢀcatalytic reactivity for the transesterification of plant oil with methanol .
However, SrꢀFe oxide is quickly poisoned by the free fatty acid usually present in
waste cooking oil, because saponification takes place on the catalyst’s surface. To the
best of our knowledge, sulfated alkaline earth metalꢀferric oxides have not yet been
explored as an acid catalyst for use in esterification reactions. Three magnetic alkaline
earth metalꢀferric oxides were thus adopted in this work for the grafting of sulfate
groups to study the resulting improvements in catalytic properties, and to simplify the
separation procedure of the used catalyst from the reaction medium. The catalytic
properties of sulfated composite oxides were examined in the esterification of oleic
acid with methanol, and their acid properties were well characterized. Of particular
interest is that sulfated strontiumꢀferric oxide is a superacid, and shows high reactivity
and stability in the esterification process.
2
2
. Experimental
.1. Catalyst preparation
Three alkaline earth metalꢀferric composite oxides (denoted as composite oxides),
including strontiumꢀferric oxide (SrꢀFe oxide), calciumꢀferric oxide (CaꢀFe oxide) and
3

Green Chemistry
Page 4 of 27
View Article Online
DOI: 10.1039/C5GC00188A
magnesiumꢀferric oxide (MgꢀFe oxide), were prepared by coꢀprecipitation. Iron (III)
nitrate nonanhydrate (purity 99%), strontium nitrate (purity 98%), calcium (II) nitrate
tetrahydrate (purity 98%), tetraethyl silicate (purity 99.96%) were purchased from
Showa, Inc, while magnesium nitrate hexahydrate (purity 98%) was obtained from
AlfaꢀAesar. All the reagents were commercial available and used as received. In a
typical preparation, alkaline earth metal nitrate and iron (III) nitrate nonanhydrate in
the desired molar ratio were preꢀmixed in 80 ml of deionized water. 100 ml of 15 wt%
ammonia hydroxide (from Fisher Science Inc.) was subsequently added into the
solution drop by drop under vigorous stirring, and then aged at room temperature for
1
2 hours. The solid products which were obtained from the solutions by completely
o
o
evaporating the liquid phases of the reactions at 85 C were then calcined at 1100 C
2
4
ꢀ
for 10 hours. Sulfated strontiumꢀferric oxide (SO /SrꢀFe oxide), sulfated
2
ꢀ
calciumꢀferric oxide (SO /CaꢀFe oxide) and sulfated magnesiumꢀferric oxide
4
2
ꢀ
(SO /MgꢀFe oxide) were prepared by mixing 5g composite oxide with 100ml of 2M
4
H SO aqueous solution at room temperature for 2 hours, and then the sulfated solids
2
4
o
were heatꢀtreated at 400 C for 2 hours to remove weakly bonded sulfate groups or
unbonded sulfuric acid.
2
ꢀ
2ꢀ
4
2ꢀ
Three reference samples, SO /Fe O , SO /ZrO and SO /SrO, were also
4
2
3
2
4
employed for the present experiments. Fe O was prepared by precipitation of ferric
2
3
hydroxide in a solution consisting of 38.17g iron (III) nitrate and 100 ml water by
adding 100 ml 15% ammonium hydroxide into the solution drop by drop, and then
o
calcined at 1100 C for 10 hours. Subsequently, 1 g of the Fe O was dispersed in 20
2
3
ml 2M H SO aqueous solution under stirring for 2 hours. ZrO particles were
2
4
2
15
obtained following the method published by Fu et al. . ZrO was prepared by
2
precipitation of zirconium hydroxide in the solution containing 10.31g zirconium (VI)
oxychloride octahydrate (ZrOCl ·8H O) and 160ml H O by adding 30ml 30%
2
2
2
o
ammonia hydroxide into the solution drop by drop, followed by calcinating at 600 C
for 8 hours. Sulfate groups were then grafted on the resulting ZrO by immersing 1 g
2
ZrO in 20 ml 0.5M H SO aqueous solution under stirring for 2 hours. The
2
2
4
2
4
ꢀ
2ꢀ
preparation method for obtaining SO /SrO is similar to that of SO /ZrO. 50 ml 30%
4
ammonia hydroxide was added into the solution consist of 29.62 g strontium nitrate
and 50 ml H O drop by drop. SrO was obtained by calcining the resulting particles
2
o
obtained from the solution at 1100 C for 10 hours. Sulfate groups were incorporated
on SrO surface by mixing 1g SrO in 20ml 2M H SO aqueous solution under stirring
2
4
2
ꢀ
for 2 hours. The XRD pattern of SO /SrO is shown in Figure S1. All sulfated
4
o
reference samples were calcined at 400 C for 2 hours prior to use.
SiO powder for the KorosꢀNowak test was synthesized using a modified Ströber
2
2
6
process . In general, 10 ml tetraethyl silicate (98% from Seedchem Co.) was added in
4

Page 5 of 27
Green Chemistry
View Article Online
DOI: 10.1039/C5GC00188A
the mixture consisting of 66 ml isoꢀpropanol (99% from Sigmaaldrich Co.), 4.1ml
ammonium hydroxide (30% from Showa, Inc.) and 10ml deꢀionized water under
vigorous magnetic stirring. The reaction of which was allowed to proceed at room
o
temperature (25 C) for 10 hours. The resulting particles were separated from the
reaction medium by vacuum filtration using filter paper and subsequently washed
o
with deꢀionized water twice, and were then dried at 110 C in an oven for 12 hours.
2
.2. Catalyst characterization
The specific surface areas of the sample catalysts were determined through the
isothermal model of BrunauerꢀEmmettꢀTeller (BET) using a PMI Sorptometer
CBETꢀ201A. The XRD patterns were obtained by using a Rigaku Ultima IV Xꢀray
diffractometer with Cu K radiation operated at 40 kV and 40 mA in the range of
α
o
diffraction 2θ=10ꢀ90 with a sweep rate of 1 degree/minute. The elemental
composition and surface morphology of the catalyst samples were analyzed with a
scanning electron microscope (SEM) equipped with an energy dispersive
spectroscope (EDS) (JEOL Co. Ltd. JSMꢀ7000F). A JASCO Asia Portal FT/IRꢀ4100
spectrometer was utilized to obtain the FTꢀIR spectra of the samples by using the
conventional KBrꢀpellet technique. Thermogravimetry analysis (TGA) was performed
o
on a Perkin Elmer TGA7. About 5 mg of sample was heated at 10 C/min from room
temperature to desired temperature under a flowing stream of air while the associated
weight loss was recorded.
2
.3. Total acidity determination
The total acidity for all synthesized catalysts was determined by the nꢀbutyl amine
2
7
titration method . In a typical procedure, 50 ml 0.025M solution of nꢀbutyl amine
from Tedia) in toluene (from Echo) was preꢀmixed with 0.5g of catalyst overnight at
room temperature. The excess base was then back titrated by trichloroacetic acid
from Merck) using neutral red as an indicator. Nꢀbutyl amine is a strong base, and it
(
(
was expected that nꢀbutyl amine adsorbed on acid sites with various acid strengths.
By back titrating the free nꢀbutyl amine left in the solution after the acid sites
saturated by nꢀbutyl amine, the total number of acid sites present on the catalyst
surface can be estimated.
2
.4. Acid properties
The acidic properties of the sample catalysts were determined by the Hammett
indicators and an infrared spectroscopic investigation of pyridine adsorbed on the
samples. The indicators used in the present work were Dicinnamalacetone (pK =ꢀ3.0),
a
anthraquinone (pK =ꢀ8.2), pꢀNitrotoluene (pK =ꢀ11.35), pꢀNitrochlorobenzene
a
a
5

Green Chemistry
Page 6 of 27
View Article Online
DOI: 10.1039/C5GC00188A
(
pK =ꢀ12.70) and 2,4ꢀDinitrotoluene (pK =ꢀ13.75). In the typical procedure for
a
a
o
determining the acidic strength, 0.1g sample was preꢀtreated at 120 C for 5 hours, and
.1 wt% Hammett indicator diluted in toluene was then mixed with the sample in a
0
sealed sample vial. The acidic strength of the sample was determined by the change in
color of the indicators with different pK . The nature of acid sites (Brönsted or Lewis)
a
was distinguished by infrared spectroscopic investigation of pyridine adsorbed on the
o
sample. The sample in a U shape sample tube was preꢀtreated in N stream at 400 C
2
for 2 hours, and then connected with the pyridine vapor. After this, the IR spectra of
the pyridine vapor treated samples were recorded in a JASCO Asia Portal FT/IRꢀ4100
spectrometer using KBr pellets.
2
.5. Esterificaiton activity
The catalytic reactivity of the synthesized catalysts was evaluated in the
esterification of oleic acid (from Aldrich) with methanol (from Echo) in a molar ratio
o
of 1:4 at 100 C. The test reaction was carried out in a sealed 100 ml batch reactor
equipped with magnetic stirring. Oleic acid was preꢀmixed with a weighted sample
catalyst (10 wt% based on the weight of oleic oil), followed by methanol. The
progress of the reaction was monitored by periodically withdrawing samples from the
reaction mixture. The conversion of oleic acid was analyzed on a gas chromatograph
TM
(
Shimadzu GCꢀ2010) equipped with an Agilent capillary column DBꢀ5HT (15 m ×
0
2
.53 mm × 0.5 μm) and a flame ionization detector using N as the carrier gas. The
catalytic reactivity of the catalysts were evaluated by turnover frequency (TOF). TOF
is defined as the turnover number (TON) per unit time, and TON is defined as mole of
oleic acid converted in the reaction per mole of active site.
2
.6. Recycling test
The reusability as well as the stability of the catalysts were tested by carrying out
o
repeated cycles of the reaction at 100 C for 2 hours. After performing the reaction
over a catalyst, the catalyst was washed with hexane and subsequently calcined at
o
4
00 C under air atmosphere. The recovered catalyst was then reloaded in the reactor
with fresh reagents.
3
3
3
. Results and Discussion
.1 Characterization
.1.1 XRD pattern
The crystalline phases of three sulfated composite oxide samples were carried
6

Page 7 of 27
Green Chemistry
View Article Online
DOI: 10.1039/C5GC00188A
out by power Xꢀray diffraction (XRD), and their XRD patterns are shown in Figure 1.
It reveals that these samples are crystalline materials, and no intense bands relating to
the crystalline forms of the alkaline earth metal oxides are observed in the XRD
patterns, indicating that the alkaline earth metal ions highly disperse in the catalysts.
2ꢀ
2ꢀ
The obtained phases of SO /CaꢀFe oxide and SO /MgꢀFe oxide are indexed as
4
4
spinel CaFe O (JCPD No: 32ꢀ0168) and spinel MgFe O (JCPD No: 17ꢀ0465),
2
4
2
4
2
4
ꢀ
respectively, while the XRD pattern of SO /SrꢀFe oxide is indexed as Sr Fe O in
7
10 22
the standard data (JCPD No: 26ꢀ0980). The atomic ratio of ferric ion to alkaline earth
2
ꢀ
2ꢀ
metal ion (Fe/M) determined by EDS for SO /SrꢀFe oxide, SO /CaꢀFe oxide and
4
4
2
ꢀ
SO /MgꢀFe oxide are 1.45, 2.45 and 3.35, respectively. Obviously, the Fe/M ratio
4
determined by EDS are different from the stoichiometry of the phases detected by
XRD for all sulfated composite oxides, implying the presence of another crystalline
phase. Except for the crystalline phases of the composite oxides that were identified,
another crystalline phase present in the XRD patterns of all sulfated composite oxide
samples is indexed as Fe O (JCPD No: 33ꢀ0664). The XRD traces for the
2
3
2
ꢀ
SO /SrꢀFe oxide samples with various Fe to Sr atomic ratios are shown in Figure 2.
4
2ꢀ
2ꢀ
When the Fe/Sr ratio is in the range of 1.45 (SO /SrꢀFe oxideꢀ1) to 3.33 (SO /SrꢀFe
4
4
oxideꢀ3), a fall in the Sr moiety does not lead to a change in the crystalline phase of
2
ꢀ
2ꢀ
SO /SrꢀFe oxide. When Fe/Sr ratio is as high as 34.58 (SO /SrꢀFe oxideꢀ4), only
4
4
the crystalline phase of Fe O (JCPD No: 33ꢀ0664) is identified.
2
3
(
a)
(
b)
c)
(
10
20
30
40
50
60
70
80
90
2θ (degrees)
7

Green Chemistry
Page 8 of 27
View Article Online
DOI: 10.1039/C5GC00188A
2
ꢀ
2ꢀ
4
Figure 1. Powder Xꢀray patterns for (a) SO /SrꢀFe oxide; (b) SO /CaꢀFe oxide and
4
2
ꢀ
(
c) SO /MgꢀFe oxide
4
2ꢀ
SO4 /SrꢀFe oxideꢀ1 (Fe/Sr=1.45)
2
ꢀ
SO4 /SrꢀFe oxideꢀ2 (Fe/Sr=2.10)
2ꢀ
SO4 /SrꢀFe oxideꢀ3 (Fe/Sr=3.33)
2
ꢀ
SO4 /SrꢀFe oxideꢀ4 (Fe/Sr=34.58)
1
0
20
30
40
50
60
70
80
90
2θ (degrees)
2
4
ꢀ
Figure 2. Powder Xꢀray patterns for SO /SrꢀFe oxide samples with various Sr to Fe
2
ꢀ
atomic ratio. The atomic ratio of Fe to Sr for all SO /SrꢀFe oxide samples were
4
determined by EDS analysis.
3
.1.2 FT-IR spectra of sulfated composite oxides
The sulfate ions coordinated to the composite oxides were identified by FTꢀIR.
The spectra are shown in Figure 3. With respect to the structure of sulfate ions, the
sulfate ion in free form has T symmetry (Scheme 1(a)), with an intense IR band at
d
ꢀ1
around 1100 cm . When a sulfate ion chemically bonds to the composite oxide’s
2
9
surface, the sulfate ion has C symmetry (Scheme 1(b) and 1(c)) . As seen in Figure
2
v
2
ꢀ
3
9
(a), the IR spectrum of SO /SrꢀFe oxide shows four IR bands in the range of 1230 ꢀ
4
ꢀ1
90 cm . However, these IR bands were absent in the spectrum of the SrꢀFe
composite oxide without sulfate modification, indicating that the IR bands in the
ꢀ1
range of 1230ꢀ990 cm are associated with the covalently bonded sulfate ions, and the
coordination of sulfate ions on the catalyst surface leads to a change in its
ꢀ1
ꢀ1
configurations. The bands at 1225 cm and 1130 cm can be assigned to the
asymmetric and symmetric stretching of the S=O bond, respectively. In addition, the
ꢀ1
ꢀ1
bands at 1076 cm and 995 cm are attributed to the asymmetric and symmetric
11,28,29
stretching of SꢀO bonds
. With regard to the structure of the adsorbed sulfate
8

Page 9 of 27
Green Chemistry
View Article Online
DOI: 10.1039/C5GC00188A
group on metal oxides, there are two possible forms that have been reported in the
3
0,31,32
literatures, bridged bidentate and chelating bidentate
. According to the IR bands
2ꢀ
2ꢀ
of the sulfate ions on SO /SrꢀFe oxide, it can be proposed that SO groups
4
4
covalently bonded to SrꢀFe oxide surfaces by the bridged bidentate form, in which
2
8
sulfate group bridges across two metal ions (Scheme 1(c)). The absence of the band
ꢀ1
around 1400 cm reveals that a further reaction between adjacent sulfate groups,
2
ꢀ
33
which would result in the formation of polynuclear sulfate S O , did not take place .
2
7
ꢀ1
An additional band at 1636 cm corresponds to OꢀH bending of molecular water. The
2
4
ꢀ
2ꢀ
4
IR spectrum of SO /CaꢀFe oxide and SO /MgꢀFe oxide are also shown in Figure 3
(
spectrum (b) and (c)). Their IR spectra show similar IR bands to the IR spectrum of
2
ꢀ
ꢀ1
SO /SrꢀFe oxide in the range of 1230ꢀ900 cm , indicating the presence of sulfate
4
groups with the bridged bidentate structure on their surfaces.
The relative thermal stability of the coordinated sulfate ions on three sulfated
o
composite oxides was also examined by FTꢀIR after preꢀtreating the samples at 700 C.
The IR spectra are also shown in Figure 3 (from spectrum (a') to (c')). It is obvious
o
that preꢀtreating the samples at 700 C leads to a change in their IR spectra, especially
2
ꢀ
2ꢀ
o
for SO /MgꢀFe oxide. After preꢀtreating SO /MgꢀFe oxide at 700 C for 2 hours, the
4
4
2
ꢀ
IR spectrum of preꢀtreated SO /MgꢀFe oxide does not show any IR bands related to
4
sulfate ions, indicating a complete removal of the bonded sulfate groups (Figure 3
2
ꢀ
2ꢀ
spectrum (c')). In contrast, preꢀtreating SO /SrꢀFe oxide and SO /CaꢀFe oxide at
4
4
o
7
00 C for 2 hours only diminishes the band intensities of sulfate ions (Figure 3
spectrum (a') and spectrum (b')). These results reveal that the residual sulfated ions on
2
ꢀ
2ꢀ
4
the SO /SrꢀFe oxide and SO /CaꢀFe oxide samples have high thermal stability, and
4
the thermal stability of the adsorbed sulfate ions is associated with the alkaline earth
metal ions implanted on the catalysts.
2
-
O
O
O
O
O
O
O
S
S
S
O
O
O
O
O
O
M
M
M
M
M
O
O
O
O
(
a)
(b)
(c)
Scheme 1. The structure of (a) free sulfate ion with T symmetry, (b) bonded
d
sulfate ion with chelate bidentate configuration, and (c) bonded sulfate ion with
29
bridged bidentate configuration .
9

Green Chemistry
Page 10 of 27
View Article Online
DOI: 10.1039/C5GC00188A
1
076
9
1
130
1
225
95
1
636
(a)
(a')
(b)
(b')
(c)
(c')
2000 1800 1600 1400 1200 1000 800
ꢀ
1
Wavenumber (cm )
2
ꢀ
2ꢀ
Figure 3. Infrared spectra of (a) SO /SrꢀFe oxide; (b) SO /CaꢀFe oxide and (c)
4
4
2
ꢀ
2ꢀ
SO /MgꢀFe oxide, and the infrared spectra of heatꢀtreated (a') SO /SrꢀFe oxide, (b')
4
4
2
ꢀ
2ꢀ
o
SO /CaꢀFe oxide and (c') SO /MgꢀFe oxide at 700 C for 2 hours.
4
4
3
.1.3 Textural properties of the coordinated sulfate ions on the catalysts
The TGA analyses of the sulfated composite oxides were carried out in air, and
the profiles are shown in Figure 4. The initial weight loss on the TGA profiles is
o
assigned to desorption of physically adsorbed water (100ꢀ200 C). Further, increasing
o
o
the temperature from 250 C to 600 C leads to gradual weight loss, associated with the
removal of surface hydroxyl groups and carbonates, since most adsorbed hydroxyl
o
28,34
groups and carbonates become thermally unstable before 600 C on metal oxides
.
o
The significant weight loss that started at 600 C for all the catalyst samples is due to
o
the decomposition of adsorbed sulfate ions (600ꢀ700 C). This is a typical feature for
3
5
the TGA profiles of sulfated metal oxides . Based on the TGA profiles, when the
o
temperature increased from 600 to 700 C, the percentages of the weight of the sulfate
2
ꢀ
2ꢀ
4
2ꢀ
ions eliminated from SO /SrꢀFe oxide, SO /CaꢀFe oxide and SO /MgꢀFe oxide
4
4
10

Page 11 of 27
Green Chemistry
View Article Online
DOI: 10.1039/C5GC00188A
are estimated to be 8.8% (w/w), 5.5% (w/w) and 2.5% (w/w), respectively.
Table 1 represents the acid site density estimated by TGA analysis (DTGA) and by
2
ꢀ
the nꢀbutyl amine titration method (DBAT). The IR spectrum of SO /MgꢀFe oxide
4
2
ꢀ
o
shown in Figure 3 indicates that heatꢀtreating SO /MgꢀFe oxide at 700 C leads to
4
complete removal of adsorbed sulfate ions. According to the TGA profiles of
2
ꢀ
SO /MgꢀFe oxide (Figure 4), the density of the sulfate groups (S ) grafted on
4
ion
2
ꢀ
SO /MgꢀFe oxide can thus be calculated. It is assumed previously that the sulfate
4
group binds to MgꢀFe oxide in the bridged bidentate form, illustrating that the density
of the acid sites is about twoꢀfold the density of the bonded sulfate ions. Therefore,
the acid site density can be estimated from S . As seen in Table 1, the acid site
ion
density determined by TGA analysis (DTGA) is close to that determined by the nꢀbutyl
2
4
ꢀ
amine titration method (DBAT) for SO /MgꢀFe oxide. This result is in agreement with
the configuration of the adsorbed sulfate groups deduced from the IR bands shown in
Figure 3, in which the sulfate group binds to the composite oxide by the bridged
bidentate form.
2
4
ꢀ
The proposed configurations of the adsorbed sulfate groups on SO /SrꢀFe oxide
2
4
ꢀ
and SO /CaꢀFe oxide are also deduced from the IR spectra shown in Figure 3 and
the difference between Sion and DBAT shown in Table 1. As seen in Figure 3, the IR
2
4
ꢀ
spectra show that a small fraction of sulfate ions remained on SO /SrꢀFe oxide and
2
ꢀ
o
SO /CaꢀFe oxide after 700 C calcination. Figure 4 shows that no obvious weight
4
2
ꢀ
2ꢀ
loss was found on the TGA profiles of SO /SrꢀFe oxide and SO /CaꢀFe oxide when
4
4
o
the temperature is over 700 C. The residual sulfate ions on the heatꢀtreated catalysts
can be classified as the sulfate groups with high thermal stability. DBAT represents the
density of the active sites actually present on catalyst surface. With respect to the
2
4
ꢀ
2ꢀ
configuration of the bonded sulfate ions on SO /SrꢀFe oxide and SO /CaꢀFe oxide,
4
Sion should be close to DBAT if the sulfate groups bond to the catalyst surfaces by
2
4
ꢀ
2ꢀ
chelating bidentate form. However, the DBAT of SO /SrꢀFe oxide or SO /CaꢀFe
4
oxide is about 3 folds higher than the S , suggesting that most sulfate ions bond to
ion
SrꢀFe oxide and CaꢀFe oxide by bridged bidentate form. According to the proposed
configuration of bonded sulfate group, the density of the sulfate groups with high
ꢀ
2
ꢀ2
2
thermal stability can thus be calculated as 4.29×10 and 3.69×10 mmol/m on
2
ꢀ
2ꢀ
SO /SrꢀFe oxide and SO /CaꢀFe oxide, respectively. The distribution of Lewis and
4
4
Brönsted acid sites on the catalysts was investigated via pyridine FTꢀIR analysis
(
Figure S2). It was found that both Lewis and Brönsted acid site are present on
2
ꢀ
2ꢀ
SO /SrꢀFe oxide and SO /CaꢀFe oxide. Their configurations are proposed as
4
4
indicated in Scheme 2(a). The ratio of the Brönsted acid sites to Lewis acid sites
N /N ) estimated from the integral area under the specific bands shown in Figure S2
(
B
L
2
ꢀ
2ꢀ
discloses that Brönsted acid sites dominate on SO /SrꢀFe oxide and SO /CaꢀFe
4
4
11

Green Chemistry
Page 12 of 27
View Article Online
DOI: 10.1039/C5GC00188A
2
ꢀ
oxide. Therefore, the configuration of most active sites present on SO /SrꢀFe oxide
4
2
ꢀ
and SO /CaꢀFe oxide can be proposed as the model shown in Scheme 2(b).
4
O
O
O
O
O
O
O
O
+
S
S
H
H
H
+
Lewis acid site
+
H
H
O
O
O
H
+
δ
Bronested acid site
M
M
M
M
M
M
O
O
O
O
O
O
O
O
M=metal cation
M=metal cation
(a)
(b)
Scheme 2. Schematic diagram of the sulfate group grafted on an alkaline earth
metalꢀferric oxide.
0
2
4
6
8
(c)
(b)
10
12
(a)
14
16
100
200
300
400
500
600
700
800
o
Temperature ( C)
2
ꢀ
2ꢀ
Figure 4. TGA profiles in air of (a) SO /SrꢀFe oxide, (b) SO /CaꢀFe oxide and (c)
4
4
2
ꢀ
SO /MgꢀFe oxide
4
Table 1. Acidic density of sulfated composite oxides determined by TGA (DTGA) and
by the nꢀbutyl amine titration (DBAT
)
TGA analysis
nꢀbutyl amine titration
a
Area
b
c
d
e
catalyst
W
S
content
Sion density
D
TGA
Acidity
D
BAT
2
(
m /g)
ꢀ2
ꢀ2
2
ꢀ2
2
ꢀ2
2
(
×10 g/g cat.)
9.00±0.28
5.35±0.21
2.85±0.50
(×10 mmol/m )
(×10 mmol/m )
(mmol/g)
(×10 mmol/m )
28.38±0.07
17.80±0.75
12.43±0.46
2
ꢀ
f
SO4 /SrꢀFe oxide
9.47±0.21
9.90±0.09
19.80
2.69±0.07
1.90±0.02
0.60±0.11
2
ꢀ
f
SO4 /CaꢀFe oxide 10.7±0.54
5.21±0.06
10.42
2
ꢀ
SO4 /MgꢀFe oxide 4.87±1.04
a
6.13±0.25
12.26
BET specific surface areas of sulfate modified catalysts
12

Page 13 of 27
Green Chemistry
View Article Online
DOI: 10.1039/C5GC00188A
b
c
d
e
sulfate ion weight per gram catalyst
sulfate ion density on catalyst surface
acid site density estimated via TGA analysis
acid site density evaluated by using the nꢀbutyl amine titration method
ps.
the associated error and error bar in the tables and the figures were obtained by
using the equation of Standard Deviation (SD)
2
∑(X − M )
SD =
n −1
Where X refers to the individual data, M is the mean of the data and n is number of
data points.
3
.1.4 Acid properties
The acidic strength is expressed by the Hammett acidity function, H . It was
0
3
6
reported by Gillespie et al. that the acidic strength of 100% H SO was H =ꢀ11.93 .
2
4
0
When the H of any solid acid is less than ꢀ11.93, the solid acid is known as a
0
superacid. As the results show in Table 2, all the sulfated composite oxides exhibit
2ꢀ
4
remarkably high acidity, and particularly SO /SrꢀFe oxide can be claimed to
presence of superacid sites. The results of the Hammett indicator analysis demonstrate
that there is a direct correlation between the acidic strength of the active sites and the
alkaline earth metal ions incorporated to the catalysts. This observation is in
2
ꢀ
agreement with the results of the pyridine FTꢀIR analysis, where SO /SrꢀFe oxide
4
shows higher binding ability for pyridine than the others. It seems that the acidic
strength of the three sulfated composite oxides increases along with the difference in
electronegativity between oxygen anions and the adjacent metal cations. The presence
2
4
ꢀ
of superacid sites only on SO /SrꢀFe oxide may be because the surface sulfate ions
bound to the metal ions with a low electronegativity (Sr). The inductive effect of S=O
of the sulfate ions makes the Lewis acid strength of Sr stronger, since it is easier for
the bonded sulfate ion to pull covalent electrons away from the Sr ions. Once the
Lewis site is converted to a Brönsted acid site via the covalent binding of water to the
same Sr ion, the high acid strength character enables the Brönsted acid site to transfer
+
a neutral Hammett base with a lower H into its conjugate acid (BH ).
0
To have a better understanding of the correlation between the incorporated
alkaline earth metal cations and the acid site density, a series of strontiumꢀferric
composite oxides with various Sr to Fe atomic ratios were prepared. The surface
atomic ratios measured by EDS for each SrꢀFe composite oxide sample are shown in
Table 3. In Table 3, the density of the acid sites is proportional to that of the Sr ions
incorporated on catalyst surface, revealing that the surface Sr ions facilitate the
binding of sulfated groups. Thus, when the catalyst surface is rich in Sr ions, more
13

Green Chemistry
Page 14 of 27
View Article Online
DOI: 10.1039/C5GC00188A
acid sites are created after sulfate modification.
Table 2. Acidic strength of the sulfated composite oxides classified by the
Hammett indicators
3
0
7
H
a
b
Catalyst
EN
D
EN
ꢀ3.00 ꢀ8.20 ꢀ11.35 ꢀ12.70 ꢀ13.75
2
ꢀ
SO
4
/SrꢀFe oxide
0.963
1.034
1.293
2.647
2.576
2.317
+
+
+
+
+
+
+
+
±
+
-
-
-
-
2
ꢀ
4
SO /CaꢀFe oxide
2
ꢀ
SO
4
/MgꢀFe oxide
-
+
a.
, acidic color; ±, slightly acidic color; ꢀ, basic color
electronegativity of metal ions
the difference in electronegativity between oxygen ions and the adjacent metal
ions of the support of a sulfated catalyst.
b.
2
ꢀ
Table 3. Acidic density of the SO /SrꢀFe oxide samples with various Sr to Fe atomic
4
ratios.
a
b
Atomic Ratio
Surface area
Total acidity
(mmol/g)
D
BAT
Catalyst
2
ꢀ2
2
Sr : Fe
(m /g)
(×10 mmol/m )
2
2
ꢀ
SO
4
/SrꢀFe oxideꢀ1
1 : 1.45
9.82±0.10
2.68±0.10
27.29±0.73
ꢀ
SO4 /SrꢀFe oxideꢀ2
1 : 2.10
1 : 3.33
1 : 34.58
9.58±0.06
8.62±0.44
7.21±0.33
2.38±0.11
0.91±0.08
0.33±0.04
24.79±0.96
10.55±0.45
4.50±0.29
2
ꢀ
SO4 /SrꢀFe oxideꢀ3
2
ꢀ
SO4 /SrꢀFe oxideꢀ4
a
the atomic ratio of Sr to Fe was calculated using EDS analysis
b
acid site density evaluated by using the nꢀbutyl amine titration method
3
.2 IR studies of adsorbed oleic acid on catalyst surface
Before sulfate modification, the composite oxides were inactive for the
esterification of oleic acid with methanol. The IR spectra of the composite oxides
filtrated from the reaction mixture exhibit the characteristic IR bands of physically
and chemically adsorbed oleic acid (Figure S3). As directly confirmed by the IR
spectra shown in Figure S3, the inactivation of the composite oxides for the
esterification of oleic acid can be ascribed to the fact that the catalyst surface was
saturated by the chemical and physical adsorbed oleic acids, thereby blocking the
active sites present on the catalysts, as the scheme shown in Figure S4.
The IR spectra shown in Figure 5 were taken from the sulfate modified
composite oxides filtrated out from the reaction mixture of oleic acid and methanol
2
ꢀ
after 2 hours of the reaction. In the IR spectrum of the used SO
4
/MgꢀFe oxide, the
bands attributed to CꢀH stretching vibration appear as a weak shoulder of the broad
14

Page 15 of 27
Green Chemistry
View Article Online
DOI: 10.1039/C5GC00188A
ꢀ1
bands in the region of 3700ꢀ2800 cm , and the characteristic IR band relating to
ꢀ1
chemically adsorbed oleic acid at 1743 cm is too low to be seen, indicating that
grafting sulfate groups efficiently inhibits the bonding of oleic acid on its surface. In
2
ꢀ
contrast to the IR spectrum of the used SO /MgꢀFe oxide, there is a clear band at
4
ꢀ
1
2ꢀ
2ꢀ
1
743 cm in the IR spectrum of the used SO /SrꢀFe oxide and the used SO /CaꢀFe
4 4
2
ꢀ
oxide, indicating that more oleic acid chemically adsorbed on the used SO /SrꢀFe
4
2
ꢀ
2ꢀ
4
oxide and used SO /CaꢀFe oxide rather than on used SO /MgꢀFe oxide. Moreover,
4
2
4
ꢀ
the IR spectrum of the used SO /MgꢀFe oxide exhibits an intense broad band in the
ꢀ1
ꢀ1
region of 3700ꢀ2800 cm , and a narrow band at 1636 cm , suggesting that its surface
is still rich in hydrogen bonded hydroxyl groups (HꢀOH) and physically adsorbed
2
ꢀ
water. For the IR spectrum of SO /SrꢀFe oxide, the broad band associated with
4
2
ꢀ
HꢀOH observed in the IR spectrum of a fresh SO /SrꢀFe oxide is displaced by a
4
ꢀ
1
2ꢀ
sharp band at 3465 cm in the IR spectrum of a used SO /SrꢀFe oxide (Figure 5).
4
3
4
This could be contributed by the isolated hydroxyl groups (isolated OH) . The
transfer in the configuration of surface hydroxyl groups can be interpreted as that the
distance of adjacent hydroxyl groups is enlarged after the reaction. When the reaction
2
ꢀ
is conducted over a SO /SrꢀFe oxide, oleic acids in considerable amount will
4
substitute for a portion of the hydroxyl groups originally bonded to the metal ions
without sulfate modification. As the adjacent hydroxyl groups are divided by bonded
oleic acids, the hydrogenꢀbonding interaction is no longer maintained. It can be
assumed that oleic acid and the hydroxyl group compete for the binding sites, the Sr
2
ꢀ
ions in particular, on SO /SrꢀFe oxide during the reaction, resulting in the
4
elimination of hydrogen bonded hydroxyl groups, as seen in Figure S5. Therefore,
2
4
ꢀ
o
calcining used SO /SrꢀFe oxide at 400 C to remove bonded oleic acid and then
exposing the treated catalyst to air would thus lead to the recovery of the HꢀOH on the
2
ꢀ
SO /SrꢀFe oxide surface (Figure 5). The following conclusion can be proposed with
4
2
4
ꢀ
regard to the distribution of adsorbed oleic acid on the SO /SrꢀFe oxide:
incorporating Sr ions would facilitate the adsorption of oleic acid, resulting in the
creation of the hydrophobic regions with densely bonded oleic acid.
15

Green Chemistry
Page 16 of 27
View Article Online
DOI: 10.1039/C5GC00188A
3
015
2928
2
857
1
636
2
-
UsedSO
4
/Mg-Fe oxide
1
743
2
-
4
UsedSO /Ca-Fe oxide
1
636
2
2
-
-
Fresh SO
4
4
/Sr-Fe oxide
/Sr-Fe oxide
2
928
3
465
2
857
1743
3
015
UsedSO
2
-
UsedSO
4
/Sr-Fe oxide
o
calcinedat 400 C
4000 3800 3600 3400 3200 3000 2800 2600 2400 2200 2000 1800 1600 1400 1200 1000 800
ꢀ
1
Wavenumber (cm )
2
ꢀ
Figure 5. IR spectra of the used sulfated composite oxides, fresh SO /SrꢀFe oxide
4
2
4
ꢀ
o
and the used SO /SrꢀFe oxide calcined at 400 C for 2 hours.
3
.3 Esterification catalytic reactivity
The catalytic reactivity of the catalysts was evaluated in the esterification of oleic
o
acid with methanol at 100 C. As seen the initial rates (r ) shown in Table 4, all the
0
2
ꢀ
sulfated composite oxides exhibit higher reactivity than SO /Fe O , indicating that
4
2
3
the incorporated alkaline earth metal ions play a key role in the promotion of catalytic
2
4
ꢀ
2ꢀ
4
reactivity. It was found that both SO /ZrO and SO /SrꢀFe oxide exhibited higher
2
catalytic activities for the reaction in comparison with others, and in particular the
2
ꢀ
2ꢀ
high catalytic reactivity of SO /SrꢀFe oxide is comparable with that of SO /ZrO
2
4
4
(Table 4).
Nevertheless, when the reactivity of the active sites was investigated in terms of
16

Page 17 of 27
Green Chemistry
View Article Online
DOI: 10.1039/C5GC00188A
2
ꢀ
turnover frequency (TOF), the results show that the sites present on SO /MgꢀFe
4
2
ꢀ
oxide show a similar rate per active site to those on SO /SrꢀFe oxide (Table 4). These
4
results are obviously contrary to those of the Hammett indicator analysis, in which the
2
ꢀ
acid strength of SO /SrꢀFe oxide is about two orders of magnitude higher than that
4
2
4
ꢀ
of SO /MgꢀFe oxide (Table 2). To explain these contradictions, it is worth to note
again that incorporating Sr cations can facilitate the bonding of oleic acid on the
catalyst. Melero et al. reported that both Lewis and Brönsted acid sites are able to
3
8
catalyze the progress of the esterification of fatty acids to produce biodiesel .
Brönsted acid sites catalyze the esterification of fatty acids via the protonation of the
acid group (ꢀCOOH) to give oxonium ions, while the Lewis acid sites catalyze the
esterificaiton of fatty acids through the coordination of acid groups on the active sites.
The ꢀCOOH moiety of an oleic acid is hydrophilic. As a region on the catalyst surface
is covered by dense oleic acid, this makes it difficult for the ꢀCOOH moiety of an
oleic acid to approach the active sites located in this hydrophobic region, further
resulting in the deactivation of the active sites. The region with densely bonded oleic
acid is denoted as high diffusion barrier zone (Figure 6).
Table 4. Catalytic reactivity of sulfated catalyst samples for the catalysis of the
esterification of oleic acid with methanol
r0a
Conversion of oleic acid
(%)
ꢀ1
ꢀ1
TOF (min )b
ꢀ1
Catalyst
(mM
·s
·
g of
5
min
120 min
catalyst)
2
ꢀ
ꢀ
c
SO
4
/SrꢀFe oxide
ꢀ
39.45±1.34
13.09±1.55
8.74±0.37
36.50±2.12
5.77±1.03
36.4±0.57
33.41±1.41
28.50±2.83
10.60±0.14
96.97±3.35
77.72±0.87
72.46±1.22
98.56±0.34
63.56±1.35
97.86±0.65
96.61±1.85
87.04±3.06
67.88±5.12
3.44±0.12
1.05±0.06
0.76±0.01
3.27±0.19
0.65±0.01
3.18±0.05
2.92±0.12
2.49±0.25
0.93±0.01
1.06±0.04
0.44±0.03
1.03±0.02
3.11±0.18
0.62±0.01
0.96±0.01
1.00±0.04
2.22±0.22
2.31±0.03
2
SO
4
/CaꢀFe oxide
2
SO
4
/MgꢀFe oxide
2
ꢀ
4 2
SO /ZrO
2
ꢀ
4 2 3
SO /Fe O
2
2
2
2
ꢀ
ꢀ
ꢀ
ꢀ
SO
SO
SO
SO
4
/SrꢀFe oxideꢀ1
/SrꢀFe oxideꢀ2
/SrꢀFe oxideꢀ3
/SrꢀFe oxideꢀ4
4
4
4
a
initial rate
b
initial turnover frequency. TOF is calculated based on the following equation.
M FFA
TOF =
Msite ⋅t
Where MFFA and Msite are mole of the oleic acid consumed in the reaction per
unit time (t) and mole of active sites, respectively. Msite is obtained from the
result of nꢀbutyl amine titration for each catalyst.
17

Green Chemistry
Page 18 of 27
View Article Online
DOI: 10.1039/C5GC00188A
c
surface Sr:Fe atomic ratio of 1:1.17
O
O
O
OH
OR
OH
O
OH
H
2
O
ROH
Bronsted
acid site
O
O
O
O
O
O
O
O
O
+
O
O
O
O
+
H
H
S
+
H
S
+
H
S
H
H
S
O
O
H
O O
H
O
O
O
O
H
O
O
O
H
O
O
O
O
H
O
O
O O
H
O
H
H
O
O O
H
O
O
O
H
O
M1 O M1 O M1 O M1
O
M1 O M
O
M1 O M1 OM
O M1 O M1
O
M1 O M1 O M2 O M2 O M2 O M1 O M1 O
M2 O M2 O M
1
O
M2 O M1 O M1 O
2
1
Lewis acid site
High diffusion barrier zone
Low diffusion barrier zone
M
M
1
2
=Sr cation
=Fe cation
Figure 6. Schematic diagram for explaining the deactivation of the active sites located
in the high diffusion barrier zone.
2
4
-
3
.4 Effect of the implanted iron cations on the reactivity of SO /Sr-Fe oxide
In order to study the deactivation mechanism of the active sites present on the
2
4
ꢀ
sulfated composite oxide surfaces, a series of SO /SrꢀFe oxides with various Sr to Fe
2
ꢀ
ratio and SO /SrO were used as the catalysts to individually catalyze the
4
o
esterification at 100 C. In terms of TOF, the decrease in the surface Sr ion density
accompanied an increase in the rate per active site for the catalysts with various Fe/Sr
2
ꢀ
atomic ratio (Table 4). When the atomic ratio of Fe/Sr reaches 34.58 (SO /SrꢀFe
4
2
ꢀ
2ꢀ
4
oxideꢀ4), the TOF of SO /SrꢀFe oxideꢀ4 is comparable to that of SO /ZrO . After
4
2
the reaction was completed, the density of the oleic acid bonded on the used
2
ꢀ
2ꢀ
SO /SrꢀFe oxide and SO /SrO samples were measured using TGA, whose profiles
4
4
are shown in Figure S6 and Figure S7, and TGA results are summarized in Table 5.
Obviously, the densities of the bonded oleic acid shown in Table 5 disclose that the
content of the bonded oleic acids is proportional to the density of the Sr ions on the
catalysts.
2
ꢀ
The influence of bonded oleic acid on the reactivity of SO /SrꢀFe oxide was
4
2
4
ꢀ
further proved by carrying out the esterification over the SO /SrꢀFe oxide preꢀtreated
o
2ꢀ
2ꢀ
at 700 C (denoted as preꢀtreated SO /SrꢀFe oxide) and the used SO /SrꢀFe oxide
4
4
o
recovered at 400 C, respectively. It can be concluded from the IR spectra (Figure 3)
and the acidic density data (Table 1) that a portion of bonded sulfate groups remains
2
ꢀ
on preꢀtreated SO /SrꢀFe oxide, and the density of the residual sulfate ions is
4
ꢀ2
2
4
.29×10 mmol/m . In comparison to the catalytic reactivity of the catalysts for the
esterification, a dramatic decrease in the catalytic reactivity was found on preꢀtreated
2
ꢀ
SO /SrꢀFe oxide (Figure 7), while no obvious reactivity loss was observed on the
4
2
4
ꢀ
o
used SO /SrꢀFe oxide recovered at 400 C (Figure 8). These results suggest that the
18

Page 19 of 27
Green Chemistry
View Article Online
DOI: 10.1039/C5GC00188A
2
ꢀ
deactivation of preꢀtreated SO /SrꢀFe oxide is not due to the leaching of the grafted
4
2
4
ꢀ
sulfate ions on preꢀtreated SO /SrꢀFe oxide. Investigating the reactivity of the active
sites in terms of TOF (representing the reactivity per active site of a catalyst) shows
2
4
ꢀ
that the TOF of SO /SrꢀFe oxide is about five times higher than that of preꢀtreated
2
ꢀ
ꢀ1
SO /SrꢀFe oxide (TOF=0.20±0.05 min ). Considering the nature of the active sites
4
2
ꢀ
2ꢀ
present on SO /SrꢀFe oxide and preꢀtreated SO /SrꢀFe oxide, they should have
4
4
similar catalytic reactivity for the esterification because these two catalysts have the
same crystalline phase and similar surface characteristics (e.g., surface configuration,
specific surface area and Fe/Sr ratio), except for the density of the oleic acids bonded
on the surfaces during the reaction. It was noted previously that the bonded sulfate
groups can prevent the adsorption of oleic acid. In other words, partially removing the
o
covalently bonded sulfate ions at 700 C would create more Sr ions without bonded
sulfate ions, and further lead to adsorb more oleic acids during the reaction. These
results demonstrate that the decrease in the rate per active site (TOF) for preꢀtreated
2
ꢀ
SO /SrꢀFe oxide can be related to the increase in bonded oleic acid density. This
4
finding is in agreement with our previous assumption that the deactivation of the
2
4
ꢀ
active sites on a SO /SrꢀFe oxide is due to the creation of the high diffusion barrier
2
4
ꢀ
zones on the SO /SrꢀFe oxide with high Sr/Fe ratio (Figure 6). TOF is defined as mol
of oleic acid (MFFA) converted in the reaction per mol of active site (M ) per time (t).
site
When the active sites located in high diffusion barrier zone are inactive for the
reaction, it leads to the overestimation in the density of the active sites actually
catalyzing the reaction, and subsequently causes the underestimation in the TOF of
2
4
ꢀ
the catalysts. It can be imaged that most active sites on SO /SrꢀFe oxideꢀ4 surface
2
4
ꢀ
located in low diffusion barrier zone since SO /SrꢀFe oxideꢀ4 surface features low
2
4
ꢀ
Sr/Fe ratio. This explains why the TOF of SO /SrꢀFe oxideꢀ4 is greater than the TOF
2
ꢀ
of the SO /SrꢀFe oxides with a low Fe/Sr ratio.
4
The influence of the implanted Fe ions on the configuration of the covalently
bonded sulfate ions as well as on the catalytic reactivity was investigated by carrying
2
ꢀ
2ꢀ
4
out the esterification over SO /SrO and SO /SrꢀFe oxideꢀ1. The IR spectra of the
4
catalysts after reaction are shown in Figure S8, showing that a unique IR band at 1455
ꢀ
1
2ꢀ
cm only appears in the IR spectrum of used SO /SrO. This IR band reveals the
4
2
ꢀ
presence of the polynuclear sulfate (S O ) produced from the reaction between two
2
7
33
adjacent sulfate groups . Since the bonded sulfated groups in large amount still
2
ꢀ
o
remained on the used SO /SrO which was recovered at 800 C (Figure S8), the
4
ꢀ2
2
bonded sulfated ion density of >13.53×10 mmol/m was obtained from the weight
o
o
2ꢀ
loss from 400 C to 800 C in the TGA profile of SO /SrO (Figure S7). According to
4
the density of the boned sulfate ions, it was found that there are more bonded sulfate
2
4
ꢀ
2ꢀ
ions present on SO /SrO rather than on SO /SrꢀFe oxideꢀ1. However, the density of
4
19

Green Chemistry
Page 20 of 27
View Article Online
DOI: 10.1039/C5GC00188A
2
ꢀ
ꢀ2
2
the acid sites on SO /SrO (1.56±0.35×10 mmol/m ) is about 1/17 of that on
4
2
ꢀ
2ꢀ
ꢀ1 ꢀ1
SO /SrꢀFe oxideꢀ1. As seen in Figure S9, SO /SrO (r =0.55±0.14 mM·s ·g of
4
4
0
2
ꢀ
catalyst) also exhibits much lower reactivity than SO /SrꢀFe oxideꢀ1 (Figure S9 and
4
2
7
ꢀ
Table 4). It can be concluded from these results that the formation of S O will lead
2
to the deactivation of the adjacent active sites, and this conclusion is in agreement
3
3
2ꢀ
4
with the results reported in previous study . On the SO /SrꢀFe oxide surface, since
sulfate ion tends to bond on Sr cation rather on Fe ion, the formation of polynuclear
sulfate species can be efficiently prohibited when the incorporated Sr ions are divided
by incorporated Fe ions on the catalyst surface. On the other hand, Sr ion has higher
binding ability for oleic acid during the reaction in comparison to Fe ion. Incorporated
2
4
ꢀ
Fe ions can disperses the distribution of bonded oleic acid on SO /SrꢀFe oxide,
leading to the inhibition of the creation of high diffusion barrier zone.
1
00
90
80
70
60
50
40
30
20
10
0
0
20
40
60
80
100
120
Reaction time (min)
Figure 7. Conversion versus reaction time for oleic acid esterification catalyzed by (▲)
2
ꢀ
2ꢀ
o
SO /SrꢀFe oxide and by (●) the SO /SrꢀFe oxide preꢀtreated at 700 C for 2 hours.
4
4
o
The oleic acid esterification was carried out at 100 C.
20

Page 21 of 27
Green Chemistry
View Article Online
DOI: 10.1039/C5GC00188A
100
8
6
4
2
0
0
0
0
0
0
20
40
60
80
100
120
140
Reaction time (min)
Figure 8. Conversion versus reaction time for oleic acid esterification catalyzed by (●)
2
ꢀ
2ꢀ
fresh SO /SrꢀFe oxide and by (▲) recovered SO /SrꢀFe oxide obtained by
4
4
2
ꢀ
o
heatꢀtreating the used SO /SrꢀFe oxide at 400 C. The oleic acid esterification was
4
o
carried out at 100 C.
2
ꢀ
2ꢀ
Table 5. Density of bonded oleic acid on used SO /SrꢀFe oxides and used SO /SrO
4
4
Density of bonded oleic acid
Catalyst
ꢀ
2
2
(
×10 mmol/m )
2
2
ꢀ
ꢀ
Used SO
4
/SrꢀFe oxideꢀ1
4.30
1.41
1.34
0.12
7.77
Used SO4 /SrꢀFe oxideꢀ2
2
ꢀ
Used SO4 /SrꢀFe oxideꢀ3
2
ꢀ
Used SO4 /SrꢀFe oxideꢀ4
2ꢀ
Used SO4 /SrO
3
.5 Reusability study
As noted in the literature, a dramatic loss in catalytic reactivity for a recovered
2
ꢀ
SO
4
/ZrO
2
is ascribed to the significant leaching of adsorbed sulfate groups in the
2,22,23
reaction mixture
. It was found that the conversion of oleic acid decreased from
2
ꢀ
9
8.85% to 52.14% after SO /ZrO was recycled two times (Figure 9). This decline in
4 2
catalytic reactivity for a sulfated catalyst may be due to sulfate groups leaching in the
reaction mixture or the substrate poisoning the active sites. In this case, the decline in
activity could be ascribed to the low stability of the surface sulfate groups, because
2
4
ꢀ
o
removing the adsorbed species by heatꢀtreating the used SO
/ZrO
2
at 400 C did not
recover all catalytic reactivity.
2
ꢀ
As seen in Figure 9, the catalytic reactivity of SO
4
/MgꢀFe oxide continuously
21

Green Chemistry
Page 22 of 27
View Article Online
DOI: 10.1039/C5GC00188A
declined with an increase in the number of recycles, while the reactivity of
2
ꢀ
SO /CaꢀFe oxide was almost fully recovered during the first few cycles, and a
4
th
significant decline in reactivity was found after the 4 cycle. Interestingly,
2
ꢀ
SO /SrꢀFe oxide can be reused multiple times without an obvious decrease in
4
2
ꢀ
catalytic reactivity, and 84% catalytic reactivity was recovered after SO /SrꢀFe oxide
4
was recycled 13 times (Figure 10). It is likely that the stability of a sulfated composite
oxide catalyst is related to the alkaline earth metal cations implanted in them. The
strength of the covalent bond between metal alkaline earth metal cations and sulfate
groups is proportional to the bond polarity, as well as the difference in the
3
9
electronegativity of metal cations and oxygen anions . As the alkaline earth metal
cations have a lower degree of electronegativity, such as Sr, the high bond strength of
the covalent bonds between sulfated groups and alkaline earth metal cations may
2
4
ꢀ
explain why the sulfate groups on SO /SrꢀFe oxide have more stability in the
2
ꢀ
2ꢀ
reaction mixture rather than the sulfate groups on SO /CaꢀFe oxide and SO /MgꢀFe
4
4
th
oxide. After the 14 cycle, a serious aggregation of the catalysts was observed when
th
the recovered catalyst was reꢀdispersed in the reaction mixture in the 15 cycle. The
serious aggregation of the recovered catalyst is due to that the surface sulfate groups
in considerable amount were replaced by covalently bonded oleic acids, leading to a
significant change in hydrophilic/hydrophobic property on the catalyst surface.
2
ꢀ
O
SO /ZrO
SO42ꢀ/Zr 2
4 2
1
00
SꢀCF
SꢀMF
80
60
40
20
0
Fresh
1
2
3
4
Number of Reuses
Figure 9. Conversion of oleic acid over recycled catalysts. Reaction conditions:
amount of catalyst 0.8g, oleic acid to methanol molar ratio 1:4, reaction temperature
o
1
00 C, reaction time 2 hours.
22

Page 23 of 27
Green Chemistry
View Article Online
DOI: 10.1039/C5GC00188A
100
80
60
40
20
0
Fresh 1
2
3
4
5
6
7
8
9
10 11 12 13 14 15
Number of Reuses
2
ꢀ
Figure 10. Conversion of oleic acid over recycled SO /SrꢀFe oxide. Reaction
4
conditions: amount of catalyst 0.8g, oleic acid to methanol molar ratio 1:4, reaction
o
temperature 100 C, reaction time 2 hours.
3
.6 Kinetic study
The kinetic parameters of the esterification of oleic acid with methanol over
2
ꢀ
SO /SrꢀFe oxide were studied using an excess of methanol at different temperatures.
4
The results of KorosꢀNowak test indicate that catalytic reactions took place in the
kinetic region on the active sites present in the low diffusion barrier zone (Figure S10).
The plot of ꢀln(C /Ca ) versus reaction time shows a linear relationship, indicating
a
0
that the reaction is a pseudo first order one with respect to the concentration of oleic
acid (C ), as seen Figure S11. The activation energy (E ) was determined by fitting the
a
a
o
rate constants at 70, 80, 90, 100 and 110 C using Arrhenius plot as given blow:
−
RT
Ea
Ea
k = Aexp(
) or ln k = ln A −
RT
where k is the rate constant, T is the reaction temperature in Kelvin, A is
preꢀexponential factor, R is the universal gas constant and E is the activation energy.
a
A plot of ln(k) versus 1/T was plotted to determine the E based on the slope of
a
the resulting graph shown in Figure S12. The activation energy was found to be
ꢀ
1
40
2
8.53±0.72 kJmol . Satyarthi et al. reported that the activation energy of the
ꢀ1
sulfated zirconia in the esterification of oleic acid with methanol is 24.40 kJmol ,
2
4
ꢀ
where is close to the activation energy of SO /SrꢀFe oxide. The low activation
energy can be attributed to the promotion in the acid strength due to the high
induction effect of the sulfated ion grafted on Sr cation when the esterification of oleic
acid is catalyzed over the active sites located in low diffusion barrier zone on
2
ꢀ
SO /SrꢀFe oxide. By comparison with the kinetic parameters of the currently used
4
23

Green Chemistry
Page 24 of 27
View Article Online
DOI: 10.1039/C5GC00188A
2
ꢀ
catalysts in Table S1, SO /SrꢀFe oxide gives a fairly good performance in catalyzing
4
the esterification of fatty acid with short chain alcohol.
4
. Conclusion
This study examined the potential of using alkaline earth metalꢀferric composite
oxide to enhance the stability of grafted sulfate ions and the catalytic reactivity of the
active sites in the esterification of oleic acid with methanol for the production of
biodiesel. The results show that acidic strength of the sulfated composite oxides was
associated with the electronegativity of the incorporated alkaline earth cation. In
2
ꢀ
particular, SO /SrꢀFe oxide was classified as a solid superacid, and its acidity
4
function (H ) determined by Hammett indicators was < ꢀ12.70 due to the low
0
electronegativity of Sr cations in comparison with others. It was found that the acid
site density and the catalytic reactivity for the esterification are associated with the
2
ꢀ
distribution of the Sr ions incorporated on SO /SrꢀFe oxide. Examining the catalytic
4
2
ꢀ
reactivity in term of TOF, the higher TOF value was obtained from the SO /SrꢀFe
4
oxide with higher surface Fe/Sr atomic ratio. About 85% reactivity could be recovered
2
4
ꢀ
after SO /SrꢀFe oxide was recycled 13 times in the esterification of oleic acid with
methanol because of the high chemical/thermal stability of the sulfate ions bonded on
Sr cations. It can be concluded from these results that implanting Sr ions facilitated
the grafting of sulfate ions to create more acid sites, but also facilitated the binding of
2
ꢀ
oleic acid to form high diffusion barrier zones over SO /SrꢀFe oxide, leading to the
4
deactivation of the active site circulated by densely bounded oleic acid. The catalytic
2ꢀ
2ꢀ
4
o
study of SO /SrO and preꢀtreated SO /SrꢀFe at 700 C provides direct evidences to
4
prove that the incorporation of Fe ions on catalyst surface could efficiently suppress
the deactivation of the active sites locating in the area with densely bonded oleic acid
and the further reaction between adjacently bonded sulfate ions to form polynuclear
sulfate.
Acknowledgements
Financial support for this work was provided by Taiwan’s National Science
Council (Grants No. NSC 103ꢀ2221ꢀEꢀ126ꢀ017ꢀMY2 and MOST 103ꢀ3113ꢀEꢀ006
ꢀ006 ꢀ).
References
1
P.D. Patil, V.G. Gude, S. Deng, Ind. Eng. Chem. Res., 2009, 48, 10850.
E. Lotero, Y. Liu, D.E. Lopez, K. Suwannakarn, D.A. Bruce, J.G. Goodwin, Ind.
2
Eng. Chem. Res., 2005, 44, 5353.
K.ꢀH. Chung, D.ꢀR. Chang, B.ꢀG. Park, Bioresour. Technol., 2008, 99, 7438.
A.S. Ramadhas, S. Jayraj, C. Muraleedharan, Fuel, 2005, 84, 335.
3
4
24

Page 25 of 27
Green Chemistry
View Article Online
DOI: 10.1039/C5GC00188A
5
6
J. Van Gerpen, M. Canakci, Am. Soc. Agric. Eng., 2001, 44, 1429.
E. Lotero, Y. Liu, D.E. Lopez, K. Suwannakaran, D.A. Bruce, J.G. Goodwin Jr, Ind.
Eng. Chem. Res., 2005, 44, 5353.
M. Di Serio, M. Cozzolino, M. Giordano, R. Tesser, P. Patrono, E. Santacesaria, Ind.
Eng. Chem. Res., 2007, 46, 6379.
S.A. Ardizzone, P. Beltrame, G. Zuretti, Appl. Catal. A, 2006, 314, 240.
C.M. Garcia, S. Teixeira, L.L. Marciniuk, U. Schuchardt, Bioresource Technol.,
7
8
9
2
008, 99, 6608.
J.L. RoperoꢀVega, A. AldanaꢀPérez, R. Gómez, M.E. NiñoꢀGómez, Appl. Catal. A,
010, 379, 24.
L.K. Noda, R.M. de Almeida, L.F.D. Probst, N.S. Gon
005, 225, 39.
J.I. Moreno, R. Jaimes, R. G mez, M.E. NiñoꢀGómez, Gatal. Today, 2011, 172, 34.
Y. Du, S. Liu, Y. Zhang, C. Yin, Y. Di, F.ꢀS. Xiao, Catal. Letters, 2006, 108, 155.
H. Yang, H. Song, H. Zhang, P. Chen, Z. Zhao, J. Mol. Catal. A, 2014, 381, 54.
B. Fu, L. Gao, L. Niu, R. Wei and G. Xiao, Energ. Fuel., 2009, 23, 569.
L.S. Ling, H. Hamdan, J. Non-Cryst Solids, 2008, 354, 3939.
D. Prasetyoko, Z. Ramli, S. Endud, H. Nur, J. Mol. Catal. A, 2005, 241, 118.
T. Wakayama, H. Matsuhashi, J. Mol. Catal. A, 2005, 239, 32.
B.M. Reddy, M.K. Chem. Rev., 2009, 109, 2185.
G.D. Yadav, J.J. Nair, Micropor. Mesopor. Mater., 1999, 33, 1.
K. Arata Solid Superacids. Adv. Catal., 1990, 37, 165ꢀ211.
1
1
0
1
2
çalves, J. Mol. Catal. A,
2
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
3
4
5
6
7
8
9
0
1
2
3
4
5
ó
S. Ardizzone, C.L. Bianchi, V. Ragaini, B. Vercelli, Catal. Lett., 1999, 62, 59.
A. Corma, H. García, Catal. Today, 1997, 38, 257.
J. Ni, F.C. Meunier, Appl. Catal. A, 2007, 333, 122.
C.ꢀL. Chen, C.ꢀC. Huang, D.ꢀT. Tran, J.ꢀS. Chang, Bioresour. Technol., 2012, 113
,
8
.
2
6
H.ꢀC. Wang, C.ꢀY. Wu, C.ꢀC. Chung, M.ꢀH. Lai and T.ꢀW. Chung, Ind. Eng. Chem.
Res., 2006, 45, 8043
H. Ranjan Sahu and G. Ranga Rao, Bull. Mater. Sci., 2000, 23, 349.
M.L. GuzmánꢀCastillo, E. LópezꢀSalinas, J.J. Fripiat, J. SánchezꢀValente, F.
HernándezꢀBeltrán, A. RodríguezꢀHernández, J. NavarreteꢀBolaños, J. Catal.,
2
2
7
8
2
003, 220, 317.
2
3
9
0
R.M. de Almeida, L.K. Noda, N.S. Gonçalves, S.M.P. Meneghetti, M.R.
Meneghetti, Appl. Catal. A, 2008, 347, 100.
A. Mantilla, G. Ferrat, A. L
ópezꢀOrtega, E. Romero, F. Tzompantzi, M. Torres, E.
Ort zꢀIslas, R. G mez, J. Mol. Catal. A, 2005, 228, 333.
X. Li, W. Huang, Energy Source, Part A, 2009, 31, 1666.
A.S. Khder, E.A. ElꢀSharkawy, S.A. ElꢀHakam, A.I. Ahmed, Catal. Commun.,
í
ó
3
3
1
2
2
008, 9, 769.
3
3
3
3
4
5
G.D. Yadav, A.D. Murkute, J. Catal., 2004, 224, 218.
C.ꢀC. Huang, K.L. Hohn, J.R. Schlup, J. Phys. Chem. C, 2009, 113, 11050.
S. Tominaka, T. Momma, B. Scrosati, T. Osaka, J. Power Sources, 2010, 195
,
4
065.
3
3
3
3
6
7
8
9
R.J. Gillespie, T.E. Peel, E.A. Robinson, J. Am. Chem. Soc., 1971, 93, 5083.
M. Yurdakoç, M. Akçay, Y. Tonbul, K. Yurdakoç, Turk. J. Chem., 1999, 23, 319.
J.A. Melero, J. Iglesias, G. Morales, Green Chem., 2009, 11, 1285.
M. S. Chandrasekharaiah, J. Phys. Chem., 1964, 68, 2020.
25

Green Chemistry
Page 26 of 27
View Article Online
DOI: 10.1039/C5GC00188A
4
0
J.K. Satyarthi, D. Srinivas, P. Ratnasamy, Energy Fuels, 2010, 24, 2154.
26

Page 27 of 27
Green Chemistry
O
O
O
View Article Online
D
OROI: 10.1039/C5GC00188A
OH
OH
O
OH
H O
2
ROH
Bronsted
acid site
O
O
O
O
O
O
O
O
O
+
O
O
O
O
+
H
H
S
+
H
S
+
H
S
H
H
S
O
O
H
O O
H
O
O
O
O
H
O
O
O
H
O
O
O
O
H
O
O
O O
H
O
H
H
O
O O
H
O
O
O
H
O
M1 O M1 O M1 O M1 O M1 O M
O M1 O M1 OM1 O M1 O M1
O
M1 O M1 O M2 O M2 O M2 O M1 O M1 O M2 O M2 O M
O M2 O M1 O M1 O
1
2
Lewis acid site
High diffusion barrier zone
Low diffusion barrier zone
M =Sr cation
1
M =Fe cation
2
Schematic diagram for explaining the deactivation of the active sites located in the
high diffusion barrier zone.