Sulfonated niobia and pillared clay as catalysts in etherification reaction of glycerol

Article abstract of DOI:10.1016/j.apcata.2014.03.037

Sulfonated niobia (HY-340 CBMM) and pillared clay (Fluka) were tested in the catalytic conversion of glycerol by etherification reactions. The solids were treated with concentrated fuming sulfuric acid (AS100 and NS100), and a 30% aqueous solution of this acid (AS30 and NS30). Both the presence of sulfur and the increase in the acidity of the solids demonstrate the suitability of the sulfonation process, especially in samples treated with concentrated fuming sulfuric acid. The best catalyst for the reaction of etherification with tert-butyl alcohol was AS100, with a glycerol conversion of 95% after 5 h at 393 K and yield of 60.3, 33.2 and 5.4%, respectively for mono-tert-butyl-glycerol (MTBG), di-tert-butyl-glycerol (DTBG) and tri-tert-butyl-glycerol (TTBG).

Full text of DOI:10.1016/j.apcata.2014.03.037

Applied Catalysis A: General 478 (2014) 98–106
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Sulfonated niobia and pillared clay as catalysts in etherification
reaction of glycerol
Patrícia A. Celdeira, Maraisa Gon c¸ alves, Flávia C.A. Figueiredo, Sandra Maria Dal Bosco,
∗
Dalmo Mandelli, Wagner A. Carvalho
Centro de Ciências Naturais e Humanas, Universidade Federal do ABC, Rua Santa Adélia, 166, Santo André CEP 09210-170, SP, Brazil
a r t i c l e i n f o
a b s t r a c t
Article history:
Sulfonated niobia (HY-340 CBMM) and pillared clay (Fluka) were tested in the catalytic conversion of
glycerol by etherification reactions. The solids were treated with concentrated fuming sulfuric acid (AS100
and NS100), and a 30% aqueous solution of this acid (AS30 and NS30). Both the presence of sulfur and the
increase in the acidity of the solids demonstrate the suitability of the sulfonation process, especially in
samples treated with concentrated fuming sulfuric acid. The best catalyst for the reaction of etherification
with tert-butyl alcohol was AS100, with a glycerol conversion of 95% after 5 h at 393 K and yield of
Received 8 November 2013
Received in revised form 25 March 2014
Accepted 28 March 2014
Available online 6 April 2014
Keywords:
Glycerol
Etherification
Sulfonation
Niobia
6
0.3, 33.2 and 5.4%, respectively for mono-tert-butyl-glycerol (MTBG), di-tert-butyl-glycerol (DTBG) and
tri-tert-butyl-glycerol (TTBG).
©
2014 Elsevier B.V. All rights reserved.
Pillared clay
1
. Introduction
as a result of polymerization reactions. MTBG is the main product
obtained in glycerol etherification, as a result of the electrophilic
attack of the tert-butyl cation (a tertiary carbocation) preferably on
the primary carbon of glycerol due to steric hindrance and electro-
static effects exerted by OH glycerol groups. However, DTBG and
TTBG are preferred as potential oxygenate additives to diesel fuels
because of their ease of blending and non-polar properties.
Several processes have been used for the chemical transfor-
mation of glycerol, a valuable by-product of biodiesel production
1]. Among these processes, the conversion of glycerol into oxy-
[
genated fuels through etherification reactions has recently been
explored [2–5]. This approach represents a promising and eco-
nomic alternative, since it makes use of a by-product of biodiesel
production allowing increased yield of a biodiesel manufacturing
process. Ethers, in particular, may be used as additives to diesel,
since they have the ability to promote cleaner burning in internal
combustion engines. Mixtures of tert-butyl ethers of glycerol with
a high content of diethers, and particularly triethers, are known as
potential additives for biodiesel [6,7]. These ethers can act reducing
emissions of particulate matter, hydrocarbons, carbon monoxide,
and aldehydes, and reduce the viscosity of the biodiesel, favoring
their use. This topic is of great importance because of the grow-
ing demand for new additives, specifically for biodiesel, which are
biodegradable, non-toxic and renewable.
Most studies uses isobutylene (IB) as the etherification agent
[8–10]. However, it should be noted that the use of TBA, in substi-
tution of IB, avoids both the need of solvents and the mass transfer
limitation phenomena related to the complex three-phase system
[11]. Moreover, TBA is a by-product of the large-scale propylene
oxide production and it also can be produced from biosources, like
starch or lignocellulosic biomass [12,13]. A highly acidic solid acid
catalyst is required to initiate the reaction between two alcohols
[14]. Different types of acidic heterogeneous solid catalysts such as
sulfuric acid, sulfonated styrene-divinyl-benzene copolymers and
p-toluene-sulfonic acid, have been studied in glycerol conversion
and their high activity is caused by the Brønsted acid sites of the
sulfonic group. The reaction between glycerol and TBA at 1:4 molar
ratio can produce mono, di and tri butyl glycerol ethers. Reactions
are typically promoted between 343 and 373 K, with TBA:glycerol
molar ratio of 4:1. Gu et al. [14] reported that glycerol and TBA
at 1:1 molar ratio does not undergo self-etherification. Glycerol
conversion reaches values close to 100% after 5–8 h of reaction,
with selectivities to DTBG + TTBG as high as 80% [11,15]. The high-
est yield for glycerol ethers (DBGE + TBGE) was obtained with an
amount of catalyst equivalent to 7.5 wt% with respect to the mass
Glycerol etherification with tert-butyl alcohol (TBA) is an acid
catalyzed reaction, resulting in a mixture of mono-tert-butyl-
glycerol (MTBG), di-tert-butyl-glycerol (DTBG) and tri-tert-butyl-
glycerol (TTBG). Some unwanted by-products can also be formed
∗ Corresponding author. Tel.: +55 11 4996 8386; fax: +55 11 4996 8386.
E-mail addresses: wagner.carvalho@ufabc.edu.br,
wagner.carmense@gmail.com (W.A. Carvalho).
http://dx.doi.org/10.1016/j.apcata.2014.03.037
0
926-860X/© 2014 Elsevier B.V. All rights reserved.

P.A. Celdeira et al. / Applied Catalysis A: General 478 (2014) 98–106
99
of glycerol [11]. In general, the most suitable catalysts are com-
niobia treated with fuming sulfuric acid: NS100 and (vi) niobia
treated with 30% aqueous sulfuric acid solution: NS30.
®
mercial resins such as Amberlyst , but the major limitation on the
®
use of these resins is their thermal stability (393 K for Amberlyst -
®
1
5, reaching up to 463 K for Amberlyst -70). Furthermore, both
2.2. Characterization
activity and selectivity losses were observed during consecutive
tests [8,16].
Textural properties of the solids were obtained by nitrogen
Another possibility of active catalysts in these reactions is rep-
resented by niobia and pillared clays. Niobium-containing catalysts
have attracted great interest in heterogeneous catalysis as a cata-
lyst and support for a variety of important reactions. Niobia has a
high acidity, which, curiously, is maintained in aqueous environ-
ments and this particular property is favorable to the dehydration
reaction in aqueous phase [17–20]. Various attempts have been
made to develop catalysts from pillared smectite clays. Due to
their small particle size (<2 ␮m) and high intercalation capacity,
some clay minerals present high surface area, useful in adsorption
and catalysis processes [21]. Moreover, many specimens show cat-
alytic properties, especially after having been submitted to small
modifications in their composition or structure. Thermal stability,
associated with the ease of access of reagent molecules into the
interlayer catalytic sites, make the use of pillared clays as catalysts
increasingly attractive [22]. Compared with most oxides commonly
used as supports, it is possible to regard pillars as responsible for
some form selectivity in the reactions, since a two-dimensional net-
work of interconnected micropores is created which can adsorb
mono ethers which undergo further etherification [23]. The active
sites of the catalyst adsorb both fresh glycerol and the product
adsorption measurements at 77 K in an Autosorb-1MP device
Quantachrome Instruments). The “apparent” surface area was esti-
mated according to the BET equation, based on the adsorption
(
data in the partial pressure range (P/P ) from 0.05 to 0.25. Prior
0
to the measurements, the samples were outgassed at 423 K and
−
3
1
.3 × 10 Pa for 4 h, to remove moisture.
Chemical functional groups present in the samples surface,
especially sulfonic acid groups in the treated samples, were
determined by infrared spectroscopy using a Varian 3100 FT-IR
Spectrometer. The analyses were performed mixing dried samples
with potassium bromide (KBr) in a 1:30 weight ratio and ground
into fine powder. This mixture was dried at 373 K for 24 h and thin
pellets were made in manual equipment. The spectra were then
−
1
acquired at this temperature by accumulating 100 scans at 4 cm
−
1
resolution in the range of 400–4000 cm
.
NH3 temperature programmed desorption was used to deter-
mine the acid properties of the catalytic materials. The experiments
were conducted on a Quantachrome ChemBet 3000 device. The cat-
alysts (100 mg) were pretreated at 423 K for 30 min and then cooled
to 373 K under a He flow. The pretreated samples were saturated
with 5% NH /He for 1 h at 373 K, with subsequent flushing with
3
(
MTBG), undergo further etherification and produce DTBG and
helium at 373 K for 2 h to remove the physisorbed ammonia. TPD
analysis was carried out from 373 K to 1073 K at a heating rate of
TTBG. So the extent of bulk ethers adsorption depends upon the
porosity of the catalyst.
−
1
1
0 K min
X-ray diffraction patterns were obtained using a Philips X’Pert
.
Coupled with the fact that studies mainly focus on catalyst
development for etherification of glycerol with isobutylene, sul-
fonic commercial resins with low thermal stability have been
reported as the most active catalysts for glycerol etherification.
In this work we investigated the possibility of promoting the
acidity of two inorganic solids, niobia and pillared clay, by a sim-
ple procedure of sulfonic acid-functionalization. Besides not using
expensive reagents for sulfonation, the solids exhibit high thermal
stability, enabling their use in more drastic reaction conditions.
The activity and selectivity of these catalysts in the etherification
of glycerol was also evaluated. Sulfonated materials have been
characterized by Fourier Transform Infrared spectroscopy (FTIR),
nitrogen adsorption, Temperature-Programmed Desorption (TPD),
X-ray Diffraction (XRD) and Energy Dispersive X-ray Spectrome-
try (EDS). The solids have been tested as catalyst for etherification
of glycerol. The effect of the temperature, the TBA:glycerol molar
ratio as well as the reusability of the catalysts have also been eval-
PW3373 diffractometer, with Cu K␣ radiation and an excitation
power of 2 kW (40 kV and 50 mA).
The amount of sulfur on the carbon surface was determined by
EDS. Analyses were performed on a JEOL JSM-6701F field emission
scanning electron microscope operating at 10.0 kV and 10.0 mA.
2.3. Catalytic tests
Glycerol (99.5%—Sigma-Aldrich) etherification with tert-butyl
alcohol (TBA, 99.7%—Sigma-Aldrich) was carried out in a 300 mL
stainless steel batch reactor with a mechanical stirrer and under
autogenous pressure. The system was previously purged 2–3 times
with N2 in order to remove the air and the experiments were
performed under inert atmosphere, typically at 393 K, with 5 wt%
of catalyst and TBA:glycerol molar ratio varying from 4:1 to 8:1.
The reaction time was measured from the moment the system
achieved operating temperature (20 min after closing the reac-
tor). Samples were analyzed by gas chromatography (CG/FID,
Agilent 7890A, DB-Wax 30 m × 0.25 mm × 0.25 ␮m) using ace-
tonitrile as internal standard. The identification of the reaction
products was done by a gas chromatograph coupled to a mass
spectrometer (GC–MS, Shimadzu QP-2010Plus, SGE BP-20-strong
Wax 30 m × 0.25 mm × 0.25 ␮m) and confirmed according to the
method described by Jamróz et al. [24]. Replicates of the reactions
yielded data presenting coefficients of variation under 5%. The coef-
ficient of variation for GC/FID replicate quantification was below 3%.
The mass balance of the reactions analyzed presented values above
95%.
®
uated and compared to the commercial resins Amberlyst -15 and
®
Amberlyst -70.
2
. Materials and methods
2.1. Catalysts preparation
Samples of aluminum pillared clay (Fluka) and niobia (HY-340
CBMM) were sulfonated by treatment with fuming sulfuric acid
20% free SO ) or a 30% aqueous sulfuric acid solution. Treatments
(
3
were carried out using 100 mL of sulfuric acid and 10 g of each solid
kept under stirring at ambient temperature for 10 h. All samples
were filtered and repeatedly washed with warm water until neu-
trality and then dried in an oven at 393 K for 24 h. Solids were
identified as (i) parent pillared clay: A, (ii) pillared clay treated
with fuming sulfuric acid: AS100, (iii) pillared clay treated with
Glycerol conversion (%), product yield (%) and product selectiv-
ity (%) were calculated using the following equations:
moles of glycerol reacted
moles of glycerol taken
Glycerol conversion (%) =
× 100%
3
0% aqueous sulfuric acid solution: AS30, (iv) parent niobia: N, (v)
(1)

1
00
P.A. Celdeira et al. / Applied Catalysis A: General 478 (2014) 98–106
Fig. 1. Nitrogen adsorption/desorption isotherms of pillared clay (left) and niobia (right) samples.
typical of microporous materials with small fractions of mesopores,
as shown in Table 1. The micropore volume (VN2) was determined
Product yield (%) = moles of
a product obtained
moles of glycerol taken
× 100%
from the N isotherms by applying the Dubinin–Radushkevich (DR)
2
equation; the mesopore volume (Vmeso) was calculated by subtrac-
(2)
ting VN2 from the total volume of N adsorbed at a relative pressure
2
(
P/P ) of 0.97 (V0.97). No significant changes were observed in nio-
0
bia after sulfuric acid treatment, indicating that the material has a
higher stability than that observed in pillared clays.
The clay XRD diffractograms exhibit the pattern corresponding
to smectite. Parent pillared clay presents the 001 peak at 2ꢀ = 4.9
Product selectivity (%) =
moles of product obtained
moles of glycerol reacted
a
×
100%
(3)
(
Fig. S1). Thus, the spacing between the layers of material should
be d = 18.0 A˚ . After acid treatment, it is possible to observe two
3
. Results and discussion
.1. Catalysts characterization
The textural properties of the solids were evaluated by nitrogen
main changes. The peak 0 0 1 is shifted to a higher angle, so that
d = 15.0 A˚ in AS30 and d = 12.8 A˚ in AS100. In addition, there is a
3
reduction in the intensity of this signal in the solid AS100. These
results corroborate the results of adsorption/desorption of N , sug-
2
gesting the partial destruction of the pillars and/or lamellae in the
clay. This structural change is observed mainly after the treatment
with fuming sulfuric acid. There were no reflections related to the
niobia (Fig. S2), suggesting the occurrence of a material consisting of
amorphous pentavalent niobium oxide, without apparent modifi-
cations after the acid treatment.
adsorption/desorption isotherms at 77 K, shown in Fig. 1.
Isotherms of pillared clays present a H3 hysteresis loop, accord-
ing to the IUPAC classification. They present two asymptotic
branches to vertical P/P = 1, without any indication of plateaus at
0
P/P0 high values. Such isotherms are classified as Type I and IV,
which can be obtained with aggregates of plate-like particles. Thus,
the thin and flexible montmorillonite particles cause the presence
of non-rigid slit-shaped pores. After the acid treatment there occurs
a decrease in the specific surface area, mainly for the sample treated
with fuming sulfuric acid (Table 1). At least two changes occur in the
material subjected to acid treatment: (i) the partial closure of the
micropores caused by the inserted groups, (ii) the partial destruc-
tion of the pillars and/or lamellae of the material. Both changes
may limit access to the porous structure, thus reducing the specific
surface area of the solid. The results of X-ray diffraction, presented
later, should confirm the change made by this treatment.
IR spectra of pillared clay based materials (Fig. S3) indi-
cate the presence of the characteristic bands expected in clays:
−
1
−1
O H stretching, 3450 cm assigned
3635 cm assigned to the Al
to H H stretching (absorbed water), 1635 cm assigned to O
deformation, 1130–1050 cm
−
1
O
H
−
1
assigned to Si O stretching, and
−
1
770 cm assigned to Si O deformation [25]. Bands around 1035
−
1
and 1186 cm , assigned respectively to the SO2 asymmetric and
symmetric stretching vibration mode [26], could confirm the pres-
ence of SO3H groups. However, these bands may overlap the Si
O
bond stretching bands, characteristic of the clay. For niobia based
materials (Fig. S4) the presence of broad bands was identified at
− −1
1
N2 adsorption isotherms of niobia-based materials have basi-
cally the same profile, characterized by a mixture of Type I and IV,
3400 cm and 1625 cm , characteristic of surface O H groups,
−
1
and a band in the region of 650 cm , characteristic of the Nb
O
bond in amorphous materials [27]. In the spectra of the NS100
−
1
and NS30 samples there is a band between 1100 and 1170 cm
,
Table 1
BET surface area, total pore volume (V0.97), mesopore volume (Vmeso) and micropore
which can be attributed to the sulfonated group inserted by acid
treatment.
ꢀ
ꢁ
volume VN2 of the material analyzed.
The analysis of energy dispersive X-ray with elemental map-
ping provided data about the elemental composition. The mapping
of the major constituents of clays indicates the presence of sili-
con, aluminum and oxygen, in addition to magnesium (Fig. S5).
After acid treatment, the samples also exhibit sulfur whose content
A
AS30
120
0.18
0.05
AS100
N
NS30
175
0.15
0.07
NS100
2
−1
)
Surface area (m
g
250
100
0.18
0.04
130
150
0.12
0.06
V0.97 (cm3
g
g
−1
)
0.21
0.11
0.10
0.05
3
−1
)
Vmeso (cm
VN2 (cm3
ꢀ
ꢁ
−1
g
)
0.10
0.13
0.14
0.05
0.08
0.06
−2
−2
increases from 0.6 ± 0.1 ␮mol m
(0.24%) to 1.0 ± 0.2 ␮mol m

P.A. Celdeira et al. / Applied Catalysis A: General 478 (2014) 98–106
101
Fig. 2. NH3–TPD curves of the catalysts based on pillared clay (left) and niobia (right).
(
0.46%) in AS30 and AS100, respectively (Fig. S6). The mapping of
treatments have double the acidity of the material, although no
significant changes were observed in the acidity of the samples
niobia samples indicated only the presence of niobium and oxygen,
as expected (Fig. S7). After the acid treatment there is a homoge-
neous distribution of sulfur on the particle surface (Fig. S8). NS30
−
1
AS30 and AS100 (2.3 mmol g in both treated samples, compared
−
1
to 1.1 mmol g in the parent clay). However, it must be considered
that the textural analysis and X-ray diffraction results indicated a
possible change in the structure of this solid after treatment with
fuming sulfuric acid. The treatment may have affected the structure
of the pillars, causing partial collapse of the lamellae and reduc-
ing access to the porous structure. Thus, although the most drastic
treatment led to an increase in the concentration of sulfonic acid
groups on the solid surface, the total acidity was maintained.
−2
presented 0.6 ± 0.2 ␮mol m (0.34%) of sulfur, while in the NS100
−
2
sample the concentration reached 5.4 ± 0.4 ␮mol m
(2.6%). As
observed in the clay, no changes were observed in the morphology
of the solid after treatment.
The concentration of sulfur found in the sulfonated niobia with
fuming sulfuric acid is higher than the values found in the literature
for other inorganic materials. While the sulfur surface density in
−
2
sulfonated carbon reaches values as high as 1300 ␮mol m [28,29],
inorganic solids, among which are composites containing silica
3.2. Glycerol etherification
−
2
and alumina, have sulfur content typically less than 2.2 ␮mol m
9,30,31].
Clays have been subjected to sulfonation processes. However,
these procedures typically make use of silylation agents such as
3-mercaptopropyl)trimethoxysilane, which are then oxidized and
acidified. However, the sulfonated clay presented only 0.7% sulfur
32]. In the present study, sulfur adsorption onto the clay surface
[
A blank reaction carried out without addition of solid catalyst
indicated that the glycerol conversion was negligible. In order to
test the activity and selectivity of niobia and pillared clay-based
catalysts, reaction time and amount of catalyst were evaluated
maintaining the TBA:glycerol ratio of 4:1 and the temperature
at 393 K. The major product formed was MTBG, and the equilib-
rium was not achieved throughout the 10 h reaction time. After
5 h of reaction there was a slight decrease in MTBG yield, while
the amounts of DTBG and TTBG increased in the reaction, suggest-
ing the occurrence of consecutive tert-butylation reactions. Similar
results were obtained in reactions catalyzed by sulfonated carbons
[39]. Yields of major products, MTBG, DTBG and TTBG after 5 h of
reaction using different catalysts are presented in Fig. 3. Isobuty-
lene derived products (IB) from TBA protonation and subsequent
dehydration were also identified in the reaction carried out in this
work (concentrations less than 5%). MTBG was the main product for
all catalysts evaluated, ranging from 60 to 70% for clays, and 45%
for niobia. The only active niobia-based catalyst was NS100, with
yield of 45%, 3% and 4% for MTBG, DTBG and TTBG, respectively. For
both systems, the catalysts subjected to acid treatment gave higher
initial rates of product formation, especially those treated with
fuming sulfuric acid. The catalyst AS100, in which the concentra-
tion of strong acid sites (sulfonic groups) is increased, also presents
the highest yield for the more substituted products DTBG + TTBG
(approximately 39%).
(
[
must be regulated by the electrostatic attractions of the sulfate ions
and solid surface, similarly to that observed by Grubb et al. [33] for
the phosphate immobilization on fly ash. At acidic pH, the con-
dition is more favorable for sulfate adsorption onto the clay and
niobia surfaces, since surface structural hydroxyl groups became
positively charged and their adsorption capacity with sulfate would
increase. The point of zero charge, PZC, values of pillared clay (5.6)
and niobia (2.0) confirm this assumption [34,35]. The proposed sites
of anion adsorption on silicate layers are positively charged alu-
minol (AlOH2+) and silanol (SiOH2+) functional groups at exposed
crystal edges [36]. Since the aluminol group, AlOH, is considered
the main active site for anions interaction [37,38], sulfate adsorp-
tion mechanisms on aluminum (hydr)oxides present on the pillars
can be inferred on pillared clays. In a similar way, (NbOH2+) sites
widely present on the surface of niobia are responsible for sulfate
adsorption [35].
The acidity of the solid surface was evaluated by the measure-
ment of ammonia desorbed and the results are presented in Fig. 2.
Niobia treatment increases the amount of acidic groups on the solid
−
1
−1
surface, from 0.5 mmol g in the parent material to 1.5 mmol g
The concentration and strength of acid sites is essential for the
−
1
®
in the NS30 and 4.3 mmol g in the NS100. This value is 10 times
higher than that obtained in untreated niobia, and it is close to
the value found in the sulfonic resin Amberlyst® 15, commonly
used as a catalyst in etherification of glycerol. This resin has a
etherification reaction. Different types of Amberlyst -type resins
®
®
®
(Amberlyst -15, Amberlyst -35 and Amberlyst -36, with acidity
−1
®
in the range from 4.2 to 5.6 mmol g ) and Lewatit -K262 (Bayer)
showed higher activity than other acidic catalysts, which is related
to the nature of their acid sites. These resins were also tested
under the same reaction conditions, providing values which are in
−1
−1
referring to sul-
total acidity of 5.1 mmol g
and 4.7 mmol g
fonic groups (data provided by the manufacturer). In the clay, the

1
02
P.A. Celdeira et al. / Applied Catalysis A: General 478 (2014) 98–106
Fig. 3. Glycerol etherification reaction with TBA using clays and niobia-based solids, and commercial resins as catalyst (TBA:glycerol molar ratio of 4:1, 5 h, 393 K, 5 wt%
catalyst).
agreement with the literature [40,41]. According to Fig. 3, treat-
ment with fuming sulfuric acid was efficient, allowing to obtain
a pillared clay based catalyst with higher activity and selectivity
compared to the commercial resins.
of these two products was identified in the reaction medium. Fur-
thermore, Li et al. [44] suggest that the sulfonic acid groups (and
not the solid matrix) are responsible for the adsorption of glyc-
erol. According to the authors, the strong interaction (hydrogen
like bonding) between the sulfonic acid groups and glycerol (Fig. 4)
favors the adsorption and conversion of glycerol to its ethers. MTBG
can also be adsorbed on the catalyst surface and undergoes further
etherification processes producing DTBG and TTBG.
As glycerol is a bulky and highly viscous compound, a high pore
volume, especially of mesopores, is desirable in the catalyst. Thus
diffusional resistance is minimized, facilitating the adsorption pro-
cess. Similarly, the adsorption of bulky ethers such as MTBG and
DTBG and their subsequent conversion to more substituted ethers
depends on the porosity of the catalyst. Therefore, among the tested
catalysts, the solid which combined the presence of a higher num-
ber of sulfonated sites in a mesoporous structure (AS100) was the
most active and selective in the etherification of glycerol.
Etherification reactions probably occur via nucleophilic attack
of glycerol molecules on tert-butyl cations. Some authors have
suggested that electrostatic effects promoted by the glycerol hydro-
xyls, the solvation and steric hindrance caused by the hydroxyls
or even the interaction with the catalyst surface could target
the attack on the primary hydroxyls [11,42]. However, evidence
to support these effects was not obtained in this work, since
the quantification method used does not allow to distinguish
between the 3-tert-butoxy-1,2-propandiol and 2-tert-butoxy-1,3-
propandiol products, resulting from the attack on the terminal and
central hydroxyl groups, respectively.
The first step of the reaction of the alcohols in the presence of an
acid catalyst should be the protonation of a hydroxyl group. Several
authors have shown that the preferential attack occurs in the TBA,
with the subsequent formation of a tertiary carbocation [11,43].
This carbocation would be responsible for the attack on the glyc-
erol molecule, resulting in formation of respective ethers (Fig. 4).
The results obtained in this work support this hypothesis. If the pro-
tonation occurred in a terminal hydroxyl group of glycerol, acetol
would be formed as a dehydration product [42], while the central
hydroxyl attack would lead to the production of acrolein. Neither
3.2.1. Effect of reaction temperature
The performance of AS100 was studied at various reaction
temperatures. At 363 K the conversion reached only 30%. The
etherification of glycerol is a reversible reaction, whose rate is
greatly reduced at low temperatures, as observed by Klepacova
et al. [15,45]. At 363 K the greater selectivity for DTBG (64%) was
obtained, which is consistent with a decrease in the reversibility of
the reaction. At 393 K yields are considerably higher, up to 95% con-
version of glycerol. At 413 K yields decrease, limiting the conversion
of glycerol to 66%. The reversibility of the reactions as well as the
formation of isobutylene and diisobutylene from TBA contribute
for this low conversion. Furthermore, Lee et al. [46] showed that,
although the adsorption of glycerol occurs spontaneously (ꢁG < 0),
the process decreases at high temperatures, thus reducing its avail-
ability to the etherification. At 393 K, it is clearly observable that
the yields of DTBG and TTBG increase at the expense of MTBG,
stating that MTBG is the primary product and that, at this temper-
ature, its reaction with TBA is favored (Fig. 5). A similar behavior
was observed with NS100 as a catalyst (Fig. 5b). However, the sys-
tem reaches equilibrium only after 9 h at 393 K or 3 h at 413 K.
The slower production of MTBG also inhibits further substituted
products to be obtained. At higher temperature it is also possible
to observe a reduction in MTBG yield after 5 h reaction without
proportional increase in other products. As product concentration
Fig. 4. Illustrative mechanism of MTBG production catalyzed by acidic sites on the
pillars of sulfonated pillared clay.

P.A. Celdeira et al. / Applied Catalysis A: General 478 (2014) 98–106
103
Fig. 5. Kinetics of glycerol etherification (MTBG = square, DTBG = triangle, TTBG = circle) catalyzed by AS100 (left) and NS100 (right). TBA:glycerol molar ratio of 4:1, 393 K,
wt% catalyst.
5
decreases, more glycerol is identified in the reaction medium, indi-
cating that the backward reaction is being favored.
temperature, there is a pronounced consumption of all main
®
products, especially at 423 K. At this temperature, Amberlyst -70
®
Concerning the reaction temperature, some different results are
presented in the literature. Viswanadham and Saxena [13] observed
presented the same behavior. However, at 393 K Amberlyst -70 is
still stable, with continuous production throughout the reaction,
providing yields of 52.5% of MTBG and 29.3% of DTBG + TTBG after
10 h reaction time.
®
®
that Amberlyst -15 or Amberlyst -35 provided the highest yield at
48 K, while a higher reaction temperature caused inferior product
3
®
quality with lower DTBG + TTBG formation. The authors also indi-
cated that the best results with a zeolite based catalysts (BEA) can
be obtained at 363–383 K. On the other hand, Kiatkittipong et al.
Amberlyst -15 is a styrene-divinylbenzene copolymer, and its
acidity is associated to the presence of sulfonic groups ( SO H).
3
−
1
The resin exhibits Brønsted acid sites concentration of 4.7 mmol g
[
47] demonstrated that DTBG + TTBG yield, in a reaction catalyzed
and a maximum operating temperature of 393 K (in non-aqueous
media), specified by the manufacturer. This limit is higher for
Amberlyst -70, which is stable up to 463 K, but exhibiting lower
®
by Amberlyst -15, is enhanced by increasing the temperature from
3
®
63 K to 393 K, which is in agreement with the results obtained
−
1
®
in this study. In this case, the authors promoted the continuous
water removal from the reaction medium, which indicates that the
influence of water is higher with increasing reaction temperature.
Indeed, glycerol etherification has been proposed as an equilibrium
limited reaction and the presence of water molecules might favor
the backward reaction [48].
acidity (2.6 mmol g ). According to Fig. 5, Amberlyst -15 provided
lower yields than AS100 and NS100, even being more acidic. In
addition, it should be considered that at 393 K an abrupt reduction
in the concentration of products was observed after 5 h reac-
tion time. Besides its low thermal stability, water shows a much
®
stronger adsorption on Amberlyst -15 catalyst than ethers [49],
The influence of temperature can also be evaluated regarding
TBA [50], and glycerol [51]. Thus, it is possible that both cat-
alytic activity and selectivity are being affected by the presence
of water in the reaction medium. Water is considered the most
polar component in these systems, competing with TBA and glyc-
erol for the active sites, which affects the reaction equilibrium and
®
®
the results obtained with Amberlyst -15 and Amberlyst -70 ion-
exchange resins as catalysts (Fig. 6) [4].
®
Amberlyst -15 showed a maximum yield of MTBG (50.9%) and
DTBG + TTBG (25.9%) after 7 h of reaction at 363 K. Above this
Fig. 6. Kinetics of glycerol etherification catalyzed by Amberlyst®-15 (left) and Amberlyst®-70 (right) at different temperatures. TBA:glycerol molar ratio of 4:1, 393 K, 5 wt%
catalyst.

1
04
P.A. Celdeira et al. / Applied Catalysis A: General 478 (2014) 98–106
Fig. 7. Selectivity of the etherification reaction catalyzed by AS100 as a function of TBA:glycerol molar ratio (393 K, 5 h, 5 wt% catalyst).
prevents the formation of DTBG and TTBG. Indeed, Kiatkittipong
et al. [47] demonstrated that the Langmuir–Hinshelwood activ-
ity based model, which includes the effect of the strongest water
adsorption, has been found to be the best kinetic model to fit
the experimental results of glycerol etherification catalyzed by
3.2.3. Catalyst stability and recycling
The stability of such system was evaluated by reuse of AS100
and NS100 catalysts after separation from the reaction medium.
After each reaction, the catalyst was separated by filtration, washed
with distilled water and dried in an oven for 12 h at 373 K. The
results are shown in Fig. 8, where solid bars correspond to cata-
lysts washed with water between cycles, and crosshatched bars
represent catalysts subjected to acid washing between cycles.
The conversion of glycerol catalyzed by AS100 was maintained
until the third cycle at around 85%, with a further reduction to 80%.
The selectivity undergoes the greatest change following the cycles.
In the first cycle, the selectivity for DTBG + TTBG reaches 33.7%,
decreasing to 22.4% in the fourth cycle. Niobia is even more sen-
sitive to consecutive cycles. The conversion decreased from 97.5%
in the first cycle to 55.7% in the fourth cycle, while the selectivity
to DTBG + TTBG is reduced from 23.6% to 17.4% in the same range.
Several authors have observed loss of activity of the catalysts in
reaction cycles [54–56], attributing this factor to mass transfer lim-
itations in solids or deactivations in the active sites. In the systems
studied in this work, the site deactivation supposition is more feasi-
ble since, when the catalyst was subjected to an acid wash between
®
Amberlyst -15. This behavior was not observed in a similar way
for the catalysts prepared in this work in reactions conducted up to
3
93 K. This fact is possibly related to a greater stability of the cata-
lysts facing adsorption of water generated in the reaction medium.
3.2.2. Effect of tert-butyl alcohol:glycerol molar ratio
The influence of the TBA:glycerol molar ratio was evaluated in
the range of 4:1–8:1 using the catalyst AS100. Glycerol conversion
increases with the TBA:glycerol molar ratio, ranging from 84.7%
(
4:1) to 97.7% (8:1). Furthermore, the system reaches equilibrium
more quickly (about 2 h) using the highest TBA:glycerol molar ratio.
On the other hand, the product selectivity is not significantly altered
(Fig. 7). Similar results were obtained by Yuan et al. [52] and Pico
et al. [53], and indicate that the molar ratio of 4:1 provides the best
conditions for obtaining products without using a large excess of
TBA.
Fig. 8. Yield of the etherification reaction catalyzed by AS100 (left) and NS100 (right) during four reuses (TBA:glycerol molar ratio 6:1, 393 K, 5 h, 5 wt% catalyst).

P.A. Celdeira et al. / Applied Catalysis A: General 478 (2014) 98–106
105
Fig. 9. Yield of the etherification reaction catalyzed by AS100 (left) and NS100 (right) with removal of the catalyst after 1 h reaction time (TBA:glycerol molar ratio 6:1, 393 K,
h, 5 wt% catalyst).
5
cycles, both the activity and selectivity were maintained (a slight
decrease is observed with NS100).
products significantly reduced. The recycling of AS100 is effective
through an acid wash between cycles, indicating the occurrence of
deactivation of the active sites during the reaction. On the other
hand, contribution of homogeneous catalysis in these systems was
not identified, indicating that the changes observed in the stabil-
ity test are related to deactivation of acid sites of the catalysts. The
catalyst activity can be easily restored by an acid washing of the
solid.
The possibility of homogeneous catalysis in systems containing
the most active catalysts in the etherification of glycerol, AS100
and NS100, was evaluated according to the procedure proposed
by Sheldon et al. [57]. After 2 h reaction time, the catalyst was
removed from the reaction medium by filtration, and the reaction
was continued. The results are shown in Fig. 9. In these figures, filled
markers correspond to the yields of the reaction without removal
of the catalyst.
Acknowledgments
There was a stabilization in product formation after removal of
the catalyst. A slight reduction in MTBG yield can be attributed to
its conversion to more substituted products, promoted by temper-
ature. The end of catalytic activity after the solid removal is clearly
seen in the case of NS100, where the MTBG yield is stopped while
in the system containing the catalyst this product is continuously
formed. Thus, there are indications that the reaction in homoge-
neous phase does not occur, except the contribution of temperature
to the equilibrium of the reaction medium.
This work was supported by Funda c¸ a˜ o de Amparo a` Pesquisa
do Estado de S a˜ o Paulo, FAPESP (Project No. 2011/14626-3,
2
2
0011/22264-4 and 2009/07125-8), CAPES-PRODOC (Project No.
647/2010) and CNPq (Project No. 551070/2010-2).
Appendix A. Supplementary data
Fig. S1 X-ray diffractograms of the samples A, AS30 and AS100.
Fig. S2 X-ray diffractograms of the samples N, NS30 and NS100.
Fig. S3 FTIR spectra of the samples A, AS30 and AS100.
Fig. S4 FTIR spectra of the samples N, NS30 and NS100.
Fig. S5 Typical SEM/EDX micrographs of A (10 kV, mag. 10,000×).
Fig. S6 Typical SEM/EDX micrographs of AS30 and AS100 (10 kV,
mag. 250×).
4
. Conclusions
Sulfonic acid groups were efficiently inserted into pillared clay
and niobia by treatment with sulfuric acid (concentrated or 30%
aqueous solution). The acid treatment caused a decrease in the spe-
cific surface area of the clay, especially for the sample treated with
fuming sulfuric acid, which is related to the partial destruction of
the pillars and/or lamellae of the material. The sulfur content of the
solids increases in the samples treated with fuming sulfuric acid,
reaching 0.46% and 2.6% (respectively for pillared clay and niobia).
Fig. S7 Typical SEM/EDX micrographs of N (10 kV, mag. 10,000×).
Fig. S8 Typical SEM/EDX micrographs of NS30 and NS100 (10 kV,
mag. 1000×).
References
−
1
The total acidity of AS100 (2.3 mmol g ) was maintained after the
treatment due to the reduced access to the porous structure of the
solid.
[
[
1] R.A. Sheldon, Green Chem. 16 (2014) 950–963.
2] P. Guerrero-Urbaneja, C. García-Sancho, R. Moreno-Tost, J. Mérida-Robles, J.
Santamaría-González, A. Jiménez-López, P. Maireles-Torres, Appl. Catal., A:
Gen. 470 (2014) 199–207.
In the glycerol etherification with TBA the main product formed
was MTBG (60–70% for pillared clay and 45% for niobia), and both
DTBG and TTBG were formed in consecutive tert-butylation reac-
tions. The temperature significantly influences the etherification
reaction and at 393 K the glycerol conversion reached 95%. By
increasing the TBA:glycerol molar ratio the conversion of glycerol
increased, but the product selectivity is not significantly altered,
indicating that the 4:1 molar ratio provides the most suitable con-
ditions for obtaining products without using a large excess of TBA.
The stability test of the catalyst indicated maintenance of AS100
activity until the third cycle, but the selectivity to more substituted
[
[
3] M.D. González, P. Salagre, R. Mokaya, Y. Cesteros, Catal. Today 227 (2014)
171–178.
4] M. Gon c¸ alves, V.C. Souza, T.S. Galhardo, M. Mantovani, F.C.A. Figueiredo, D.
Mandelli, W.A. Carvalho, Ind. Eng. Chem. Res. 52 (2013) 2832–2839.
[5] M. Ayoub, A.Z. Abdullah, Catal. Commun. 34 (2013) 22–25.
[
6] A. Behr, H. Schmidke, C. Lohr, M. Schneider, US Patent 4,222,183 to Henkel
KgaA, 1994.
[
7] D.S. Bradin, US Patent 5,578,090 to BRI, Macon, 1996.
[8] J.A. Melero, G. Vicente, G. Morales, M. Paniagua, J.M. Moreno, R. Roldan, A.
Ezquerro, C. Perez, Appl. Catal., A: Gen. 346 (2008) 44–51.
[
9] M.D. González, Y. Cesteros, J. Llorca, P. Salagre, J. Catal. 290 (2012) 202–209.
[
10] F. Frusteri, L. Frusteri, C. Cannilla, G. Bonura, Bioresour. Technol. 118 (2012)
350–358.

1
06
P.A. Celdeira et al. / Applied Catalysis A: General 478 (2014) 98–106
[
11] F. Frusteri, F. Arena, G. Bonura, C. Cannilla, L. Spadaro, O. Di Blasi, Appl. Catal.,
[35] Y. Zha, The Rational Preparation of Niobia Promoted and Supported Platinum
Catalysts, ProQuest, Chicago, IL, 2007.
A: Gen. 367 (2009) 77–83.
[
[
[
12] A. Quitain, H. Itoh, S. Goto, J. Chem. Eng. Jpn. 32 (1999) 280–287.
13] N. Viswanadham, S.K. Saxena, Fuel 103 (2013) 980–986.
14] Y. Gu, A. Azzouzi, Y. Pouilloux, F. Jerome, J. Barrault, Green Chem. 10 (2008)
[36] B.A. Manning, S. Goldberg, Clays Clay Miner. 44 (1996) 609–623.
[37] M.W. Kamiyango, W.R.L. Masamba, S.M.I. Sajidu, E. Fabiano, Phys. Chem. Earth
34 (2009) 850–856.
1
64–167.
[38] P. Zarzycki, F. Thomas, J. Colloid Interface Sci. 302 (2006) 547–559.
[39] T.S. Galhardo, N. Simone, M. Gon c¸ alves, F.C.A. Figueiredo, D. Mandelli, W.A.
Carvalho, ACS Sustainable Chem. Eng. 1 (2013) 1381–1389.
[40] F. Ancillotti, V. Fattore, Fuel Process. Technol. 57 (1998) 163–194.
[41] A. Gicquel, B. Torck, J. Catal. 83 (1983) 9–18.
[
[
[
15] K. Klepacova, D. Mravec, M. Bajus, Chem. Pap. 60 (2006) 224–230.
16] R.S. Karinen, A.O.I. Krause, Appl. Catal., A: Gen. 306 (2006) 128–133.
17] M. Ziolek, P. Decyk, I. Sobczak, M. Trejda, J. Florek, H. Golinska, W. Klimas, A.
Wojtaszek, Appl. Catal., A: Gen. 391 (2011) 194–204.
[
18] N.H. Pham, J.P.T. Yomair, J.C.S. Ruiz, W. Dong, J.A. Dumesic, A.K. Datye, Appl.
[42] V.L.C. Gon c¸ alves, B.P. Pinto, J.C. Silva, C.J.A. Mota, Catal. Today 133-135 (2008)
673–677.
Catal., A: Gen. 397 (2011) 153–162.
[
[
19] I. Sobczak, K. Jagodzinska, M. Ziolek, Catal. Today 158 (2010) 121–129.
20] R. Rodrigues, N. Isoda, M. Gon c¸ alves, F.C.A. Figueiredo, D. Mandelli, W.A. Car-
valho, Chem. Eng. J. 198–199 (2012) 457–467.
[43] K. Klepacova, D. Mravec, A. Kaszonyi, M. Bajus, Appl. Catal., A: Gen. 328 (2007)
1–13.
[44] B.W. Chen, X. Wang, C. Ma, L.R. Wang, Energy Fuels 26 (2012) 7060–7067.
[45] K.M. Klepácová, D.E. Hájeková, M. Bajus, Pet. Coal 45 (2003) 54–57.
[46] H.J. Lee, D. Seung, K.S. Jung, H. Kim, I.N. Filimonov, Appl. Catal., A: Gen. 390
(2010) 235–244.
[47] W. Kiatkittipong, P. Intaracharoen, N. Laosiripojana, C. Chaisuk, P. Praserthdam,
S. Assabumrungrat, Comput. Chem. Eng. 35 (2011) 2034–2043.
[48] M. Ayoub, M.S. Khayoon, A.Z. Abdullah, Bioresour. Technol. 112 (2012)
308–312.
[
[
21] T.J. Pinnavaia, Science 220 (1983) 365–371.
22] F.C.A. Figueiredo, E. Jordão, W.A. Carvalho, Catal. Today 107/108 (2005)
2
23–229.
23] F.C.A. Figueiredo, E. Jordão, W.A. Carvalho, Appl. Catal., A: Gen. 351 (2008)
59–266.
24] M.E. Jamróz, M. Jarosz, J. Witowska-Jarosz, E. Bednarek, W. Tecza, M.H. Jamróz,
J.C. Dobrowolski, J. Kijenski, Spectrochim. Acta, A: Mol. Biomol. Spectrosc. 67
[
[
2
(
2007) 980–988.
[49] F. Aiouache, S. Goto, Chem. Eng. Sci. 58 (2003) 2065–2077.
[50] M.L. Honkela, T. Ouni, A.O.I. Krause, Ind. Eng. Chem. Res. 43 (2004)
4060–4065.
[51] D. Gelosa, M. Ramaioli, G. Valente, M. Morbidelli, Ind. Eng. Chem. Res. 42 (2003)
6536–6544.
[52] Z. Yuan, S. Xia, P. Chen, Z. Hou, X. Zheng, Energy Fuels 25 (2011) 3186–3191.
[53] M.P. Pico, A. Romero, S. Rodríguez, A. Santos, Ind. Eng. Chem. Res. 51 (2012)
9500–9509.
[54] M.S. Khayoon, B.H. Hameed, Bioresour. Technol. 102 (2011) 9229–9235.
[55] I. Dosuna-Rodriguez, C. Adriany, E.M. Gaigneaux, Catal. Today 167 (2011)
56–63.
[
[
25] M.K. Bukka, J.D. Shabta, Clays Clay Miner. 40 (1992) 92–102.
26] J.A. Sánchez, D.L. Hernández, J.A. Moreno, F. Mondragón, J.J. Fernández, Appl.
Catal., A: Gen. 405 (2011) 55–60.
27] M. Risti, S. Popovi, S. Musi, Mater. Lett. 58 (2004) 2658–2668.
28] M.Y.T. Hara, A. Takagaki, T. Takata, J.N. Kondo, K. Domen, S. Hayashi, Angew.
Chem. Int. Ed. 43 (2004) 2955–2958.
29] J. Janaun, N. Ellis, Appl. Catal., A: Gen. 394 (2011) 25–31.
30] P. Valle-Vigón, M. Sevilla, A.B. Fuertes, Appl. Surf. Sci. 261 (2012) 574–583.
31] L. Peng, A. Philippaerts, X. Ke, J. Van Noyen, F. De Clippel, G. Van Tendeloo, P.A.
Jacobs, B.F. Sels, Catal. Today 150 (2010) 140–146.
[
[
[
[
[
[
32] D.S. Moraes, R.S. Angélica, C.E.F. Costa, G.N. Rocha Filho, J.R. Zamian, Appl. Clay
[56] K. Kumbilieva, L. Petrov, Chin. J. Catal. 32 (2011) 51–59.
[57] R.A. Sheldon, M. Wallau, I.W.C.E. Arends, U. Schuchardt, Acc. Chem. Res. 31
(1998) 485–493.
Sci. 51 (2011) 209–213.
[
[
33] D.G. Grubb, M.S. Guimarães, R. Valencia, J. Hazard. Mater. 76 (2000) 217–236.
34] M. Kosmulski, Surface Charging and Points of Zero Charge, first ed., CRC Press,
Boca Raton, FL, 2009.