Etherification of glycerol with tert-butyl alcohol over sulfonated hybrid silicas

Article abstract of DOI:10.1016/j.apcata.2016.08.019

The etherification of glycerol with tert-butyl alcohol has been carried out in the liquid-phase on different sulfonic acid functionalized hybrid silicas obtained by oxidative cleavage of tetrasulfide bridges. These materials exhibited different textural properties and acidities ranging from 0.5 to 2.3?mmol?g^?1. First of all, different experimental variables, including the reactant ratio and temperature have been tested and all catalysts were essayed under optimal conditions. The best catalytic behavior of the hybrid silicas was obtained on S₅₀TS₅₀O (Y_h-GTBE?=?28%) at 75?°C and autogenous pressure. In fact, the glycerol conversion obtained on S₅₀TS₅₀O, clearly surpassed that obtained on A-15 (74 and 51%, respectively). The acidity and the hydrophilic character of the sulfonated hybrid silicas solids have been found to be key parameters for their catalytic activity whereas the surface area and porosity do not seem to be advantageous for this reaction. All catalysts were reused and subjected to an extensive characterization.

Full text of DOI:10.1016/j.apcata.2016.08.019

Accepted Manuscript
Title: Etherification of glycerol with tert-butyl alcohol over
sulfonated hybrid silicas
´
´
´
Author: R. Estevez M.I. Lopez C. Jimenez-Sanchidrian D.
Luna F.J. Romero-Salguero F.M. Bautista
PII:
S0926-860X(16)30426-4
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.apcata.2016.08.019
APCATA 15982
To appear in:
Applied Catalysis A: General
Received date:
Revised date:
Accepted date:
16-6-2016
29-7-2016
23-8-2016
´
´
´
Please cite this article as: R.Estevez, M.I.Lopez, C.Jimenez-Sanchidrian,
D.Luna, F.J.Romero-Salguero, F.M.Bautista, Etherification of glycerol with
tert-butyl alcohol over sulfonated hybrid silicas, Applied Catalysis A, General
http://dx.doi.org/10.1016/j.apcata.2016.08.019
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Etherification of glycerol with tert-butyl alcohol over
sulfonated hybrid silicas
R. Estevez, M. I. López, C. Jiménez-Sanchidrián, D. Luna, F. J. Romero-Salguero*,
F. M. Bautista*
Department of Organic Chemistry, Nanochemistry and Fine Chemistry Research
Institute (IUIQFN), Faculty of Sciences, University of Córdoba, Campus of Rabanales,
Marie Curie Building, Ctra. Nal. IV, km 396, 14014 Córdoba, Spain
* Corresponding authors. Tel.: +34 957218638; fax: +34 957212066. E-mail address:
qo2rosaf@uco.es (F. J. Romero-Salguero). Tel.: +34 957212065; fax: +34 957212066.
E-mail address: qo1baruf@uco.es (F. M. Bautista)

Graphical abstract

Highlights
 Sulfonated organosilicas were synthesized by co-condensation and oxidation.
 A higher acidity and non-porous structure would favour the catalytic activity.
 One of the acidic sulfonated organosilica showed a higher yield to h-GTBE than A-15.
 The higher the hydrophobicity, the higher the deactivation by pore blockage.
 These solids are promising catalysts for reactions that need higher temperatures.

Abstract
The etherification of glycerol with tert-butyl alcohol has been carried out in the liquid-
phase on different sulfonic acid functionalized hybrid silicas obtained by oxidative
cleavage of tetrasulfide bridges. These materials exhibited different textural properties
and acidities ranging from 0.5 to 2.3 mmol g-1. First of all, different experimental
variables, including the reactant ratio and temperature have been tested and all catalysts
were essayed under optimal conditions. The best catalytic behavior of the hybrid silicas
was obtained on S50TS50O (Yh-GTBE = 28%) at 75 ºC and autogenous pressure. In fact,
the glycerol conversion obtained on S50TS50O, clearly surpassed that obtained on A-15
(74 and 51%, respectively). The acidity and the hydrophilic character of the sulfonated
hybrid silicas solids have been found to be key parameters for their catalytic activity
whereas the surface area and porosity do not seem to be advantageous for this reaction.
All catalysts were reused and subjected to an extensive characterization.
Keywords: Sulfonated organosilicas; Amberlyst-15; etherification; glycerol; tert-butyl
alcohol

1. Introduction
In recent years, biodiesel production has sharply increased to compensate the
depletion of fossil resources. Biodiesel is renewable, biodegradable and non-toxic so it
has been assumed as a good alternative to diesel fuel. The formation of biodiesel by the
transesterification reaction of triglycerides with methanol, produces glycerol as a by-
product (10 wt.% of the total biodiesel product). Due to the large surplus of glycerol
which has been generated, the price of crude glycerol decreased from 500 US$/ton in
2001, to 100 US$/ton in 2011 [1]. This drop in the price of glycerol has resulted in an
increase in the production cost of biodiesel. In this sense, several processes which
employ glycerol as a raw material are being investigated in order to obtain added-value
products [2-7]. One promising option is the catalytic etherification of glycerol to obtain
tert-butyl glycerol ethers (TBGs), specially di-tert-butyl glycerol ethers (DTBGs) and
tri-tert-butyl glycerol ether (TTBG), the so-called high ethers (h-GTBE). These ethers
can be used as excellent diesel and biodiesel additives due to their compatible
physicochemical properties such as flash point, viscosity, cetane number, etc. [8]. These
additives can diminish the emission of particulate matter, hydrocarbons, carbon
monoxide and unregulated aldehydes in diesel vehicles [9], and act as cold flow
improvers in biodiesel, thus reducing its viscosity [10]. However, the mono-tert-butyl
glycerol ethers (MTBGs) cannot be blended with diesel due to their low solubility.
Therefore, it would be highly desirable to shift the reaction toward the formation of high
ethers.
Typically, the etherification of glycerol (G) is carried out over acid catalysts
employing isobutylene (IB) or tert-butyl alcohol (TBA) as reactants. The use of IB has
been accomplished by homogeneous catalysts (mainly p-toluenesulfonic acid) [11] and

heterogeneous catalysts such as strong acid ion-exchange resins, modified
mesostructured silicas and zeolites [12-17]. In general, etherification with IB led to
higher conversion and selectivity values to h-GTBE than etherification with TBA.
However, the use of IB is conditioned by its high cost and non-renewable nature, as
well as the need to operate under pressure to keep the IB in the liquid-phase [13]. In
fact, mass transfer effects are considered essential in the etherification with IB and
consequently the use of a solvent is required [18]. Furthermore, oligomerization of IB
has been reported to cause some decrease in ether selectivity, especially when catalysts
with a high population of strong acid sites are employed [13, 15]. The use of TBA,
which is a byproduct of polypropylene production, benefits mass transfer, suppresses IB
oligomerization and avoids using organic solvents in reaction mixtures.
By using TBA as a reactant, the etherification takes place according to the
following reactions:
(Scheme 1 near here)
As aforementioned, etherification of glycerol with TBA is catalyzed by acid
sites. Different studies have been carried out in order to improve the yield to h-GTBE.
Usually, the best results have been obtained over acid resin catalysts, mainly
Amberlyst-15 (A-15) [12, 19-21]. However, the sulfonic groups presented in the A-15
are the only part of its structure with hydrophilic character. For this reason, the water
generated in each step during the reaction would be adsorbed on these sulfonic groups
leading to a fast deactivation. In fact, Cannilla et al. reported the use of a water
permselective membrane which allowed water to be continuously removed from the
reaction mixture, improving the yield to h-GTBE on A-15 from 22 to 29% [22].
Furthermore, Ozbay et al. [23, 24] studied the activity of different acid catalysts in flow
and batch reactors, as well as the effect of temperature in the selectivity to the different

ethers. To date, apart from acid resins, very few heterogeneous catalysts have been
studied, such as sulfonated black carbons [25-28], Nafion ionomer (N-17),
tungstophosphoric heteropolyacid (HPW-17) supported separately on amorphous silica
[29] and SO3H functionalized ionic liquids [30], etc. Recently, cesium exchanged
tungstophosphoric acid supported on tin oxide and nanostructured MFI-type zeolites
have been also reported as a good catalyst for this reaction [31, 32]. González et al. [33]
studied the role of Bronsted acidity and porosity in the etherification reaction by
modifying the surface and acid characteristics of three zeolites. Fluorinated beta zeolite
with the highest amount of Bronsted acid sites showed the highest yield to h-GTBE
(26%).
Furthermore, the sulfonic resins with which the best results have been obtained,
usually present low surface areas and a lack of thermal stability and suffer from
swelling and shrinking in organic media [29]. Polysiloxanes incorporating arenesulfonic
or alkylsulfonic groups can overcome all these problems because of their higher thermal
structural stability. There are several strategies to incorporate sulfonic acid groups into
hybrid materials. A common one is the co-condensation of a tetralkoxysilane (tetraethyl
orthosilicate, TEOS, or tetramethyl orthosilicate, TMOS) or a bridged organosilane with
3-mercaptopropyltrimethoxysilane (MPTMS) with subsequent oxidation of the thiol
groups [34-36]. Also, different thioether-bridged mesoporous organosilicas containing
disulfide or tetrasulfide moieties have been synthesized in recent years. The oxidation of
their bridges is expected to yield two sulfonic acid groups in close proximity to each
other, which could lead to an enhanced catalytic activity according to some authors
[37].
On this basis, herein, we report the catalytic performance in the liquid-phase
etherification of glycerol on synthetic silica-based and organosilica-based hybrid

materials that exhibit different hydrophilic/hydrophobic character. These materials were
obtained
by
co-condensation
between
and
the
either
precursors
TEOS
bis[3-
1,4-
(triethoxysilyl)propyl]tetrasulfide
(BTEPTS)
or
bis(triethoxysilyl)benzene (BTEB), in different ratios using Brij-76 as a structure-
directing agent under acid conditions and subsequent oxidation by H2O2. Previously,
some of these materials were tested in esterification reactions, where water is also
formed, with excellent results [38]. A commercial Amberlyst-15 has also been studied
as a reference catalyst. Furthermore, the characterization of the reused materials was
subsequently studied.
2. Experimental
2.1. Preparation of catalysts
The hybrid materials were synthesized according to previously reported
procedures [38]. Briefly, the syntheses were carried out by co-condensation of BTEPTS
and either TEOS or BTEB as the silica and organosilica precursors, respectively. The
molar ratios of the precursors in the synthesis mixtures (based on Si) were 50:50 and
10:90. Briefly, Brij-76 (6 g) was added to a solution of 37% HCl (19.6 mL) in H2O (279
mL). The resulting mixture was stirred at 50 ºC for 24 h. Then, BTEPTS and either
TEOS or BTEB were added. The Si/Brij-76/HCl/H2O molar ratio in the mixture was
1:0.10:2.85:186.1. After stirring for 24 h at the same temperature, the suspension was
aged at 90 ºC for 24 h under static conditions. Later, the solid was filtered and dried in
air. The surfactant was removed by heating the as-synthesized material at reflux in an
HCl solution (1 mL of 37% HCl and 50 mL of ethanol, per gram of solid) for 12 h. The
solid was then filtered and washed with ethanol. This process was repeated three times.
The final materials were then dried in an oven at 100 ºC under vacuum.

In order to transform the tetrasulfide bridges into sulfonic acid groups, an
oxidation process with H2O2 was carried out. Thus, the hybrid material (0.3 g) was
stirred with 30% H2O2 (10 mL) at 50 ºC for 3 h. The solid was then washed with 1.2 M
H2SO4 (25 mL). Then, it was thoroughly washed with water until a neutral pH was
reached. Finally, the solid was dried at 100 ºC under vacuum.
The silica-based hybrid materials subjected to oxidation were named SxTS100-x
O
with x being the Si molar ratio of TEOS in the initial mixture. Similarly, the oxidized
organosilica-based materials were named BSxTS100-xO.
A commercial hydrogen dry A-15 was supplied by Aldrich.
2.2. Characterization of the catalysts
Textural properties were determined from the N2 adsorption-desorption
isotherms at -196 ºC, using a Micromeritics ASAP 2010 apparatus. The specific surface
area of each solid was determined by the BET method and the pore size distribution by
the DFT method with a cylindrical pore of silica as a calculation model. XRD powder
diffraction patterns were obtained using Ni-filtered Cu Kα radiation (λ = 1.5418 Å).
Finely ground samples were scanned at a speed of 2 °/min (2θ = 2-80°) using a Siemens
D-5000 diffractometer (40 kV, 30 mA). Elemental compositions were determined on an
Eurovector EA 3000 analyzer. XPS spectra were recorded with a SPECS Phoibos HAS
3500 150 MCD. The residual pressure in the analysis chamber was 5x10-9 Pa. Accurate
binding energies were determined with respect to the position of the Si 2p peak at 103.4
eV. The peaks were decomposed using a least-squares fitting routine (Casa XPS
software) with
a
Gaussian-Lorentzian (70:30) using Shirley baselines.
Thermogravimetric and differential thermal analyses were recorded on a Setaram Setsys
12 thermal analysis station by heating in air from 30-600 ºC at a rate of 10 ºC min-1.
Raman spectra for the solids were acquired with a Renishaw Raman Instrument (InVia

Raman Microscope) equipped with a Leica microscope furnished with various lenses,
monochromators and filters, in addition to a CCD. Spectra were obtained by excitation
using green laser light (532 nm) from 400 to 4000 cm-1. A total of 20 scans per
spectrum were registered in order to improve the signal-to-noise ratio.
Particle size measurements were obtained in a Mastersizer S analyzer with laser
diffraction technology using a small volume dispersion unit. Prior to the measurements,
the samples were dispersed in distilled water and sonicated for 5 min.
The acidity was determined using the following procedure. Aliquots of 50 mg of
the material were stirred with 2 M NaCl (20 g of solution) at room temperature for 24 h.
Then, the solid was filtered off and the resulting solution titrated with a 0.025M NaOH
solution using phenolphthalein as an indicator.
After the reaction, the different materials were analogously characterized.
2.3. Catalytic reaction
The liquid-phase etherification reaction between glycerol and TBA was carried
out in a round-bottom flask (100 mL) with magnetic stirring, using conventional heating
at autogenous pressure. It was found that the influence of external mass transfer was
negligible at stirring rates above 700 min-1 and therefore this rate was subsequently used
in all experiments. The internal diffusion limitations were previously studied in
reactions involving glycerol and it was found that a particle size below 0.149 mm
avoids internal diffusion [7]. In this case, a particle size below 20 µm was obtained for
every solid, making negligible the internal diffusion limitations (Fig. S1).
Firstly, experiments were performed over A-15, under different reaction
conditions, i.e., temperatures ranging from 70 to 85 ºC; at different reaction times from
1 to 17 h; TBA/G and catalyst/G ratios (wt.%) ranging from 4:1 to 8:1 and from 2.5 to
7.5%, respectively. In general, a well-defined amount of glycerol and TBA, as well as a

new charge of catalyst, were employed in each experiment. At the end of the
experiments, the reactor was cooled down by an ice bath before the sample was taken.
The reliability of the procedure was checked by repeating the experiments at least twice.
The identification of the products was carried out by GC-MS (VARIAN CP
3800, QUADRUPOLE MS 1200) equipped with a 100% ethylene glycol (30 m x 0.25
mm x 0.25 µm) capillary column Supelcowax 10, according to the characterization data
reported by Jamróz et al. [39]. The quantitative analysis was carried out by GC in a
Hewlett Packard 5890 series II, equipped with a 100% ethylene glycol (30 m x 0.25 mm
x 0.25 µm) capillary column Supelcowax 10, and a FID detector, using 4-chlorotoluene
as an internal standard [26, 27]. The analysis program was: 60 °C for 6 min heating at
20 °C/min until reaching 240 °C, with an analysis time of 25 min. Turnover frequency
values (TOF) were calculated as mmoles of produced TBGs per mmol of active species
(H+) per hour. The conversion of glycerol (XG) and product selectivity (S) were
determined by the following equation:
It should be noted that diisobutylene (DIB) or secondary reactions coming from
glycerol were never obtained under these conditions.
At the end of each reaction, the catalyst was recovered, washed with ethanol and
dried at 80 ºC. Afterwards, it was used again following the same procedure described
above.

3. Results and discussion
3.1. Characterization of catalysts
Table 1 summarizes the most relevant physicochemical properties for the
sulfonic acid functionalized hybrid silicas.
(Table 1 and Figure 1 near here)
Data from N2 adsorption-desorption isotherms evidenced that S90TS10O and
BS90TS10O showed a superposition of type I and type IV isotherms with a step at a
relative pressure around 0.2-0.3 (Fig. 1). This is typical of pores in the low mesopore
range. Data from XRD patterns showed that before the oxidation treatment, these
materials exhibited a well-defined low-angle peak (ao = 60.0 Å) characteristic of
periodic mesoporous organosilicas (PMOs), whereas after oxidation this peak became
ill-defined, indicating the existence of less ordered structures. Furthermore, S50TS50O
and BS50TS50O materials showed type III isotherms associated to non-porous materials.
In terms of the sulfur measured by elemental analysis, the S50TS50O exhibited
the highest amount of S followed by BS50TS50O. S90TS10O and BS90TS10O showed a
similar S content. Although the percentage of oxidized S decreased with the S content
(Fig. S2), the absolute amount of S in an oxidized state followed the sequence:
S50TS50O > BS50TS50O > S90TS10O > BS90TS10O. Furthermore, the same order could be
observed for the acidity determined by titration. Samples obtained using BTEB as
precursor were less acidic than samples obtained with TEOS (Table 1). As the amount
of BTEPTS increased from 10 to 50, the acidity also increased while the specific
surface area decreased. Accordingly, the surface density of acid sites was much higher
for S50TS50O and BS50TS50O than for S90TS10O and BS90TS10O.
Thermogravimetric analysis (TGA) techniques were performed to ascertain the
thermal stability of the hybrid catalysts. As can be seen in Figure 2, a weight loss below

120 ºC was observed for all materials, which was due to the removal of physisorbed
water. This weight loss did not exceed 16% in any case. The second weight loss was
centered at ca. 250 ºC and would correspond to the decomposition of the remaining
surfactant in the mesopores [40]. In addition, the S50TS50O and BS50TS50O materials
showed a peak centered at 330 ºC, probably due to the decomposition of tetrasulfide
moieties [41]. In fact, these two materials presented the highest BTEPTS to TEOS or
BTEB molar ratios (50:50). This peak was almost negligible in those materials whose
molar composition of BTEPTS was lower (90:10). With regard to the decomposition of
the propylsulfonic acid groups, this took place between 370-520 ºC, as described
previously in the literature [42]. The weight loss associated with this peak, which is
related to the content of sulfonic acid groups, was higher for S50TS50O and BS50TS50O
materials, 15 and 13% respectively, than for BS90TS10O and S90TS10O samples, whose
weight losses were in the range 4.5-6%. These data agreed with the acidities (mmol
H+/g) obtained by titration. Finally, those materials obtained by co-condensation of
BTEPTS and BTEB showed a weight loss above 520 ºC that would correspond to the
decomposition of phenylene groups from BTEB.
(Figure 2 near here)
3.2. Influence of the reaction variables
Table 2 summarizes the influence of the experimental conditions in the reaction
carried out over a commercial A-15.
(Table 2 near here)
As can be seen, an increase in the weight of the catalyst from 2.5 to 5%
produced an increase in the yield to products due to an increase in XG (8.1 to 17.4%,
respectively), while the selectivity to MTBGs and h-GTBE remained constant (83 and
17%, respectively). However, upon increasing the weight of the catalyst to 7.5%, the

YMTBG value increased but the Yh-GTBE value was practically the same, diminishing the
formation of TTBG and increasing the formation of IB. The influence of the
temperature was also important. In fact, the best results were obtained at 75 ºC (YTTBG
=
0.5%). Above 85 ºC, the formation of IB and water increased, preventing the activation
of glycerol on acid sites by competitive adsorption, as previously reported by Cannilla
et al. [22]. Comparing the TBA/G ratio, the highest yield to h-GTBE was obtained with
a 4:1 ratio. Accordingly, the reaction conditions selected were 75 ºC, TBA/G ratio of
4:1 and 5 wt.% in relation to the initial amount of glycerol. Once the experimental
conditions were selected, the reaction was carried out at different reaction times (1, 3, 6
and 17 h) on all the catalysts. It should be noted that blank experiments showed that the
mixture of TBA/G did not react in the absence of the catalyst under the present
experimental conditions.
3.3. Catalytic activity
(Figure 3 near here)
As can be seen in Figure 3, the glycerol conversion increased with time for all
the sulfonic acid functionalized hybrid silicas. At short reaction times, all the synthetic
solids showed low conversion values except for the S50TS50O catalyst (XG = 17 %) at 1
h, which was also the most acidic among the hybrid materials. The highest conversions
obtained on A-15 at 1 and 3 h, especially at 3 h, could be due to its stronger acid sites.
In fact, it is well-known that arenesulfonic groups are more acidic than alkylsulfonic
ones [41]. Furthermore, the highest conversion obtained over the most acid catalyst
agrees with the reaction mechanism described in the literature, which proposes a fast
protonation of TBA on acid sites, giving rise to a tertiary carbocation that reacts with
glycerol. Moreover, IB, coming from the dehydration of the alcohol, and water were

formed. Whereas IB can react with glycerol to form the different ethers, the water
generated can compete with TBA and glycerol for the acid sites on the catalyst surface.
When A-15 is employed, the sulfonic groups are the only part of the structure with a
hydrophilic character. For this reason, the water generated during the reaction would be
adsorbed on these sulfonic groups leading to a faster deactivation [29, 33]. In fact, the
highest conversion on A-15 was reached at 3 h and this value remained constant until 17
h of reaction time (Fig. 3). However, Sh-GTBE increased with time even though no change
in conversion was observed. A possible explanation would be that glycerol would not
be able to shift the water adsorbed in the acid sites and so it would not react further.
Nevertheless, the MTBGs molecules would remain close to the acid sites because of
their hydrophilic and hydrophobic moieties, allowing their subsequent transformation in
h-GTBE. By contrast, the hybrid silicas did not show the same catalytic behavior, even
S50TS50O catalyst that exhibited a similar density of acid sites (0.13 mmol H+/m2) to
that of A-15 (0.12 mmol H+/m2). The conversion smoothly increased with reaction time
and equilibrium was not reached even at 17 hours. The more gradual variation of
conversion with time for the hybrid silicas with respect to A-15 can be related to the
more hydrophilic character of the silicas and organosilicas, which can partially adsorb
the water generated during the reaction on the silanol groups. It is worth noting that
over S50TS50O the conversion increased considerably from 6 to 17 h. This is due to the
production of DTBGs and TTBG and their contribution to the overall conversion, i.e.,
produced TBGs (Fig. S3).
Clearly, the acidity of the catalysts strongly influences their catalytic
performance. Thus, an acceptable correlation was observed when representing
conversion at 3 h against the acidity of the catalysts (Fig. S4). However, similar results
of XG with time were obtained for BS50TS50O and S90TS10O although the acidity of the

BS50TS50O was twice that than that of S90TS10O. This could be explained again by the
different hydrophilic character of both solids.
The selectivity to MTBG over S50TS50O decreased with reaction time, i.e. as the
conversion increased (Fig. 4), whereas the selectivity to h-GTBE increased. Indeed, the
selectivity to TTBG was rather low throughout the reaction. The rest of the solids
behaved in a similar way. Likewise, at a similar conversion (~17%), all catalysts gave
approximately the same selectivity except for A-15, which exhibited higher selectivity
to h-GTBE (Fig. S5). In all the catalytic runs, the 1-MTBG/2-MTBG ratio was always
higher than 9, whereas the 1,3-DTBG/1,2-DTBG ratio was always higher than 4. These
data agreed with those found in literature, not only in the etherification of glycerol with
TBA but also with IB [22, 43].
(Figure 4 near here)
Table 3 gives the catalytic activity for all catalysts after 17 h. The S90TS10O and
BS90TS10O materials, which exhibited similar acidities and also textural properties in
the micro-mesoporous range, gave similar selectivity to h-GTBE (< 7%). The MTBGs
produced inside the pores of these materials would hinder the diffusion of the reactants
through the pores, avoiding the formation of h-GBTE. Furthermore, BS50TS50O gave a
higher selectivity to h-GTBE than S90TS10O (12 and 7%, respectively), despite
producing similar glycerol conversion, as aforementioned. The non-porous structure of
BS50TS50O would favour the production of h-GTBE on the external surface. Thus,
whereas the activity of the solids is influenced mainly by the acidity and textural
properties, its deactivation is strongly influenced by the hydrophobic character.
Accordingly, S50TS50O with the highest acidity and also with a non-porous structure
gave the highest yield to h-GTBE (18%).
(Table 3 near here)

Interestingly, S50TS50O allowed us to detect the presence of TTBG. In addition,
the yield to h-GTBE obtained over S50TS50O catalyst (18%) was the same as that
obtained over A-15. Due to the high catalytic activity obtained on S50TS50O and the
high increase in the glycerol conversion from 10 to 17 h, an experiment was performed
at a longer reaction time (24 h). Glycerol conversion and yield to h-GTBE (ca. 28%)
further increased with reaction time (Table 3). This result is very remarkable because
the Yh-GTBE obtained over A-15 has rarely been matched in this reaction. These results
obtained over A-15 were also in agreement with those found in the literature under
similar experimental conditions [12, 19, 22]. In spite of its higher acidity and its large
pore size, the water generated during the reaction would affect the catalytic performance
of A-15, as aforementioned. It should also be noted that the yield to h-GTBE obtained
on S50TS50O was similar or even higher than that found in literature with other solids
under similar experimental conditions [20, 22, 25-27, 31-33]. Some of these solids
exhibited a low surface areas and high acidity, as was the case for S50TS50O. For a better
comparison, TOF was calculated for every catalyst (Table 4). As can be seen, at similar
reaction times, S90TS10O and S50TS50O exhibited the highest TOF values for the solids
studied (entries 6-10). Furthermore, at a similar reaction time, these two solids (entries
7-8) exhibited higher or similar TOF values to other solids reported in literature with
higher acidity, even at lower reaction temperatures (entries 12 and 13). High values of
TOF (9.9 and 35.3 h-1) were achieved over a highly acid black carbon obtained from
coffee ground wastes and a nanostructured MFI zeolite, respectively. However, the Yh-
was in both cases around 20%. The selectivities to h-GTBE exhibited by these
GTBE
solids were similar but the conversions were lower than that obtained with our best
catalyst. González et al. achieved a TOF value of 14.7 with a fluorinated beta zeolite
(entry 15) [33]. Although this value is much higher than that obtained on S50TS50O at

these experimental conditions, the nature of the active sites in this material was clearly
different to the sulfonated materials. Furthermore, the overall yield to h-GTBE over
fluorinated beta zeolite was slightly inferior to that obtained over S50TS50O. Very
recently, Srinivas et al. reported the highest value of Yh-GTBE (40%) to date (entry 16),
over a cesium exchange tungstophosphoric acid supported on tin oxide. The high value
of yield achieved, could be due to the high amount of catalyst employed, 27 wt.% in
relation to the initial amount of glycerol, which was practically six times higher than the
weight employed in this work.
(Table 4 near here)
Concerning the reaction temperature, different results are presented in literature.
Some authors have previously reported that an increase in the reaction temperature did
not involve an increase in the yield to h-GTBE on A-15 [12]. In addition, it should be
noted that its maximum operating temperature is ca. 120 ºC. Carvalho et al. [28]
reported an increase in the yield to h-GTBE in the range of 60 to 120 ºC. However, at
150 ºC, the glycerol conversion decreased drastically to 4% and the selectivity to
MTBG reached 97.5%. They explained this fact by the reversible reaction that took
place at higher temperatures, promoting the thermal decomposition or recombination of
reaction intermediates. In order to check the influence of the temperature in the catalytic
behavior of the hybrid materials, some reactions were carried out over S50TS50O. At 100
ºC and 10 hours, a higher conversion (68%), a higher Sh-GTBE (19%) as well as a higher
TOF value (6.5 h-1) were achieved compared with those obtained at 75 ºC (XG = 41%;
Sh-GTBE = 16%; TOF = 3.9 h-1). These results are in agreement with different studies
which employ solids with low surface areas and high acidities [25, 27, 28] but also
contrasted with those obtained on A-15 (Table 2). Under these conditions, a sample was
taken at longer reaction time (24 h). The XG obtained on S50TS50O was almost the same

(68%) as that obtained at 10 h (71%), although the selectivity to h-GTBE increased
from 19 to 25%. Furthermore, DIB was also detected, probably due to a higher rate of
dehydration of TBA to IB favored by both, the higher temperature and the higher
reaction time. In fact, Cannilla et al. [44] reported the formation of DIB mainly at
longer reaction times and higher temperatures, from 6 to 27 h and 70 to 80 ºC,
respectively. In addition, Srinivas et al. also observed the formation of polymeric
products at temperatures around 110 ºC [31]. Since the selectivity to h-GTBE continued
increasing with time at 100 ºC, it does not seem that reversible reactions occur. Thus,
the stabilization in the glycerol conversion with time was probably due to the formation
of DIB. In addition, one more reaction was carried out for 24 h, increasing the
temperature to 120 ºC. Under these conditions, the XG attained (20%) was much lower
than at 100 ºC. The high temperature together with the high acidity of the S50TS50O
catalyst, promoted a fast dehydration of TBA at the beginning of the reaction,
increasing the formation of IB and subsequently the formation of DIB, thus making the
formation of reaction products difficult. Furthermore, this temperature could also
promote the thermal decomposition or recombination of reaction intermediates,
changing the selectivity and the glycerol conversion, as it was previously reported [8].
Despite these facts, the good results obtained at 10 h and 100 ºC allow us to
confirm that these materials are very attractive for acid-catalyzed reactions that need
higher temperatures, such as the etherification of glycerol with n-butanol, which yields
oxygenated additives for biodiesel fuels and has recently aroused great interest [45].
3.4. Reusability and catalyst stability
All solids have been reused once. The catalytic behavior of the reused materials
was compared with that of the fresh ones (Figure 5). It should be emphasized that no
activation treatment was carried out after use. Thus, after reaction, the solid was filtered

off and washed with ethanol at room temperature. As can be seen, the highest loss of
activity was observed with S50TS50O catalyst, which was also the most active in this
reaction at 75 ºC and 17 h (Yh-GTBE = 10%). This loss of activity was mainly due to a
decrease in the glycerol conversion (74 and 57% for the fresh and used catalysts,
respectively) whereas the selectivity to h-GTBE remained almost constant (20 and 17%,
respectively). The non-porous structure would avoid the decrease in the selectivity to h-
GTBE. This behavior has been previously reported by other authors who describe that
after washing with water at room temperature, the activity of the solids decreased
mainly between the first and second run [27]. This was explained by the adsorption of
the chemical reactants that takes place on the active sites of the catalysts and which
were not removed by water washing. Celdeira et al. [26] reported a decrease in the
activity of their catalysts after the first cycle, subjecting them to an acid washing in
order to reactivate the acid sites of the catalysts. On the contrary, the activity of the
other solids described in this work remained almost constant (Fig. 5).
(Figure 5 near here)
With the hybrid silicas, a decrease on the surface areas was observed on all used
solids after the reaction (Table 1), being specially pronounced for BS50TS50O and
BS90TS10O. This significant loss of surface area in BSxTS100-xO samples could be
explained by their higher hydrophobicity, which facilitates the adsorption of the more
hydrophobic reaction products, thus resulting in pore blockage. At the same time, a
decrease on the acidity was also observed for all the reused solids. However, elemental
analyses indicated that the reused solids had the same measured S content as the fresh
ones (Table 1). Additionally, XPS experiments revealed that the oxidized S (peak at
approximately 169 eV) remained almost constant after the reaction (Table 1 and Fig.
S2). Also, TGA analyses of the reused solids showed a very similar profile to those of

the fresh ones except for BS90TS10O catalyst (Fig. 2 and Fig. S6). Indeed, the reused
materials lost the same weight between 370-520 ºC than the fresh ones, which would
indicate a similar amount of sulfonic acid groups, in agreement with S analyses (Table
1). However, a higher weight loss was observed in the range of 220-300 ºC for all
catalysts (around 1%), particularly for BS90TS10O (4%). It could be assigned to the loss
of glycerol or some reaction products that remained strongly adsorbed on the solid after
the reaction. This adsorption would result in blockage of the pores and so it would
explain the loss of surface area experienced by these materials, particularly BS90TS10O.
Therefore, all these results would support the fact that there was no leaching of
the sulfonic groups during the reaction. Thus, it could be considered that some active
sites may be blocked by adsorbed products and/or unreacted glycerol. In order to
confirm that several products and/or unreacted glycerol remained in the pores of the
catalyst after reaction, the used S50TS50O and BS90TS10O materials were subjected after
filtration to extraction with ethanol for 1 h under reflux. Then, ethanol was distilled and
the final residue was analysed by GC. The chromatographic analysis of the extract from
S50TS50O clearly showed a peak attributed to glycerol. The analysis of this sample by
GC-MS showed another peak that corresponded to MTBG, whose presence in the
surface of the solid would decrease the accessibility of the reactants to the acid sites.
However, there is no evidence of the other products in the chromatogram of the extract.
The high hydrophilic character of the S50TS50O would be the reason for the higher
adsorption of the most hydrophilic compounds (glycerol, TBA and MTBGs). The
presence of these compounds would not affect the conversion because they can furtherly
react to form the h-GTBE. Furthermore, taking into account the high catalytic
performance of this solid, it is expected a higher amount of MTBGs. Thus, a simple
washing with ethanol could be insufficient to remove the product and/or the reactants

which remain on the surface of the catalyst. On the contrary, a higher amount of
adsorbed compounds was detected in the extract from BS90TS10O (Fig. S7), whose
chromatogram exhibited the peak of glycerol, two different peaks of the MTBG isomers
and also the peak associated to 1,3-DTBG, thus confirming the pore blockage not only
by the unreacted glycerol but also by several reaction products. This fact was also
confirmed by comparison of its Raman spectrum with that of the commercial 1-MTBG
(Fig. S8). As expected, the surface, the pore volume and the acidity of the catalyst
slightly increased after extraction with ethanol (Table 1). In addition, a further reaction
was carried out with the regenerated S50TS50O, attaining a glycerol conversion of 70%
and a selectivity value to h-GTBE of 23%, which were very similar to the values
obtained on the fresh catalyst. All these observations suggest that the main reason for
the loss of activity is the pore blockage and/or the deactivation of the acid sites rather
than leaching of the active sites. As discussed previously, this effect is indeed more
evident for the more hydrophobic materials.
4. Conclusions
Different silica-based and organosilica-based hybrid materials have been
successfully employed in the catalytic etherification of glycerol with tert-butyl alcohol.
In fact, the catalytic performance over the S50TS50O catalyst exceeded that obtained on a
highly active commercial sulfonic resin A-15 after 24 h at 75 ºC with a higher overall
conversion and a higher yield to h-GTBE (ca. 28%). The high density of acid sites of
S50TS50O, together with its textural and chemical properties would explain the high
catalytic performance obtained on this catalyst. The competitive adsorption of the water
generated during the reaction on the acid sites of A-15 would explain that conversion
did not increase after 3 h of reaction time. However, the presence of silanol groups on
the hybrid silicas precludes the inhibition effect of water and consequently the

conversion increased with time on these catalysts. Furthermore, the sulfonated hybrid
silicas exhibited a higher thermal stability than A-15, making these materials very
attractive for acid-catalyzed reactions that require higher reaction temperatures.
Additionally, on the basis of the results obtained from the characterization of the
reused catalysts, it could be concluded that there was no leaching of the active sites. A
decrease in the surface area and the pore volume was associated with the blockage of
the pores by unreacted glycerol and reaction products. Moreover, this effect was more
evident for the most hydrophobic hybrid materials.
Acknowledgements
Subsidies granted by the Spanish Ministry of Education, Culture and Sport
(Project ENE2011-27017), Spanish Ministry of Economy and Competitiveness
(MAT2013-44463-R), Junta de Andalucía (Projects P11-TEP-7723 and P10-FQM-6181
and Research Groups FQM-162 and FQM-346) and FEDER funds are gratefully
acknowledged. The technical support and facilities from Córdoba University´s SCAI are
greatly appreciated. R. Estévez is indebted to the Junta de Andalucía for the contract
associated to PIF Project (P11-TEP-7723).

References
[1]
C.A. Quispe, C.J. Coronado, J.A. Carvalho Jr, Renew. Sust. Energ. Rev. 27
(2013) 475-493.
[2]
[3]
A. Corma, S. Iborra, A. Velty, Chem. Rev. 107 (2007) 2411-2502.
A. Behr, J. Eilting, K. Irawadi, J. Leschinski, F. Lindner, Green Chem. 10
(2008) 13-30.
[4]
M. Pagliaro, R. Ciriminna, H. Kimura, M. Rossi, C. Della Pina, Angew. Chem.
Int. Edit. 46 (2007) 4434-4440.
[5]
C.-H.C. Zhou, J.N. Beltramini, Y.-X. Fan, G.M. Lu, Chem. Soci. Rev. 37 (2008)
527-549.
[6]
N. Rahmat, A.Z. Abdullah, A.R. Mohamed, Renew. Sust. Energ. Rev. 14 (2010)
987-1000.
[7]
R. Estevez, S. Lopez-Pedrajas, F. Blanco-Bonilla, D. Luna, F. Bautista, Chem.
Eng. J. 282 (2015) 179-186.
[8]
C. Beatrice, G. Di Blasio, M. Lazzaro, C. Cannilla, G. Bonura, F. Frusteri, F.
Asdrubali, G. Baldinelli, A. Presciutti, F. Fantozzi, Appl. Energ. 102 (2013) 63-71.
[9]
M. Izquierdo, M. Hermo, X. Ariza, Renew. Sust. Energ. Rev. 16 (2012) 6717-6724.
[10] H. Noureddini, U.S. Patent 6015440 (2000).
J. Izquierdo, M. Montiel, I. Palés, P. Outón, M. Galán, L. Jutglar, M. Villarrubia,
[11] A. Behr, L. Obendorf, Eng. Life Sci. 2 (2002) 185-189.
[12] K. Klepáčová, D. Mravec, M. Bajus, Appl. Catal. A-Gen. 294 (2005) 141-147.
[13] J. Melero, G. Vicente, G. Morales, M. Paniagua, J. Moreno, R. Roldán, A.
Ezquerro, C. Pérez, Appl. Catal. A-Gen. 346 (2008) 44-51.
[14] M.D. González, P. Salagre, R. Mokaya, Y. Cesteros, Catal. Today 227 (2014)
171-178.

[15] R. Karinen, A. Krause, Appl. Catal. A-Gen. 306 (2006) 128-133.
[16] M.D. González, Y. Cesteros, J. Llorca, P. Salagre, J. Catal. 290 (2012) 202-209.
[17] M.D. González, P. Salagre, E. Taboada, J. Llorca, E. Molins, Y. Cesteros, Appl.
Catal. B-Environ 136 (2013) 287-293.
[18] K. Klepáčová, D. Mravec, A. Kaszonyi, M. Bajus, Appl. Catal. A-Gen. 328
(2007) 1-13.
[19] K. Klepáčová, D. Mravec, M. Bajus, Chem. Pap. 60 (2006) 224-230.
[20] M.P. Pico, J.M. Rosas, S. Rodríguez, A. Santos, A. Romero, J. Chem. Technol.
Biot. 88 (2013) 2027-2038.
[21] N. Ozbay, N. Oktar, G. Dogu, T. Dogu, Int. J. Chem. React. Eng. 8 (2010).
[22] C. Cannilla, G. Bonura, L. Frusteri, F. Frusteri, Environ. Sci. Technol. 48 (2014)
6019-6026.
[23] N. Ozbay, N. Oktar, G. Dogu, T. Dogu, Top. Catal. 56 (2013) 1790-1803.
[24] N. Ozbay, N. Oktar, G. Dogu, T. Dogu, Ind. Eng. Chem. Res. 51 (2011) 8788-
8795.
[25] M. Gonçalves, M. Mantovani, W.A. Carvalho, R. Rodrigues, D. Mandelli, J.S.
Albero, Chem. Eng. J. 256 (2014) 468-474.
[26] P.A. Celdeira, M. Goncalves, F.C. Figueiredo, S.M. Dal Bosco, D. Mandelli,
W.A. Carvalho, Appl. Catal. A-Gen 478 (2014) 98-106.
[27] M. Gonçalves, F.C. Soler, N. Isoda, W.A. Carvalho, D. Mandelli, J. Sepúlveda,
J. Taiwan Inst. Chem. E. 60 (2016) 294-301.
[28] T.S. Galhardo, . imone, M. Gon alves, F.C. Figueiredo, D. Mandelli, .A.
Carvalho, ACS Sustainable Chem. Eng. 1 (2013) 1381-1389.
[29] F. Frusteri, F. Arena, G. Bonura, C. Cannilla, L. Spadaro, O. Di Blasi, Appl.
Catal. A-Gen 367 (2009) 77-83.

[30] X. Liao, S.-G. Wang, X. Xiang, Y. Zhu, X. She, Y. Li, Fuel Process. Technol.
96 (2012) 74-79.
[31] M. Srinivas, G. Raveendra, G. Parameswaram, P.S. Prasad, N. Lingaiah, J. Mol.
Catal. A-Chem. 413 (2016) 7-14.
[32] N. Simone, W.A. Carvalho, D. Mandelli, R. Ryoo, J. Mol. Catal. A-Chem.
(2016) 7-14.
[33] M.D. González, Y. Cesteros, P. Salagre, Appl. Catal. A-Gen. 450 (2013) 178-
188.
[34] S. Hamoudi, S. Kaliaguine, Micropor. Mesopor. Mat. 59 (2003) 195-204.
[35] Q. Yang, J. Liu, J. Yang, M.P. Kapoor, S. Inagaki, C. Li, J. Catal. 228 (2004)
265-272.
[36] D. Esquivel, E. De Canck, C. Jimenez-Sanchidrian, P. Van Der Voort, F. J
Romero-Salguero, Curr. Org. Chem. 18 (2014) 1280-1295.
[37] V. Dufaud, M.E. Davis, J. Am.Chem. Soc. 125 (2003) 9403-9413.
[38] M.I. López, D. Esquivel, C. Jiménez‐Sanchidrián, F.J. Romero‐Salguero,
ChemCatChem 5 (2013) 1002-1010.
[39] M.E. Jamróz, M. Jarosz, J. Witowska-Jarosz, E. Bednarek, . Tęcza, M.H.
Jamróz, J.C. Dobrowolski, J. Kijeński, pectrochim. Acta Part A. 67 (2007) 980-988.
[40] D. Esquivel, C. Jiménez-Sanchidrián, F.J. Romero-Salguero, J. Mat. Chem. 21
(2011) 724-733.
[41] I.K. Mbaraka, B.H. Shanks, J. Catal. 244 (2006) 78-85.
[42] D. Margolese, J. Melero, S. Christiansen, B. Chmelka, G. Stucky, Chem. Mat.
12 (2000) 2448-2459.
[43] F. Frusteri, L. Frusteri, C. Cannilla, G. Bonura, Bioresource Technol. 118 (2012)
350-358.

[44] C. Cannilla, G. Bonura, L. Frusteri, F. Frusteri, Open Chem. 12 (2014) 1248-
1254.
[45] K.Y. Nandiwale, S.E. Patil, V.V. Bokade, Energ. Technol. 2 (2014) 446-452.

Figure and Scheme Legends
Figure 1 Nitrogen adsorption-desorption isotherms of materials S90TS10O and
BS90TS10O fresh and reused. Adsorption (○) and desorption (●) branches.
Figure 2 TG analysis for the hybrid organosilicas and A-15. TGA (solid line); DTA
(dashed line).
Figure 3 Study of the variation of conversion with time of reaction on the catalysts
studied. Reaction conditions: 5.0 wt.% of catalyst referred to the starting amount of
glycerol, TBA/G molar ratio = 4, reaction temperature = 75 ºC.
Figure 4 Evolution of selectivity values to the different products with reaction time on
S50TS50O at the same experimental conditions as in Figure 2.
Figure 5 Comparison of the conversion of glycerol and polyether yields on the fresh and
reused solids studied at the same experimental conditions as in Figure 2 after 17 h of
reaction time.
cheme 1 Reaction pathway for the etherification of glycerol with tert-butyl alcohol.
ater formed in every step has been omitted.