HMF etherification using NH4-exchanged zeolites

Article abstract of DOI:10.1039/c5nj03461b

The properties of BEA, MFI and Silicalite-1 zeolites in the ammonium and protonic forms are studied in the etherification of HMF (5-hydroxymethylfurfural) in anhydrous ethanol and compared with FTIR data on ammonium ion siting and displacement by competitive adsorption, as well as data on ammonium ion dissolution in aqueous solution. For the first time it is demonstrated that ammonium-exchanged zeolites are active and show better performances (particularly for the BEA structure) in the acid-catalyzed etherification reaction. This behavior is associated to a reversible dissociation of NH₄⁺ ions, which is favored by the BEA zeolite structure. A critical condition for enhanced catalytic performances is that dissociated ammonia remains in the zeolite cages, and may be reversibly re-adsorbed. It is thus likely that the dissociated ammonia participates in the reaction or induces a confinement effect.

Full text of DOI:10.1039/c5nj03461b

NJC
View Article Online
View Journal
PAPER
HMF etherification using NH -exchanged
4
zeolites†‡
Cite this:DOI: 10.1039/c5nj03461b
a
a
a
a
Katia Barbera, Paola Lanzafame,* Siglinda Perathoner, Gabriele Centi,
b
b
b
Massimo Migliori, Alfredo Aloise and Girolamo Giordano
The properties of BEA, MFI and Silicalite-1 zeolites in the ammonium and protonic forms are studied in
the etherification of HMF (5-hydroxymethylfurfural) in anhydrous ethanol and compared with FTIR data
on ammonium ion siting and displacement by competitive adsorption, as well as data on ammonium ion
dissolution in aqueous solution. For the first time it is demonstrated that ammonium-exchanged zeolites are
Received (in Montpellier, France)
1
6th December 2015,
active and show better performances (particularly for the BEA structure) in the acid-catalyzed etherification
+
4
Accepted 8th February 2016
reaction. This behavior is associated to a reversible dissociation of NH
ions, which is favored by the BEA
zeolite structure. A critical condition for enhanced catalytic performances is that dissociated ammonia
remains in the zeolite cages, and may be reversibly re-adsorbed. It is thus likely that the dissociated
ammonia participates in the reaction or induces a confinement effect.
DOI: 10.1039/c5nj03461b
www.rsc.org/njc
2
0,21
In previous publications
we reported the performance of
1
. Introduction
mesoporous zeotype materials (Al-MCM-41, SBA-15) on the
Zeolites, in addition to their pivotal role as heterogeneous etherification of HMF with ethanol to produce biodiesel compo-
catalysts in acid-catalysis for the production of high quality nents, demonstrating the important role of the nature, strength
1
fuels and chemicals, find increasing applications in biomass and localization of acidic sites in the products selectivity. More
2
–5
22
transformation reactions. The increasing number of papers recently, the activity of Silicalite-1 (MFI structure) was studied
focused on the use of zeolites for the conversion of biomass in the same reaction and the role of the surface acidity induced
6–14
derivatives
shows that these materials will play a determining by defects was highlighted. Weak acid sites associated to zeolite
role in the future of bio-refinery.
15–19
However, the characteristics defects (hydroxyl nests) were shown to be active in the etheri-
of biomass platform molecules (rich in oxygen) and the operation fication reaction, a reaction that is of more general interest for
conditions (typically in the liquid phase) are rather different from zeolites applied to sustainable chemistry/energy topics, from
2
3
those required for refinery applications (still the main use of the formation of dimethylether (DME) from methanol to the
2
4,25
zeolites). It is thus necessary to extend the concepts of the use of formation of various additives for fuels.
zeolites beyond the current uses which are essentially related to A general dilemma for these reactions is that strong acid
their acid-properties (Brønsted and, to a minor extent, Lewis sites would enhance the performance but at the same time
acidity) when other transition metals are not present.
One concept not explored is the use of the NH
catalyze secondary reactions. Blocking strong acid sites with a
-exchanged molecule forming a stable, but reversible complex, could be a
4
zeolites. These materials represent the intermediate typically way to control the acidity and accessibility of these strong sites,
used in preparing the acid form by thermal decomposition. In in a somewhat similar way to the reversible blocking of very
high-temperature applications, as typically occurs for oil-refinery reactive sites in enzymes being one of the key characteristics of
uses, they are not stable, but for liquid phase operations in mild the enzyme itself. In liquid-phase acid-catalyzed reactions, the
conditions, as often encountered in biomass platform transfor- ammonium form of a zeolite can thus enable control and tune the
mations, they can be potentially utilized. We were thus interested catalytic performance, but this aspect, as mentioned, was never
in investigating this new area of the catalytic chemistry of zeolites. reported in the literature. The etherification of HMF with ethanol,
26,27
besides its applicative interest,
is a good model reaction to
a
b
University of Messina, Section Industrial Chemistry, ERIC aisbl and CASPE-INSTM,
V.le F. Stagno d’Alcontres 31, 98166 Messina, Italy. E-mail: planzafame@unime.it
Department of Environmental and Chemical Engineering, University of Calabria,
via Bucci, 87036 Rende, Italy
investigate this aspect of zeolite reactivity due to the strong
dependence of the selectivity and catalytic behavior on the acidity.
It may be observed that the catalytic behavior associated to
the modification of the intrinsic reactivity of the zeolite active
sites by counter-ions was never investigated in detail, although
†
Paper dedicated to a friend, François Fajula, in recognition of his pioneering
contribution to the advances in zeolite science and technology.
‡
Electronic supplementary information (ESI) available. See DOI: 10.1039/c5nj03461b there is emerging evidence for cooperative effects between
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2016
New J. Chem.

View Article Online
Paper
NJC
framework and extra-framework sites as well as transition state 2. Experimental
confinement being involved in determining the acidity of
2.1 Materials
2
8–30
zeolites.
However, a more complex catalytic chemistry is
BEA and ZSM-5 (MFI) zeolites in the ammonium form were
provided respectively by Zeolysts Int. (CP814E*) and Alsi Penta
present. We observe that the charge-balancing extra-framework
cations occupy well-defined crystallographic positions, exposed
at the surface of the intra-crystalline cages and channels where
they can act as active centers. Moreover, they are the main sources
of intra-zeolitic electric fields, whose strength depends on the
cation radius and charge. Their properties, and the electrostatic
fields they generate, render the counterions capable of forming
adducts with molecules moving inside the framework, resulting
in a strong electronic perturbation of the adsorbed molecules and
finally in their chemical activation. In proton-exchanged zeolites,
bridging Si(OH)Al groups confer a remarkable Brønsted acidity to
(
SM-27). The Silicalite-1 sample (NH
4
-Sil) was obtained from the
2
2
synthesized Na-Silicalite-1 after treatment with an aqueous
solution of NH
with NH
+
4
NO
3
/NH
4
OH (pH = 10.5) to ion-exchange Na
-form sample
+
4
. To generate the acidic forms, each NH
4
was calcined in air with the following temperature program:
À1
ramp 2 1C min to 500 1C, hold the isotherm for 5 h and then
cool down to room temperature. We denote the resulting acidic
forms as H-MFI, H-BEA, and H-Sil. For BEA and MFI type zeolites
2 2 3
the SiO /Al O ratio ranges from 22 to 25.
31
the framework. These species can form hydrogen-bonded
adducts with molecules trapped inside the zeolite voids in appro-
priate conditions. Such adducts can further evolve to protonated
species. The formation of hydrogen-bonded precursors and
protonation are usually the first stages of Brønsted acid-catalyzed
chain reactions, which, together with shape selectivity, make
zeolite heterogeneous catalysts of paramount importance in
petrochemistry. The local environment of both Lewis and
Brønsted acidic centers affects not only their availability to
2
.2 Catalyst characterization
FTIR measurements of all calcined samples were carried out
on self-supporting wafers and the spectra were collected at a
resolution of 4 cm with a Bruker Equinox 55FTIR spectro-
photometer equipped with a MCT detector. The self-supporting
wafers were prepared and activated under vacuum (10 Torr)
for 1 hour at 400 1C in an IR cell allowing in situ thermal treatments,
and a basic probe molecule dosage (acetonitrile, D
D
À1
À4
3
-acetonitrile,
2
O, and pyridine). Then samples were outgassed both at room
reactants, but also the stabilization of reaction intermediates/
+
temperature and 150 1C in order to evacuate the physisorbed
molecules.
transitional states. The presence of a cation such as NH
4
,
+
3
which may reversibly dissociate to NH and H , further influences
Strong Lewis and Brønsted acidic site quantification was
performed by integrating the area underneath the bands at 1450
and 1545 cm after pyridine dosage. The integrated molar
this intra-crystalline reactivity of zeolites, thus offering a better
pathway to develop new selective routes on one hand, and to
understand this chemistry on the other hand.
À1
extinction coefficients (IMEC) were calculated, according to
Beer’s law, by dosing pyridine at increasing concentrations into
each sample under study after evacuation at 150 1C under high
There are thus many relevant motivations to investigate the
+
catalytic chemistry of NH
4
-exchanged zeolites, which go beyond
the interest for the specific reactions and catalysis itself, because
they represent other elements in understanding the supra-
2
2
vacuum, as described elsewhere.
3
2
A high performance ammonia ion selective electrode (ISE)
from Thermo Scientific Orion was used to accurately determine
molecular chemistry in the void-space of zeolites. To contribute
to this knowledge, we report here the role of cationic species
+
+
+
the dissolved ammonia in aqueous solution due to NH4 release
(
NH4 and H ) in two tridimensional zeolites (BEA and MFI)
from the zeolites. For each experiment, the electrode was
immersed in a solution containing 34 mL of distilled water,
having a similar Si/Al ratio. In fact, it is important to compare
the behavior of NH -exchanged zeolites with that of the corre-
4
+
6
80 mL of Ionic Strength Adjuster (ISA) and 1 g of NH
then ammonia release was monitored in the temperature range
5–50 1C, and in isothermal conditions at 50 1C for 3 hours.
4
-zeolite;
sponding acid forms, to better determine the role of the
ammonium ion on the catalytic performances. For the same
reason, it is important to analyze the effect of zeolite cages with
different sizes, being reasonable that a confinement effect would
be present in this dynamic equilibrium between the dissociated
and un-dissociated ammonium ion forms. MFI is a 10-member
ring zeolite, while BEA is a 12-member ring zeolite. Both zeolites
have a tridimensional pore structure, but they present different
2
Prior to the experiment, the electrode response was calibrated
using NH4 standard solutions at the different working tem-
peratures in the range 25–50 1C.
+
2
.3 HMF etherification reaction
channel dimensions, 0.47 Â 0.46 nm for MFI and 0.59 Â 0.59 nm The HMF etherification with ethanol was carried out in a Parr
for BEA. The catalytic performances for the above cited reaction autoclave reactor (Teflon-lined) provided with a Parr 4848
using these two types of zeolites (with comparable Si/Al ratios) controller. Reaction tests were carried out at 140 1C for 30 min,
were compared and in situ techniques were used to investigate and 1, 3 and 5 h under autogeneous pressure, strictly following
+
22
the dynamic equilibrium of NH4 in aqueous media. In order to the procedure described elsewhere. The reaction products of
clarify the role of the zeolite structure in relation to the presence interest, 5-(ethoxymethyl)furan-2-carbaldehyde (EMF) and ethyl
of strong acid sites (related to aluminum substitution of Si ions 4-oxopentanoate (EOP) were quantified by GC-FID analyses.
2
in the zeolite framework), a Silicalite-1 zeolite, i.e. with the MFI The H O formation during the catalytic tests was monitored
structure, but without aluminum, was also utilized. In this using a Methrom 831 KF Coulombometer instrument at room
sample, the acidity is related primarily to defects in the structure. temperature.
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2016

View Article Online
NJC
Paper
3
. Results and discussion
The analysis of the selectivities to products at 50% of HMF
conversion, and in kinetic regime, allows to provide further
indications on the differences in specific reactivity of active
sites (Fig. 1). The ammonium exchanged forms of all zeolites are
more selective toward EMF production than the corresponding
protonic form. These differences are more pronounced at least
for the more active and selective BEA-type zeolites. There is thus
a complex behavior, which cannot be related only to zeolite
structure but may depend on a dissociative equilibrium of the
3.1 Catalytic results and comparison with zeolite acidity
In the reaction conditions used for the catalytic tests (140 1C,
autogeneous pressure, ethanol as reactant and reaction medium),
the main reaction products obtained were EMF and EOP, while
other by-products, such as 2,5-furancarboxaldehyde, 2,5-dimethyl-
furan, 5-(methyl)furan-2-carbaldehyde, ethyl formate, ethyl acetate
and diethyl ether, were detected in trace amounts. Attention was
focused on EMF and EOP formation, both deriving directly from
HMF conversion on Lewis and Brønsted acid centers, respectively.
Moreover, the formation of water (see ESI‡), produced in situ by
ethanol condensation, was observed as already assessed for
similar systems. The catalytic behavior of the two zeolites
with different topologies (BEA and MFI) in the corresponding
NH
compared with those of the NH
materials as references for an Al-free MFI structure (Table 1).
All systems show a high productivity in EMF and EOP, with
+
+
NH
4
ion. The characterization of the acidity for the NH
4
zeolites didn’t show substantial differences in terms of the
number and strength of acidic sites (Table 2). It should be
remarked that pyridine may compete with ammonia for chemi-
2
0
+
sorption on the acid sites of NH
4
-exchanged zeolites. While
+
this is what likely happens in BEA, only some NH
4
is shifted in
+
+
4
and H forms was followed, and their performance was
NH -MFI, indicating that some of these sites (particularly
4
+
+
4
exchanged and H Silicalite-1
Lewis-acid type, LAS) are located in zeolitic pockets that are
not accessible to pyridine and probably also HMF.
+
total HMF conversion after 5 h of reaction observed only for 3.2 Measurement of the concentration of NH4 released in
NH
4
-BEA, H-BEA and H-MFI. This catalytic behavior is in agree- water solution
2
1,22
ment with that observed previously in similar systems,
where
+
In order to measure the amount of NH
4
ions released in
a trend of EMF and EOP formation was demonstrated with the
relative amount of Lewis and Brønsted acidic sites, detected after
pyridine desorption at 150 1C (close to the reaction temperature),
respectively, as reported in Table 2.
The space-time yield to EMF is slightly higher in the ammonium
form for BEA and Silicalite-1 than the corresponding H forms,
solution, specific tests were performed. The determination of
+
the NH
4
ions released directly during the catalytic tests was not
possible under the reaction conditions (140 1C and anhydrous
ethanol as solvent) because the membrane used in the ion-
selective electrode works only in aqueous media. These tests
+
+
were performed by suspending the NH4 forms of the BEA,
while for MFI it shows an opposite trend. The space-time yield
MFI and Silicalite-1 zeolites in water and analyzing the amount
to EOP is decreased in NH
to protonic form, while nearly constant in Silicalite-1 samples
Table 1).
4
-exchanged zeolites with respect
+
of NH
4
ions released in the solution with an ion-selective
electrode in the range 25–50 1C. This type of measurement thus
provides an indication of the relevance of this effect, although
under different conditions to the catalytic tests because water is
formed during the reaction (at the end of reaction is typically in
the order of thousands of ppm, see ESI‡).
(
Table 1 Comparison of catalytic activity for all the samples after 5 hours
of reaction. T = 140 1C, HMF = 2.5 mmol, EtOH = 3.4 mL
Fig. 2a shows the variation in the values of the potential
(proportional to the NH4 concentration in solution) of the
electrode (reported in mV) monitored during the experiments
À1 catÀ1
Space-time yield (mmol h
g
)
+
HMF
conversion (%)
Samples
NH -BEA
EMF
EOP
at different working temperatures. The value of the potential
of NH -Sil is positive (corresponding to a very low NH
4 4
4
100
100
60
100
55
3.52
3.11
1.29
2.28
0.92
0.82
0.44
0.62
0.44
0.63
0.15
0.13
+
H-BEA
NH -MFI
H-MFI
NH -Sil
H-Sil
4
concentration in the solution, about 20 ppm) for the full
temperature range studied, indicating a negligible presence
4
+
51
of NH4 in the aqueous solution. For NH -BEA and NH -MFI
4
4
the potential values are always in the negative range, but
two different trends could be observed. A progressive decrease
of the potential was observed for both zeolites in the
Table 2 Amount of strong Lewis (LAS)/Brønsted (BAS) acid sites for all the
samples after pyridine desorption at 150 1C
2
5–40 1C temperature range, but for NH
4
-BEA the initial value
catÀ1
FT-IR pyridine adsorption (mmol g
)
of the potential (À35 mV, corresponding to B1000 ppm of
a
b
+
Samples
NH -BEA
LAS
BAS
NH
released in solution) reaches an inflection point at
4
around 45 1C and returns to the initial value at 50 1C.
4
265
274
78
169
5.5
4
741
744
758
837
0.36
0.17
H-BEA
NH -MFI
H-MFI
NH -Sil
NH -MFI shows instead a constant reduction in potential for
4
4
the entire investigated temperature range, starting from À65 mV
(corresponding to 2500 ppm). Within the limits of the extra-
4
H-Sil
polation of the results to higher temperatures and the use
of ethanol as solvent, it may be first concluded that there
is no relationship between the dissociation constant of the
a
À1
b
Calculated using band at 1445 cm , 150 1C. Calculated using band
À1
at 1545 cm , 150 1C.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2016
New J. Chem.

View Article Online
Paper
NJC
Fig. 1 Product selectivity distribution obtained at 50% HMF conversion for (a) NH
4
-series and (b) H-forms. Reaction conditions: T = 140 1C,
HMF = 2.5 mmol, EtOH = 3.4 mL.
+
4
Fig. 2 Variation of NH
T = 50 1C.
concentration, expressed as mV for the different samples in (a) the T range 25–50 1C and (b) under isothermal conditions:
+
+
NH
4
-exchanged zeolites (i.e. the amount of NH
4
ions released catalytic performances of these two samples are related to these
to the solution) and their activity in HMF conversion (see space- different characteristics.
time yields in Table 1). The second observation is that there is a
reversible de-ammoniation reaction for the NH -BEA zeolite
that is not apparently present (in the investigated range) for
4
+
3
4
.3 Siting of NH ions and competitive adsorption of bases
NH
4
-MFI. As confirmation, measurements carried out under Few literature data contain analysis of the local siting of
isothermal conditions for 160 minutes at 50 1C reveal that only ammonium ions, for which monodentate, bidentate, or tridentate
+
33,34
the BEA zeolite was able to fully restore the initial NH
Fig. 2b). There is thus a clear difference in the behavior of BEA
4
content structures might be formed (Scheme 1).
(
For the sake of brevity, we will describe in detail only the
and MFI in terms of reversible de-ammoniation that is likely profiles of the IR spectra recorded for the NH -BEA zeolite and
4
related to the different locations of the acid sites (demonstrated remark only on the differences observed in the other samples.
also with the FTIR data, Table 2) and to the ‘‘confinement’’ Zeolite BEA is characterized by two sets of perpendicular straight
effect of the zeolitic cage. It is likely that the differences in the 12-MR channels, running along the [100] and [010] directions.
4 s
Scheme 1 Possible configurations of NH -zeolite bonding. (a) bidentate, (b) tridentate with C3v symmetry, and (c) tridentate with C symmetry.
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2016

View Article Online
NJC
Paper
Fig. 3 FT-IR spectra of NH
at 150 1C (dashed line). (a) Hydroxyl region (b) NH
4
-BEA zeolite before and after base probe molecule adsorption (full lines), desorption at room temperature (dotted line) and
+
4
bending region.
They intersect in a third kind of curved channel along the [001]
direction, whose geometry is governed by the stacking sequence, interaction status of NH
We may suggest that this is direct proof of a change in the
+
À1
4
ions. In the 1800–1350 cm fre-
and create large cavities where bulky molecules can freely locate quency region, it is possible to identify the presence of two
3
5
without causing a blockage. BEA zeolites present high accessi- groups of IR-active bending modes: a triplet of peaks between
À1
+
bility for all the directions of molecules up to a 5.95 Å diameter, 1550 and 1400 cm
associated to bidentate NH4 species
thus also making it possible for bulky molecules to interact with (which have two NH groups involved in H bonding, and hence
all the acidic centers present in the catalyst.
Fig. 3 shows the IR spectra of the dehydrated parent NH
have a C2v symmetry), and an intense triplet in the range
-BEA 1700–1600 cm , typical of tridentate species (having three
À1
4
after the dosage of different base probe molecules. The spectra NH groups H-linked to the zeolite and a single unperturbed
were recorded after base saturation at room temperature (full line), NH, with the symmetry C3v or C , depending on the position of
evacuation at room temperature (dotted line) or at 150 1C under the oxygen to which the third H is linked). The free NH groups
s
vacuum (dashed line).
point into the channels and cavities and are not appreciably
band in the perturbed by the interaction with the opposite walls.
À1
NH -BEA displays an intensive 3745 cm
4
OH stretching frequency region (Fig. 3a), related to isolated
The interaction of NH -BEA with each base probe molecule
4
terminal SiOH groups located on the external and internal deserves an individual explanation (Fig. 3b). The IR spectra
3
6
À1
surfaces (low-frequency tail). No feature at 3610 cm is observed, obtained after interaction with acetonitrile (pK
a
= 25) display a
+
À1
due to the presence of NH
4
ions bonded to Brønsted acid broad adsorption band centered at around 1700 cm with a
À1
centers. The broad and complex absorption covering the range right shoulder, and another one appears at 1350 cm . During
300–2800 cm , can be associated to stretching vibrations of evacuation at room temperature and 150 1C, these bands are
À1
3
N–H in different environments, i.e. monodentate, bidentate and eroded until the initial spectrum is completely restored. The
+
37
tridentate hydrogen-bonded NH4 configurations. As also docu- profiles obtained with D -CN are similar to those previously
3
33,38
mented from ab initio calculations,
the negatively charged described. The situation upon D O (pK = 16.6) dosage is
2 a
walls of the zeolites stabilize bidentate and tridentate structures. different: saturation leads to the appearance of an intense band
À1
In agreement, we do not observe any absorption related to at 1700 cm , which is removed during outgassing, but the
monodentate configurations, which usually present a steep features related to the tridentate species decrease in intensity
À1
À1
feature at 3375 cm . The dosage of the different base probes (triplet), with a corresponding increase of a band at ca. 1590 cm
.
À1
leads to the erosion of the bands associated with hydroxyl In parallel, the band at 1550 cm is irreversibly eroded and the
groups, and when further outgassing at room temperature is peak at 1490 cm doubles in intensity. Finally, the band at
carried out, a full restoration of the starting configuration is 1450 cm is highly perturbed, but fully restored, upon out-
À1
À1
achieved only for samples treated with acetonitriles. In the case gassing at room temperature. The final spectrum obtained after
of interaction with D -CN, the restoration of the silanol band is outgassing at 150 1C reveals an overall decrease of the features
3
partially impeded by the formation of Si-OD species (band at related to tridentate species, and an increase of those typical of
À1
2
757 cm , not shown here). This effect is more pronounced bidentate species, with a modification in the relative intensities
upon dosage of D
2
O.
2
after interaction with D O.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2016
New J. Chem.

View Article Online
Paper
NJC
Fig. 4 FT-IR spectra of (a) NH
temperature (dotted line) and at 150 1C (dashed line) in the NH
4
-MFI and (b) NH
4
-Sil samples before and after base probe molecule adsorption (full lines), desorption at room
+
4
bending region.
+
The same experiments carried out over the NH
4 4
-MFI zeolite BEA zeolite structure. FTIR data for NH siting and shift by
(Fig. 4a) reveal two substantial differences, with respect to the competitive adsorption reveal a significant difference between
observations for NH
4
-BEA.
the BEA and MFI forms.
Whereas in this case it is also possible to clearly observe
Although further studies are necessary to better understand
the typical features of both bidendate and tridentate-bonded this complex chemistry related to the dissociation of ammonium
+
NH4 species in the starting materials, the interactions of both ions in solution, its dependence on the zeolite structure and the
acetonitrile and D
3
-CN lead to a red-shift perturbation of relationship between these phenomena and the catalytic behavior,
À1
the bands at 1620 and 1550 cm , with a full recovery upon we may conclude that the FTIR data for the characterization of
degassing at room temperature. This would suggest a higher these species and their stability, release of ammonia in solution
instability of the two configurations, even in the presence of a and catalytic behavior are consistent.
very weak base. This behavior is even more pronounced in the
We demonstrate for the first time that ammonium-exchanged
presence of D O, where an overall attenuation of the bands zeolites are active and show better performances (particularly
2
(
both of bidentate and tridentate species) is registered, as a proof BEA) in acid-catalyzed etherification reactions. A critical condi-
+
of an irreversible reduction in the amount of NH4 ions inter- tion is that dissociated ammonia remains in the zeolite cages
acting with acidic Brønsted centers. and may be reversible re-adsorbed. Thus it is likely that the
In the case of the NH -Sil sample (Fig. 4b), the features dissociated ammonia participates in the reaction or induces a
species in different environments cannot be confinement effect. The exact nature of these aspects is under
envisaged, probably due to their concentrations being below further investigation.
the detection limits. All these observations are direct proof of In conclusion, the relation of the reversible dissociation of
4
+
related to NH
4
+
+
the strength of the NH4 interaction with acidic centers, which NH4 ions in the zeolite intra-cages with the zeolite structure
+
in turn has been highlighted during the monitoring of NH4
release in aqueous solution (Fig. 2).
itself is responsible for the different catalytic reactivity of the
samples, and may lead to the better performances with respect to
the corresponding protonic form. This has been demonstrated for
the HMF etherification with ethanol, a reaction of industrial
interest for the production of bio-diesel additives. However, it is
4
. Conclusions
BEA, MFI and Silicalite-1 zeolites in the ammonium and pro- likely of more general relevance to understand the catalytic
tonic forms have been tested in the etherification of HMF in chemistry of zeolites in these types of reactions and the possibility
anhydrous ethanol. The study of the sample acidity confirms of enhancing the behavior by cooperative effects.
that the selectivity to EMF and EOP is driven by the different
natures and amounts of Lewis and Brønsted acid sites, respec-
tively. In particular, the ammonium-exchanged form of zeolites
References
presents a higher selectivity in comparison with the corres-
ponding protonic form. The ammonium-exchanged form of the
zeolite is active in the acid-catalyzed etherification reaction of
HMF in anhydrous ethanol, leading to an improved selectivity
and activity, in particular for the BEA zeolite form. The behavior
is not related to ammonia dissociation in solution, but is likely
associated to a reversible dissociation, which is favored by the
1 A. Corma, Chem. Rev., 1997, 97, 2373–2420.
ˇ
2 J. Cejka, G. Centi, J. Perez-Pariente and W. J. Roth,
Catal. Today, 2012, 179, 2–15.
3 L. Wang and F.-S. Xiao, Green Chem., 2015, 17, 24–39.
4 D. Kubicka and O. Kikhtyanin, Catal. Today, 2015, 243, 10–22.
5 P. A. Jacobs, M. Dusselier and B. F. Sels, Angew. Chem., Int. Ed.,
2014, 53, 8621–8626.
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2016

View Article Online
NJC
Paper
6
7
8
P. Lanzafame, D. M. Temi, S. Perathoner, A. N. Spadaro and 22 P. Lanzafame, K. Barbera, S. Perathoner, G. Centi, A. Aloise,
G. Centi, Catal. Today, 2012, 179, 178–184.
A. A. Lappas, S. Bezergianni and I. A. Vasalos, Catal. Today,
M. Migliori, A. Macario, J. B. Nagy and G. Giordano, J. Catal.,
2015, 330, 558–568.
2
009, 145, 55–66.
23 E. Catizzone, A. Aloise, M. Migliori and G. Giordano,
Appl. Catal., A, 2015, 502, 215–220.
M. M. Antunes, S. Lima, P. Neves, A. L. Magalh ˜a es, E. Fazio,
A. Fernandes, F. Neri, C. M. Silva, S. M. Rocha, M. F. Ribeiro, 24 B. P. Pinto, J. T. de Lyra, J. A. C. Nascimento and C. J. A.
M. Pillinger, A. Urakawa and A. A. Valente, J. Catal., 2015,
29, 522–537.
K. S. Arias, S. I. Al-Resayes, M. J. Climent, A. Corma and
Mota, Fuel, 2016, 168, 76–80.
25 A. Bohre, B. Saha and M. M. Abu-Omar, ChemSusChem,
2015, 8, 4022–4029.
3
9
S. Iborra, ChemSusChem, 2013, 6, 123–131.
0 K. S. Arias, M. J. Climent, A. Corma and S. Iborra,
ChemSusChem, 2014, 7, 210–220.
26 R. J. van Putten, J. C. van der Waal, E. de Jong, C. B.
Rasrendra, H. J. Heeres and J. G. de Vries, Chem. Rev.,
2013, 113, 1499–1597.
1
1
1
1 J. Faria, M. Pilar Ruiz and D. E. Resasco, ACS Catal., 2015, 5, 27 J. Luo, J. Yu, R. J. Gorte, E. Mahmoud, D. G. Vlachos and
761–4771. M. A. Smith, Catal. Sci. Technol., 2014, 4, 3074–3081.
2 J. Jae, G. A. Tompsett, A. J. Foster, K. D. Hammond, S. M. 28 A. Comas-Vives, M. Valla, C. Coperet and P. Sautet, ACS
4
Auerbach, R. F. Lobo and G. W. Huber, J. Catal., 2011, 279,
57–268.
3 J. S. Kruger, V. Choudhary, V. Nikolakis and D. G. Vlachos,
ACS Catal., 2013, 3, 1279–1291.
4 E. Taarning, C. M. Osmundsen, X. Yang, B. Voss, S. I. Andersen
and C. H. Christensen, Energy Environ. Sci., 2011, 4, 793–804.
5 B. Kamm, Angew. Chem., Int. Ed., 2007, 46, 5056–5058.
Cent. Sci., 2015, 1, 313–319.
29 A. J. Jones, R. T. Carr, S. I. Zones and E. Iglesia, J. Catal.,
2014, 312, 58–68.
30 A. J. Jones, S. I. Zones and E. Iglesia, J. Phys. Chem. C, 2014,
118, 17787–17800.
31 E. Kassab, J. Fouquet, M. Allavena and E. M. Evleth, J. Phys.
Chem., 1993, 97, 9034–9039.
2
1
1
1
1
6 P. Lanzafame, G. Centi and S. Perathoner, Catal. Today, 32 J. Jiou, K. Chiravuri, A. Gudapati and J. J. Gassensmith,
014, 234, 2–12. Curr. Org. Chem., 2014, 18, 2002–2009.
7 S. Abate, P. Lanzafame, S. Perathoner and G. Centi, 33 A. Zecchina, L. Marchese, S. Bordiga, C. Paz `e and E. Gianotti,
ChemSusChem, 2015, 8, 2854–2866. J. Phys. Chem. B, 1997, 101, 10128–10135.
8 G. Centi, P. Lanzafame and S. Perathoner, Catal. Today, 34 W. Wu and E. Weitz, Appl. Surf. Sci., 2014, 316, 405–415.
2
1
1
1
2
2
2
011, 167, 14–30.
35 R. Buzzoni, S. Bordiga, G. Ricchiardi, C. Lamberti and
A. Zecchina, Langmuir, 1996, 12, 930–940.
36 K. Barbera, F. Bonino, S. Bordiga, T. V. W. Janssens and
P. Beato, J. Catal., 2011, 280, 196–205.
9 S. Abate, K. Barbera, G. Centi, P. Lanzafame and S. Perathoner,
Catal. Sci. Technol., 2016, DOI: 10.1039/C5CY02184G.
0 P. Lanzafame, D. M. Temi, S. Perathoner, G. Centi, A. Macario,
A. Aloise and G. Giordano, Catal. Today, 2011, 175, 435–441.
1 K. Barbera, P. Lanzafame, A. Pistone, S. Millesi, G. Malandrino,
37 K. Suzuchi, T. Noda, N. Katada and M. Niwa, J. Catal., 2007,
250, 151–160.
A. Gulino, S. Perathoner and G. Centi, J. Catal., 2015, 323, 38 E. H. Teunissen, R. A. van Santen, A. P. J. Jansen and F. B.
9–32. van Duijneveldt, J. Phys. Chem., 1993, 97, 203–210.
1
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2016
New J. Chem.