Efficient glucose dehydration to HMF onto Nb-BEA catalysts

Article abstract of DOI:10.1016/j.cattod.2018.08.004

The one-pot production of HMF from glucose was investigated in pure hot water and biphasic water/methylisobutylketone (MIBK) solvent using mesoporous Nb(0.02 and 0.05 mol%)-Beta zeolites obtained by a post synthesis methodology. The mesoporous Nb-Beta zeolites present residual framework Al-acid sites, extra-framework isolated Nb(V) and Nb₂O₅ pore-encapsulated clusters in which Nb(V)O-H exhibit moderate strength Br?nsted acidity. After optimization, the dehydration of glucose onto the Nb-modified Beta-zeolites occurred with a selectivity of 84.3% in HMF for a glucose conversion of 97.4%. This result has been obtained in a biphasic water/MIBK solvent and in the presence of NaCl, at 180 °C, after 12 h.

Full text of DOI:10.1016/j.cattod.2018.08.004

Accepted Manuscript
Title: Efficient glucose dehydration to HMF onto Nb-BEA
catalysts
Authors: Natalia Candu, Magdi El Fergani, Marian Verziu,
Bogdan Cojocaru, Bogdan Jurca, Nicoleta Apostol, Cristian
Teodorescu, Vasile I. Parvulescu, Simona M. Coman
PII:
S0920-5861(18)31368-3
DOI:
Reference:
https://doi.org/10.1016/j.cattod.2018.08.004
CATTOD 11609
To appear in:
Catalysis Today
Received date:
Revised date:
Accepted date:
20-3-2018
13-7-2018
4-8-2018
Please cite this article as: Candu N, El Fergani M, Verziu M, Cojocaru
B, Jurca B, Apostol N, Teodorescu C, Parvulescu VI, Coman SM, Efficient
glucose dehydration to HMF onto Nb-BEA catalysts, Catalysis Today (2018),
https://doi.org/10.1016/j.cattod.2018.08.004
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.

Efficient glucose dehydration to HMF onto Nb-BEA catalysts
Natalia Candu,1 Magdi El Fergani,1 Marian Verziu,1 Bogdan Cojocaru,1 Bogdan Jurca2, Nicoleta Apostol3,
Cristian Teodorescu3, Vasile I. Parvulescu,1* Simona M. Coman1*
1 Department of Organic Chemistry, Biochemistry and Catalysis, Faculty of Chemistry, University of
Bucharest, Regina Elisabeta Blvd., no. 4-12, Bucharest 030016, Romania
2 Centre of Research “Catalysts and Catalytic Processes”, Faculty of Chemistry, University of Bucharest, Bdul
Regina Elisabeta 4-12, Bucharest 030016, Romania
3 National Institute of Materials Physics, 405 Atomiștilor Str., Măgurele, Ilfov, Romania
*e-mail: vasile.parvulescu@chimie.unibuc.ro; simona.coman@chimie.unibuc.ro
Graphical Abstract
The one-pot production of HMF from glucose in biphasic water/MIBK solvent using Nb(0.02
and 0.05moles%)-Beta zeolites occurs with a selectivity to HMF of 84.3% for a glucose
conversion of 97.4%.

Highlights




Niobium-Beta zeolites were prepared via a post synthesis methodology
Efficient one-pot transformation of glucose to 5-hydroxymethylfurfural (HMF)
The catalysts are stable in hot water, at 180C
Working in biphasic water/MIBK solvent improves the selectivity to HMF

Abstract
The one-pot production of HMF from glucose was investigated in pure hot water and biphasic
water/methylisobutylketone (MIBK) solvent using mesoporous Nb(0.02 and 0.05moles%)-Beta zeolites
obtained by a post synthesis methodology. The mesoporous Nb-Beta zeolites present residual framework Al-
acid sites, extra-framework isolated Nb(V) and Nb2O5 pore-encapsulated clusters in which Nb(V)O-H exhibit
moderate strength Brønsted acidity. After optimization, the dehydration of glucose onto the Nb-modified
Beta-zeolites occurred with a selectivity of 84.3% in HMF for a glucose conversion of 97.4%. This result has
been obtained in a biphasic water/MIBK solvent and in the presence of NaCl, at 180C, after 12h.
Keywords: Nb-beta zeolites; Nb2O5 pore-encapsulated clusters; glucose; 5-hydroxymethylfurfural; biphasic
water/MIBK solvent.
1. Introduction
In the recent years, both academia and industry made great efforts to reduce the humanity reliance
on the fossil resources and the unprecedented environmental pollution. Among other directions, a special
attention has been focused on the valorisation of the neutral carbon inside the new biorefinery concept
looking for the production of chemicals and fuels from renewable resources and feedstocks. In this context,
the biomass, an inexpensive, renewable and abundant feedstock, evolved as a potential unique carbon
resource candidate for the production of various organic chemicals (ie, platform molecules) [1, 2]. As in the
case of petroleum-based industry, for a fast, green and sustainable development, catalysis must be placed in
the core of the biorefinery technologies. However, due to the complexity and diverse reactivity of the
biomass molecules, to be developed in a relatively short time, the renewable feedstock conversion needs
novel catalytic systems [3, 4]. In this respect, a faster alternative might be the development of strategies able

to adapt the already known structural and chemical properties of the catalysts used in the oil-based industry
to the biomass-based industry. Synthetic zeolites are potential candidates among these catalysts and several
properties as the particularly well dispersed high Brønsted acidity in the crystalline matrix combined with the
controlled porosity at the molecular dimensions recommend they in this attempt [5, 6]. However, in contrast
to the crude oil refinery, where thermally and hydrothermally stable zeolites are required, biorefinery
demands catalysts with high stability in hot liquid water since most of the biomass transformation processes
involve hydrolysis reactions.
Among the different biomass products, the production of 5-hydroxymethylfurfural (HMF) elicited a
special interest because it has been identified by the United States Department of Energy as one of the top
ten most valuable chemical platforms [7]. HMF may serve as intermediate for the production of a large
number of other value added chemicals being an alternative to formaldehyde in the production of resins. It
could also replace terephthalic acid for the production of polyesters after oxidation to 2,5-furandicarboxylic
acid (FDCA) or it can be converted into 2,5-dimethylfuran (DMF) as a liquid biofuel [8], and the examples can
be continued. However, despite this large versatile application its low product selectivity connected to
important difficulties in the separation led to high production costs. This generates a major drawback for a
sustainable utilization of HFM as a potential substitute for the petroleum derivatives.
To date, most of the biomass derived routes focused on its acid-catalyzed production directly from
glucose (a cheaper and large available raw material by comparison with fructose). This process requires the
concerted participation of both the Lewis acid (for glucose isomerization to fructose) and Brønsted acid (for
fructose dehydration to HMF) sites. In this respect, zeolites represent important catalysts candidates [8].
Adapting the textural properties of zeolites through the dealumination or desilication (ie generating
a hierarchical connection of micro-, meso- and macroposity [9]) may open new opportunities for the use of
zeolites as highly efficient catalysts for various processes. This can be achieved either by the treatment with
mineral acids or through the insertion of a heteroatom in the hydrophobic dealuminated zeolite framework
[10]. Occasionally, their stability may be even preserved after the addition of small quantities of mineral
acids. As an example, Otomo et al. [11] reported a dealuminated Beta-zeolite as an effective and stable

catalyst for the direct transformation of glucose to HMF. The reaction occurred with a selectivity of 55% for
a conversion of glucose of 78%. On the other side, the heteroatom insertion route can also provide an
improved stability in hot water. Based on this, in the last years, the heteroatomic Lewis acid- [12, 13] and
bifunctional-zeolite catalysts [14] emerged as an expanding field for the production of the platform molecules
[9, 15, 16].
For this purpose, several heteroatoms (e.g., Ti, Sn, Zr, V or Nb) have been inserted by using either
direct or post-synthetic strategies. The direct synthesis includes the conventional hydrothermal
crystallization [17], dry-gel conversion [18], fluoride modification [19] and microwave-assisted [20] methods.
However, most of these direct synthesis strategies are complicated and time-consuming, sometimes
resulting in a lack of reproducibility and feasibility for a large-scale synthesis. Therefore, other alternative
post-synthesis strategies emerged as important [21-26].
Niobium compounds are obviously stable in water and display remarkable catalytic acid properties
as either bulk or component of the catalysts. However, to date only scarce examples of Nb based zeolites
have been reported. Corma et al. [27] reported, for instance, an efficient direct synthesis of Nb incorporated
in the Beta zeolite framework in a fluoride medium, as a way to produce solid Lewis acids for the organic
reactions conducted in water. The impregnation of zeolites only generates extra framework niobium oxide
species [28-30]. In the same context, we reported the synthesis of niobium-modified aluminum hydroxide
fluorides, prepared via the direct fluorolytic sol-gel synthesis [31, 32] and niobia-based magnetic carriers
catalysts [33, 34]. By adjusting the preparation conditions (e.g., precursor nature, calcination temperature
and loading of niobia) it was possible to produce efficient catalysts for the synthesis of lactic acid, through
the direct hydrolysis of cellulose and starch.
More recently, we focused our interest to a post synthesis strategy for the production of highly
dispersed Nb species (0.02-0.05 moles%) into the zeolite matrix [35]. This strategy supposes a successive
zeolite dealumination and impregnation with Nb ethoxide. The novel catalysts (Nb(0.05)-β37.5)
demonstrated a high catalytic efficiency in glucose oxidation to succinic acid serving as an attractive

alternative to Ru-based catalysts. They were able to produce succinic acid in water with a selectivity of 84%,
for a total conversion of glucose at 180C under 18 bar of oxygen.
In this study, we expanded the scope of the use of Nb-Beta zeolites to the direct synthesis of HMF
from glucose. The reactions were carried out in a biphasic H2O/methyl-isobutyl-ketone (MIBK) system
affording the direct dehydration of glucose to HMF with selectivities up to 84.3% for a conversion of glucose
of 97.4%.
2. Experimental Section
2.1. Catalyst preparation
All the synthesis reagents were of analytic purity from Sigma-Aldrich and used as received.
The catalysts were synthesized as reported elsewhere [35] through a two-step post synthesis
methodology, involving a dealumination of H-Beta zeolite followed by its impregnation with Nb
ethoxide [36]. The mesoporous Beta zeolites (purchased from ZEOLYST International Company)
considered in this study had a Si/Al ratio of 12 (Nb-Beta12) and 18 (Nb-Beta18), respectively, and
the concentration of deposited niobium was of 0.02 and 0.05 moles% Nb. The prepared samples were
denoted as Nb (0.02)-Beta12, Nb (0.05)-Beta12, Nb (0.02)-Beta18, and Nb (0.05)-Beta18,
respectively.
2.2. Catalyst characterization
Powder X-ray diffraction (XRD), diffuse reflectance infrared Fourier transform spectroscopy (DRIFT),
and ICP-OES have been previously applied to study the structure and composition of the bulk materials and
were described elsewhere [35]. Textural characteristics (surface area, pore volume and pore diameter) were
determined from the adsorption-desorption isotherms of nitrogen at -196 °C using a Micromeritics ASAP
2020 Surface Area and Porosity Analyzer. TG-DTA analyses were recorded using a Shimadzu apparatus in a
Pt crucible. The heating rate was of 5º and 10°C min-1, respectively, starting from room temperature till 900
°C under a nitrogen flow of 10 mL min-1. XPS measurements were performed at a normal emission in a Specs
setup, by using the Al Kα monochromated radiation (hν = 1486.7 eV) of an X-ray gun, operating with 300 W

(12 kV/25 mA) power. A flood gun with an electron acceleration at 1 eV and electron current of 100 μA was
used in order to avoid the charge effects. Photoelectrons were collected in a normal emission by using a
Phoibos 150 analyzer, operating with a pass energy of 30 eV. The XPS spectra were fitted by using Voigt
profiles combined with their primitive functions, for inelastic backgrounds [37].
NH3-TPD measurements were carried out using an AutoChem II 2920 station. The samples (10-30
mg), placed in a U-shaped quartz reactor with an inner diameter of 0.5 cm, were pre-treated under He (Purity
5.0) at 80ºC for 1 h, and then exposed to a flow of 1vol% NH3 in helium, for 1 h. After that, the samples were
purged with a flow of He (50 mL min-1) for 20 min at 25ºC in order to remove the weakly adsorbed species.
TPD analysis has been carried out with a heating rate of 10ºC min-1 till 500 ºC. The desorbed products were
on-line analyzed by GC-TCD chromatography. The desorbed NH3, expressed as mmoles of NH3 per gram of
catalyst, was determined using a calibration curve.
2.3. Catalytic tests
The activity tests in the batch mode were carried out as follows: 0.03 g of catalyst were added to a
solution of 0.18 g (1 mmole) glucose in 5 mL water. After closing, the reactor was heated up to 180°C, under
stirring (1200 rpm), for 12-24 h. Additional catalytic tests were made in the same conditions but in a biphasic
solvent (3.5 mL aqueous solution of 20% NaCl and 1.5 mL MIBK). To investigate the chemical stability of the
catalysts, the content of the leached metal (ie, Nb and/or Al) into the reaction liquid, was determined by ICP-
OES (Agilent Technologies, 700 Series). Another catalytic test has been carried out with the same scope: 0.03
g Nb-Beta zeolite catalyst was boiled in 5mL water, at 180°C, for 3h, under stirring. After 3h, the catalyst was
separated from the liquid by centrifugation, and washed and dried at room temperature, then used in the
reaction test in the following conditions: 0.18 g (1 mmole) glucose in 5 mL water, 180°C and 24h.
For recycling tests, the recovered catalyst was washed with distilled water and centrifuged three
times, then dried at 80C and calcined at 450C in static air atmosphere for 8h.
2.4. Products analysis

After reaction, the catalyst was centrifuged and the products were separated by distillation under
vacuum. The reaction mixture was silylated, diluted with 1 mL of toluene and analyzed by GC-FID
chromatography (GC-Shimadzu apparatus). The product identification was made by comparison with
reference standard samples and by using a GC-MS Carlo Erba Instruments QMD 1000 equipped with a Factor
Four VF-5HT column.
3. Results and discussion
3.1. Catalysts characterization
The insertion of the transition metals into the crystalline framework positions of zeolites is
not an easy task because it depends on different factors such as the rigidity of the zeolitic crystalline
framework, the ionic radius of the metal and the synthesis conditions [38]. However, even if
successful, the confirmation of an isomorphous substitution is difficult and, for many years, the idea
of the substitution of the tetrahedral framework positions by transition metals met much criticism.
Nevertheless, recent reports have shown that such a substitution is indeed possible for several
metallosilicate molecular sieves. However, the insertion of the transition metal into the zeolite or
mesoporous molecular sieve structure is not dictated by the cation size or the state of oxidation of
metal cation in a simple way [39].
The synthesis and characteristic properties of the Nb-Beta zeolites were recently presented in more
details [35]. In brief, the insertion of Nb (0.02 and 0.05 mol%) into the Beta zeolite (Si/Al=12 and 18,
respectively) framework is not replacing the tetrahedral positions occupied by aluminum. The described post
synthesis methodology affords a very high dispersion of the attached Nb(V) species. Indeed, upon the
incorporation of Nb into SiBeta18, the d302 spacing increased from 3.917 Å (SiBeta18) to 3.929 Å (for Nb(0.02)-
Beta18) and 3.926 Å (for Nb(0.05)-Beta18) indicating an expansion of the siliceous Beta zeolite matrix [35].
However, comparing to d302 spacing values, reported by Dzwigaj et al. [36] for a similar material
(0.02mol%Nb-SiBeta) prepared by a post synthesis methodology, the values determined in our work [35]
seem to be enough small to explain the unsuccessful insertion of Nb in the tetrahedral positions of the Beta
matrix. The small expansion of the Beta zeolite matrix is, therefore, most probably due to its distortion caused

the non-framework pore encapsulated Nb(V) nano-oxides. This is in accordance to the difference between
the size of the niobium ions (Nb5+ 0.591 Å, Nb4+ 0.768 Å and Nb3+ 0.795Å) and that of Al3+ (0.239 Å). Also the
length of the Nb-O bond (1.89 Å for tetracoordinated Nb(V) species in Beta zeolite) does not fit that of the
Si-O bond (typically 1.60-1.65 Å in zeolites) [40].
Indeed, in accordance to the FTIR spectroscopy results, the most probable state of niobium is that
corresponding to Nb(V) in Nb(V)O-H where niobium is linked to the zeolite walls via Nb-O-Si bonds [35]. In
accordance to literature reports [36], the geometry of Nb is dependent on the synthesis method.
Differently to zeolites prepared through a direct synthesis procedure where tetrahedral species
correspond to a niobyl bond (-Nb=O) [27] the post synthesis procedure leads only to extra framework
Nb(V)O-H species.
The presence of extra framework Nb(V)O-H species is also suggested by the textural
characterization (Table 1) on the basis of N2 adsorption-desorption isotherms at 77 K (Figure 1).
Figure 1.
N2 adsorption-desorption isotherms were characteristic to a mesoporous texture (Type
IV) with BET surface areas of 536 and 438 m2g-1 for H-Beta18 and Nb(0.02)-Beta18, and of
458 and 391 m2g-1 for H-Beta12 and Nb(0.02)-Beta12, respectively. The mesoporosity, as it
could be inferred from XRD [35] and in accordance to the previous work, is the effect of
packing of small zeolite nanocrystals.
Table 1.
However, mesopores calculated using the BJH method (desorption branch, Figure 1) showed no
effect of the addition of niobium (Table 1). Thus, the decrease of the surface areas was not accompanied by
changes in the pore size.
XPS spectroscopy confirmed the presence of Nb(V) (BE of 207.3 eV) [41] in the exchanged
BEA zeolites (Figure 2). Moreover, the absence of the binding energies characteristics to Al(2p) (BE
of 73-76 eV) confirms the successful dealumination of the Beta zeolite, at least in the superficial level.

Figure 2.
Also, the XP spectra in the region of the O(1s) level (Figure 3) show bands confirming the presence
of two different oxygen species, ie the band at 533.2 eV associated to oxygen in SiO2 [42] and that at 531 eV
associated to M-O-Si bonds [43, 44]. The band of Si(2p) at 103.5 eV corresponds to Si(IV) in zeolite
frameworks (Figure 4).
Figure 3.
Figure 4.
The acidity of the new materials was characterized by NH3-TPD measurements. Desorption peaks
were measured at two temperature regions, bellow and above 300ºC, which refer as low-temperature (LT)
and high-temperature (HT) regions [45]. The nature of the acid sites from the two regions was controversial
over the years. However, it is accepted that HT desorption peaks describe the presence of strong Lewis acid
sites (ie, extra-framework aluminum) while the LT desorption peaks describe ammonia released from
+
protonated NH4 cations, ie Brønsted acid sites. This association is also supported by the IR spectroscopy
results which show the possible formation of protonated ammonia associations over H-zeolites [46]. In
accordance to these assignments and the characterization results obtained from the other described
techniques, the catalytic properties of Nb-zeolites should be closely related to the contribution of both
Brønsted and Lewis acidity (Table 2). The NH3-TPD results indicate that irrespective of the amount of inserted
Nb, Nb-Beta zeolites contain a concentration of the acid sites that is higher than that of the parent zeolite.
However, due to the strength of the bond, ammonia desorbs from NbO-H sites at slightly higher
temperatures than from SiO-H. In fact, these results show that the desorbed amount of ammonia is
proportional with the amount of inserted niobium.

Table 2.
However, even if not being replaced by Nb, a partial extraction of tetrahedrally coordinated Al from
the zeolite framework and its deposition on the zeolite surface as Al−O extra-framework species can not be
neglected. The residual extra-framework aluminum species also exhibit a strong acidity and prevent the
solubilization of the zeolite framework in hot water.
3.2 Catalytic tests
For the HMF synthesis from mono-/di-/polysaccharides (such as fructose, glucose, sucrose,
cellobiose, cellulose, etc.) a large variety of heterogeneous catalysts like zeolites [47], ion-exchange resins
[48], heteropoly acid salts [49] and sulfated and phosphoric acid-treated metal oxides [50, 51] were
investigated. However, besides HMF, a large spectrum of other water soluble side products and unwanted
HMF aldol-condensation by-products, generically called "humins", are usually produced in large amounts.
These humins not only decrease the selectivity to HMF but also block the active sites of the catalysts, and
intermediate calcination steps are necessary to reactivate the catalysts after each cycle. Moreover, the use
of fructose as substrate is not economical because of its low availability and high cost, glucose being
considered a more desirable substrate for the synthesis of HMF.
The transformation of saccharides to HMF takes place in two steps, where the isomerisation of
glucose to fructose requires the presence of Lewis acid sites and the dehydration of fructose to HMF the
presence of Brønsted acids (Scheme 1).
Scheme 1
.

Alumosilicates catalyze the dehydration of fructose in pure water with low selectivity to HMF
(25-40 wt%) at 145 C, a step that parallels the subsequent rehydration of HMF to levulinic and formic

acids, onto the Brønsted acid sites [52].
The comparison of the present results with those reported in the literature is difficult due to
the differences in the experimental conditions. However, the preliminary results obtained at 180C,
in water, clearly show a superior selectivity to HMF of the Nb-Beta comparing to the parent H-Beta
zeolites (Table 3). As higher the amount of Nb inserted in the zeolites framework, as higher was the
HMF selectivity. Also, these results suggest that, indeed, the textural properties of the catalysts and
the concentration of the acid sites are two important parameters of this reaction. Accordingly, using
Beta12-based catalysts, characterised by a monomodal mesoporous texture, with pores of 3.5 nm
(Table 1), and a higher concentration of acid sites (Table 2), resulted lower conversions of glucose
and selectivities to HMF, and higher selectivities to levulinic acid. These results indicate a further
dehydration of the HMF intermediate to levulinic acid. Contrarily, on Beta18-based catalysts,
characterised by a bimodal mesoporous texture (3.9 and 6.5 nm) and lower acid sites concentration,
both the conversation of glucose and selectivities to HMF were slightly higher.
Table 3.
It is already well known that in water, at 150-200C, the hydrolysis of the siloxane bonds (Si−O−Si)
represents the main degradation mechanism. This is opposed to dealumination, which dominates processes
occurring under steaming conditions [53, 54]. However, in line with the recent reports of Sels et al. [55], the
stabilization of zeolites in hot water, occurred with some structural changes. Thus, the XRD pattern of the
spent Nb-Beta displays a new diffraction line centered at 2 of 18, that is most probably related to the
formation of an aluminum amorphous phase (aluminum oxihydroxide AlOOH boehmite, JCPDS 00-008-0096)
onto the surface of the zeolite (Figure 5 exemplifies the XRD pattern of Nb(0.02)-Beta18 catalyst).

Figure 5.
However, no consensus exists where these species result from the aluminum zeolite framework or
not [56]. However, there are reports indicating that in the resulting amorphous phase, Al remains
tetrahedrally coordinated while the micropore volume and concentration of accessible acid sites of zeolite is
dramatically reduced [57]. In our case, ICP-OES measurements indicated practically no leaching of aluminum
(< 0.2 ppm). Also, leaching of niobium was under the detection limit indicating a high stability of the Nb
species deposited in the Beta zeolite matrix in hot water conditions. Moreover, the presence of Nb seems to
stabilize the zeolite structure in these harsh conditions.
The role of the potential leached aluminum species even at a level below 0.2 ppm is important to be
demonstrated. To answer this question, some additional glucose dehydration tests were carried out using
pre-boiled Nb-Beta zeolites in pure water, at 180°C, for 3h. Interestingly enough, the pre-boiled catalysts led
to a higher catalytic efficiency (Table 4) compared to fresh materials (see Table 3). After the catalytic reaction,
no aluminum species were evidenced in the zeolite channels (Figure 6) of the spent catalysts. However,
although the topology of the Beta zeolite has been preserved, the cristallinity of the catalysts was further
decreased.
Table 4.
Figure 6
.
Low HMF yields corresponded to high LA yields (Tables 3 and 4). Therefore it is of the importance to
repress the HMF rehydration (Scheme 1). Also, HMF is assumed to react with sugars and sugar intermediates
to produce the so-called humins. Indeed, TG-DTA of the fresh (Figure 7A) and spent (Figure 7B) catalysts
indicate the formation of high amounts of humins during the reaction. Their presence is also causing the
deactivation of the catalyst.

Figure 7
.
A method to suppress the undesired side reactions consists in a combination of glucose dehydration
under aqueous conditions with an in situ extraction of the HMF in an organic phase [8]. With this aim
Maireles-Torres et al. [58] evaluated the microporous alumosilicates as acid catalysts and H2O/MIBK biphasic
system as solvent, for the dehydration of glucose to HMF. The measured conversions were higher than 85%
but the selectivity to HMF was lower than 30%, at 195C. In another example, Nikolla et al. [59] reported
promising results for glucose dehydration with a bifunctional catalyst combining Sn- and Ti-Beta zeolites with
HCl. Sn-Beta afforded an optimum HMF yield of 57% for a 79% conversion while Ti-Beta an yield of 53% for
a 76% conversion.
Using the same approach, ie a biphasic system consisting of MIBK as extraction solvent and aqueous
NaCl salt, the Nb-Beta zeolites utilized in this study were able to produce HMF in higher selectivity by using
MIBK as extraction solvent and aqueous NaCl salt, at 180C and 12 h (Table 5).
Table 5
As Table 5 shows, the optimum catalyst features for the highest selectivity to HMF (ie, 84.3%
selectivities to MHF for a conversion of glucose of 97.4%) are created in Nb(0.05)-Beta 18 sample. The MIBK
layer of the biphasic system was able to extract continuously the HMF produced in the aqueous phase (Figure
8). A lower concentration of HMF in the aqueous phase limits the rate of side reactions and thereby improves
the HMF yields [60].
Figure 8.

Indeed, the amounts of humins and organic compounds accumulated on the solid catalysts after the
reaction in a biphasic solvent were lower than after the reaction in the aqueous phase, as also confirmed by
the TG-DTA analysis (Figure 9 versus Figure 6).
Figure 9.
The stability of the Nb(0.05)-Beta 18 catalyst was also investigated (Figure 10).
.
Figure 10.
As Figure 10 shows, after four runs, the conversion of glucose and the selectivity to
HMF were unchanged indicating a high stability and very good recyclability of the catalyst.
4. Conclusions
Nb-Beta zeolite catalysts prepared via a two-step method, namely dealuminisation followed
by the insertion of Nb catalyse the glucose dehydration to HMF with selectivities to MHF of 84.3% for a
conversion of glucose of 97.4%, in a biphasic water/MIBK solvent, at 180C and after 12h. The presence
of Nb(V)O-H species where niobium is linked by Nb-OSi bonds to the zeolitic walls affords a high
increase of the acid sites concentration with moderate strengths. Important to notice, it also
stabilises the zeolite, most probably by polarizing the framework. The acidity of the zeolite in
association to the mesoporous texture generates an efficient environment for the dehydration of
glucose to HMF. The catalysts can be recovered by centrifugation and regenerated by simple
calcination, the catalytic efficiency remaining fairly constant during four consecutive runs. The system
presented here is still under investigation by our group for further optimizations.

Acknowledgments
Prof. Simona M. Coman kindly acknowledges UEFISCDI for the financial support (project PN-III-P4-ID-
PCE-2016-0533, Nr. 116/2017).
References
[1] H. Kobayashi, H. Ohta, A. Fukuoka, Catal. Sci. Technol. 2 (2012) 869-883.
[2] R. A. Gross, B. Kalra, Science 297 (2002) 803-807.
[3] J. Q. Bond, D. M. Alonso, D. Wang, R. M. West, J. A. Dumesic, Science 327
1114.
(2010) 1110-
[4] B. H. Shanks, Ind. Eng. Chem. Res. 49 (2010) 10212-10217.
[5] W. Vermeiren, J. P. Gilson, Top. Catal., 52 (2009), 1131-1161.
[6] R. A. Meyers, Handbook of Petroleum Refining Processes, 3rd ed., McGraw Hill, New York, 2004, ISBN 0-
07-139109-6.
[7] T. Werpy, G. Peterson, Top Value Added Chemicals from Biomass, vol. 1, Pacific
Northwestern National Laboratory, 2004.
[8] R.-J. van Putten, J. C. van der Waal, E. de Jong, C. B. Rasrendra, H. J. Heeres, J. G. de Vries,
Chem. Rev. 113 (2013) 1499-1597.
[9] P. A. Jacobs, M. Dusselier, B. F. Sels, Angew. Chem. Int. Ed. 53 (2014) 8621-8626.
[10] M. Moliner, Dalton Trans. 43 (2014) 4197-4208.
[11] R. Otomo, T. Yokoi, J. N. Kondo, T. Tatsumi. App Catal A 470 (2014) 318-326.
[12] Y. Román-Leshkov, M. E. Davis, ACS Catal. 1 (2011) 1566-1580.
[13] L. Bui, H. Luo, W. R. Gunther, Y. Román-Leshkov, Angew. Chem. 125 (2013) 8180 –
8183; Angew. Chem. Int. Ed. 52 (2013) 8022-8025.

[14] E. D’Hondt, S. Van de Vyver, B. F. Sels, P. A. Jacobs, Chem. Commun. (2008) 6011-6012.
[15] D. Kubička, I. Kubičková, J. Čejka, Catal. Rev. 55 (2013) 1-78;
[16] E. Taarning, C. M. Osmundsen, X. Yang, B. Voss, S. I. Andersen, C. H. Christensen, Energy
Environ. Sci. 4 (2011) 793-804.
[17] M. A. Camblor, A. Corma, A. Martínez, J. Pérez-Pariente J. Chem. Soc. Chem. Commun.
(1992) 589-590.
[18] N. Jappar, Q. H. Xia and T. Tatsumi, J. Catal. 180 (1998) 132-141.
[19] W. R. Gunther, Y. R. Wang, Y. W. Ji, V. K. Michaelis, S. T. Hunt, R. G. Griffin, Y. Román-
Leshkov, Nat. Commun. 3 (2012) 1109.
[20] K. M. Jinka, S. C. Lee, S. E. Park, R. V. Jasra, Stud. Surf. Sci. Catal. 174 (2008) 1187-1190.
[21] J. P. Nogier, Y. Millot, P. P. Man, T. Shishido, M. Che, S. Dzwigaj, J. Phys. Chem. C. 113
(2009) 4885-4889.
[22] P. Li, G. Q. Liu, H. H. Wu, Y. M. Liu, J. G. Jiang, P. Wu, J. Phys. Chem. C. 115 (2011)
3663.
[23] C. Hammond, S. Conrad, I. Hermans, Angew. Chem. Int. Ed. 51 (2012) 11736-11739.
[24] P. Wolf, C. Hammond, S. Conrad, I. Hermans, Dalton Trans. 43 (2014) 4514-4519.
[25] J. Dijkmans, D. Gabriëls, M. Dusselier, F. de Clippel, P. Vanelderen, K. Houthoofd, A.
Malfliet, Y. Pontikes, B. F. Sels, Green Chem. 15 (2013) 2777-2785.
[26] B. Tang, W. Dai, X. Sun, G. Wu, L. Li, N. Guan, M. Hunger, Chin. J. Catal. 36 (2015) 801-
805.
[27] A. Corma, F. X. Llabrés i Xamena, C. Prestipino, M. Renz, S. Valencia, J. Phys. Chem. C.
113 (2009) 11306-11315.
[28] I. Nowak, M. Ziolek, Chem. Rev. 99 (1999) 3603-3624.
[29] M. Ziolek, I. Nowak, H. G. Karge, Stud. Surf. Sci. Catal., 94 (1995) 270-277.
[30] I. C. L. Barros, V. S. Braga, D. S. Pinto, J. L. de Macedo, G. N. R. Fihlo, J. A. Dias, S. C.
L. Dias, Microp. Mesop. Mater. 109 (2008) 485-493.

[31] S. M. Coman, M.N. Verziu, A. Tirsoaga, B. Jurca, C. Teodorescu, V. Kuncser, V.I.
Parvulescu, G. Scholz, E. Kemnitz, ACS Catal. 5 (2015) 3013-3026.
[32] M. Verziu, M. Serano, B. Jurca, V. I. Parvulescu, S. M. Coman, G. Scholz, E. Kemnitz
Catal. Today (2017) http://dx.doi.org/10.1016/j.cattod.2017.02.051.
[33] I. Podolean, F. Anita, H. Garcia, V. I. Parvulescu, S. M. Coman, Catal. Today 279 (2017) 45-55.
[34] N. Candu, F. Anita, I. Podolean, B. Cojocaru, V. I. Parvulescu, S. M. Coman, Green Process. Synth. 6 (2017)
255-264.
[35] M. El Fergani, N. Candu, S. M. Coman, V. I. Parvulescu, Molecules 22(12) (2017) 2218.
[36] F. Tielens, T. Shishido and S. Dzwigaj, J. Phys. Chem. C, 114 (2010) 3140-3147.
[37] C. M. Teodorescu, J. M. Esteva, R. C. Karnatak, A. El Afif, Nucl. Instrum. Meth. Phys.
Res. A 345 (1994) 141-147.
[38] M. Moliner, ISRN Mater. Sci. (2012) Article ID 789525.
[39] M. Ziolek, I. Nowak, B. Kilos, I. Sobczak, P. Decyk, M. Trejda, J. C. Volta, J. Phys. Chem. Solids 65 (2004)
571-581.
[40] T. Blasco,; M. A. Camblor, A. Corma, P. Esteve, J. M. Guil, A. Martinez, J. A. Pedrigon-Melon, S. Valencia,
J. Phys. Chem. B. 102 (1998) 75-88.
[41] G. Jouve, C. Séverac, S. Cantacuzène, Thin Solid Films 287 (1996) 146-153.
[42] B. Kościelska, A. Winiarski, J.Non-Cryst. Solids 354 (2008) 4349-4353.
[43] G. M. Ingo, C. Riccucci, G. Bultrini, S. Dirè, G. Chiozzini, J. Therm. Anal. Calorim. 66,
(2001) 37-46.
[44] E. Paparazzo, Surf. Interf. Anal. 12 (1988) 115-118.
[45] H. Sato, Catal. Rev. - Sci. Eng. 39 (1997) 395-424.
[46] F. Lonyi, J. Valyon, Microp. Mesop. Mat. 47 (2001) 293-301.
[47] K.-I. Shimizu, R. Uozumi, A. Satsuma, Catal. Commun. 10 (2009) 1849-1853.
[48] X. Qi, M. Watanabe, T. M. Aida, R. L. Smith, Ind. Eng. Chem. Res. 47 (2008) 9234-9239.

[49] C. Fan, H. Guan, H. Zhang, J. Wang, S. Wang, X. Wang, Biomass Bioenerg. 35 (2011)
2659-2665.
[50] H. Yan, Y. Yang, D. Tong, X. Xiang, C. Hu, Catal. Commun. 10 (2009) 1558-1563.
[51] F. Yang, Q. Liu, X. Bai and Y. Du, Bioresource Technol. 102 (2011) 3424-3429.
[52] R. L. De Souze, H. Tu, F. Rataboul, N. Essayem, Challenges 3 (2012) 212-232.
[53] W. Lutz, Cryst. Res. Technol. 25 (1990) 921-926.
[54] C. S. Triantafillidis, A. G. Vlessidis, L. Nalbandian, N. P. Evmiridis, Microp. Mesop. Mater.
47 (2001) 369-388.
[55] T. Ennaert, J. Geboers, E. Gobechiya, Ch. M. Courtin, M. Kurttepeli, K. Houthoofd, Ch.
E.A. Kirschhock, P. C.M.M. Magusin, S. Bals, P. A. Jacobs, B. F. Sels, ACS Catal. 5 (2015)
754-768.
[56] P. Kortunov, S. Vasenkov, J. Kꢀrger, R. Valiullin, P. Gottschalk, M. Fꢁ Elía, M. Perez, M.
Stꢂcker, B. Drescher, G. McElhiney, C. Berger, R. Glꢀser, J. Weitkamp, J. Am. Chem. Soc.
127 (2005) 13055-13059.
[57] R. M. Ravenelle, F. Schüβler, A. D’Amico, N. Danilina, J. A. van Bokhoven, J. A. Lercher, C. W. Jones, C.
Sievers, J. Phys. Chem. C 114 (46) (2010) 19582–19595.
[58] M. Moreno-Recio, I. Jiménez-Morales, P. L. Arias, J. Santamaría-González, P. Maireles-
Torres, ChemistrySelect. 2 (2017) 2444-2451.
[59] E. Nikolla, Y. Román-Leshkov, M. Moliner, M. E. Davis, ACS Catal. 1 (2011) 408-410.
[60] E. I. Gürbüz, S. G. Wettstein, J. A. Dumesic, ChemSusChem. 5 (2012) 383-387.

Figure caption:
Figure 1. Adsorption-desorption of liquid nitrogen at 77 K of H-Beta18, Nb(0.02)-Beta18, H-
Beta12 and Nb(0.02)-Beta12 samples
Figure 2. XP Spectra of the Nb(3d) level for the Nb(0.02)- and Nb(0.05)-Beta 18 samples
Figure 3. XP Spectra of the O(1s) level for the Nb(0.02)- and Nb(0.05)-Beta 18 samples
Figure 4. XP Spectra of the Si(2p) level for the Nb(0.02)- and Nb(0.05)-Beta 18 samples
Figure 5. XRD patterns of the Nb(0.02)-Beta 18 samples before and after catalytic reaction
Figure 6. X-ray diffraction patterns of fresh Nb(0.02)-Beta12 and pre-boiled Nb(0.02)-Beta12
(before and after reaction)
Figure 7. TG-DTA plots of fresh (A) and spent (B) catalyst
Figure 8. Catalytic performances of Nb(0.05)-Beta18 catalyst in water and biphasic water/MIBK
solvent (Reaction conditions: 0.03 g catalyst, 0.18 g (1 mmole) glucose, 5 mL water or biphasic
solvent (3.5 mL aqueous solution of 20% NaCl and 1.5 mL MIBK), 180°C, 24 or 12 h)
Figure 9. TG-DTA plot of Nb(0.05)-Beta18 catalyst used in the synthesis of HMF in biphasic
water/MIBK solvent
Figure 10. Recycling of Nb(0.05)-Beta18 catalyst in water and biphasic water/MIBK solvent
(Reaction conditions: 0.03 g catalyst, 0.18 g (1 mmole) glucose, biphasic solvent (3.5 mL aqueous
solution of 20% NaCl and 1.5 mL MIBK), 180°C, 12 h)

Figure 1.

Figure 2.
Figure 3.

Figure 4.
Figure 5.

Figure 6
Fig 7
Figure 7.

Figure 8
Figure 9

Figure 10.

Scheme 1. Schematically representation of the HMF synthesis from glucose

Tables
Table 1. Textural properties of the incorporated Nb in molecular sieves samples
Catalyst
H-Beta18
SBET, m2/g
536
Vmicropores, cm3/g (t-plot)
Pore size (nm)
3.5
0.182
0.114
0.129
0.104
Nb(0.02)- Beta18
H-Beta12
438
3.5
458
3.9 and 6.5
3.9 and 6.4
Nb(0.02)- Beta12
391

Table 2. Acid sites concentration determined through NH3-TPD
Sample/
Tmax (C)
Acid sites concentration (mmols NH3/g catalyst)
< 300
0.066
0.130
0.132
0.064
0.124
0.126
> 300
Total
0.070
0.130
0.132
0.071
0.124
0.126
H-Beta 12
0.004
Nb(0.02)-Beta 12
Nb(0.05)-Beta 12
H-Beta 18
-
-
0.007
Nb(0.02)-Beta 18
Nb(0.05)-Beta 18
-
-