Full Papers
the micropore volume (up to ca. 25%) was observed for the
Sn/ZSM-5 zeolites, which is possibly due to the larger size of
Sn cations compared with K cations. In addition, the relatively
intense acidic conditions (pHꢀ2–3) of the ion-exchange aque-
ous suspension that were applied for the preparation of Sn/
zeolites led to a slight dealumination of the zeolite (Table S1),
which may further induce textural changes and partial destruc-
tion of the microporous network. This could also contribute to
the higher decrease in micropore volume for the Sn/zeolites,
and could also explain the observed substantial increase in the
meso/macropore volume, leading to a total pore volume
higher than that of the parent zeolite (Table 1). The textural
properties of the five metal oxides (TiO2, ZrO2/TiO2, g-Al2O3,
SiO2/Al2O3, MgO) tested were typical for these type of materi-
als. The respective N2 isotherms and pore size distribution
curves are shown in Figures S3 and S4, whereas the surface
area and pore volume data are given in Table 1. The tested ma-
terials are mostly meso/macroporous materials with low/mod-
erate surface area, at least compared with that of the ZSM-5
zeolite, but with higher total pore volume owing to the pres-
ence of the meso/macropores.
Screening of homogeneous and heterogeneous catalysts
The homogeneous and heterogeneous catalysts studied in the
present work for the glucose to fructose isomerization reaction
were selected based on their Lewis acid or basic properties. A
primary screening of all catalysts was performed at standard
conditions: 808C for 30 min and at 1008C for 60 min, at con-
stant glucose concentration. The results of the screening tests
are presented in Table 2.
All catalysts tested, except for the proton form of the ZSM-5
zeolite, catalyzed the isomerization of d-glucose into d-fruc-
tose in aqueous solution to different extents, whereas isomeri-
zation did not take place in the absence of a catalyst, that is,
in neat water.
In the case of homogeneous catalysis, NaAlO2 demonstrated
the highest glucose isomerization activity, reaching a fructose
yield above 24%, with glucose conversion and fructose selec-
tivity values of 87% and 29%, respectively. Conversely, when
NaOH was used, fructose selectivity and yield were very low,
less than 2%, despite the 99% glucose conversion, and the so-
lution obtained a dark-brown color.[28] However, in both the
above cases, the relatively high alkali base concentration and/
or intense reaction conditions (temperature, time) induced
side reactions that pushed glucose conversion over the ther-
modynamic equilibrium limit, which corresponds to ꢀ54%
conversion at 808C.[11,12] Increasing the concentration of the
alkali base catalyst and consequently the basicity of the solu-
tion, aldose–ketose isomerization is accompanied by sugar
degradation reactions, which lead to the production of organic
acids, condensation products, and humins.[15,28,34] Therefore,
mild reaction conditions and low alkali concentration are re-
quired to control these side reactions and enhance isomeriza-
tion to fructose. Indeed, when NaOH was added to the glucose
aqueous solution at the lower concentration of 0.2 wt.%, a fruc-
tose yield of 32.2%, corresponding to glucose conversion of
67% and fructose selectivity of 48.0%, was obtained. These re-
sults are in accordance with previously published detailed
works on the use of water-soluble alkali bases for glucose iso-
merization.[38–40] It has been proven that under highly alkaline
conditions and increased base concentration, the enediol inter-
mediate undergoes further non-reversible degradation reac-
tions with subsequent color formation. The rate and extent of
these side reactions are controlled mainly by the type of base
cation, base concentration, and temperature.
The catalytic materials investigated in this study were select-
ed by considering that the isomerization of glucose to fructose
is catalyzed either by Lewis acid sites or basic sites. Thus, most
of the catalysts possess Lewis acid sites, except MgO and K-ex-
changed ZSM-5 zeolites, which exhibit basic properties. The
acidic properties of the parent H-ZSM-5 zeolite were modified
by ion-exchange. As expected, after replacement of the pro-
tons by the respective metals, the Brønsted acid sites (BAS) are
reduced, with the extent of the decrease depending on the
type of metal (Sn or K). It can be seen from Table 1 that potas-
sium (K) is more drastic in lowering the BAS of H-ZSM-5, possi-
bly owing to its small size and monovalent nature, which
allows it to more easily exchange the counter-balance protons
in the medium pore ZSM-5 zeolite compared with the bulkier
and bivalent tin (Sn). With regard to Lewis acid sites (LAS), the
parent H-ZSM-5 zeolite contains a significant amount of such
sites, which are attributed to the extra-framework aluminum
species formed upon calcination of this high Al-containing
ZSM-5 sample to convert it to its proton form. The exchange
of protons by Sn cations had a minor effect on the number of
Lewis sites, offering a small increase compared with those of
the parent H-ZSM-5.[36,37] On the other hand, the Lewis acid
sites in the K-exchanged zeolites were significantly reduced,
similar to the Brønsted sites, revealing the ability of this alkali
metal to mask efficiently both types of acid sites. The four
metal oxides examined (TiO2, ZrO2/TiO2, g-Al2O3, SiO2/Al2O3) ex-
hibited only Lewis acid sites of relatively low/medium number
whereas MgO showed negligible acidity.
Zeolites are used as catalysts for a variety of reactions owing
to their unique microporous structure and their active acid
(Lewis and Brønsted) sites. In this work, H-ZSM-5 was em-
ployed in its protonated form and, as expected, did not induce
glucose isomerization, despite its considerable amount of
Lewis sites (Table 1).[41] Nonetheless, zeolites can be modified
with alkali metals to tune their acid/base properties, or with
metals such as Sn, to increase the Lewis acidity.[42] Therefore,
H-ZSM-5 was ion-exchanged with Sn and K at 50 and 100%
nominal substitution of protons (Table S1). In the case of the
Sn-exchanged zeolites, tin is not incorporated in the zeolite
framework and despite the increase in the proportion of Lewis
over Brønsted acid sites (Table 1), the catalysts did not present
With regard to the basic properties of the catalysts, the tem-
perature-programmed CO2 desorption (TPD-CO2) measure-
ments showed negligible amounts of basic sites for the K-ex-
changed zeolites, TiO2 and ZrO2/TiO2. On the other hand, both
MgO catalysts possessed considerable amounts of basic sites,
as was expected.
ChemCatChem 2016, 8, 1100 – 1110
1102
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim