Full Papers
been measured by using ammonia TPD (Figure S1). The acid
strength recorded for b-zeolites is quite low compared to that
of Amberlyst-70 or sulfated zirconia, which even evidences
some super acidity. With the purpose to demonstrate these dif-
ferences, two sets of experiments were performed. In a first
assay, a reaction medium with iso-propyl xyloside as the main
product was prepared by reacting xylose with 2-propanol in
the presence of tungstated zirconia. This material provides the
highest amount of 2-propyl xyloside among those tested in
this work. The reaction medium was then filtered and treated
in the presence of sulfated zirconia (ZrO2-SO42À) and a H-b zeo-
lite (Si/Al=19.0) as catalysts (Figure S3). In a second assay,
commercially available methyl xyloside was treated in metha-
nol in the presence of the same catalysts (Figure S4). The reac-
tion results obtained in these experiments provide evidence of
the superior performance of sulfated zirconia compared to
that of H-b zeolite (Si/Al=19.0) in the transformation of alkyl
xylosides in an alcohol medium. The higher production of fur-
fural and the rapid conversion of xylosides obtained in the
a xylose “reservoir” that is then consumed to form furfural.
Nevertheless, the use of this xylose reserve is completely differ-
ent depending on the catalyst used to promote this transfor-
mation. As described previously, Lewis acid containing materi-
als can drive the direct dehydration of xylose to furfural, in
which xylulose is a key intermediate. Although the etherifica-
tion of the pentose is the main reaction that takes place
during the early stages of the transformation, iso-propyl xylo-
side seems to be hydrolyzed back to produce xylose insofar as
the sugar is transformed into furfural. Conversely, Brønsted
acid materials, which are unable to promote xylose isomeriza-
tion to xylulose, first convert the pentose sugar into the alkyl
xyloside, which is then transformed directly into furfural to
yield lower amounts of furfural. As the tested zeolites bear
both types of acid sites, it seems that these materials can drive
the conversion of xylose into furfural through both reaction
pathways.
Although all the tested catalysts display a similar catalytic
activity, especially at low reaction temperatures, the results ob-
tained from the kinetic analysis of the reaction allow us to con-
clude that the pure Brønsted acids display a lower catalytic
performance in the transformation of xylose into furfural.
These samples yield high quantities of byproducts even at low
reaction temperatures, and the furfural-consuming side reac-
tions are heavily promoted in presence of these materials, and
thus yield low concentrations of the desired product during all
the considered reaction times. Conversely, b-zeolites display
a better catalytic performance in terms of furfural production.
These materials are able to promote the transformation of
most of the starting xylose into furfural in quite short reaction
times and keep the amount of produced byproducts low. The
reason for the better catalytic performance of these materials
seems to be the right combination of Lewis- and Brønsted-
type acid sites to promote the desired reaction. This is more
evident in the case of the H-b zeolite sample with a higher sili-
con to aluminum ratio (Si/Al=19.0), which is able to transform
65% of the starting xylose into furfural in very short reaction
times (~2 h) at 1708C with a minimal formation of side prod-
ucts compared to the other tested acid catalysts. Differences
between these materials could be ascribed to the more hydro-
phobic nature of the sample with a lower Al content, a feature
that has been described to exert a highly positive influence on
the promotion of dehydration reactions.[42]
2À
presence of the Brønsted-type ZrO2-SO4 compared to that in
the presence of H-b zeolite supports our conclusions about
the main reaction pathways that depend on the nature and
strength of the acid catalyst.
The rest of the chemical transformations have been lumped
into three different reactions, those related to the consump-
tion of xylose (Reaction 5, Scheme 1), furfural-degrading reac-
tions (Reaction 6, Scheme 1), and iso-propyl xyloside consum-
ing transformations (Reaction 7, Scheme 1). In the first group
of transformations, xylose can be consumed through a wide
variety of transformations to yield random polymers known as
humins. This reaction seems to take place at a similar rate in
the presence of all the tested catalysts. The second group of
transformations include ketal formation by the acetalization of
furfural with the alcohol solvent,[36] random polymerization to
humin-like polymers, and furfural reduction through H-transfer
reactions to produce furfuryl alcohol and, subsequently, levu-
linic acid and alkyl levulinates.[33] As a result of the high reactiv-
ity of furfural and the different possibilities of this chemical to
react, no differences can be established between the different
tested materials. However, both the degradation of xylose and
furfural seem to occur at a substantially lower reaction rates
than those reported for homogeneous mineral acids in water
medium,[11] which could show the beneficial effect of the use
of 2-propanol as the reaction medium. Finally, the third group
of reactions consists of the degradation of iso-propyl xyloside,
although this is unlikely to be produced under the conditions
studied, which is inferred from the negligible values calculated
for the kinetic constant assigned to this transformation during
the fitting of the proposed kinetic model to the experimental
data. In this way, the formation of iso-propyl xylose during the
early stages of the reaction can be considered not only as an
alternative route to transform xylose into furfural but also as
an effective way to protect the xylose from undergoing an un-
desired transformation towards the formation of byproducts.
In other words, the etherification of xylose with the alcohol
solvent leads to the protection of the starting feedstock, which
prevents its consumption through side reactions and yields
Conclusions
The dehydration of xylose to furfural in the presence of hetero-
geneous acid catalysts can provide high furfural yields if the re-
action solvent and the physicochemical properties of the solid
acids used as catalysts are selected properly. With regard to
the solvent, the length of the alkyl chain in alcohols influences
the extent of side reactions strongly and thus 2-propanol per-
forms better than methanol or ethanol. As for the influence of
the physicochemical properties of solid acids, both the acid ca-
pacity and the acid strength play an important role in the
transformation considered. However, it is the nature of the
acid sites (Lewis/Brønsted) that exerts a significant influence
ChemCatChem 2016, 8, 2089 – 2099
2097
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim