1
82
S. Sato et al. / Journal of Molecular Catalysis A: Chemical 221 (2004) 177–183
in the dehydration of 1,3-butanediol, does not proceed over
CeO2, andthatmonoalcoholssuchas1-butanoland2-butanol
were not dehydrated to butenes over CeO2 at low tempera-
the intrinsic catalytic activity of CeO2 for the dehydration of
1,3-butanediol or not.
Then, we examined another TPR experiment for samples
with different treatment (Fig. 3). Contact of water with the re-
duced CeO2 surface regenerates the reduction peak at around
◦
tures <375 C. 1-Butene is a major product in the dehydration
◦
of 2-butanol over CeO2 at ≥375 C. It is also known that lan-
◦ 3+
500 C, and it induces the oxidation of Ce cations into
thanide oxides including CeO2 catalyze Hoffman elimination
to produce ␣-olefin [23].
4
+
Ce . This is consistent with the reports of Otsuka et al.
[24,25]: a reduced CeO2 powder is oxidized to CeO2 by wa-
Diols are more reactive than monoalcohols, and they are
dehydrated to either unsaturated alcohols or saturated ketones
◦
ter even at 300 C, and the oxidation by water occurs not
(
Tables 2–4). 1,3-Diols are basically reactive in the dehydra-
only on the surface but also in the bulk. The reduction peak
at <600 C would be concerned with surface CeO2 because
◦
tion to produce unsaturated alcohols. 1,4-Butanediol has spe-
cific reactivity: CeO2 catalyzes dehydration of 1,4-butanediol
to 3-buten-1-ol, which is not further dehydrated to 1,3-
reduction at high temperatures reduces both the reduction
peak (Fig. 3) and the specific surface area (Table 5). Since the
◦
3+
4+
butadiene, at 425 C. Polyols such as 1,2- and 2,3-butanediol,
surface Ce is readily oxidized to Ce by H2O produced
in the dehydration, it is reasonable that the CeO2 surface is
maintained at high oxidation state in the working state of the
dehydration.
and 1,2,3-propanetriol are dominantly dehydrated into ke-
tone through the enol form. Butanone would be directly
formed via 2-buten-2-ol in the dehydration of 2,3-butanediol:
3
-buten-2-ol is not an intermediate because it is observed as
a stable product in the reaction of 1,3-butanediol over CeO2
4.3. Reaction model in the dehydration of 1,3-diol
(Table 2).
over CeO2
CeO2 works with different catalytic functions depending
on the temperature as well as on the reactant (Tables 1–4). It
can be summarized that CeO2 catalyzes dehydration of 1,3-
ZrO2 and CeO2 have different catalytic features in dehy-
dration of 4-methylpentan-1-ol, while ZrO2 is much more
acidic and basic than CeO2 [26]. ZrO2 has a fluorite (CaF2)
structure similar to CeO2, and it also has ability for the dehy-
dration of 1,3-butanediol, while 3-buten-1-ol is formed in ad-
dition to 3-buten-2-ol, trans-, and cis-2-buten-1-ol (Table 2).
The product distribution over ZrO2 is explained by an-
other mechanism that an OH anion is initially eliminated;
acidic sites of ZrO2 probably activate the OH group of 1,3-
butanediol.
◦
diols to unsaturated alcohols at <400 C and dehydration of
◦
1
,4-butanediol to 3-buten-1-ol at >400 C. It also catalyzes
dehydration of 1,2- and 2,3-diols to aldehyde and ketone.
4
.2. Activation of alcohols over CeO2 and its
redox property
We suggested that the catalytic activity of CeO2 in
the activation of alcohols would be correlated with the
In contrast, CeO2 produces only 3-buten-2-ol and
2-buten-1-ol without producing 3-buten-1-ol. If an OH
group on the 3-position is eliminated first as an OH anion,
3-buten-1-ol has to be formed together with 2-buten-1-ol.
Thus, the product distribution cannot be explained by the
acid-catalyzed mechanism mentioned above. We speculate
that redox nature of CeO2 initially activates hydrogen
in the 2-position of 1,3-butanediol [13]: the dehydration
of 1,3-diols over CeO2 follows a radical mechanism; a
hydrogen atom located on the 2-position is abstracted first,
followed by the elimination of an OH radical (Fig. 4). Then,
redox feature [4,8,9]. In the alkylation of phenol with
◦
1
-propanol catalyzed by CeO2–MgO at 475 C, it has
been suggested that redox sites, not acid sites, activate the
alkylating reagent of 1-propanol [4]. A hydrogen atom is
eliminated from 1-propanol to produce radical species such
as 1-hydroxypropyl radical rather than propoxy radical. The
hydrogen atom can reduce Ce4 into Ce : Ce cation
oxidatively eliminates hydrogen from 1-propanol to form
proton. The 1-hydroxypropyl radical attacks the electron-rich
ortho-positions of phenol adsorbed perpendicularly on the
weak basic sites in the CeO2–MgO because the ␣-position
of 1-hydroxypropyl radical is slightly positive. A hydroxyl
radical is readily eliminated from the alkylated intermediate
to produce 2-n-propylphenol and water, together with the
+
3+
4+
it is reasonable that Ce4 is reducible by the eliminated
+
hydrogen radical in the catalytic cycle.
Here, we have a significant question: even monoalcohols,
instead of diols, may be dehydrated into olefin according to
the radical mechanism (Fig. 4). Actually, monoalcohols in-
cluding unsaturated alcohol are less reactive for dehydration
(Table 1). Dehydration of 1,3-diol proceeds at temperature as
oxidation of Ce3 to Ce
+
4+
.
In the TPR profile (Fig. 2), CeO2 is severely reduced at
◦
>
600 C. Both the average oxidation number of Ce and the
◦
catalytic activity of CeO2 in the dehydration of 1,3-diols de-
low as 325 C, while much higher temperatures are needed
◦
creased with raising the reduction temperature above 600 C
for the activation on monoalcohols. Since 1,3-diols are more
stabilized and activated than monoalcohols on the CeO2
surface, both of the OH groups in 1,3-diol could be inter-
acted with the surface of CeO2. We can correct the reaction
image of the dehydration of 1,3-diols (Fig. 5). It is remark-
able that 2-methyl-1,3-propanediol is less reactive than
(Table 5). Because SA also decreases with treatment temper-
ature, the reduced CeO2 seems to have the intrinsic catalytic
activity as active as the non-reduced one: the catalytic activ-
ity could be proportional to the SA. Unexpectedly, we still
have a question whether the oxidation state of CeO2 affects