10.1002/cctc.201700301
ChemCatChem
COMMUNICATION
been observed in other vanadium-catalyzed reactions, such as
benzene hydroxylation.[10] Similar to those previous studies, the
maximum in catalytic performance of our 5V/SiO2 catalyst
correlates well with its maximum in surface monomeric
VO4/polymeric VOx species at 5% loading. Only trace amounts
of butanol (<0.1%) can be detected on the vanadium catalysts,
most likely due to its rapid dehydration to butene on the acid
sites. The ratio of the three isomers is different on these three
vanadia catalysts and the values differ significantly from the
equilibrium distribution, suggesting that they are the primary
products and secondary isomerization of butene does not
occur.[11] It is interesting to note that more than 10% selectivity to
acetaldehyde and diacetyl (from double dehydrogenation of
2,3BDO) is observed on these vanadia catalysts while it is not
observed on the basic MgO catalyst and only trace amounts are
observed on strong acidic catalysts, Al2O3 and P/ZSM5(30). The
acetaldehyde is likely to have been produced from the C-C
cleavage of acetoin on the weak acid sites, since the MgO
catalyst discussed previously produces acetoin instead, which
can be ascribed to the absence of acid sites.
temperature has a significant effect the selectivity to butene.
100% conversion is reached at 350 °C and above, while butene
selectivity increases from 24.6% to 45.2% as the reaction
temperature is increased from 250 °C to 500 °C.
To verify the transfer hydrogenation mechanism, in-situ
diffuse reflectance infrared fourier transform spectroscopy
(DRIFTS) was conducted on the 5V/SiO2 catalyst with
chemisorbed 2,3BDO (Fig. 3). When 2,3-butanediol is adsorbed
onto the catalyst, the C-O stretch bands (1292 cm-1), O-H stretch
band (3200-3600 cm-1), C-H stretch bands (2902, 2942, 2987
cm-1) of 2,3BDO is readily observed at 30 °C. At 100 °C, the
C=O stretch band (1700 cm-1) of MEK is observed. At 300-500
°C, alkene bands such as the =C-H stretch at 3217 cm-1, =C-H
bend (920 cm-1) and the C=C stretch at 1640 cm-1 indicate the
presence of butene.[15]
This combination of dehydration (MEK, MPA),
dehydrogenation (acetaldehyde, acetoin and diacetyl) and
hydrogenation (2-butanol) products on the vanadium catalysts in
absence of external
H
source suggests that transfer
hydrogenation between 2,3-butanediol and MEK occurs on the
vanadium catalysts.[12] MEK gains two hydrogen atoms to form
2-butanol, which then dehydrates to form butene, while 2,3-
butanediol loses two hydrogen atoms to form acetoin, which can
either lose another two hydrogen atoms to form diacetyl or
undergo C-C cleavage to form two molecules of acetaldehyde.
Transfer hydrogenation reactions involving different ketones
(hydrogen acceptor) and alcohols (hydrogen donor) in both
liquid and gas phases have been previously demonstrated over
various catalysts, even between MEK and 2-propanol.[12a,13]
Since 2,3BDO is a vicinal diol (hydrogen donor) that dehydrates
to form MEK (a hydrogen acceptor ketone), the use of
secondary hydrogen donors for MEK is not required, allowing
2,3BDO to be directly deoxydehydrated to butene on the VOx
sites without external reductants. The high yield of MEK
suggests that this transfer hydrogenation is the rate-limiting step
in the reaction pathway. Furthermore, no butane was observed,
which suggests that transfer hydrogenation in this reaction
pathway is preferential to C=O hydrogenation instead of C=C
hydrogenation. The low ratio of acetaldehyde to butene, for
example in the 5V/SiO2 catalyst (13.7% acetaldehyde versus
37% butene), suggests that acetaldehyde could undergo further
reactions. Acetaldehyde could undergo steam reforming by
reacting with water molecules produced from the two
dehydration reactions. Indeed, previous studies of other silica-
supported metal oxide have shown that mainly hydrogen and
carbon dioxide, with trace amounts of CH4, CO and propylene,
can be produced.[1b, 14] This is supported by the detection of CO2
in the outlet stream using the TCD detector of the GC, as well as
0.5% and 1.7% of CH4 at higher reaction temperatures, 450 °C
and 500 °C, respectively for the 5V/SiO2 catalyst. The hydrogen
produced from acetaldehyde steam reforming could have further
contributed to the hydrogenation of the MEK intermediate.
Figure 3. In-situ FTIR (DRIFTS) results of 2,3BDO adsorbed on 5V/SiO2.
Table 2. Major products from intermediates testing at 400 °C
Catalyst
5V/SiO2
5V/SiO2
5V/SiO2
5Pt/SiO2
Feed
Major Products Yield (%)
2,3BDO + H2
MEK + H2
2-butanol
MEK +H2
37% butene, 41% MEK, 14% acetaldehyde
3% butene, 0.6% 2-butanol
100 % butene
19% 2-butanol, 11% butane, 1% butene
Additionally, various reaction intermediates were used as
feedstock on the 5V/SiO2 catalyst and on a 5Pt/SiO2 standard
hydrogenation catalyst (Table 2) to verify the reaction pathways.
When the reaction was performed under a gaseous mixture of
N2 and H2 with a molar ratio of 1:1 using the same gas flow rate
(40ml/min), the 5V/SiO2 catalyst produces the same product
distribution as when the reaction was carried out in pure N2.
Furthermore, when MEK and H2 are used as feedstock, a very
low conversion of MEK to butene and 2-butanol is observed,
which can be easily explained since the activation of H2 on
5V/SiO2 is very weak. When the 5V/SiO2 catalyst was tested
using 2-butanol as feedstock, 100% selectivity to butene (with
Further investigation of the effect of temperature on the
high-performing 5V/SiO2 catalyst (Fig. 2) indicates that
This article is protected by copyright. All rights reserved.