T. Tabanelli et al.
Applied Catalysis A, General 619 (2021) 118139
ammonia/oxygen 5/13/13, mol%); the results are shown in Figs. 2 and
S2.
decrease in the number of moles of reactants which may interact per
unit time with the catalyst surface, whereas the number of oxygen
moles increases. This basically corresponds to a surface saturation,
observed at higher ethanol concentration.
At 370 ◦C (Fig. 2), reactant conversion was lower than 15 %. As for
the distribution of products in function of the W/F ratio, our results
indicate that the only primary products were ethylene and acetaldehyde;
the selectivity to ethylene then underwent only a minor decrease when
the W/F ratio was increased, whereas that to acetaldehyde rapidly
declined, with a corresponding increase in selectivity to acetonitrile, CO
+ CO2, HCN, and some undetected heavier compounds as well; however,
the selectivity to the latter products reached a maximum value at 0.2 g s
mlꢀ 1 W/F ratio and then declined.
b) The selectivity to acetonitrile showed either a maximum value at an
intermediate temperature or continuously decreasing values;
generally speaking, the decrease corresponded to an increased for-
mation of CO + CO2, whereas the presence of a maximum value was
due to a relatively higher formation of undetected compounds
(“heavy” compounds) at a lower temperature. The best selectivity
was seen in the lowest concentrations of ethanol in the feed; the
greater difference was seen when the concentration of ethanol was
increased from 5% to 7.5 %, and this was mainly due to the greater
formation of heavy compounds.
These experiments confirm the kinetic relationship between acetal-
dehyde and acetonitrile, suggesting that this mechanism occurs by the
reaction of aldehyde with ammonia and the generation of the ethani-
mine intermediate compound. Furthermore, our data clearly highlight
that the catalyst acidity is detrimental for catalytic behavior, since
c) The selectivity to acetaldehyde declined when the temperature was
raised; the greatest selectivity was shown with tests conducted at the
highest ethanol concentration. This may occur because, under con-
ditions of surface saturation, the reactions involving acetaldehyde
are slower than on a “clean” surface. Moreover, under these condi-
tions acetaldehyde was less efficiently transformed into acetonitrile,
and underwent side reactions to form heavier compounds.
◦
ethylene formation was already significant at 370 C. In this case, the
formation of N2, deriving from ammonia combustion, was negligible,
because of the low temperature used.
◦
When our experiments were conducted at 440 C (Figure S2), the
same reaction network was inferred, with acetaldehyde and ethylene as
the only primary products. It is worth noting that the initial selectivity to
CO + CO2 (i.e. the selectivity extrapolated to nil conversion) was close to
zero; this means that ethanol did not undergo a direct reaction of
combustion even at such a relatively high temperature. Once again, the
rapid decline in acetaldehyde selectivity corresponds to the increased
selectivity to acetonitrile, HCN, CO + CO2 and heavy compounds.
One major difference with respect to experiments conducted at a
lower temperature is that at a W/F ratio above 0.1–0.2 g s mlꢀ 1, the
selectivity to both acetonitrile and heavy compounds decreased.
Therefore, at high temperature, acetonitrile is not a stable compound
and undergoes consecutive oxidation to COx.
d) The selectivity to ethylene was not much affected by ethanol partial
pressure; this indicates that ethanol dehydration to ethylene
occurred on sites which were different from those responsible for
ethanol (oxi)dehydrogenation into acetaldehyde. In these sites also,
however, a saturation effect was observed when ethanol concentra-
tion was raised, because the overall rate of ethylene formation
reached a plateau.
e) In all experiments, the selectivity to CO + CO2 increased in parallel
with the temperature rise; however, the variation seen differed
depending on the ethanol partial pressure. In fact, in experiments
conducted using 2 and 5% ethanol in the feed, the selectivity to CO +
CO2 was relatively low at low temperature, but then the rise
observed with an increase in temperature was very rapid.
Conversely, in experiments conducted using 7.5 and 13 % ethanol in
the feed, the selectivity to CO + CO2 was slightly higher at a lower
temperature, compared to experiments at lower ethanol concentra-
tion, but then the increase seen in parallel with the temperature in-
crease was not so significant. As a result, at high temperature and
ethanol concentration, the selectivity to CO + CO2 was much lower
than that observed under leaner ethanol concentration. This effect
can be explained by taking into account the surface saturation due to
the adsorbed C2 molecules. A saturation implies a lower availability
of oxidizing sites (in other words, it can be considered a surface
“over-reduction”), which are supposed to be responsible for the
combustion to carbon oxides. Therefore, under these “saturated
surface” conditions, the catalyst is less selective to combustion
compounds, but more selective to heavier, condensation compounds.
f) The effect of ethanol concentration on the selectivity to N2 was sig-
nificant. The greater the ethanol concentration, the lower the
amount of ammonia which was oxidized into molecular nitrogen.
This is not just attributable to the fact that the reaction between the
intermediately formed acetaldehyde and ammonia was quicker
compared to the parallel reaction of ammonia combustion when
there was a greater concentration of adsorbed acetaldehyde. Indeed,
an important contribution may derive, once again, from V over-
reduction occurring under surface saturation conditions, which
made the combustion of ammonia kinetically less significant than
when the catalyst surface was cleaner.
These experiments demonstrate that the relatively low selectivity
obtained with the VPP catalyst is related not only to the important
parallel contribution of ethanol dehydration into ethylene, but also to
the fact that the key reaction intermediate undergoes consecutive
transformations to both the desired compound and by-products – i.e.
CO, CO2, and HCN – and to heavy compounds as well. Lastly, even
acetonitrile undergoes consecutive combustions when the reaction is
conducted at high temperature.
3.2. Ethanol ammoxidation to acetonitrile: the role of reactant partial
pressure
The control of selectivity in partial oxidation reactions, when con-
ducted with mixed oxide catalysts (and especially with the VPP), is
closely related to the redox properties of the active metal ion and its
average oxidation state under steady conditions, the latter being
affected in turn by the gas-phase composition. Therefore, we conducted
a series of experiments in which we changed the partial pressure of
ethanol, while keeping the inlet concentration of oxygen and ammonia
constant; ethanol molar fractions equal to 0.02, 0.05, 0.075 and 0.13
were used. Results are summarized in Fig. 3. The following effects were
noted:
a) The conversion of ethanol, which in all cases increased over the
range of temperature examined, showed a decreasing trend in cor-
respondence to an increased partial pressure of ethanol in the feed.
This is a clear indication of a surface saturation effect; in fact, the
rank of the overall integral rate of ethanol transformation, measured
at 400 ◦C, was: 2.5 % ethanol < 5% ethanol ≈ 7.5 % ethanol ≈ 13 %
ethanol. An alternative explanation for the phenomenon observed is
that under relatively high concentrations of ethanol and ammonia,
when both compounds act as reducing species in the redox cycle, the
catalyst evolves to a more reduced steady state, e.g. with predomi-
nance of V4+ as compared to V5+ sites. This implies that there is a
In conclusion, a major outcome of these experiments is that the best
yields to acetonitrile are obtained at either 2% (27 % at 400 ◦C and 22 %
at 420 ◦C) or 5% ethanol in the feed (18 % at 400 ◦C and 23 % at 420 ◦C),
but the best acetonitrile productivity was obtained with 5% ethanol in
the feed, at 420 ◦C. Therefore, further experiments were conducted
using the feed composition of 5 vol% ethanol, 13 % ammonia, and 13 %
4