8
42
USACHEV et al.
bons were discussed in detail in [12]. The oxidative of Al O was practically unchanged, although S to eth-
2
3
dehydrogenation of alkanes with gaseous oxygen ylene was increased to the level of 63–70%. Further
allows one to recover the lattice oxygen in the reduced increase of the Al O content (up to 90%) led to a nat-
2
3
catalyst oxides. However, the presence of O in the
2
ural increase in the conversion of ethane to ~20%
reaction mixture results in a number of problems: (a) a (during first pulse) with a simultaneous increase of S
deep oxidation of olefins, which have higher reactivity for ethylene to 75%. The value of S in a series of five
as compared to alkanes; (b) the explosion hazard of pulses of ethane increased to 88.8%. It is important
such mixtures, which requires the introduction of inert that system 10%MoO –90%Al O was hardly
3
2
3
diluents in the reaction zone, complicating the selec-
tion of target products. In this regard, another
approach is proposed involving separate dehydrogena-
changed the nature of the oxidation of ethane to eth-
ylene after three oxidative regenerations at 600°C. This
indicates its high stability in a series of reaction–
regeneration cycles. At the same time, according to the
tion steps in the absence of O and the oxidation of the
2
reduced catalyst by air oxygen [3, 4, 13]. The cyclic first pulse data the fresh sample is more active in the
operation mode of the reaction–regeneration process deep oxidation as compared with the regenerated sam-
opens the possibility to increase the selectivity for ole- ple: total S to CO is decreased from 25.4 to 14.7%.
x
fins and their concentration in the reaction products.
In alternate supply of ethane and air to the reactor and
regenerator with a fixed catalyst layer results in tech-
nological complications associated with the switching
of gas flows. The system with a microspherical catalyst
circulating between the reactor and regenerator is free
from these disadvantages [4]. In [4] we developed a
procedure for preparing a microspherical Mo-con-
taining catalyst, on which the selectivity for ethylene
exceeded 90% and made it promising for use.
In this study, we investigated the dependence of
molybdenum oxide properties on the nature and con-
centration of the modifiers (aluminum, gallium, and
yttrium oxides) in oxidative transformations of ethane.
This may indicate a strengthening of the interaction
between the components of the system in high-tem-
perature treatments.
Interesting results were obtained for the catalyst
containing only 5% of MoO (Table 1). This sample in
3
the first pulse reaches maximum conversion of ethane
(
50%) at S to ethylene being 82.5%, i.e. the ethylene
yield exceeded 40%. This is probably due to the high
dispersion of MoO particles. It should also be noted
3
that with increasing the number of pulses, the conver-
sion is reduced and in the fifth pulse, it is equal to 24%
with S being 90%. Obviously, this is due to the con-
sumption of lattice oxygen, which is indicated by a
decrease in S for CO by 2.5 times. Moreover, methane
x
appears in reaction products, whose formation can
take place at the reduced Mo-component.
EXPERIMENTAL
In the next series of Mo catalysts, Ga O was stud-
2
3
Catalyst Preparation and Testing
ied as the second component (Table 2). The conver-
Ammonium heptamolybdate (AHM) and crystal
hydrates of nitrates of aluminum, gallium, and yttrium
sion of ethane on individual oxide Ga O was previ-
2
3
ously studied. It is clear that Ga O has a quite high
2
3
(analytically pure) were used as reagents. Mixture
oxidizing power: during five pulses, the ethane con-
version varied from 55 to 44%, although the selectivity
for ethylene was low (26–31%). In the deep oxidation
products, carbon monoxide dominated, the selectivity
for which was 21–27%. It should be emphasized that
the basic product of the ethane conversion was meth-
thereof in certain proportions were triturated in a por-
celain mortar, heated in an air atmosphere to 600°C at
a rate of 100 K/h, and allowed to stand at this tempera-
ture for 1 h. The resulting mixed oxides were ground in
a porcelain mortar, and the resulting powders were
pressed into tablets (3 × 8 × 30 mm), which were
crushed; the fraction of particles with dimensions of
ane (the selectivity for CH reached ~50%).
4
These results are obviously determined by the abil-
5
0–160 μm was collected. Samples were tested in a
ity of Ga O to reduce in the given experimental con-
pulse setup described in [3]. The methods for analysis
of the resulting products and a formula for calculating
the ethane conversion and selectivities for ethylene,
methane, and carbon oxides are also reported in [3].
2
3
ditions, leading to oxidation, dehydrogenation, and
hydrogenolysis of ethane. It is possible that the latter
reaction proceeds according to the scheme: 2C H =
2
6
C H + 2CH In the number of oxide systems MoO –
2
4
4.
3
Ga O containing 82–10% MoO (Table 2), the fol-
2
3
3
RESULTS AND DISCUSSION
lowing general rules are observed. As compared to
Results of the ethane conversion over MoO – simple oxide Ga O , the conversion of ethane is twice
3
3
2
3
Al O systems in a series of consecutive pulses are decreased using the system with 82.2% MoO with
2
3
shown in Table 1. For comparison, the data for indi- simultaneous almost threefold increase in the ethylene
vidual oxide MoO are also shown. A low conversion selectivity because of the complete suppression of
3
of ethane (1.1–0.7%) with the selectivity (S) for eth- methane formation. Further increase in MoO con-
3
ylene of 49–70% was observed for this sample. The tent to ~70% causes an increase in the conversion of
activity of the 6MoO –Al O system containing 10.6% ethane (to 40%), while the selectivity for ethylene
3
2
3
PETROLEUM CHEMISTRY
Vol. 56
No. 9
2016