Gas-phase oxidative cracking of ethane in a nitrogen atmosphere

Article abstract of DOI:10.1134/S0023158413040113

The gas-phase oxidative cracking of ethane in very dilute gas mixtures ([C₂H₆]₀ ≤ 10%) has been investigated at atmospheric pressure, temperatures of up to 750 C, reaction times of up to 4 s, and initial reactant ratios of [O₂]₀: [C₂H ₆]₀ = 0.2-1.

Full text of DOI:10.1134/S0023158413040113

ISSN 0023ꢀ1584, Kinetics and Catalysis, 2013, Vol. 54, No. 4, pp. 383–393. © Pleiades Publishing, Ltd., 2013.
Original Russian Text © R.N. Magomedov, A.Yu. Proshina, V.S. Arutyunov, 2013, published in Kinetika i Kataliz, 2013, Vol. 54, No. 4, pp. 401–412.
GasꢀPhase Oxidative Cracking of Ethane in a Nitrogen Atmosphere
R. N. Magomedov*, A. Yu. Proshina, and V. S. Arutyunov
Semenov Institute of Chemical Physics, Russian Academy of Sciences, Moscow, 119991 Russia
*
eꢀmail: tetrationat@rambler.ru
Received October 30, 2012
Abstract—The gasꢀphase oxidative cracking of ethane in very dilute gas mixtures ([C H ]
≤
10%) has been
2
6 0
investigated at atmospheric pressure, temperatures of up to 750
°
C, reaction times of up to 4 s, and initial reacꢀ
tant ratios of [O ] : [C H ] = 0.2–1.
2
0
2
6 0
DOI: 10.1134/S0023158413040113
The rapid growth of the world’s accessible product selectivity and yield still have not allowed any
resources of “unconventional” natural gas and the industrially attractive process to be designed. It is,
consequent marked drop of its price relative to the therefore, necessary to carry out a comprehensive
price of petroleum makes natural gas a very attractive kinetic study of the gasꢀphase partial oxidation of С –С
2
5
petrochemical feedstock [1]. As a consequence, there alkanes, which, along with methane, are main compoꢀ
is increasing interest in the efficient conversion of light nents of natural gases. It is particularly important to
hydrocarbons into base petrochemical products. Use thoroughly elucidate the role of oxygen, which both
of a feedstock consisting of light alkanes, which are increases the conversion and reaction rate and inhibits
less reactive than liquid hydrocarbons, needs higher the formation of heavy products, including soot and its
temperatures and, accordingly, intensive heat fluxes to precursors. The mechanism of action of minor oxygen
effect endothermic processes. Since the most efficient admixtures, which exert the strongest effect on pyrolꢀ
way of supplying energy to a reacting hydrocarbon is ysis and inhibit the formation of heavy products [7, 8],
partial oxidation of the hydrocarbon just in the reactor, remains unclear. In addition, enhancing the carbon
the partial oxidation of light alkanes has been attractꢀ oxides selectivity, oxygen thereby reduces the yield of
ing increasing attention in recent years.
valuable products, such as olefins. This makes it very
important to optimize its concentration for the proꢀ
cess.
Because of the relatively low reactivity of the light
alkanes, even their catalytic reforming generally
occurs at fairly high temperatures, when a considerꢀ
The issue of the mutual effect of light alkanes in
able contribution can be made by freeꢀradical chain their simultaneous oxidation in complicated mixtures,
gasꢀphase reactions, particularly when the reactor has such as natural gases, is still practically uninvestigated.
a large empty space [2]. This opens up the possibility Because of the complexity of the branchedꢀchain freeꢀ
of carrying out noncatalytic oxidation processes, radical oxidation of even individual alkanes [5, 6],
which have certain advantages over, especially in the their mutual influence can give rise to hardly predictꢀ
reforming of complicated hydrocarbon stocks, such as able nonlinear effects.
real natural and associated gases, whose components
In recent years, there have been a large number of
vary widely in their reactivity. For example, we have studies dealing with the gasꢀphase partial oxidation of
recently demonstrated the possibility of carrying out methane [5, 6]. The partial oxidation of its homoꢀ
selective gasꢀphase oxycracking of heavy, С3+ compoꢀ logues has been much less adequately investigated.
nents of associated petroleum gas (APG) into lighter Most of the studies in the gasꢀphase oxidative cracking
hydrocarbons and highꢀoctane compounds, mainly of light methane homologues have been aimed at obtainꢀ
into ethylene, hydrogen, and methane. The small ing lower olefins [2]. Works in this field were started as
amount of oxygen (air) present in the gas mixture early as the 1940s–1950s [9]. However, there have been
accelerates the process and, in addition, prevents any only a few studies carried out under conditions typical for
noticeable formation of heavy condensation products, natural and associated gases, in which the С2+ alkane
including soot. After this oxidative reforming, APG is concentration in methane is relatively low.
usable as a fuel for local power plants, primarily for
field gasꢀturbine and gasꢀpiston generators [3, 4]. This
opens up a promising way of utilizing APG.
We are carrying out comparative kinetic studies of
the gasꢀphase oxidative cracking of light, С –С
2
5
alkanes in inert gas atmospheres and methane. We
However, although the noncatalytic partial oxidaꢀ believe that along with the investigation of the partial
tion of light hydrocarbons is energetically and technoꢀ oxidation of light alkanes at high pressures [10, 11],
logically attractive [5, 6], the insufficiently high target these studies will provide the necessary information
383

384
MAGOMEDOV et al.
6
Gas to be analyzed
8
9
7
4
1
2
To fume hood
3
5
11
Gas to be analyzed
10
Fig. 1. Experimental setup: (
) valve, ( ) laboratory transformer, (
thermocouple, (10) threeꢀpoint KTKhAꢀtype thermocouple, and (11) gas meter.
1
) cylinder with an ethane + oxygen + diluent gas mixture, (
2
) pressure regulator, (
3) rotameter,
(4
5
6
) pressure gage, ( ) quartz reactor, ( ) fourꢀchannel temperature controller, (9
7
8
) KTKhAꢀtype
for a detailed kinetic analysis of the partial gasꢀphase ple mounted in an axial quartz well 5 mm in diameter.
oxidation of light alkanes and for gaining an adequate A movable thermocouple, in a separate quartz well,
insight into the mechanism of the processes occurring allowed the temperature profile in the reactor to be
in the wide range of conditions that is important for monitored. The power of the electric heaters was conꢀ
practical implementation.
trolled using a Termodatꢀ13K5 fourꢀchannel temperꢀ
ature controller, which responded to the signals from
the thermocouples located in the operation zones of
the corresponding heaters. Because our experiments
were performed on very dilute ethane–oxygen mixꢀ
tures, the reactor conditions were nearꢀisothermal.
The first methane homologue that is least typical in
respect of partial oxidation is ethane. Like methane, it
has no –СН2 groups, which determine many characꢀ
teristic properties of heavier alkanes. It differs from the
heavier alkanes in that its C–H and C–C bonds are
stronger, imparting specific kinetic features to the proꢀ
cess in which it is involved. In a complicated hydrocarꢀ
bon mixture, it is simultaneously a reactant and a
product of partial oxidation of both methane and its
heavier homologues. This makes the analysis of the
oxidation processes involving ethane very difficult.
The surface areaꢀtoꢀvolume ratio (S/V) for the
reactor, with the surface area of the quartz wells for
–1
thermocouples taken into account, was 5.4 cm . The
experiments were carried out at 550 to 750 С using
°
premixed gases with a preset composition. The flow
rate of the gas mixture was measured with a rotameter
Here, we report the kinetics of the oxidative crackꢀ and was varied in the 2.4–25 mL/s range (under stanꢀ
ing of very dilute ethaneꢀcontaining mixtures at atmoꢀ dard conditions). At these flow rates, the residence
spheric pressure and a low oxygen concentration.
EXPERIMENTAL
time of the mixture in the isothermal part of the reacꢀ
tor (tr) at 700 was 0.35–3.6 s. The ethane concenꢀ
tration in the mixture in most experiments was 5 mol %,
°
С
and the oxygen concentration was 1–5 mol %, so the
[
O ]/[C H6] ratio was 0.2 to 1. In order to investigate
2 2
Experiments were performed at atmospheric presꢀ
sure using a setup based on a flow reactor (Fig. 1). The
inner diameter of the heated cylindrical quartz reactor
was 14 mm, and the length of the isothermal highꢀ
temperature zone was 200 mm. The reactor was
heated using three independent electric heaters, and
the ethane concentration effect on the process, we
additionally conducted experiments at ethane conꢀ
centrations of 2 and 10 mol % and a fixed oxygenꢀtoꢀ
ethane ratio of 1 : 2. The diluent gas was nitrogen.
The composition of the initial gas mixture and
this allowed a fairly uniform temperature profile to be reaction products was determined on an M3700 chroꢀ
maintained in the highꢀtemperature zone. The temꢀ matograph with three detectors. H , O , N2, and CO
2
2
perature in the operation zones of the three heaters were quantified using a 2 m
×
3 mm column packed
was measured with a KTKhA threeꢀpoint thermocouꢀ with molecular sieve CaA, argon as the carrier gas, and a
KINETICS AND CATALYSIS Vol. 54
No. 4
2013

GASꢀPHASE OXIDATIVE CRACKING OF ETHANE IN A NITROGEN ATMOSPHERE
385
thermalꢀconductivity detector. СО2 was determined on
a column packed with Porapak Q (3 m mm) using
helium as the carrier gas and a thermalꢀconductivity
KINETIC MODELING OF THE PROCESS
×
3
The experiments were supplemented with kinetic
simulation in terms of a detailed model of the partial
detector. The С –С hydrocarbons were analyzed using
1
3
oxidation of С –С2 hydrocarbons. The model was
1
an RtꢀQꢀBOND capillary column (30 m
helium as the carrier gas, and a flame ionization detecꢀ
tor. The analyses were carried out at 80 . The chroꢀ
×
0.53 mm),
specially developed for the temperature range 600–
200 K, pressure range 1–100 atm, and a wide range
1
°
С
of [О ]/[С Н6] ratios. In accordance with the princiꢀ
2
2
matograph was calibrated against reference gas mixꢀ
ples of construction of this model, which were detailed
in an earlier publication [12], the numerical values of
all kinetic parameters were taken from independent
literary sources, mainly from the NIST kinetic dataꢀ
base [13], and were not varied in the modeling. The
base variant of the model was presented in an earlier
(1) work [14]. In the simulation of the results of this study,
the initial model [14] was augmented with the reacꢀ
tions presented in Table 1. In addition, because reacꢀ
tions involving hydrogen peroxide play a significant
(2) role under the conditions examined, its formation and
decomposition were represented in terms of the soꢀ
called Lindemann mechanism [15], which explicitly
takes into account the formation of the excited state
via which the reaction proceeds both toward deactivaꢀ
tures.
The following quantities were derived from the
analytical data:
conversion of the reactants,
0
(Ci − C K )
i
N
xi =
×100%,
0
Ci
selectivity on the carbon basis,
prod
C
С
Ci
n
i
S =
× 100%,
C
i
prod
Ci
ni
∑
i
selectivity on the oxygen basis,
prod
O
O
i
Ci n K
i
N
tion and toward decomposition (Table 2). This
enabled us to improve the agreement between the
experimental and simulated data. Together with the
added reactions, the modified model includes 490 eleꢀ
mentary steps.
S =
× 100%,
(3)
0
2
(C − C K )
O2
O2
N
water vapor concentration,
O
O
S
= 100 − S ,
Н2О
∑
i
i
Calculations were carried out for isothermal condiꢀ
tions using the Kintecus V450 program [16], which
allows kinetic modeling of oxidation, combustion, etc.
(4)
O
H2O
0
S
(C − C K ) × 2
O2 O2 N
CН2О
=
,
K ×100%
N
selectivity on the hydrogen basis,
RESULTS
prod
Ci
H
H
Si
n
i
Reactant Conversion
=
× 100%.
H
(5)
prod
Ci
ni
∑
i
The temperature at which the radical chain oxidaꢀ
tive conversion of hydrocarbons begins to proceed at a
noticeable rate generally depends on the composition
of the reaction mixture and on the process conditions.
According to numerous studies [2], the oxidative
0
Here, Ci is the initial concentration of the
i
th reactant
(
mol %), Ci is the concentration of the ith reactant in
prod
the gas at the reactor outlet (mol %), Ci is the conꢀ
centration of the th product in the gas at the reactor
pyrolysis of ethane below 600
appreciable rate only in the presence of a catalyst or on
outlet (mol %), ni is the number of carbon atoms in the reactor surface.
°
С takes place at an
i
C
H
The ratio of the empty volume of the reactor to the
the molecule, ni is the number of hydrogen atoms in
O
volume occupied by a catalyst or solid material and the
chemical composition and properties of the catalyst
determine the relative contributions from the homoꢀ
geneous and heterogeneous reactions to the oxidation
process, thus exerting a direct effect on the composiꢀ
tion of the product mixture [2]. In the gasꢀphase proꢀ
cess, the interaction of alkyl radicals with oxygen and
the molecule, ni is the number of oxygen atoms in the
molecule, and KN is the coefficient accounting for the
change of the total number of moles in the reaction.
The carbon balance was struck with 4% accuracy.
±
The water vapor concentration was calculated from
oxygen imbalance using Eq. (4). The hydrogen balꢀ
ance, with the calculated water vapor concentration
oxygenꢀcontaining radicals (НО , ОН, О) yields oxyꢀ
genate products, whose further oxidation mainly
2
taken into account, was struck with 8% accuracy.
±
The experiments were carried out using the followꢀ yields CO. Conversely, the total oxidation reactions
ing bottled gases: specialꢀpurity nitrogen (grade 1, occurring on the catalyst surface considerably raise the
9
(
9.999%), helium (grade A, 99.995%), methane yield of СО2 and, accordingly, reduce the CO : СО2
99.99%), and ethane (99.99%). Before use, ethane + ratio. As a consequence, the oxygen conversion in the
oxygen + diluent gas mixture were kept in a 10ꢀL cylꢀ presence of a catalyst is typically higher than in the
inder for at least 12 h.
empty space of the reactor, where a pure gasꢀphase
KINETICS AND CATALYSIS Vol. 54
No. 4
2013

386
MAGOMEDOV et al.
Table 1. Reactions added to the base model [14]
, cm mol s–1
3
–1
E
, kcal/mol
n
Entry
Reaction
A
1
2
3
4
5
6
7
8
9
C H + O
→
CH + HCO
1.32
2.64
5.06
1.99
3.13
2.6
×
×
×
×
×
×
×
×
×
108
0.43
8.07
6.88
4.75
14.17
96.58
0
1.55
2
4
3
CH + OH
→
CH OH + H
1019
1011
1013
106
–1.8
3
2
CH + CO
→
CH CO
0
3
3
CH OOH + O
→
OH + CH O2
0
3
3
C H + HO
→
C H + O
2
1.61
2
2
2
2
3
C H + N
→
C H + H + N
2
1017
1013
1013
1018
1013
1013
1012
1010
1013
1013
1013
1012
103
10–12
1011
104
1013
1014
1011
1010
1012
0
2
4
2
2
3
CH CO + HO2
→
CH CHO + O2
3
0
3
3
CH CO + OH
→
CH CHO + O
4.22
1.38
0
0
3
3
C H + C H
→
C H + C H + H
2
1.04
23.05
69.71
–0.43
–1.08
39.94
60.41
0
–1.66
2
5
2
5
2
4
2
4
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
CH + CH4
→
H + C H
5
0
3
2
2
H +CH O
→
CH OH
1.38
1.03
4.44
8
×
×
×
×
×
×
×
×
×
×
×
×
×
×
×
×
0
2
2
3
C H + O
→
C H + OH
0.2
0
2
3
2
2
CH CO + O2
→
CH O + CO2
3
3
CH + CH4
→
C H + H
0
3
2
6
CH + CO
→
C H + OH
3.8
0
3
2
2
CH + CH CO
→
C H + CO
3.3
0
3
3
2
6
CH O + CH CO
→
CH O + CH CHO
6.03
4.837
6.2
0
0
3
3
2
3
CH OH + CH CO
→
CH CHO + CH OH
12.34
0
3
3
3
3
2
CO + CH O
→
CH O + HCO
0
3
2
C H + C H
→
C H + C H
3
9.64
1.81
2.41
1.51
1.81
8.13
1.26
0
0
2
6
2
2
2
5
2
C H + CH CO
→
CH CHO + C H
5
17.53
68.36
90.62
12.92
3.68
8.51
2.75
0
2
6
3
3
2
C H + C H
→
C H + C H
3
2
4
2
2
2
3
2
C H + CO
→
HCO + C H3
0
2
4
2
CH O + CH CO
→
CH CHO + HCO
0
2
3
3
C H + CH CHO
→
C H + CH CO
0
2
3
3
2
4
3
C H + CH CHO
→
C H + CH CO
2
0
2
5
3
6
3
Note: The kinetic parameters refer to the threeꢀparameter form of the Arrhenius equation: k = /RT).
A Tn exp(–E
Table 2. Hydrogen peroxide formation and disappearance reactions
3
–1
Entry
Reaction
HOOH + N2
OH + OH
HOOH
H O + N
A
, cm mol s–1
E
, kcal/mol
n
1
2
3
4
H O + N
2
→
1.29
3
×
×
×
×
1033
1014
1012
1024
53.26
–4.86
2
2
HOOH
OH + OH
HOOH + N2
→
21
0
0
→
9.05
5.8
–0.37
–3
→
0
2
2
2
Note: HOOH stands for the activated form of the H O molecule.
2
2
process takes place. For example, complete oxygen ethane conversion) [17]. At the same time, the oxygen
conversion in the oxidative dehydrogenation of ethane conversion in the absence of a catalyst at 750 and an
°
С
in the presence of a mixed oxide of an alkalineꢀearth ethane conversion of 60% was only 80%; that is, there
metal and a rareꢀearth metal was achieved at a rather was a fairly close coincidence between the conversions
low temperature of 600°С (even at a relatively low of the reactants.
KINETICS AND CATALYSIS Vol. 54
No. 4
2013

GASꢀPHASE OXIDATIVE CRACKING OF ETHANE IN A NITROGEN ATMOSPHERE
387
The temperature dependences of the oxygen and
ethane conversions (Fig. 2) indicate that the converꢀ
sions of the reactants coincide quantitatively at similar
concentrations of the reactants in the mixture, which
is typical of the gasꢀphase process. This coincidence is
observed up to the temperature at which complete
oxygen conversion is reached. At higher temperatures,
the rate of increase of ethane conversion with temperꢀ
ature decreases. Apparently, the increase in ethane
conversion at these temperatures is due to the increasꢀ
ing contribution from the thermal cracking and ethane
dehydrogenation reactions. At high temperatures, the
considerable contribution from these reactions, which
are characterized by a high activation energy, reduces
the accelerating effect of oxygen on ethane conversion.
Note that the temperature interval within which
the barely noticeable reaction develops to the extent of
almost complete oxygen conversion in combination
(
а)
100
9
0
0
8
O = 5%
2
O = 1%
2
70
6
0
0
O = 2.5%
2
5
40
1
2
3
3
0
0
2
10
0
550575600625650675700725750775
Temperature, °C
with a high ethane conversion is narrower than 50
°
С.
(b)
O = 5%
100
This critical ignition of the reaction mixture in the narꢀ
row temperature range is characteristic of branched
chain processes. The critical character of the crossover
of the reaction to the branched chain mechanism is also
indicated by the fact that the apparent activation energy
determined by formal data processing in this temperaꢀ
ture region is as high as >100 kcal/mol.
At oxygen concentrations of [O2]0 = 2.5 and 5%,
the amount of oxygen in the mixture is far below the
amount corresponding to the stoichiometry of total
ethane oxidation and oxygen is almost entirely conꢀ
2
90
80
70
60
50
40
30
O = 2.5%
2
O = 1%
2
1
2
3
20
sumed (Fig. 2a). However, at [O2] = 1% the residual
0
10
oxygen concentration is stably at a ~0.05% level.
Nearly the same incomplete oxygen conversion was
observed in experiments on partial methane oxidation
0
5
50575600625650675700725750775
Temperature, °C
[
5, 6].
In addition to qualitatively accounting for the critꢀ
Fig. 2. (a) Oxygen and (b) ethane conversions versus temꢀ
ical character of the transition of the reaction to the
ignition regime under these conditions, the kinetic
model provides a quite satisfactory quantitative fit to
experimental data. In view of the accepted indepenꢀ
dent model construction principle [12], it can be
stated that the existing theoretical conceptions are
perature ([С Н6]0 = 5%, tr = 2 s). The points and dashed
2
lines represent experimental data, and the solid lines repreꢀ
sent the results of modeling. [O ] = (1) 1, (2) 2.5, and (3) 5%.
2
The practically complete conversion of oxygen is
quite consistent with the observed kinetics of the proꢀ accompanied by only partial ethane conversion,
cess. The hitherto unexplained discrepancies between whose value decreases with a decreasing initial oxygen
the experimental and simulated data include the fact concentration ([О ] /[С Н ] ratio). As the initial
2
0
2
6 0
that, at all oxygen concentrations, the calculated oxyꢀ oxygen concentration is increased, the oxygen conꢀ
gen conversion reaches 100%. Furthermore, accordꢀ sumption per converted ethane molecule increases
ing to the results of the kinetic modeling, an increase monotonically. Accordingly, the ethane conversion at
in the oxygen concentration increases both the temꢀ the maximum oxygen conversion temperature
perature (Fig. 2a) and time necessary for 100% conꢀ increases less rapidly than the initial oxygen concenꢀ
version (Fig. 3a). In fact, as the oxygen concentration tration (Fig. 4). This dependence of ethane conversion
is raised from 2.5 to 5% in the experiment, the oxygen on the oxygen concentration is likely due to the
conversion somewhat increases and the time required increase in the amount of oxygen consumed in the
for the conversion to reach its quasiꢀsteadyꢀstate value oxidation of the resulting products. Jallais et al. [18]
decreases (Fig. 3a).
observed an intensive gasꢀphase oxidation of rich
Raising the oxygen concentration in the mixture ethylene + air mixtures with an equivalence ratio of 3
both increases the maximum ethane conversion at a to 10 at atmospheric pressure. This process occurred
given temperature and shortens the time to maximum in approximately the same temperature range (starting
conversion (Fig. 3b).
at 550 С) and yielded mainly the same products (СО,
°
KINETICS AND CATALYSIS Vol. 54
No. 4
2013

388
MAGOMEDOV et al.
(а)
100
1
00
9
0
0
9
8
7
6
5
4
3
2
1
0
0
0
0
0
0
0
0
0
0
O = 1%
2
8
O = 2.5%
2
70
O = 5%
2
60
50
40
1
2
30
2
0
0
0
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Residence time, s
1
(
b)
0
1
2
3
[O ], mol %
4
5
6
100
O = 5%
2
2
90
80
70
60
50
40
30
20
10
0
O = 2.5%
2
Fig. 4. Ethane conversion at the complete oxygen converꢀ
sion temperature versus the initial oxygen concentration
[O ] ([С Н6]0 = 5%, tr = 2 s). The points represent experꢀ
imental data, and the solid lines represent the results of
modeling.
2
0
2
O = 1%
2
lower than, the conversion of the hydrocarbon itself
[20, 21].
1
2
As the oxygen concentration is increased, the ratio
between the amounts of reacted ethane and oxygen at
the complete oxygen conversion temperature decreases
monotonically from ~2 at [О ] /[С Н ] = 0.2, which
2
0
2
6 0
0
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Residence time, s
formally corresponds to the overall reaction
С Н + 0.5О → С Н + Н О,
2
(I)
2
6
2
4
2
to ~1 at [О ] /[С Н ] = 1. This may be due to the
Fig. 3. (a) Oxygen and (b) ethane conversions versus resꢀ
2 0
2
6 0
idence time of the reaction mixture in the reactor
increasing contribution from the overall reactions
С Н + О 2СО + 3Н2,
С Н + Н ,
(
[
С Н6]0 = 5%, 700
°
C). The points and dashed lines repꢀ
resent experimental data, and the solid lines represent the
results of modeling. [O ] = ( ) 2.5 and ( ) 5%.
2
→
(II)
(III)
(IV)
2
6
2
1
2
С Н6
2
→
2
4
2
2
0
С Н + О
2
→
2СО + 2Н2.
2
4
The kinetic model provides a good fit to the observed
dependence of ethane conversion on the initial oxygen
concentration (Fig. 4).
СН , СО2), so it was described well by the gasꢀphase
model. In experiments reported by Heracleous and Lemꢀ
4
onidou [19], gasꢀphase ethylene oxidation (С Н : О =
2
4
2
It can be seen in Fig. 2 that, at a constant ethane
1
: 1) in a mixture heavily diluted with helium started
concentration of [С Н ] = 5% and oxygen concenꢀ
2
6 0
above 670 and mainly yielded water and carbon monꢀ
°
С
trations of [О2]0 < 2.5%, the reaction onset and comꢀ
plete oxygen conversion temperatures increase as
oxide with a CO selectivity of about 90%.
[
О ] is raised. This can formally be attributed to the
As the temperature is raised above the complete
oxygen conversion temperature, the ethane converꢀ
sion increases at a comparatively low rate, remaining
below the oxygen conversion even at the equimolar
oxygenꢀtoꢀethane ratio. This observation is in good
agreement with the literature [8, 17, 20], according to
which ethane differs significantly from its heavier
2 0
inhibition of freeꢀradical ethane oxidation with oxyꢀ
gen. A similar effect was observed in experiments
reported by Heracleous and Lemonidou [19], in
which gasꢀphase ethane dehydrogenation under very
similar conditions in the absence of oxygen began
4
0
°
С
below than the same reaction in the presence of
oxygen; that is, oxygen inhibited this process as well.
homologues. In the oxidative cracking of the latter, the However, as the temperature was elevated, the ethane
oxygen conversion is typically equal to, or slightly conversion in the presence of oxygen increased signifꢀ
KINETICS AND CATALYSIS Vol. 54
No. 4
2013

GASꢀPHASE OXIDATIVE CRACKING OF ETHANE IN A NITROGEN ATMOSPHERE
389
icantly and became 75% at 700 С, while the ethane
°
(а)
6
.0
conversion in the absence of oxygen rose only to 44%.
A similar inhibiting effect of oxygen on a branched
chain process was observed for partial methane oxidaꢀ
tion at high pressures [5, 6].
С Н
2
O2
6
Н
2
СO
5
.0
.0
С Н
СН4
СO2
2
Н О
4
2
4
Temperature Effect on the Product Composition
The dependences of the reactant and main product
concentrations on the reaction temperature are preꢀ
sented in Fig. 5. As can be seen from this plot, the priꢀ
mary oxypyrolysis products at high temperatures are
ethylene and, apparently, water, whose concentration
was calculated from the oxygen imbalance. Under our
experimental conditions, both products appear once
the conversion of the reactants begins. The other
products form at higher temperatures and, except
hydrogen, at significantly lower concentrations.
3.0
2
.0
.0
0
1
The results of our modeling demonstrate the forꢀ
mation of hydrogen peroxide along with water. Hydroꢀ
gen peroxide was not detected experimentally, but,
according to the results of modeling, it forms mainly at
low temperatures. Its maximum concentration is
5
75 600 625 650 675 700 725 750 775
Temperature, °C
6.0
(b)
С Н
2
0.44% at 650
°
С. At higher temperatures, it decomꢀ
6
5
4
3
.0
.0
.0
poses rapidly, and this is accompanied by a sharp
increase in the ethane conversion rate. This observaꢀ
tion is in good agreement with the conception that, in
alkane oxidation above 500 С, hydrogen peroxide
°
Н О + Н О
2
С Н
2
2
2
4
decomposition becomes one of the main chain
branching reactions [22]. For this reason, the experiꢀ
mental data were compared with the calculated sum of
the water and hydrogen peroxide concentrations.
The model quite adequately describes not only the
qualitative composition of the products and its temperꢀ
ature dependence, but also the quantitative product
composition. The exception is the hydrogen concentraꢀ
tion underestimated by a factor of about 2 and the
slightly underestimated CO and СН4 concentrations.
The causes of this discrepancy are to be cleared up.
The increase of the Н2 : CO ratio caused by a rise in
temperature is likely due to the increase in the contriꢀ
bution from the thermal dissociation of the C–H bond
in the ethyl radical, whose rate in the temperature
range examined is well above the rate of hydrogen
atom abstraction by oxygen [19]. After the entire oxyꢀ
gen is reacted, the ethylene concentration continues
to increase along with the hydrogen concentration.
This is apparently due to the thermal dehydrogenation
of ethane via reaction (III) at high temperatures. An
important product of the process is methane, which
O2
Н2
2.0
СO
1.0
СН4
СO2
0
5
75 600 625 650 675 700 725 750 775
Temperature, °C
Fig. 5. Reactant and product concentrations versus temꢀ
perature ([С Н ] = 5%, [O ] = 2.5%, tr = 2 s): (a) experꢀ
iment; (b) model.
2
6 0
2 0
water. (In our modeling, the total amount of water and
hydrogen peroxide was considered instead of the
amount of water alone.) As the temperature is elevated
to 700 С, at which complete oxygen conversion is
°
forms above 650 С, when the breaking of the C–C
°
reached, both the calculated and observed ethylene
formation selectivities on the carbon basis decrease
from 100 to ~70%. As the temperature is further
raised, these selectivities change insignificantly. In the
same temperature range, the CO and СН4 formation
selectivities are 18–20 and 7–8%, respectively; as the
bond in ethane and the oxidative cracking of ethylene
occur at a high rate [19].
The temperature dependence of the ethane oxidaꢀ
tion product formation selectivities (Fig. 6) is consisꢀ
tent with the aboveꢀconsidered temperature depenꢀ
dence of the product concentrations. Below 650 С
°
and, accordingly, at low reactant conversions, the only temperature is further increased, they also remain
oxypyrolysis products are practically ethylene and practically invariable. At the same time, throughout
KINETICS AND CATALYSIS Vol. 54
No. 4
2013

390
MAGOMEDOV et al.
tures there is a larger contribution from parallel reacꢀ
tions, namely, thermal cracking and dehydrogenation
and deep ethane and ethylene oxidation.
(а)
1
00
9
8
7
6
5
4
3
2
1
0
0
0
0
0
0
0
0
0
Residence Time Effect on the Product Composition
Until the consumption of the entire oxygen, the
concentrations of all reaction products increase
monotonically with time (Fig. 7). The ethylene and
water concentrations, which, as follows from Fig. 7,
are primary ethane conversion products, almost coinꢀ
cide up to this point in time. This is consistent with
overall reaction (I). Hydrogen also forms in large
amounts even at short reaction times owing to the high
rate of thermal dehydrogenation at the given temperaꢀ
С Н
2
4
СO
СН4
СO2
ture (700 С). The other products form after some time
°
delay and in smaller amounts as a result of the cracking
and deep oxidation of ethane and the resulting ethylꢀ
ene. This is particularly evident from the results of
kinetic modeling. The model quite adequately fits the
experimental data (Fig. 7b). The exception is the fact
that, because the calculated oxidation rate is overestiꢀ
mated, oxygen conversion in the model is complete in
an approximately two times shorter time than in the
experiment. In addition, the model predicts that oxyꢀ
gen conversion will come to completion rapidly, while
in the experiment, this process is more prolonged and,
moreover, the aboveꢀmentioned residual oxygen conꢀ
centration persists.
After the conversion of the entire oxygen (which
takes a little more than 1 s under the given conditions),
the process practically stops and the concentrations of
the products varies insignificantly in the time interval
examined, even though a rather large amount of
ethane is still present. This indicates that the ethane
pyrolysis rate in the absence of oxygen is low at the
given temperature. Only a slight increase in the hydroꢀ
gen and ethylene concentrations is observed because
of the corresponding decrease in the ethane concenꢀ
tration.
0
6
00 625 650 675 700 725 750 775
Temperature, °C
8
0
(b)
7
6
5
4
3
2
1
0
0
0
0
0
0
0
0
Н2
Н2
С Н
O
x
y
6
00 625 650 675 700 725 750 775
Temperature, °C
The constancy of the water vapor concentration
after the disappearance of oxygen indicates that there
is no appreciable contribution from the gasꢀphase
steam reforming of ethane under these conditions.
This finding is consistent with the data of Schadel et al.
Fig. 6. Product formation selectivities on the (a) carbon
and (b) hydrogen basis versus temperature ([С Н6]0 = 5%,
2
[
O ] = 2.5%, tr = 2 s). The points and dashed lines repreꢀ
2 0
sent experimental data, and the solid lines represent the
results of modeling.
[
23], according to which the water vapor conversion in
the absence of a catalyst below 900 , when the gasꢀ
°
С
phase pyrolysis of ethane and its heavier homologues
proceeds at a high rate, is extremely low and does not
exceed 5%.
the temperature range examined, the Н О formation
2
selectivity decreases almost linearly, and this is
accompanied by the corresponding increase in the Н2
formation selectivity. As was noted above, this can be
due to the acceleration of the thermal dehydrogenaꢀ
tion of ethane relative to its oxidative dehydrogenaꢀ
tion, as well as due to oxygen redistribution in favor of
СОх formation.
Oxygen Concentration Effect
on the Product Composition
Along with temperature, the initial oxygen concenꢀ
tration ([О ] /[С Н ] ratio) is the most important
2
0
2
6 0
Thus, while the dominant reaction at low temperaꢀ factor both in the process rate and in the product comꢀ
tures and, correspondingly, low ethane conversions is position. Other conditions (ethane concentration,
oxidative ethane dehydrogenation, at high temperaꢀ reaction time, and temperature) being equal, an
KINETICS AND CATALYSIS Vol. 54
No. 4
2013

GASꢀPHASE OXIDATIVE CRACKING OF ETHANE IN A NITROGEN ATMOSPHERE
391
increase in the initial oxygen concentration causes a
monotonic decrease in the ethylene, total hydrocarꢀ
bon, and hydrogen formation selectivities, enhancing
the oxygenꢀcontaining product (CO and water) forꢀ
mation selectivities (Fig. 8).
6
.0
.0
(а)
С Н
СO
СН4
O2
2
Н О
6
5
2
С Н
2
Н2
4
Although the ethylene formation selectivity
decreases with an increasing ethane conversion, the
ethylene yield is practically independent of the oxygen
concentration in the [О ] /[С Н ] range examined
СO2
4.0
2
0
2
6 0
when the ethane conversion is below 50% (Fig. 9).
Kinetic modeling demonstrated that, at ethane converꢀ
sions of up to 40%, an increase in the oxygen concenꢀ
tration from 1 to 2.5% even raises the ethylene formaꢀ
tion selectivity. This may be due to the acceleration of
3.0
2.0
1.0
the dehydrogenation reactions in the presence of О2
.
Ethane Concentration Effect
In order to see whether secondary processes exert
any significant effect on the oxidative conversion of
ethane, we carried out experiments at different initial
ethane concentrations and a fixed reactant ratio of
0
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Residence time, s
6
5
4
3
2
.0
(b)
[
О ] /[С Н ] = 0.5. Raising the ethane concentration
2 0 2 6 0
from 2 to 10% lowered the onset temperature of oxyꢀ
gen conversion (ignition temperature) and the comꢀ
plete oxygen conversion temperature by approxiꢀ
.0
.0
.0
.0
С Н
2
6
mately 75°C. This is quite consistent with the existing
understanding of the hydrocarbon concentration
effect on ignition temperature of the hydrocarbon.
Our kinetic modeling qualitatively confirmed the
experimental data; however, the calculated effect of
the ethane concentration on the process onset temꢀ
perature is not as strong as the observed effect. Note
that the initial ethane concentration had almost no
effect on the product composition.
Н О + Н О
2
2
2
С Н
2
O2
4
Н2
СO
DISCUSSION
1.0
0
СН4
Most of the works dealing with gasꢀphase oxidative
cracking of light alkanes consider the role of homogeꢀ
neous reactions in the catalytic processes occurring at
fairly high temperatures [2]. Heracleous and lemꢀ
onidou [19] investigated the contribution from gasꢀ
phase reactions to the oxidative dehydrogenation of
СO2
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Residence time, s
Fig. 7. Reactant and product concentrations versus the
residence time of the reaction mixture in the reactor
([С Н ] = 5%, [O ] = 2.5%): (a) experiment; (b) model.
ethane in the presence of a MoO /Al O catalyst by
3
2
3
conducting a number of experiments in an empty
Uꢀshaped quartz reactor. The temperature in the
reactor was elevated at a rate of 15 deg/min to 700 С,
and the system was held at this temperature for 20 min
to complete the process. The gas composition at the
reactor outlet was determined by mass spectrometry.
The ethane and oxygen concentrations in the feed,
2
6 0
2 0
°
ene—formed with >95% selectivity. As the temperaꢀ
ture was raised, the selectivity decreased; nevertheless,
С, when the ethane conversion was
0%. The main byꢀproduct was CO. The CO formaꢀ
it was 60% at 700
°
8
which was diluted with helium, were 2%. The resiꢀ tion selectivity was up to 30%, while only small
dence time of the mixture in the reactor space conꢀ amounts of СО were detected and the CO formation
verted to standard conditions was 6 s. The process selectivity was below 2%. Thus, in spite of some differꢀ
commenced at 650
and yielded almost equal ence between the feed compositions, the process onset
amounts of С Н4 and Н2. At 655
, CO, Н О, and temperature and product composition data obtained
2
2
°
С
°
С
2
2
smaller amounts of СН4 and СО2 appeared. At low by Heracleous and lemonidou [19] are close to our
ethane conversions, the primary product—ethylꢀ results.
KINETICS AND CATALYSIS Vol. 54
No. 4
2013

392
MAGOMEDOV et al.
100
(а)
100
9
8
7
6
5
4
3
2
1
0
0
0
0
0
0
0
0
0
9
0
0
O = 1%
2
8
С Н
2
4
СO
O = 2.5%
2
1
2
3
СН4
70
60
СO2
O = 5%
2
5
0
0
0
0.2
0.4
0.6
O ]/[C H ]
0.8
1.0
1.2
4
0
10 20 30 40 50 60 70 80 90 100
Ethane conversion, %
[
2
2
6
70
60
50
40
30
20
10
(b)
Fig. 9. Ethylene formation selectivity versus ethane conꢀ
version ([С Н6]0 = 5%, 700 C, tr = 0.3–3.9 s). The points
represent experimental data, and the solid lines represent
the results of modeling. [O ] = ( ) 1, ( ) 2.5, and ( ) 5%.
°
2
1
2
3
2
0
С H
x
tent with our data. Noticeable conversions, similar for
the two reactants, were observed starting at >575
and took values of 80 and 90% for ethane and oxygen,
y
°
С
H O
2
Н2
respectively, at [О ] /[С Н ] = 1 and the maximum
2
0
2
6 0
experimental temperature (650
same ethane conversion was observed an approxiꢀ
mately 25 C higher temperature. This discrepancy is
°
С). In our study, the
°
rationally explained by the small differences between
the experimental conditions. The most abundant carꢀ
bonꢀcontaining products in the study by Machli et al.
[
20] were ethylene and carbon oxides (mainly CO),
whose yield at 650 was 56 and 20%, respectively.
°
С
0
0.2
0.4
0.6
O ]/[C H ]
0.8
1.0
1.2
These data are in good agreement with the ethylene
yield observed by us at a similar ethane conversion and
the same reactant ratio.
[
2
2
6
Fig. 8. Product formation selectivities on the (a) carbon
and (b) hydrogen basis versus the [O ] /[С Н6]0 ratio
The effects of the diluent gas nature and reactor
surface on the oxidative cracking of dilute ethane mixꢀ
tures were considered in our other publication [24].
2
0
2
([С Н6]0 = 5%, 700°C, tr = 2 s). The points and dashed
2
lines represent experimental data, and the solid lines repꢀ
resent the results of modeling.
CONCLUSIONS
The gasꢀphase oxypyrolysis of ethane is definitely a
branched chain process passing to the ignition regime
From the standpoint of experimental conditions,
our study is most similar to that of Machli et al. [20],
in which the reaction time was 2.4 s, the initial ethane
concentration in helium was [С Н ] = 6%, and the
at
∼
650 С. In the reaction between ethane and oxygen
°
in their equimolar mixture, the ratio of the consumed
amounts of the reactants is close to unity.
2
6 0
[
О ] /[С Н ] ratio was varied between 0.2 and 1.
2 0 2 6 0
Raising the oxygen concentration and the oxygenꢀ
toꢀethane ratio leads to an increase in the ethane conꢀ
version. At the same time, the ratio of the amounts of
reacted oxygen and ethane at the complete oxygen
Although the temperature at which the oxidative
dehydrogenation of ethane was studied in that work
was somewhat lower and was varied in a narrower
range (575–650
°
С) and the reactor was filled with conversion temperature increase monotonically with
quartz chips, the results of that study are quite consisꢀ an increasing О2 concentration because of the increasꢀ
KINETICS AND CATALYSIS Vol. 54
No. 4
2013

GASꢀPHASE OXIDATIVE CRACKING OF ETHANE IN A NITROGEN ATMOSPHERE
393
ing amount of oxygen spent on the formation of secꢀ
ondary oxidation products.
6. Arutyunov, V.S., Okislitel’naya konversiya prirodnogo
gaza (Oxidative Reforming of Natural Gas), Moscow:
KRASAND, 2011.
After 100% oxygen conversion is reached, the
ethane conversion continues to increase with an
increasing temperature. This is apparently due to the
increasing contribution from thermal pyrolysis.
7
. Huynh, L.K., Shoaibi, A.Al., Carstensen, H.ꢀH., and
Dean, A.M., Prepr. Pap.—Am. Chem. Soc., Div. Fuel
Chem., 2009, vol. 54, no. 1, p. 56.
8. Choudhary, V.R. and Mulla, S.A.R., AIChE J., 1997,
Raising the ethane concentration lowers the onset
temperature of the reaction, and increasing the oxygen
concentration slightly elevates this temperature.
vol. 43, p. 1545.
9. Deanesly, R.M. and Watkins, C.H., Chem. Eng. Prog.,
1951, vol. 47, no. 3, p. 134.
The primary oxypyrolysis products are ethylene
and water, but, as the ethane conversion increases,
their formation selectivity decreases because of the
increasing CO, hydrogen, and methane formation
selectivities.
1
1
1
0. Sheverdenkin, E.V., Arutyunov, V.S., Rudakov, V.M.,
Savchenko, V.I., and Sokolov, O.V., Theor. Found
Chem. Eng., 2004, vol. 38, no. 3, p. 311.
1. Arutyunov, V.S., Rudakov, V.M., Savchenko, V.I., and
Sheverdenkin, E.V., Theor. Found Chem. Eng., 2005,
vol. 39, no. 5, p. 487.
An increase in the oxygenꢀtoꢀethane ratio diminꢀ
ishes the ethylene formation selectivity because of the
increase in the CO formation selectivity. The ethylene
yield at low ethane conversions is practically indepenꢀ
dent of the initial oxygen concentration.
2. Sinev, M., Arutyunov, V., and Romanets, A., Adv.
Chem. Eng., 2007, vol. 32, p. 171.
13. http://kinetics.nist.gov
The kinetic model of the gasꢀphase oxidation of C1 14. Arutyunov, V.S., Dubovitskiy, V.A., Karnaukh, A.A.,
and Sinev, M.Yu., Proc. 15th Seminar on New Trends in
Research of Energetic Materials, Pardubice, Czech
Republic, 2012, part II, p. 407.
and C hydrocarbons, based on independent kinetic
data, quite adequately describes the observed regularꢀ
ities.
2
1
5. Kondrat’ev, V.N. and Nikitin, E.E., Kinetika i mekhaꢀ
nizm gazofaznykh reaktsii (Kinetics and Mechanisms of
GasꢀPhase Reactions), Moscow: Nauka, 1974.
ACKNOWLEDGMENTS
16. http://www.kintecus.com/
This study was supported by the Presidium of the
Russian Academy of Sciences through Fundamental 17. Mulla, S.A.R., Buyevskaya, O.V., and Baerns, M., Appl.
Research Program No. 3 (“Energetic Aspects of the
Deep Processing of Fossil and Renewable Hydrocarꢀ
bon Raw Materials”) and by the Division of Chemistry
and Materials Science of the Russian Academy of Sciꢀ
ences through Program No. 7 (“Development of the
Scientific Foundations of Environmentally Friendly
and ResourceꢀSaving Chemical Technologies”).
Catal., A, 2002, vol. 226, p. 73.
1
8. Jallais, S., Bonneau, L., Auzanneau, M., Naudet, V.,
and BockelꢀMacal, S., Ind. Eng. Chem. Res., 2002,
vol. 41, p. 5659.
1
9. Heracleous, E. and Lemonidou, A.A., Appl. Catal., A
004, vol. 269, p. 123.
,
2
2
0. Machli, M., Boudouris, C., Gaab, S., Find, J., Lemꢀ
onidou, A.A., and Lercher, J.A., Catal. Today, 2006,
vol. 112, p. 53.
REFERENCES
2
1. Beretta, A., Piovesan, L., and Forzatti, P., J. Catal.
999, vol. 184, p. 455.
,
1
2
3
. Asche, F., Oglend, A., and Osmundsen, P., Energy
Policy, 2012, vol. 47, p. 117.
1
2
2. Herbinet, O., Husson, B., Serinyel, Z., Cord, M.,
Warth, V., Fournet, R., Glaude, P.ꢀA., Sirjean, B., Batꢀ
tinꢀLeclerc, F., Wang, Z., Xie, M., Cheng, Z., and
Qi, F., Combust. Flame, 2012, vol. 159, p. 3455.
. Arutyunov, V.S. and Magomedov, R.N., Russ. Chem.
Rev., 2012, vol. 81, no. 9, p. 790.
. Arutyunov, V.S., Bykhovskii, M.Ya., Sinev, M.Yu., and
Korchak, V.N., Russ. Chem. Bull., 2010, vol. 59, no. 8,
p. 1528.
23. Schadel, B.T., Duisberg, M., and Deutschmann, O.,
Catal. Today, 2009, vol. 142, p. 42.
4
. Arutyunov, V.S., Sinev, M.Yu., Shmelev, V.M., and
Kiryushin, A.A., Gazokhimiya, 2010, no. 1 (11), p. 16.
2
4. Magomedov, R.N., Proshina, A.Yu., Peshnev, B.V., and
Arutyunov, V.S., Kinet. Catal. (in press).
5
. Arutyunov, V.S. and Krylov, O.V., Okislitel’nye prevrashꢀ
cheniya metana (Methane Oxidation Reactions), Mosꢀ
cow: Nauka, 1998.
Translated by D. Zvukov
KINETICS AND CATALYSIS Vol. 54
No. 4
2013