You are on page 1of 8

Chemical Engineering Journal 238 (2014) 916

Contents lists available at ScienceDirect

Chemical Engineering Journal


journal homepage: www.elsevier.com/locate/cej

Oxidative conversion of light alkanes diluted by nitrogen, helium


or methane
V.S. Arutyunov , R.N. Magomedov, A.Yu. Proshina, L.N. Strekova
Semenov Institute of Chemical Physics, Russ. Acad. Sci., Kosygina 4, Moscow 119991, Russia

h i g h l i g h t s

g r a p h i c a l a b s t r a c t

 Gas-phase oxidation of diluted C2C5

a r t i c l e

i n f o

Article history:
Available online 12 October 2013
Keywords:
Natural gas
Light alkanes
Oxidative conversion
Partial oxidation
Oxycracking
Olens

100
90

Hydrocarbon conversion, %

alkanes proceeds via chain-branching


mechanism.
 Both hydrocarbon and oxygen
convert almost similarly in space and
time.
 Oxygen signicantly increases the
rate of the conversion and decreases
its time.
 Oxygen inhibits the formation of
heavy products.
 The nature of diluting gas
signicantly inuences on the
conversion.

80
70
60
50

C5H12
C2H6

40

C4H10

30

C3H8

20

10
0
450

500

550

600

650

700

750

800

Temperature, C

a b s t r a c t
Abundant resources and low prices make natural gas very attractive as petrochemical feedstock and
increase the interest to its oxidative conversion to basic petrochemicals. The paper presents results of
gas-phase oxidative conversion of light C2C5 alkanes diluted by nitrogen, helium or methane
([CnH2n+2]0 6 5%) at low oxygen [O2]0/[CnH2n+2]0 = 0.21, atmospheric pressure and temperatures
500750 C. Even signicantly diluted, light alkanes display a chain-branching behavior at oxidation with
a sharp transition trough a self-ignition limit. Oxygen signicantly increases conversion and decreases
reaction time, especially at intermediate temperatures below that of thermal cracking. The nature of
diluting gas signicantly inuences on the process.
2013 Elsevier B.V. All rights reserved.

1. Introduction
Recent development in unconventional gas production has multiplied its resources and decreased the price vefold relative to that
of petroleum [1]. Abundant resources and low prices make unconventional natural gas very attractive as petrochemical feedstock
and increase the interest to its oxidative conversion to basic
petrochemicals. But the oxidative conversion of light alkanes is
much complex technological task than conversion of liquid hydrocarbons. It needs high temperatures and intensive heat uxes.
Because the most efcient way to introduce heat into hydrocarbon
Corresponding author.
E-mail address: arutyunov@chph.ras.ru (V.S. Arutyunov).
1385-8947/$ - see front matter 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.cej.2013.10.009

gas is the oxidation of small fraction of hydrocarbons to CO2 and


water, it explains the rising interest to the partial oxidation of light
C2C5 hydrocarbons that compose the principal admixture to
methane in natural gas.
At high temperatures, which are necessary even for the catalytic
activation of light alkanes, gas-phase processes play signicant role
and in some cases even predominate. It is not only makes very
important thorough elucidation of the role of gas-phase reactions
at catalytic conversion of light hydrocarbons but, as well, opens
the possibility for homogeneous gas-phase processes which need
no catalyst, thus signicantly facilitates the conversion of complex
mixtures of hydrocarbons with very different reactivity such as
real natural gases. Especially it is important to elucidate the role
of oxygen because it not only increases the rate of reaction and

10

V.S. Arutyunov et al. / Chemical Engineering Journal 238 (2014) 916

decreases its temperature, but, as well, inhibits the formation of


heavy products including soot and decreases the selectivity to such
important products as olens.
Temperature is the main parameter that governs the oxidation
of light alkanes and determines the formation of principal products
at their partial oxidation. But temperature limits of formation of
such principal products as oxygenates, olens and syngas. at partial oxidation of methane and its heavier homologues signicantly
differ [2,3] which makes it possible selectively oxidize C3+ alkanes
just in the natural gas without conversion of methane itself [4,5].
The aim of this study is the investigation of the kinetics of
gas-phase oxidation of light C2C5 alkanes diluted by inert gas or
methane ([CnH2n+2]0 6 5%) at low oxygen [O2]0/[CnH2n+2]0 = 0.21,
atmospheric pressure and in the temperature range 500750 C.
We believe that this investigation will provide valuable material
for more detailed understanding of the mechanism of the oxidation
of light alkanes in a wide range of conditions interesting for the
practical realization.

2. Experimental part
Experiments were conducted at atmospheric pressure in a ow
quartz reactor with i.d. 14 mm and the length of isothermal zone
200 mm. Three independent electrical heaters with automatic regulation of temperature by separate thermocouples inside the rector
provide almost uniform temperature prole. Taking into account
the surface of quartz pockets for thermocouples, the surface to volume ratio S/V was 5.4 mm1. To elucidate the inuence of the inner surface of the reactor, in some experiments the second
reactor with i.d. 8 mm and S/V ratio 13.3 mm1 was used. In several experiments a thin (4 mm) layer of the industrial catalyst
IPM-0.2 for isomerization of pentanehexane fraction with 0.2
Pt/cAl2O3 (fraction 0.50.8 mm) was placed at the entrance of
the reactor. The contact time with catalyst (0.06 s) was at least
20 times less than the complete reaction time.
Experiments were conducted in the temperature range
500750 C with the use of three diluting gases: nitrogen, helium
or methane. Prepared in advance hydrocarbonoxygenbuffer gas
mixture was fed from cylinder trough the gas controller with a
gas ow in a range from 2.4 to 25 ml/s which at 700 C corresponds
to the residence time from 0.35 to 3.6 s. The most of experiments
were made at alkane concentration in buffer gas 5 mol.% with the
mole ratio [O2]0/[CnH2n+2]0 from 0.2 to 1. But in some experiments
aimed on the investigation of the inuence of alkane concentration,
it was varied from 2 to 9.2 mol.% at constant ratio
[O2]0/[CnH2n+2]0 = 0.5. Taking into account the low concentration
of hydrocarbon, experimental conditions can be considered as almost isothermal, with experimental heating in the most cases below 5 C.
Products were analyzed by GC equipped with several columns
and detectors. Carbon was usually balanced within 4%. The concentration of water was calculated from the oxygen balance. More
detailed description of experimental equipment and conditions is
given elsewhere [6,7].
Experimental investigation of the oxidation of ethane was
accompanied by the kinetic modeling with the use of the model
for partial oxidation of C1C2 hydrocarbons, specially developed
for the temperature range 6001200 K and pressures from 1 to
100 atm, according to principles described in [8]. This model
includes 454 elementary reactions [6] with all kinetic parameters
taken from independent literature sources, predominantly NIST
database [9]. These parameters were not varied or adjusted
anymore during the course of calculations. All calculations were
conducted for isothermal conditions with the use of the package
Kintecus V450 [10].

3. Results and discussion


The temperature of the noticeable radical-chain oxidative conversion of any particular hydrocarbon strongly depends on the
composition of reaction mixture and reaction conditions. Due to
numerous literature data [11], below 600 C the noticeable oxidative conversion of ethane proceeds only in the presence of catalyst
or on the surface of the reactor. Among the main products, except
of ethylene, there are CO2 and H2O. At that the rise of the conversion with temperature is slow enough. If there is a signicant void
volume, above 600650 C the fast increase of the gas-phase process is observed which is accompanied by the decrease of the
selectivity to CO2 and H2O and the increase of the selectivity to
CO, H2 and CH4.
The relative input of heterogeneous and homogeneous reactions depends on the properties of catalyst, its amount and the void
volume of reactor, thus determining the composition of products
[11]. At gas-phase oxidation the interaction of alkyl radicals with
oxygen and oxygen containing radicals lead to the formation of
oxygenates with their subsequent oxidation mainly to CO. In the
presence of catalyst, on the contrary, the deep oxidation on its surface leads to signicant yield of CO2 and, as a result, low ratio
CO/CO2. As a consequence, the catalytic oxidation (oxycracking,
oxypyrolysis) of ethane on the active catalysts usually reveals the
strong quantitative and time difference in the conversion of oxygen
and ethane. For example, in our experiments in the presence of a
thin layer of Pt catalyst at the entrance of the reactor, it was observed almost complete conversion of oxygen (95%) already at
low temperatures beginning from 300 C (Fig. 1). Complete conversion of oxygen was accompanied by very low conversion of ethane
(10%) corresponding to its stoichiometric deep oxidation to CO2
and H2O, the only detectable products at these conditions. Because
oxygen practically completely consumes in the thin layer of the
catalyst, further conversion proceeds at oxygen-free conditions.
So the further increase of the conversion of ethane was observed
only at temperatures above 650 C (Fig. 1b) which corresponds to
its thermal gas-phase cracking with increasing the selectivity to
ethylene.
Very different behavior was observed in a void reactor. The conversion of both reagents begins only at temperatures above 600 C
and proceeds very similarly (Fig. 1). In this case oxygen presents in
the signicant part of the reactor thus providing real oxidative conditions. Very similar observations were made in [12], where complete conversion of oxygen at the oxidative dehydrogenation of
ethane in the presence of mixed alkaline-earth and rare-earth oxides was attained already at 600 C. This conversion of oxygen was
accompanied by relatively low conversion of ethane. But in the absence of catalyst even at 750 C the conversion of oxygen was only
80% with that of ethane 60%, thus both reagents were converted
very similarly.
Although quartz surface of reactor walls cannot be considered
as absolutely inert, experiments in reactors with different diameters (different surface to volume ratio S/V) have shown very similar
results. The only difference in the reactor with smaller diameter
and thus of about 2.5 times higher ratio S/V was the temperature
shift of about 50 C to a higher temperatures for attaining the same
conversions, which can be interpreted as the retarding effect of the
surface, presumably, due to the surface termination of radicals.
The similarity in the conversion of both reagents that was observed at high temperatures up to the temperature of complete
oxygen conversion (Fig. 2) is one of the main features of the gasphase oxidation of light alkanes. At higher temperatures the rate
of the increase of the conversion of ethane with the increase of
temperature signicantly drops. This further increase is presumably the sequence of the increasing input of the thermal cracking
and thermal dehydrogenation of ethane. As it follows from Fig. 2,

11

V.S. Arutyunov et al. / Chemical Engineering Journal 238 (2014) 916

100

90

80

Ethane conversion, %

Oxygen conversion, %

90

100

()

70
60
50
40
30
20

80
70
60
50
40
30
20
10

10
0
100

(b)

200

300

400

500

600

700

0
100

800

200

300

400

500

600

700

800

Temperature,

Temperature,

Fig. 1. Oxygen (a) and ethane (b) conversion as function of temperature in the presence (j) and the absence (d) of Pt catalyst ([C2H6] = 5%, [O2]0 = 2.5%, balanced by N2).
Residence time in the catalyst zone 0.06 s, residence time in the void volume 0.80.9 s.

Oxygen conversion, %

90

100

(a)

90

80
70
60
50

2 =1%

2 =5%

2 =2,5%

40
30

Ethane conversion, %

100

80

(b)

2 =5%
2 =2,5%

70
60
2 =1%

50
40
30

20

20

10

10

0
550 575 600 625 650 675 700 725 750 775

0
550 575 600 625 650 675 700 725 750 775

Temperature,

Temperature,

Fig. 2. Temperature dependence of conversion of oxygen (a) and ethane (b) ([O2] = 1% (), 2.5% (d) and 5% (N)), [C2H6]0 = 5%, tr = 2 s, balanced by N2). Symbols and dashed
lines experimental results, solid lines results of kinetic simulation.

oxygen increases the conversion of ethane. As it will be shown below, it also decreases the time of the attaining of maximal ethane
conversion at the same temperature. But the rising input at high
temperatures of thermal reactions with high activation energy
leads to the decreasing of the accelerating inuence of oxygen on
the conversion with temperature.
The most important feature of the gas-phase oxidation of light
alkanes is the sharp increase of the conversion in a narrow (less
than 50 C) temperature interval (Fig. 2). Such critical transition
from negligible to very fast reaction (ignition of reaction) is an
evident consequence of branched-chain nature of the oxidative
conversion of light alkanes which takes place even at very low concentration of reagents. The critical transition to self-ignition region
is conrmed by very high, more than 100 kcal/mol, value of formal
activation energy in this temperature range. Thus, chain-branching
is a principal feature that makes this process fast enough even
without the use of catalyst.
At oxygen concentration [O2]0 = 2.5% and 5%, which is signicantly less than stoichiometric ratio, it consumes practically completely (Fig. 2). But at [O2]0 = 1% the residual concentration
[O2] = 0.05% usually was observed, which means that the reaction
stops on some unclear reasons before the full consumption of
decient component. Approximately the same incomplete conver-

sion of oxygen was experimentally observed at the partial oxidation of methane [2,3].
Kinetic simulation not only well describes the critical transition
to self-ignition at these conditions, but gives satisfactory quantitative description of all main experimental results. Taking into account the principally independent design of the model [6,8] it is
possible to conclude that for this reaction the existing theoretical
description well enough corresponds to real kinetics. Among the
still unclear discrepancies between calculated and experimental
results are the incomplete conversion of oxygen (Fig. 2a) and
experimental decrease of the time of achievement of complete
conversion with increase of oxygen. Simulation gives the opposite
dependence with the increasing of this time.
Temperature dependence of the selectivity of the formation of
principal products is presented in Fig. 3. At low conversion of reagents at temperatures below 650 C practically solely oxidation
products are ethylene and water. (Simulation represents the sum
of water and hydrogen peroxide both of which were not detected
experimentally). With the increase of temperature to 700 C at
which under these conditions oxygen is completely converted,
both experimental and calculated carbon selectivity to ethylene
drops from 100% to 70% and with the further increase of
temperature changes insignicantly. In this temperature interval

12

V.S. Arutyunov et al. / Chemical Engineering Journal 238 (2014) 916

100

80

()

90

Selectivity on , %

80

Selectivity on , %

(b)

70

70
60
50
40
30
20

60
50
40
30
20
10

10
0
600

625

650

675

700

725

750

0
600

775

625

650

Temperature,

675

700

725

750

775

Temperature,

Fig. 3. Carbon (a) and hydrogen (b) selectivity of the formation of principal products as a function of temperature: C2H4 (d), CO (j), CH4 (), CO2 (N), H2 (), H2O (4), C[Hy
(h). Symbols and dashed lines experimental results, solid lines results of kinetic simulation. ([C2H6]0 = 5%, [O2]0 = 2.5%, balanced by N2, tr = 2 s).

carbon selectivities of the formation of CO and CH4 gain 1820%


and 78%, respectively, also without signicantly change at further
increase of temperature. Simultaneously, in the whole considered
temperature range almost linearly decrease of the hydrogen selectivity of the formation of H2O was observed, accompanied by the
corresponding increase of the selectivity of the formation of H2.
As it was mention above, it can be attributed to the increase of
the rate of the thermal dehydrogenation of ethane relative to the
rate of its oxidative dehydrogenation, as well as to the redistribution of oxygen in the favor of the formation of COx.
Thus, at low temperatures and, correspondingly, low conversions it takes place mainly oxidative dehydrogenation of ethane.
At higher temperatures noticeably increases the input of parallel
reactions of its thermal cracking and dehydrogenation, as well as
that of reactions of deep oxidation of ethane and ethylene. But with
the exclusion of the redistribution of hydrogen between water and
molecular hydrogen above 700750 C, the selectivity of the formation of other products changes insignicantly (Fig. 3).
Up to the complete consumption of oxygen the concentrations
of all reaction products monotonously increase with time (Fig. 4).
Concentrations of ethylene and water, which is evidently the primary reaction products, up to this moment are practically identical, in the compliance with the gross-reaction

C2 H6 0:5O2 ! C2 H4 H2 O

6,0

()

5,0

Concentrations, % mol.

Concentrations, % mol.

6,0

Hydrogen also forms in signicant amounts already at short


time of reaction due to the high rate of thermal dehydrogenation
at this temperature (700 C). All other products form at lower concentration and with some delay, presumably, as a result of cracking
and deep oxidation of forming ethylene. This time delay is especially evident at kinetic simulation (Fig. 4b). Simulation describes
experimental results well enough with the exception of shorter
time for the complete oxygen conversion, which is the result of
some higher calculated rate of the oxidation. Besides, simulation
predicts very sharp termination of oxygen conversion, while experiments demonstrate overextended oxygen conversion and even
some its residual concentration, as it was mentioned above.
Experiments show that the increase in the initial oxygen
concentration causes a monotonic decrease in the ethylene, total
hydrocarbon, and hydrogen formation selectivities, enhancing the
oxygen-containing product (CO and water) formation selectivities.
After the complete conversion of oxygen that under these
conditions needs slightly above one second, the reaction practically
stops. The concentrations of formed products in the considered
time range change insignicantly, in spite of the substantial
amount of unconverted ethane. Only some rise of the concentrations of hydrogen and ethylene at the expense of that of ethane
can be observed. It can be the consequences of the low rate of
the ethane pyrolysis without oxygen at these conditions.
Steady concentration of water after complete conversion of
oxygen means the absence of the noticeable input of the gas-phase

4,0
3,0
2,0
1,0
0,0
0,0

0,5

1,0

1,5

2,0

2,5

Residence time, s

3,0

3,5

4,0

(b)

5,0
4,0
3,0
2,0
1,0
0,0
0,0

0,5

1,0

1,5

2,0

2,5

3,0

3,5

4,0

Residence time, s

Fig. 4. Concentration of reagents and products as a function of the residence time (a) experimental results: C2H6 (Formate this caption
CO (j), CH4 (), CO2 (N); (b) results of kinetic simulation ([C2H6]0 = 5%, [O2]0 = 2.5%, balanced by N2, T = 700 C).

), O2 (s), C2H4 (d), H2O (4), H2 (N),

13

V.S. Arutyunov et al. / Chemical Engineering Journal 238 (2014) 916

steam conversion of ethane at these conditions. It corresponds to


[13], where it was shown that at temperatures below 900 C, above
which the active gas-phase pyrolysis of ethane and its heavier
homologues in the absence of catalyst takes place, the conversion
of steam is very low and does not exceed 5%.
Our experiments have shown that the oxidative cracking of all
diluted C2C5 alkanes proceeds by the similar branched-chain
mechanism. So it is natural that the properties of diluting gas plays
signicant role in this process. Fig. 5 presents the temperature
dependence for oxidative cracking of ethane and propane diluted
by different gases.
The replacement of nitrogen as diluting gas by helium, which
has higher thermal conductivity, evidently increases the temperature of the reaction. Nitrogen has higher molar thermal capacity
and thus more effective in quenching of self-ignition of stoichiometric hydrocarbonair mixtures. But in small ow reactor and in
strongly diluted rich alkaneoxygen mixtures with low thermal effect the higher thermal conductivity of helium, presumably, plays
more important role. It can slow down low but nevertheless
important self-heating of the process, thus increasing the temperature necessary for the attaining of the high conversion of oxygen
(Fig. 5a). The different inuence of these gases on some of threeand monomolecular reactions also cannot be excluded.
Of special interest is the investigation of the oxidation of light
hydrocarbons in a large excess of methane which corresponding
to real natural and associated gases. Methane has signicantly
higher thermal capacity than nitrogen or helium. The investigation
of self-ignition delays of C1C5 alkanes in shock tube [14] has
shown that self-ignition delay for methane about order of magnitude higher than that for C2C5 alkanes. It indicates signicantly
lower reactivity of methane in comparison even with ethane. At
reaction time of about one second noticeable pyrolysis of methane
was observed only at temperatures above 1000 C [15,16]. But in
conditions of parallel radical oxypyrolysis of heavier alkanes methane cannot be considered as inert gas even at signicantly lower
temperatures, so its inuence is more complex for interpretation.
Fig. 5 shows that the presence of methane cardinally changes
the temperature dependence of the process. It practically completely diminishes the sharp temperature threshold of the reaction
for both ethane (Fig. 5a) and propane (Fig. 5b). At that, at low temperatures the presence of methane increases the conversion of
oxygen, which formally can be considered as the promotion of
the oxidation of ethane by methane. We believe that it is due to
the very effective chain-branching at oxidation of methane, as it

100

100

()

90

80

Oxygen conversion, %

Oxygen conversion, %

90

was suggested in [2,3]. But at the same time methane decreases


the conversion at high temperatures, which is presumably due to
its high thermal capacity that move this chemical system out of
the self-ignition region. It is worth to note that this transition from
promotion to inhibition by methane takes place at temperatures in
the vicinity of 650 C. It is just the temperature at which the concentration of methyl peroxide radicals, which provide very effective low-temperature chain-branching in methane [2,3], becomes
negligible.
In the absence of a sharp transition trough a chain-branching
limit (limit of self-ignition) it is possible to estimate the effective
activation energy of the process. In the assumption of the rst
order of the reaction on oxygen the effective activation energy of
the oxidation of ethane in a large excess of methane is
EC2 H6 35 kcal/mol. The same value for the oxidation of propane
in a large excess of methane is EC3 H8 32 kcal/mol. These gures
well correspond to the effective activation energy 30 kcal/mol
which is typical for the high temperature oxidation of alkanes,
heaver, than methane [17].
The kinetics the oxidation of ethane in a large excess of methane also demonstrates the absence of self-ignition. The conversion
of oxygen at T = 700 C increases almost linearly with time and signicantly slower than in nitrogen or helium. The composition of
products at the oxidation of ethane in a large excess of methane
is very similar to that with inert diluting gases. But with the increase of temperature there appears propylene which forms in signicantly higher concentrations than at the oxidation in inert
diluents, presumably, due to the interaction of formed ethylene
with methyl radicals. Concentrations of reagents and products
during almost all reaction change linearly with time, at that
D[C2H6]  D[O2]. But the carbon content in products exceeds that
in converted ethane which evidences of the partial conversion of
methane.
In similar way oxygen steadily increases the conversion and decreases the time of conversion of all C2C5 alkanes as it can be seen
for n-butane in Fig. 6.
But the effect of oxygen on the conversion of light alkanes
strongly depends on temperature. Oxygen especially increases
the conversion at intermediate (600700 C) temperatures, but
much lesser the conversion at lower temperatures, when catalytic
oxidation usually prevails, and at higher temperatures, when
noticeable input of thermal cracking takes place (Fig. 7a). This gure presents the conversion of n-butane on oxygen at various temperatures and xed time enough for complete oxygen conversion

70
60
50
40
30
20
10

(b)

80
70
60
50
40
30
20
10

0
550 575 600 625 650 675 700 725 750 775

Temperature,

0
500

550

600

650

700

750

800

Temperature,

Fig. 5. Conversion of oxygen as a function of temperature at the oxidation of ethane (a) and propane (b) in different diluting gases: N2 (d), He (N) and CH4 (j) ([CxHy]0 = 4.5
5%, [O2]0 = 2.42.5%, tr = 2 s).

14

V.S. Arutyunov et al. / Chemical Engineering Journal 238 (2014) 916

3,0

Time of 90%-butane conversion, s

100

Butane conversion, %

90
80
70
60
50
40
30
20
10
0
0,0

()
0,5

1,0

1,5

2,0

2,5

3,0

3,5

2,5
2,0
1,5
1,0
0,5

(b)
0,0

4,0

Concentration of oxygen, %

Residence time, s

Fig. 6. Conversion of n-butane as a function of the residence time at different [O2]0: (a) ([O2] = 0% (j), 1% (), 2.5% (d) and 5% (N)) and the time of 90%-n-butane conversion as
a function of initial concentration of oxygen (T = 750oC, [n-C4H10] = 5%, balanced by N2).

and obtaining practically stable concentration of main reaction


products (see Fig. 4).
The concentration of oxygen (the ratio [CnH2n+2]/[O2]) is also
very important. The most efciently oxygen increases the rate of
conversion at its low concentrations when [CnH2n+2]/[O2] > 2
(Fig. 7b). With the increase of oxygen its relative promoting effect
decreases.
It is naturally that the heavier is alkane the more diverse is the
set of obtained oxidation products. At high temperature oxidation
of n-butane except of C3C4 olens the formation of such products
as C4H6 (butadiene) and acetylene were observed, which along
with C3H4 (propadiene and methyl acetylene) to some extent can
be considered as soot precursors [18]. Experiments have shown
that the increase of oxygen evidently decreases the formation of
butadiene and to some extent the formation of acetylene. Thus it
can be concluded that oxygen really inhibits the formation of heavy condensation products leading to formation of soot and tar.
At relatively low temperature (600 C) the increase of oxygen
from 1% to 10% weakly inuenced on the carbon selectivity of
formation of main products of oxidative cracking of n-pentane,
except of some decrease of C3H6, C2H6 and CH4 in the favor of
CO. At that the selectivity of the formation of ethylene (30%) in
this range of oxygen concentrations was practically constant. But

100

90

80

Butane conversion, %

Butane conversion, %

100

()

90

70
60
50
40
30
20

(b)

80
70
60
50
40
30
20
10

10
0

at higher temperature (750 C) the selectivity of the formation of


ethylene also linearly dropped from 47% to 39% with the increase of oxygen, which was accompanied by a fast linear increase
of CO from 3% to 35%. It evidences the increasing input of secondary processes of the deeper oxidation of the reagent and products.
At that, at oxygen concentrations less than 5% the carbon disbalance increased, presumably, due to the increase of the formation
of undetectable heavy products. At the oxidation of pentaneoxygen mixtures diluted by nitrogen with oxygen concentrations more
than 5% (C5H12/O2 6 1), there was revealed the NTC (Negative
Temperature Coefcient) region at temperatures from 350 to
500 C. At lower concentrations of oxygen (C5H12/O2 P 2) only
the transition to self-ignition at 500 C was observed.
The increase of the number of carbon atoms in alkane steadily decreases the temperature of the fast gas-phase conversion
(self-ignition limit). Due to temperature range of fast gas-phase
conversion three main groups can be separated: methane, ethane
butane, pentane and heavier (Fig. 8). As it was mentioned
above the gas-phase conversion of methane needs temperatures
above 1000 C [15,16]. So, it makes possible selective gas-phase
oxidation of C3+ alkanes in complex gas mixtures, such as real
natural or associated gases, without conversion of methane itself
[4,5].

Concentration of oxygen, %

0
500

550

600

650

700

750

800

Temperature,

Fig. 7. Conversion of n-butane at different temperatures as a function of initial concentration of oxygen (a) (T = 600 C (), 650 C (j),700 C (N) and 750 C (d)) and at
different initial concentration of oxygen as a function of temperature (b) ([O2] = 0% (j), 1% (), 2.5% (d) and 5% (N)) ([n-C4H10] = 5%, tr = 2 s, balanced by N2).

15

V.S. Arutyunov et al. / Chemical Engineering Journal 238 (2014) 916

100

90

80

Oxygen conversion, %

Hydrocarbon conversion, %

90

100

()

70
60
50
40
30
20
10
0
450

(b)

80
70
60
50
40
30
20
10

500

550

600

650

700

750

800

0
450

500

550

Temperature,

600

650

700

750

800

Temperature,

Fig. 8. Temperature dependences of the oxidative conversion of C2C5 alkanes (C2H6 (), C3H8 (j), n-C4H10 (N) and n-C5H12 (d)). ([CnH2n+2] = 5 mol.%, [O2] = 2.5 mol.%, tr = 2 s,
balanced by N2).

100

Hydrocarbon conversion, %

90
80
70
60
50
40
30
20
10
0
400

450

500

550

600

650

700

750

800

Temperature, C
Fig. 9. Pilot unit for the selective oxy-cracking of heavy gas-phase hydrocarbons (a) and results of pilot testing at the conversion of the typical oil associated gas (b) (C3H8 (j),
n-C4H10 (N), n-C5H12 (d), n-C6H14 ()). The composition of gas (mol.%): CH4 81.0; C2H6 1.22; C3H8 17.0; RC4H10 2.81; RC5H12 1.50; n-C6H14 1.83. The amount of
added air 15%. tr  1 s.

This possibility was used in a new technology for the processing of raw natural and associated oil gases. The signicant admixture of heavy hydrocarbons with low methane (octane) numbers
makes these gases inclined to detonation and also to tar and soot
formation at combustion. Thus, these gases are unacceptable for
the direct use as a motor fuel. As a result, many oil and gas producing sites have to specially deliver liquid motor fuel for self
power supply simultaneously aring signicant amount of associated gases.
The new technology uses the discussed above possibility of the
selective gas-phase oxycracking of heavier alkanes into lighter molecules with high octane numbers and less inclined to tar and soot
formation. The pilot testing of the unit with gas ow 20 m3/h has
shown that at temperatures of about 750 C and times of about 1
s in the presence of several percents of oxygen (in air) it is possible
to convert more than 90% of admixtures of C5+ hydrocarbons and up
to 80% that of C3C4 fraction (Fig. 9). Such treatment converts heavier gas-phase alkanes mainly into H2, C2H4, CH4, and CO with a
minor admixture of C2H6 and C3H6 thus making gas well acceptable
for the feeding gas-piston and gas turbine engines.

Besides of the technological simplicity, this gas-phase process


has other evident advantages:
(1) It let to utilize for power generation almost all initial caloric value of associated oil gas.
(2) It forms no waste products that need separate utilization.
(3) It needs no catalyst or any other active storage.
Due to our estimations, the more wide use of associated oil gas
for self power generation let to decrease its aring by 2030%.
4. Conclusions
 Gas-phase oxidation of light C2C5 alkanes even at signicant
dilution displays a chain-branching behavior with a sharp transition trough a self-ignition limit. Temperature of this transition
signicantly decreases with the increase of the number of carbon atoms in the alkane.
 If the alkane is in excess relative to the stoichiometric ratio, at
temperatures above the self-ignition limit both reagents

16




V.S. Arutyunov et al. / Chemical Engineering Journal 238 (2014) 916

convert almost similarly in space and time up to the complete


conversion of oxygen.
Oxygen signicantly increases conversion and decreases reaction time. After the complete conversion of oxygen the
further conversion of ethane and its reaction products
evidently slows down. The increasing input of the thermal
cleavage of CC bonds in C3C5 alkanes decreases at high
temperatures the rate of the increase of the conversion of
oxygen and makes it incomplete even at the highest temperatures of this study.
Oxygen especially increases the conversion at intermediate
temperatures but much lesser at lower temperatures when
the catalytic oxidation usually prevails and at higher temperatures when the noticeable input of the thermal cracking takes
place.
The nature of diluting gas signicantly inuences on the conversion. Methane promotes the conversion at low (<650 C)
temperatures but inhibits it at higher temperatures.
Oxygen inhibits the formation of heavy products including precursors of tar and soot formation.
The high rate of gas-phase oxycracking of light alkanes and the
signicant difference in the reactivity of methane and its heavier homologues can be used for the selective processing of
low admixtures of heavier alkanes in natural and associated
oil gases.

References
[1] F. Asche, A. Oglend, P. Osmundsen, Energy Policy 47 (2012) 117.
[2] V.S. Arutyunov, O.V. Krylov, Oxidative Conversion of Methane, Nauka, Moscow,
1998 (in Russian).
[3] V.S. Arutyunov, Oxidative Conversion of Natural Gas, URSS, Moscow, 2011 (in
Russian).
[4] V.S. Arutyunov, M. Ya, M. Bykhovskii, M.Yu. Sinev, V.N. Korchak, Russ. Chem.
Bull. 59 (8) (2010) 1528.
[5] V.S. Arutyunov, Russ. J. Gen. Chem. 81 (12) (2011) 2557.
[6] R.N. Magomedov, A.Yu. Proshina, V.S. Arutyunov, Kinet. Catal. 54 (4) (2013)
383.
[7] R.N. Magomedov, A.Yu. Proshina, B.V. Peshnev, V.S. Arutyunov, Kinet. Catal. 54
(4) (2013) 394.
[8] M. Sinev, V. Arutyunov, A. Romanets, Kinetic models of C1C4 alkane oxidation
as applied to processing of hydrocarbon gases: principles, approaches and
developments, in: G.B. Marin (Ed.), Advances in Chemical Engineering, vol. 32,
Elsevier, 2007, pp. 171263.
[9] http://kinetics.nist.gov.
[10] http://www.kintecus.com/.
[11] V.S. Arutyunov, R.N. Magomedov, Russ. Chem. Rev. 81 (9) (2012) 790.
[12] S.A.R. Mulla, O.V. Buyevskaya, M. Baerns, Appl. Catal. A 226 (2002) 73.
[13] B.T. Schadel, M. Duisberg, O. Deutschmann, Catal. Today 142 (2009) 42.
[14] A. Burcat, K. Scheller, A. Lifshitz, Comb. Fl. 16 (1971) 29.
[15] V.S. Arutyunov, V.I. Vedeneev, Russ. Chem. Rev. 60 (1991) 1384.
[16] Ch. Keramiotis, G. Vourliotakis, G. Skevis, M.A. Founti, C. Esarte, N.E. Sanchez,
A. Millera, R. Bilbao, M.U. Alzueta, Energy 43 (2012) 103.
[17] O. Herbinet, B. Husson, Z. Serinyel, M. Cord, V. Warth, R. Fournet, P.A. Glaude,
B. Sirjean, F. Battin-Leclerc, Z. Wang, M. Xie, Z. Cheng, F. Qi, Comb. Fl. 159
(2012) 3455.
[18] C.S. McEnally, L.D. Pfefferle, B. Atakan, K. Kohse-Hoinghaus, Prog. En. Comb.
Sci. 32 (2006) 247.