Continuous production of synthesis gas at ambient temperature from steam reforming of methane with nonthermal plasma

Article abstract of DOI:10.1246/cl.2002.1108

Nonthermal plasma steam reforming of methane was carried out with two different types of reactors such as ferroelectric packed-bed (FPR) and silent discharge (SDR) in a flow reaction system. The yields of hydrogen and carbon monoxide were much higher with

Full text of DOI:10.1246/cl.2002.1108

1
108
Chemistry Letters 2002
Continuous Production of Synthesis Gas at Ambient Temperature
from Steam Reforming of Methane with Nonthermal Plasma
ꢀ
Hajime Kabashima and Shigeru Futamura
National Institute of Advanced Industrial Science and Technology, AIST Tsukuba West, 16-1 Onogawa, Tsukuba, Ibaraki 305-8569
(Received July 10, 2002; CL-020578)
Nonthermal plasma steam reforming of methane was carried
out with two different types of reactors such as ferroelectric packed-
bed (FPR) and silent discharge (SDR) in a flow reaction system. The
yields of hydrogen and carbon monoxide were much higher with
FPR than with SDR under the same conditions. FPR could be
operated continuously for 10 h without any decrease in the yields of
hydrogen and carbon monoxide.
The both reactors employed AC power supply at 50 Hz and high
voltage up to 8.0 kV was applied for both the reactors. No
breakdowns occurred during operations within their maximum
voltages.
Methane balanced with N2 in a standard gas cylinder was
introduced to the reactor through a Teflon tubeby adjustingmethane
concentrations and flow rates with sets of mass flow controllers and
a gas mixer. Steam was supplied to the reactors by humidifying gas
(
CH4/N2) in a water-bubbling type device in a thermostatic bath.
Steam reforming of methane is an important process to produce
hydrogen and/or synthesis gas.1 Industrially, steam reforming of
methane is a process where methane reacts with excessive steam at
high temperatures (>1100 K) and high pressures (>20 atm) over a
Ni-containing catalyst. In this catalytic reforming process, large
thermal energy is needed to react methane at high temperature, and
Steam concentrations were determined by a dew point hygrometer,
and its contents were controlled within the range of 0.5–2.0%.
Steam reforming of methane was carried out at room temperature
and an atmospheric pressure by using a conventional mass flow
reaction system. H2 and methane were quantified by a TCD-GC
with a packed column of Molecular Sieve 13X. CO, CO2, ethane,
ethylene, and acetylene were analyzed by TCD- and FID-GC with a
packed column of Porapak Q + N and Molecular Sieve 13X.
In this paper, each of theproduct yields for H2, CO, CO2, and C2
hydrocarbons [eq (1)] is plotted against specific energy density
;2
1;3
2
0–40% of the raw material is consumed by combustion owing to
4
the supply of the excessive heat. Therefore, nonthermal plasma has
been applied to methane reforming at lower temperatures with
point-to-point type and dielectric barrier discharge plasma reactors
for the development of cost-effective processes of synthesis gas
production.
(
SED) given by eq (2), where ‘‘Power’’ denotes the plug-in power.
Product yield(mol%)¼100  ½Product amount(mmol)Š=
Nonthermal plasma may provide a useful reaction medium for
this reaction because the reaction temperature can be kept as low as
ambient. Recent reports have shown that the reaction temperature
½Maximum amount of product
evolved from 1% methane(mmol)Š ð1Þ
5
can be decreased to 453 K in steam reforming and ambient
temperature in carbon dioxide reforming.
SEDðkJ Lꢁ Þ¼Power(kW)=½Flow rateðL min Þ=60Š ð2Þ
1
ꢁ1
6;7
With the above-
mentioned plasma reactors, however, formation of C2 hydrocarbons
is predominat via methane coupling. Therefore, addition of an
excessive oxidizing agent such as steam and carbon dioxide is
mandatory to suppress the formation of C2 hydrocarbons.
Table 1 shows the effects of reactor and H O concentration on
2
ꢁ1
methane reforming in N at 9 kJ L of SED. Gas flow rates of FPR
2
ꢁ1
ꢁ1
and SDR were fixed at 100 mLmin and 50 mLmin , respec-
tively. With an increase in H O concentration, CH conversion and
2
4
We have already reported that a ferroelectric packed-bed
reactor (FPR) has shown the higher performance compared with a
silent discharge plasma reactor (SDR) in the hydrogen generation
from water.8 Nonthermal plasma has a potential for hydrogen-
forming reactions such as hydrocarbon reforming and water
decomposition, but its scope and limitations have not been clarified
yet. It is significant to examine the reaction behavior of
hydrocarbons and steam in nonthermal plasma from the viewpoint
of its extended application to diverse chemical processes associated
with synthesis gas utilization. Also, there have been no reports on
the steam reforming of methane at ambient temperature.
2 2
the yield of C hydrocarbons decrease, while that of CO increases
irrespective of reactors. With FPR, H yield increases with H O
2
2
concentration and a maximum is observed for CO yield. These facts
can be ascribed to the occurrence of water-gas-shift reaction
(CO þ H O ! CO þ H ). H selectivity exceeds 100% for the
2
2
2
2
H O concentration of 1.5% and 2.0% because H is derived also
2
2
8
from decomposition of H O itself and water-gas-shift reaction.
2
With SDR, CH conversion at the H O concentration of 0% and
4
2
2.0% were 6.5% and 4.4%, respectively. Also, the yields of H , CO,
2
and CO2 were much lower than with FPR under the same
conditions. For methane reforming, SDR has shown the lower
In the present work, we have studied the steam reforming of
methane for synthesis gas formation at ambient temperature in
nonthermal plasma, focusing on the effect of plasma-generating
methods and the factors governing the reaction efficiencies. A
cotinuous production of synthesis gas from methane and steam has
been also examined with FPR.
performance compared with FPR as in the case of H generation
2
from water.8 Since FPR and SDR have shown the comparable
performances in the decomposition of trichloroethylene, bromo-
9
2
methane, and tetrafluoromethane in N , almost the same plasma
intensity should be obtained in both the reactors. These facts suggest
that water activatioin is the common rate-determining step for the
steam reforming of methane and H generation from water, and that
FPR and SDR used in this research were described in detail
2
9;10
elsewhere.
5
flow rate was fixed at 50 mLmin (residence time 3 s) with SDR.
With FPR, gas flow rate ranged from 50 to
00 mLmin (residence time 8.9 to 89 s). On the other hand, gas
the reaction efficiency highly depends on the plasma-generating
method. FPR and SDR belong to the same kind of barrier discharge
plasma reactor. On the other hand, corona discharge is produced in
ꢁ1
ꢁ
1
Copyright Ó 2002 The Chemical Society of Japan

Chemistry Letters 2002
1109
Table 1. Effects of reactor and H2O concentration on steam reforming of
a
methane
H2O
CH4
Yield (mol%)c
Reactor concentration conversion
(mol%)
H2
CO CO2 C2HCsb
(%)
FPR
FPR
FPR
FPR
SDR
SDR
0
36.6
27.6
25.2
22.8
6.5
22.2 3.5
25.7 14.0
26.8 12.6 10.8
0.2
6.9
1.3
0.4
0.2
0.1
0.4
0.1
1.0
1.5
2.0
0
27.1
2.4
0.7
9.9 12.6
1.3
1.3
0.7
1.8
2.0
4.4
a
Reaction conditions: methane, 1.0%; background gas, N2; SED,
ꢁ
1
b
9
kJL
.
C2 HCs denotes the hydrocarbons such as ethane,
Figure 1. Effect of specific energy density on the CH4 conversion, H2
yield, and COx yield in N2 with FPR.
c
ethylene, and acetylene. Product yield(mol%) = 100 Â ½Product
amountðmmolÞ]/[Maximum amount of product evolved from 1%
methane (mmol)].
6
the point-to-point type of plasma reactor. Our findings clearly
show that FPR works as a much better reactor for hydrogen-forming
reactions than the other three ones. For example, Kado and
coworkers have reproted steam reforming of methane with a point-
6
to-point type of plasma reactor. The maximun H2 generation rate
with this reactor was 210 ꢀmol min in 20%-CH4/80%-H2O at
ꢁ1
4
53 K. On the other hand, the maximun H2 generation rate with FPR
1
ꢁ
was 48 ꢀmol min
in 1%-CH4/2%-H2O/97%-N2 at ambient
temperature. FPR performance is 4.4-fold lower compared to that
of Kado’s reactor. However, the H2 generation efficiency of FPR is
estimated to be higher than that of Kado’s, since the initial
concentration of CH4 is 20 times lower and the reaction temperature
is lower by 150 K with FPR.
Figure 1 shows that theCH4 conversion and theyields of H2 and
COx gradually increase with an increase in SED in N2 with FPR.
Figure 2. Time profiles of the CH4 conversion, the yields of H2 and CO,
and the selectivities of H2, CO, and COx in N2 with FPR.
When SED was set at 15 kJ L 1, CH4 conversion, H2 yield, and COx
yield were 35.4, 44.4, and 34.9%, respectively. H2 selsctivity
calculated based on the CH4 conversion exceeded 100% at SED
ꢁ
compared with SDR, suggesting the different electron temperatures
in both the reactors at the same input energy densities. For steam
higher than 6 kJ L 1. Irrespective of the SED magnitude, almost the
same CH4 conversions and COx yields were obtained, i.e., carbon
balances were higher than 98%.
ꢁ
reforming of 1%-methane in N with FPR, the optimized water
2
concentration is about 2.0%. With FPR, CO selectivity as high as
x
98% or higher is constantly obtained under the optimized
The effect of gas flow rate on the yields of H2 and CO in N2 with
conditions. This is why FPR can be operated continuously for a
long time. For steam reforming of methane at ambient temperature,
FPR may be one of the best nonthermal plasma reactors.
ꢁ1
FPR was further examined from 50 to 500 mLmin of gas flow rate
under the same condition of Figure 1. With an increase in SED, the
yields of H2 and CO gradually increased at different flow rates. The
highest yields of H2 and CO were 73.4% and 29.7%, respectively at
This work was partly supported by the Grants-in-Aid from New
Energy and Industrial Technology Development Organization of
Japan (NEDO).
ꢁ ꢁ1
1
3
0.0 kJ L of SED at 50 mLmin of gas flow rate. An interesting
trend has been observed that higher H2 yields and CO yields are
obtained at higher flow rates, i.e., shorter residence times at fixed
SEDs.
References
1
2
3
4
5
J. N. Armor, Appl. Catal., A, 176, 159 (1999).
Figure 2 shows the time profiles of CH4 conversion, the yields
of H2 and CO, and the selectivities of H2 and CO, and COx in the
steam reforming of methane in N2 with FPR. This reaction was
M. A. Pena, J. P. Gomez, and J. L. G. Fierro, Appl. Catal., A, 144, 7 (1996).
T. E. Solh, K. Jarosch, and H. I. Lasa, Appl. Catal., A, 210, 315 (2001).
S. Asaoka and M. Shimura, Petrotech, 24, 226 (2001).
ꢁ
1
S. Kado, K. Urasaki, Y. Sekine, and K. Fujimoto, Chem. Commun., 2001,
4
carried out at 12 kJ L of SED for 10 h under the conditions as the
same as forFigure 1. CH4 conversion and productselectivities could
be kept constant for 10 h. The selectivities of H2 and CO were 126%
and 58% on the average, respectively. Therefore, the molar ratio of
H2 to CO was 4.3. Also, almost all of the carbon atoms in the reacted
methane could be recovered as CO and CO2 during the continuous
operation.
15.
6
7
S. L. Yao, M. Okumoto, A. Nakayama, and E. Suzuki, Energy Fuels, 15,
1295 (2001).
L. M. Zhou, B. Xue, U. Kogelschatz, and B. Eliasson, Energy Fuels, 12, 1191
(1998).
8
9
H. Kabashima, H. Einaga, and S. Futamura, Chem. Lett., 2001, 1314.
S. Futamura, A. Zhang, and T. Yamamoto, IEEE Trans. Ind. Appl., 36, 1507
(2000).
1
0
1
H. Einaga, T. Ibusuki, and S. Futamura, IEEE Trans. Ind. Appl., 37, 1476
We have shown here the effects of plasma-generating methods
and the factors governing the reaction efficiencies for steam
reforming of methane. FPR has shown the higher performance
(2001).
S. Futamura, A. Zhang, and H. Einaga, IEEE Trans. Ind. Appl., 37, 978
(2001).
1