Journal of the American Chemical Society
Communication
contain CO concentration up to 50−70%. Bi-reforming can
2
also be used to recycle CO emissions from sources such as flue
2
gases from fossil fuel (coal, petroleum, natural gas, etc.)-
burning power plants, exhaust of cement factories, and varied
industries.
Bi-reforming to metgas is also adaptable for reforming shale
gas. Dry shale gas is essentially methane (>98%). Wet shale gas
also contains ethane, propane (which can be dehydrogenated to
ethylene and propylene), and some higher hydrocarbon
homologues (eq 5).
3
C H
+ (3n − 1)H O + CO →
n
(2n+2)
2
2
(
3n + 1)CO + (6n + 2)H → (3n + 1)CH OH
(5)
2
3
Figure 1. Single pass CH and CO conversion in bi-reforming of
4
2
Metgas can also be obtained from other natural carbon
sources including coalbed methane and methane hydrates as
well as biomass. Recent findings of significant shale gas
methane over 15-NiO-MgO catalyst at 7 atm and 830 °C: 100
−1
mL·min of CH /CO /H O with a molar ratio of 3/1.2/2.4 and
4
2
2
4
−1 −1
GHSV of 6 × 10 mL·h ·g catalyst.
(
methane) deposits considerably increased the overall reserves
of natural gas. Shale gas production is made possible by
horizontal drilling and rock fracking using high-pressure water,
chemical additives, and sand. The environmental consequences
conditions, the catalytic system is able to perform both dry and
steam reforming simultaneously with ease. This indicates that
11
of these techniques are still under debate.
steam and CO reforming on nickel catalyst involve similar
2
1
,13,18,19
Preferentially, bi-reforming is carried out under 5−30 atm of
kinetics.
The selectivity for CO and H were 100% and
2
pressure and 830-950 °C temperature, producing metgas,
98%, respectively. When the temperature was increased from
830 to 910 °C, the conversion of both methane and CO2
increased. Methane conversion increased about 15% to reach
12
which can be directly used in the methanol synthesis. Most of
9
,13
the reported studies on dry and combinations of dry
steam
and
1
4−17
reforming have been conducted at atmospheric
86% at 910 °C. The H /CO ratio decreased only very slightly
2
pressure and mostly with pure methane. In order to approach
conditions closer to practical applications, we studied bi-
reforming of methane as well as of natural gas at pressures up to
from 1.99 to 1.97. As expected, however, from the
thermodynamics of endothermic reactions, with an increase
in number of moles (Le Chatelier’s principle), the conversion
of methane decreased with increasing pressure (from 71% at 7
6
,7
30 atm. The reactions were conducted in a tubular flow
reactor system built specially for this purpose and suitable to
withstand pressures up to 100 atm and needed temperatures of
atm to about 52% at 28 atm). On the other hand, the H /CO
2
ratio increased slightly from 1.99 to 2.02 when the pressure was
increased from 7 to 28 atm. Doubling the amount of steam and
CO (CH /CO /H O with a molar ratio of 3/2.4/4.8) at 7 atm
800−1000 °C provided by external heating. All the surfaces in
contact with the catalyst and reacting gases at the high
temperatures were made of alumina to avoid any side reactions
or possible deterioration of the reactor materials. Quartz, which
is often used in reforming studies at atmospheric pressure, was
inadequate, leading to coke formation and eventually reactor
clogging. The preferred catalyst used was based on NiO
deposited on magnesium oxide, i.e., NiO/MgO, which was
active and stable over extended continuous durations (320 h),
as are related supported metal oxides. The NiO content in
NiO/MgO can be between 5 and 35%. A gas feed composition
of CH /CO /H O with a molar ratio of 3/1.2/2.4 was typically
2
4
2
2
increased the methane conversion from 71% to 85%. The
experiments were carried out in a single pass mode, but the
unreacted feed gases can be recycled to improve the overall
conversion.
When the gas hourly space velocity (GHSV) was increased
5
−1 −1
10-fold to 6 × 10 mL·min ·g , methane conversion
decreased only by about 2% but remained stable. CO2
conversion also slightly decreased, as did the H /CO ratio to
2
about 1.95−1.97.
The same 15-NiO-MgO catalyst was also tested under the
studied pressures for the bi-reforming of natural gas (composed
of methane containing ethane, propane, butane, and higher
hydrocarbons). No hydrocarbons other than unreacted
methane were detected in the products, indicating that all
higher hydrocarbons had reacted. At 7 atm, again stable CO2
4
2
2
used with nitrogen as an internal reference. Ni deposited on
MgO had been previously shown to be an effective and stable
13
catalyst for dry reforming reaction.
Reforming is frequently affected by coking, involving the
deposition of carbon in the form of soot or coke on the catalyst
(
reducing strongly its activity), as well as parts of the
and CH conversion was observed for the duration of the
experiment (160 h). The conversion of natural gas was about
4
equipment. Carbon may be formed by both CH (natural
4
gas) decomposition and CO disproportionation (Boudouard
reaction). The relative contributions depend on the reaction
conditions and catalyst used. The undesired formation of
carbon is, however, largely prevented by the presence of steam
and short residence times in the flow reactor.
As an example, after activation at 850 °C under hydrogen, a
catalyst composed of 15% NiO on MgO (15-NiO-MgO)
showed a stable activity in a typical bi-reforming reaction at 830
70%, and the H /CO ratio remained stable at around 1.9. This
2
somewhat lower H /CO ratio compared to the reaction with
2
pure methane is consistent with the presence of higher
hydrocarbons having a lower H/C ratio. The H/C ratio will
get closer to 2 as the alkane chain length increases. Whereas
methane has a H/C ratio of 4, ethane, propane, and butane
have a H/C ratio of 3, 2.7, and 2.5, respectively. Keeping the
same ratio of steam and CO to the hydrocarbon feed results in
2
°
C and 7 atm for the duration of continuous experiment up to
a lower H /CO ratio. However, when the amount of steam in
2
3
20 h (Figure 1). The H /CO ratio remained essentially 2 and
the gas feed was increased by 10%, a constant H /CO ratio of 2
2
2
also remained stable over the reaction time. Under the studied
was obtained. The H /CO ratio can therefore be easily
2
6
49
dx.doi.org/10.1021/ja311796n | J. Am. Chem. Soc. 2013, 135, 648−650