Você está na página 1de 7

148

Energy & Fuels 2006, 20, 148-154

Characterization and Performance in a Multicycle Test in a


Fixed-Bed Reactor of Silica-Supported Copper Oxide as Oxygen
Carrier for Chemical-Looping Combustion of Methane
B. M. Corbella, L. de Diego, F. Garca-Labiano, J. Adanez, and J. M. Palacios*,
Instituto de Cata lisis y Petroleoqumica (CSIC), Campus UAM-Cantoblanco, 28049-Madrid, Spain, and
Instituto de Carboqumica (CSIC), Mara de Luna 12, 50015-Zaragoza, Spain
ReceiVed July 14, 2005. ReVised Manuscript ReceiVed October 10, 2005

Chemical-looping combustion of carbonaceous compounds is a proposed two-step process for complete


CO2 capture and substantial reduction of NOx emissions. In the first stage, the reduction stage, the framework
oxygen of a reducible inorganic oxide is used for the combustion of the carbonaceous material. In the second
stage, the regeneration stage, the carrier in a reduced state is regenerated with air to recover the properties of
the fresh carrier, ready to reinitiate a new cycle. This article provides results for the performance of a copper
oxide silica-supported oxygen carrier in a 20-cycle test of chemical-looping of methane in a fixed-bed reactor
at 800 C and atmospheric pressure. The mesoporous nature of silica provided a good dispersion of the active
phase imparting a high mechanical strength to the overall carrier. Additionally, silica is stable under highly
reducing agents and inert in the two involved processes. The respective CH4, CO2, and CO breakthrough
curves in the reduction stage show that the reduction reaction rate is fast and highly selective to CO2 formation.
CO emissions are very low, only yielded at the end of the reduction stage, when the reduction stage should be
stopped to initiate a regeneration stage. Characterization studies using different techniques, such as TPR, SEMEDX, and powder XRD, reveal that CuO might decompose into Cu2O at the operating conditions used in the
reduction stage, but fortunately, the decomposition rate is so low that it has no effect on the oxygen amount
initially available for chemical-looping combustion. Copper does not promote the thermal decomposition of
methane, and deposited carbon, consequently, could not be detected in the reduced carrier. In a 20-cycle test
neither performance decay nor mechanical degradation of the oxygen carrier has been observed.

1. Introduction
Chemical looping has been proposed recently as an alternative
process for the combustion of carbonaceous materials, such as
coal, oil, or natural gas, providing complete CO2 capture together
with substantial reduction of thermal NOx emissions.1-4 Chemical-looping combustion (CLC) is a two-step process5 usually
carried out in two different reactors. In the first stage, called
the reduction stage, the carbonaceous material is burned using
the framework oxygen of a reducible inorganic oxide. If the
reduction reaction is fast enough and highly selective, complete
conversion of the combustible can be achieved, and consequently, the outlet gas from this reactor is exclusively composed
of CO2 and steam. These two gas components can easily be
separated by steam condensation, a crucial difference with
conventional combustion, and the pure CO2 obtained is ready
* To whom correspondence should be addressed. E-mail: jmpalacios@
icp.csic.es.
Instituto de Cata
lisis y Petroleoqumica.
Instituto de Carboqumica.
(1) Anheden, M.; Svederg, G. Exergy analysis of chemical-looping
combustion systems. Energy ConVers. Manage. 1998, 39, 1967-1980.
(2) Yu, J.; Corripio, A. B.; Harrison, D. P.; Copeland, R. J. Analysis of
the sorbent energy transfer system (SETS) for power generation and CO2
capture. AdV. EnViron. Res. 2003, 7, 335-345.
(3) Ishida, M.; Jin, H. Chemical-looping combustion power generation
plant system. U.S. Patent No. 5,447,024, 1995.
(4) Ishida, M.; Jin, H. A new advanced power-generation system using
chemical-looping combustion. Energy 1994, 19, 415-422.
(5) Lyngfelt, A.; Leckner, B.; Mattison, T. A fluidised-bed combustion
process with inherent CO2 separation; application of chemical-looping
combustion. Chem. Eng. Sci. 2001, 56, 3191-3113.

to be used in multiple applications or transported to be stored


in suitable locations. The oxygen carrier in the bed of the first
reactor after the reduction stage is found in a reduced state
needing, subsequently, to be transported to a second reactor for
regeneration. In the second stage, called the regeneration stage,
the carrier is regenerated with pure air recovering composition
and properties of the fresh carrier to reinitiate a new cycle. Given
that the regeneration reaction is always very fast and highly
exothermic, the outlet gas from the regeneration stage is made
of N2 exclusively. The operating temperature in both stages is
approximately the same and relatively low as compared with
that used in conventional combustion processes. Consequently,
the formation of thermal NOx in the regeneration stage of a
CLC process becomes a highly improbable event. CLC of
carbonaceous materials achieving complete CO2 capture and
substantial reduction of NOx emissions becomes an attractive
process, taking into account the severe environmental problems
involved, causing increasing trouble especially in developed
countries.
Nickel oxide6 has been one of the first inorganic oxides tested
as an oxygen carrier for a CLC of methane, as the main
component of natural gas or coal gas, at operating temperatures
in the range 500-900 C. Kinetic thermobalance tests showed
that the reactivity of nickel oxide was high in both reduction
and regeneration stages; however, it is prone to carbon deposition in the reduction stage. Nickel-based carriers in the reduced
(6) Ishida, M.; Jin, H.; Okamoto, T. A fundamental study of a new kind
of medium material for chemical-looping combustion. Energy Fuels 1996,
10, 958-963.

10.1021/ef050212n CCC: $33.50 2006 American Chemical Society


Published on Web 11/05/2005

Chemical-Looping Combustion of Methane

state of a CLC are found as metallic nickel, and it is well known


that it is a good catalyst for the thermal decomposition of
hydrocarbons. This side reaction is completely undesirable in a
CLC process since it promotes methane consumption in clear
competition with the main reaction of the reduction stage. The
deposited carbon is further converted into CO2 in the next
regeneration stage, imposing some limitations to the maximum
achievable efficiency in CO2 capture of the overall CLC process.
Carbon deposition using nickel-based oxygen carriers can be
decreased by doping nickel oxide with other inorganic oxides,7
promoting the formation of some mixed oxides by using suitable
carrier compositions8 and/or including steam in the composition
of the feeding gas of the reduction stage.9
Potentially, many other reducible inorganic oxides can be used
as oxygen carriers for a CLC process.10-12 For example, copper
oxide shows high reactivities in the reduction and regeneration
processes in thermobalance tests, and in the reduced state
metallic copper, contrary to nickel, does not enhance carbon
deposition in the reduction stage.13 However, the melting points
of some involved copper phases are not much higher than the
operating temperature used in a CLC process. Given the high
exothermiticity of the involved reactions in the regeneration
stage, temperature excursions cannot be avoided, especially in
fixed-bed reactor tests. Consequently, accumulative thermal
sintering and/or copper redistribution may be a problem in a
multicycle test with copper-based oxygen carriers. Additionally,
the thermal stability of CuO in the reduction stage may be not
high enough, and the available oxygen for CLC might be lower
than expected.
In general terms, high reactivity especially in the reduction
stage, thermal stability of all involved phases in the two stages
of a CLC process, and high mechanical strength are the main
criteria used for oxygen carrier selection. In multicycle reactor
tests of many hundreds of successive reduction-regeneration
cycles, probable accumulative thermal and chemical stresses of
the oxygen carriers can be enhanced leading to progressive
performance decay and mechanical degradation with the number
of cycles. A prior study13 indicated that to reduce these
detrimental effects, the reducible oxide as active phase must
be kept occluded in a porous matrix, presumably inert in the
two involved stages. This porous support should impart both
high dispersion of the active phase necessary to enhance
reactivity and high mechanical strength of the overall oxygen
carrier to withstand the multicycle test without progressive
mechanical degradation.
It was shown in a recent study14 that different titaniasupported copper-based oxygen carriers for a CLC of methane
(7) Jin, H.; Okamoto, T.; Ishida, M. Development of a novel chemicallooping combustion: Synthesis of a looping material with a double metal
oxide of CoO-NiO. Energy Fuels 1998, 12, 1272-1277.
(8) Jin, H.; Okamoto, T.; Ishida, M. Development of a novel chemicallooping combustion: Synthesis of a solid looping material of NiO/NiAl2O4.
Ind. Eng. Chem. Res. 1999, 38, 126-132.
(9) Ishida, M.; Jin, H.; Okamoto, T. Kinetic behaviour of solid particle
in chemical-looping combustion: Suppressing carbon deposition in reduction. Energy Fuels 1998, 12, 223-229.
(10) Adanez, J.; de Diego, L. F.; Garca-Labiano, F.; Gayan, P.; Abad,
A.; Palacios, J. M. Selection of oxygen carriers for chemical-looping
combustion. Energy Fuels 2004, 18, 371-377.
(11) Mattisson, T.; Lyngfelt, A.; Cho, P. The use of iron oxide as an
oxygen carrier in chemical-looping combustion of methane with inherent
separation of CO2. Fuel 2001, 80, 1953-1962.
(12) Cho, P.; Mattisson, T.; Lyngfelt, A. Comparison of iron-, nickel-,
copper- and manganese-based oxygen carriers for chemical-looping combustion. Fuel 2004, 83, 1215-1225.
(13) De Diego, L. F.; Garcia-Labiano, F.; Adanez, J.; Gayan, P.; Abad,
A.; Corbella, B. M.; Palacios, J. M. Development of Cu-based oxygen
carriers for chemical-looping combustion. Fuel 2004, 83, 1749-1757.

Energy & Fuels, Vol. 20, No. 1, 2006 149

exhibited a good performance in a multicycle test in a fixedbed reactor at 900 C. Titania as support provided a high
mechanical strength to the overall carrier, additionally providing
a highly dispersed copper oxide as active phase. Consequently,
the involved reactions in both the reduction and regeneration
stages were fast, highly selective, and independent of the copper
loading in the carrier. Deposited carbon became undetected in
the reduction stage; however, copper redistribution was clearly
evidenced in the last cycles of a multicycle test. This last feature
implied that a higher number of additional pores became plugged
along the test, although it did not affect carrier performance
substantially.
In this study, a high-copper loading oxygen carrier has been
prepared by wet impregnation of a saturated aqueous solution
of copper nitrate on a highly porous silica. This support like
titania also imparts a high mechanical strength to the overall
oxygen carrier. Characterization studies of fresh, reduced, and
regenerated carriers using different techniques have allowed
identification of the chemical compounds necessary to know
the involved reactions in the reduction and regeneration stages
of a CLC of methane. To prevent copper redistribution along a
multicycle test, as observed in a prior study on titania-supported
copper carriers at 900 C,14 the operating temperature has been
lowered to 800 C. Silica, as compared with titania, presents a
different pore size distribution especially in the range of
mesopores, promoting a different dispersion of the active phase
and, presumably, higher stability of some involved oxygenated
phases in the reduction stage. By replacing titania with silica,
as a more stable support against reducing agents, we expected
that it will decrease the evolution of detrimental gas compounds,
such as CO.
2. Experimental Section
2.1. Preparation of Oxygen Carriers. Highly porous powder
silica, with a specific surface area of 380 m2/g, was added to a 10
wt % of synthetic powder graphite of particles 1-2 m in diameter
as a pore-forming additive to form a powder mixture subsequently
homogenized in a planetary ball mill. Water was added to obtain
a paste of suitable consistency to be extruded in a plastic syringe
to cylindrical extrudates 2 mm in height 2 mm in diameter,
approximately. These support extrudates were softly dried overnight
at 80 C to prevent crack formation and, then, calcined at 1100 C
for 6 h to provide thermal stability at the operating temperature of
800 C used in a multicycle test of CLC in a fixed-bed reactor and
high mechanical strength to the overall carrier. Afterward, the
support extrudates were cooled, crushed, and sieved to particles
0.2-0.4 mm to facilitate successive impregnations of the active
phase. Fresh silica-supported copper-based oxygen carriers have
been prepared by wet impregnation of these silica particles with a
saturated aqueous solution of copper nitrate as active phase
precursor. Samples with maximum copper loading were obtained
by applying successive impregnations up to saturation. After each
impregnation, the sample was calcined at 500 C simply to promote
nitrate decomposition into oxide facilitating the next impregnation.
After the last impregnation achieving maximum copper content,
to get thermal stability all samples were calcined at higher
calcination temperatures as indicated. The obtained fresh samples
were denoted, for example, as CS900, where C ) CuO, S ) silica,
and the number indicates the last calcination temperature applied
to the oxygen carrier before reactor testing.
2.2. Characterization Techniques. Since previous studies13,14
revealed that macropres play the most important role in carrier
(14) Corbella, B. M.; De Diego, L.; Garca-Labiano, F.; Adanez, J.;
Palacios, J. M. The performance in a fixed bed reactor of copper-based
oxides on titania as oxygen carriers for chemical looping combustion of
methane. Energy Fuels 2005, 19, 433-441.

150 Energy & Fuels, Vol. 20, No. 1, 2006

Corbella et al.

performance, the textural properties of the oxygen carriers were


determined by porosimetry of Hg intrusion, in an apparatus
Micromeritics 9310, filling pores down to 3 nm in size. For the
identification of crystalline phases, powder X-ray diffraction (XRD)
patterns were acquired in a diffractometer Seifert 3000 using a Nifiltered Cu KR radiation together with a database including powder
patterns of most of the crystalline inorganic compounds known. A
semiquantitative analysis of the found crystalline phases was
subsequently obtained by using complete structure data from the
ICSD database and further processing by using Rietveld methods
with program Powder Cell v. 2.3. The evolution of carrier
appearance in a multicycle reactor test, the possible presence of
deposited carbon on the reduced samples, and variation of copper
distribution were studied in a scanning electron microscope (SEM)
ISI DS-130 coupled to an ultrathin window PGT Prism detector
for energy-dispersive X-ray (EDX) analysis. H2 thermoprogrammed
reduction (TPR) profiles were obtained by passing through 50 mg
of sample a flow rate of 51 cm3/min of gas containing 10 vol % of
H2 in Ar at a heating rate of 10 C/min up to 950 C. H2 evolution
was continuously monitored using a thermal conductivity detector
(TCD). CH4 profiles of the oxygen carriers in a thermal scanning
were undertaken in a Balzer apparatus using 40 mg of sample, at
a flow rate of 50 cm3/min with a gas of composition of 0.3 vol %
of CH4 in Ar and a heating rate of 10 C/min up to 950 C. The
composition of the outlet gas was determined by mass spectrometry
using a quadrupole QMS 200.
2.3. Performance Tests. The oxygen carrier performance was
studied in a 1.6 cm i.d. 34 cm height upflow fixed-bed quartz
reactor at operating temperatures of 800 C and atmospheric
pressure. For the reduction stage, a flow rate of 13 cm3/min of
pure CH4 was passed through a sample amount of 15-20 g. A
similar flow rate of pure air was used for the regeneration stage.
The inlet and outlet gases from the reactor were analyzed by gas
chromatography (GC) using a column Chrompack and TCD. The
detected compounds in the outlet gas from the reduction stage were
CH4, CO2, and CO, and their evolution in a typical test is shown
through their respective breakthrough curves. Since the available
active phase in the reactor bed along a typical multicycle test of
successive reduction-regeneration cycles may be variable, due to
repetitive sample extractions for characterization, the breakthrough
curves were plotted as a function of the dimensionless time ratio
t/to. to, which for the reduction stage is about 20 min, is the
theoretical time necessary to achieve the complete conversion of
the active phase contained in the reactor bed, taking into account
the reaction stoichiometry, mass balance of methane in the evolving
gases, and the available amount of active phase in the reactor bed.

3. Results and Discussion


3.1. Characterization of Fresh Samples. Using copper-based
oxygen carriers, one of the most important factors to be
accounted for is the apparent stability of CuO at the operating
conditions of the reduction stage, since it determines the oxygen
amount in the carrier available for CLC of methane. According
to available thermodynamics data,15 the decomposition reaction
of cupric oxide at 800 C in inert atmosphere is:

2CuO f Cu2O + 1/2O2

G ) 72 kJ/mol

(1)

Since the free energy G is positive, CuO is stable at the


operating conditions used in this study for CLC of methane.
To check experimentally possible decomposition of CuO,
three fresh samples designated 1CS500, 2CS800, and 3CS900
were prepared by successive impregnations with increasing
copper loadings and calcined at increasing temperatures of 500,
800, and 900 C, respectively. To get, if possible, the complete
conversion of CuO into Cu2O, calcination time was very long.
(15) Barin, I.; Knacke, O. Thermochemical properties of inorganic
substances; Springer-Verlag: New York, 1973.

Figure 1. Powder XRD patterns of fresh samples 1CS500, 2CS800,


and 3CS900 with increasing copper loading calcined at increasing
temperatures.
Table 1. Composition and Domain Size of Fresh Samples
Determined by XRD
cristobalita

CuO

Cu2O

sample

wt %

d (nm)

wt %

d (nm)

wt %

d (nm)

SiO2
1CS500
2CS 800
3CS900

100
90
84
17

2
2
2
35

10
13
45

103
55
40

2
38

large
889

The respective powder XRD patterns are shown in Figure 1, in


which the SiO2 pattern as carrier support has also been included
for comparison. The SiO2 pattern shows a broad reflection
located at 21.77 that can be assigned to cristobalite. It appears
broad because this compound is microcrystalline with a very
small crystal domain size. In addition to the reflection assigned
to the support, the powder XRD pattern of sample 1CS500 also
shows reflections that can be assigned exclusively to CuO. The
assigned reflections do not appear very intense because the CuO
concentration in this sample is low. The pattern of sample
2CS800 calcined at 800 C reveals the presence of some
reflections assigned to Cu2O, in addition to those due to CuO
already commented. These results indicate that CuO becomes
unstable at calcination temperatures above 800 C, in clear
disagreement with thermodynamics (eq 1). The powder XRD
pattern of sample 3CS900 shows many intense reflections,
assigned to CuO and Cu2O, respectively, because the copper
concentration in this sample is high. The presence of CuO
reveals that for CuO, although unstable, the decomposition rate
is low. From the XRD study it may be concluded that, at 800
C, CuO is unstably decomposing, at least partially, into Cu2O.
However, the rate of decomposition is apparently very low, and
probably it will not affect the oxygen availability for a CLC of
methane in a reactor test.
Additional quantitative data on the concentration of the
crystalline phases in the prepared samples and their respective
domain sizes, shown in Table 1, have been obtained through
further pattern processing by Rietveld methods. Silica used as
support is a microcrystalline material with a crystal domain size
of 2 nm that apparently is only affected at the highest calcination
temperature of 900 C. In this sample of 3CS900, the domain
size increases substantially up to 35 nm. If one assumes that
silica particles are monocrystals, a high specific surface area of
380 m2/g of this silica used as support could be associated with
interparticle voids of similar size. Crystalline domain sizes of

Chemical-Looping Combustion of Methane

Figure 2. SEM micrograph of fresh sample 2CS800.

Figure 3. SEM-EDX line profiles of Cu KR and Si KR taken from


fresh sample 2CS800.

CuO and Cu2O, however, are comparatively much higher,


suggesting that at least a part of the active phase in the fresh
samples is located on the external surface of large particles of
SiO2 that could be revealed in SEM pictures.
An SEM micrograph of sample 2CS800 calcined at 800 C
is shown in Figure 2. The smallest mesopores on the surface of
the large silica particle cannot be revealed because they are
below the microscope resolution, but the accumulation of CuO
and Cu2O particles on the external surface of the large SiO2
particle is clearly revealed. The textural properties of this sample
as determined by porosimetry of Hg intrusion reveals a pore
volume Vp of 0.31 cm3/g, specific surface area S of 40 m2/g,
and mean pore diameter dp of 17 nm. Since mercury is unable
to penetrate the smallest mesopores, the specific surface area
is, consequently, underestimated and the mean pore diameter
is overestimated. N2 adsorption measurements at 77 K would
give more accurate data of textural properties of samples;
however, in a CLC process macropores play the most important
role and the pore volume Vp determined by Hg porosimetry is
much more valuable.
SEM-EDX line profiles of Cu KR and Si KR taken from fresh
sample 2CS800 are shown in Figure 3. Si and Cu profiles run
clearly in opposition distribution, indicating they are not
chemically associated. They are in complete agreement with
XRD results in which study SiO2, CuO, and Cu2O were the
detected crystalline phases in this sample. Additionally, the Cu
profile reveals that copper-containing phases are located, at least
partially, filling large macropores on silica as revealed in the
SEM picture of Figure 2 and, partially, also filling small
mesopores not revealed by SEM but giving an almost uniform
copper distribution.

Energy & Fuels, Vol. 20, No. 1, 2006 151

Additional evidence of the presence of the two oxygenated


copper phases, CuO and Cu2O, in the fresh samples calcined at
different temperatures is gained from H2 TPR profiles shown
in Figure 4. All profiles show a reduction peak located at 260275 C that can be assigned to the reduction of CuO. The profile
of sample 3CS calcined at 900 C additionally shows a second
reduction peak located at 214 C that can be assigned to Cu2O.
Finally, sample 2CS regenerated at 800 C shows a single
reduction peak of CuO shifted toward lower temperature. Given
that possible chemical associations of Si and Cu (for example,
through the formation of possible mixed oxides) must be
discarded, since they were revealed neither by powder XRD
patterns nor by SEM-EDX line profiles, the apparent shifts of
the reduction peak of CuO suggest different locations for this
copper-containing phase. A reduction peak shifted toward higher
temperature could be assigned to more difficult-reduced copper
oxide located in small silica mesopores while a reduction peak
shifted toward lower temperature could be assigned to easyreduced copper oxide located in large silica macroporos. Profiles
in Figure 4 reveal that high calcination temperatures not only
enhance partial CuO decomposition into Cu2O but also promote
copper oxide migration to larger silica pores.
3.2. Chemical Reactions in CLC of Methane at 800 C
Using Copper-Based Oxygen Carriers. As will be shown
below, the analysis of the outlet gas from the reduction stage
before breakthrough reveals exclusively the presence of CO2
since steam is condensed out before detector is reached. After
breakthrough, CO2 drops off rapidly and some traces of CO
are apparent, being CH4 a major gas component of the outlet
gas. Powder XRD patterns taken from samples extracted from
the reactor bed at the end of the reduction stage show the
presence of metallic Cu as the only copper-containing phase.
Consequently, the involved reactions in the reduction stage of
CLC of methane at 800 C are:

4CuO + CH4 f 4Cu + CO2 + 2H2O


G ) -590 kJ/mol, H ) -217 kJ/mol (2)
3CuO + CH4 f 3Cu + CO + 2H2O
G ) -431 kJ/mol, H ) -75 kJ/mol (3)
CuO + CO f Cu + CO2
G ) -125 kJ/mol, H ) -133 kJ/mol (4)
CH4 f C + 2H2

G ) -26 kJ/mol, H ) 110 kJ/mol (5)

CuO + H2 f Cu + H2O
G ) -128 kJ/mol, H ) -101 kJ/mol (6)
Equation 2 is the main reaction in the reduction stage of CLC
of methane. Equation 3 is a less-favored side reaction leading
to partial oxidation of methane. Equation 4 explains that CO,
eventually produced in the bottom layers of an up-flow fixedbed reactor, is subsequently consumed by still unconverted top
layers. Consequently, CO emissions might only be expected at
the end of the reduction stage when the amount of available
oxygen in the reactor bed is already very low. Equation 5 is
the thermal decomposition of methane only plausible in the
absence of oxygenated Cu species. Equation 6 is the further
consumption of H2 by the top layers in an affixed-bed reactor.
H2 in the outlet gas of the reduction stage only may be expected
long after carrier breakthrough.
Similarly, before breakthrough, N2 from air used as feeding
gas in the regeneration stage is the only detected compound in
the outlet gas. Characterization studies of the regenerated carriers

152 Energy & Fuels, Vol. 20, No. 1, 2006

Corbella et al.

Figure 4. H2 TPR profiles of fresh samples: (a) 3CS calcined at 500


C. (b) 3CS calcined at 900 C. (c) 2CS regenerated at 800 C.

Figure 5. CH4 TPR profiles of fresh sample 2CS800.

show that composition of the regenerated carrier is identical to


that of the fresh carrier. Consequently, the involved chemical
reaction in the regeneration stage is:

Cu + 1/2O2 f CuO
G ) -58 kJ/mol, H ) -146 kJ/mol (7)
In this case, there are no side reactions in the regeneration stage
and the only involved reaction is fast and highly exothermic.
Consequently, carrier regeneration is usually not studied, and
the only derived effect in multicycle tests is probable accumulative thermal sintering if a good control of the operating
temperature cannot be achieved.
Profiles of the evolving gases during a thermal scan with
diluted methane are shown in Figure 5. Below 750 C, the
corresponding electrical signals detected for all gas components
are kept constant, indicating that no chemical reaction occurs
at lower temperatures. At a slightly higher temperature a
decrease of the CH4 signal together with an increase of H2O
and CO2 signals are apparent, implying that the main reaction
of CLC of methane (eq 2) takes place. No variation in the H2
signal along the thermal scan is observed, indicating there is
no thermal decomposition of methane with this type of copperbased oxygen carriers. Additionally, the O2 signal shows a peak
at 870 C, revealing that CuO apparently becomes unstable and

Figure 6. Breakthrough curves in the reduction stages of sample


2CS800 in a 20-cycle test of CLC of pure methane at 800 C. (a) CH4.
(b) CO2. (c) CO.

decomposes at this temperature into Cu2O. From this CH4


thermal scan study, we concluded the feasibility of an operating
temperature of 800 C selected for carrying out multicycle tests
of CLC of methane in a fixed-bed reactor with these types of
copper-based oxygen carriers.
3.3. Carrier Performance in a 20-Cycle Test of CLC of
Methane in a Fixed-Bed Reactor at 800 C. The respective
breakthrough curves in the reduction stage of CH4, CO2, and
CO for sample 2CS800 in a 20-cycle test of CLC of methane
at 800 C are shown in Figure 6a-c. Before breakthrough, the
only compound detected in the outlet gas is CO2, indicating
that CH4 as reactant is completely converted. Breakthrough
occurs at a t/to ratio very close to 1, and given that to was
calculated assuming CuO was the only oxygen source in the
reactor bed for CH4 combustion, it ensures that CuO is in fact
the only available oxygen source for CLC of methane. Overall
losses of available oxygen for methane combustion that could

Chemical-Looping Combustion of Methane

Energy & Fuels, Vol. 20, No. 1, 2006 153

Table 2. Evolution of the Textural Properties of Regenerated


Sample 2CS800 in a 20-Cycle Test of CLC of Methane at 800 C as
Determined by Porosimetry of Hg Intrusion
cycle

Vp (cm3/g)

S (m2/g)

dp (nm)

fresh
1
5
10
15
20

0.31
0.30
0.29
0.28
0.28
0.31

40
33
38
31
30
27

17
18
17
17
21
21

Figure 8. Evolution of powder XRD patterns of reduced sample


2CS800 along a 20-cycle CLC of pure methane at 800 C.
Table 3. Evolution of the Chemical Phases and Their Respective
Domain Sizes in the Reduced Sample 2CS800 with the Number of
Cycles along a 20-Cycle CLC of Methane at 800 C
cristobalite

Figure 7. SEM micrograph of a regenerated sample 2CS800 after a


20-cycle CLC test with methane at 800 C.

be derived from possible CuO decomposition into Cu2O must


be discarded since they should be reflected in breakthrough
occurrence at t/to about 0.5. Additionally, all curve slopes at
breakthrough are high, indicating that the carrier reduction is
very fast. CO is detected in the outlet gas only at the end of the
reduction stage when the amount of available oxygen in the
reactor bed is already very low, because at this time the reactor
bed is found almost completely converted. Finally, possible
performance decay with the number of cycles that could be
derived from changes in the concentration of some chemical
species or variations in textural properties were not evidenced
using these types of copper-based oxygen carriers tested at 800
C.
The evolution of the textural properties of sample 2CS800
along a 20-cycle performance test as determined by porosimetry
of Hg intrusion is shown in Table 2. The specific surface area,
S, decreases, and the mean pore diameter increases with the
number of cycles, suggesting that accumulative thermal sintering
is operative. However, the pore volume, Vp, mostly dependent
on large pores, is not affected substantially, and consequently,
thermal sintering although operative cannot affect carrier
performance.
An SEM micrograph of regenerated sample 2CS800 after a
20-cycle CLC test is shown in Figure 7. From comparison with
the fresh sample appearance before CLC test shown in Figure
2 it can be deduced that no substantial morphological changes
are apparent. Only a higher abundance of pores in the mesopore
range together with a better dispersion of the active phase are
observed. However, big agglomerates of the active phase similar
to those observed in the fresh sample are still visible.
Powder XRD patterns taken from reduced samples extracted
from the reactor bed at the end of the reduction stage at different
cycles of a 20-cycle CLC test are shown in Figure 8. Metallic
copper is the only copper-containing phase detected by powder
XRD, and the relative intensity of their associated reflections

Cu

quartz

cycle

wt %

d (nm)

wt %

d (nm)

wt %

d (nm)

1
5
10
15
20

88
89
78
74
73

2
2
2
2
4

15
16
19
19
17

98
85
68
74
82

7
10

82
62

seems to be independent of the number of cycles, suggesting


there are no substantial copper losses in a multicycle CLC test.
The appearance of a broad reflection at low angle, mentioned
above, assigned to cristobalite is also independent of the number
of cycles. The most significant feature in the patterns evolving
with the number of cycles is the progressive development,
especially after cycle 15, of well crystallized quartz presumably
derived from cristobalite recrystallization. Quantitative data,
obtained from further powder XRD pattern processing, on
evolving changes in composition and/or structure in the reduced
carriers in a 20-cycle test are shown in Table 3. The mean
crystalline domain size of copper is high, similar to that found
in the fresh sample, indicating that at least a part of copper
continues being located on the external surface of cristobalite
as relatively large agglomerates. Also, independent of the
number of cycles, cristobalite continues to be microcrystalline
with a mean domain size of 2 nm. But the most remarkable
feature is the apparent increase in concentration with the number
of cycles of well-crystallized quartz from partial cristobalite
recrystallization.
4. Conclusions
From the study of a high-copper oxide loading oxygen carrier
prepared by impregnation on a mesoporous silica of high
specific surface area it may be concluded that the copper oxide
as active phase is partially located filling the smallest pores of
a microcrystalline cristobalite and partially as agglomerates
placed on the external surface of large particles of cristobalite.
The breakthrough curves of the reduction stage of a 20-cycle
test of CLC of methane in a fixed-bed reactor at 800 C show
that no performance decay with the number of cycles was
apparently substantiated by the observance of no changes in
chemical or textural properties of copper carrier. The only
significant structural change was partial recrystallization of

154 Energy & Fuels, Vol. 20, No. 1, 2006

microcrystalline cristobalite into well-crystallized quartz after


a significant high number of cycles. However, copper redistribution was not observed in this study, probably prevented by a
lower operating temperature of 800 C and/or better copper
dispersion promoted by mesoporous silica.

Corbella et al.
Acknowledgment. This research was carried out with financial
support from the European Coal and Steel Community (Project
7220-PR125) and CICYT Project CTQ2004-025565/PPQ.
EF050212N

Você também pode gostar