Energy & Fuels 2009, 23, 3885–3892
3885
Chemical-Looping Combustion of Coal with Metal Oxide Oxygen Carriers Ranjani Siriwardane,*,† Hanjing Tian,†,‡ George Richards,† Thomas Simonyi,†,‡ and James Poston† National Energy Technology Laboratory, United States Department of Energy, 3610 Collins Ferry Road, Post Office Box 880, Morgantown, West Virginia 26507-0880, and Parsons, Post Office Box 618, Pittsburgh, PennsylVania 15129 ReceiVed February 23, 2009. ReVised Manuscript ReceiVed June 12, 2009
The combustion and reoxidation properties of direct coal chemical-looping combustion (CLC) over CuO, Fe2O3, Co3O4, NiO, and Mn2O3 were investigated using thermogravimetric analysis (TGA) and bench-scale fixed-bed flow reactor studies. When coal is heated in either nitrogen or carbon dioxide (CO2), 50% of weight loss was observed because of partial pyrolysis, consistent with the proximate analysis. Among various metal oxides evaluated, CuO showed the best reaction properties: CuO can initiate the reduction reaction as low as 500 °C and complete the full combustion at 700 °C. In addition, the reduced copper can be fully reoxidized by air at 700 °C. The combustion products formed during the CLC reaction of the coal/metal oxide mixture are CO2 and water, while no carbon monoxide was observed. Multicycle TGA tests and bench-scale fixed-bed flow reactor tests strongly supported the feasibility of CLC of coal by using CuO as an oxygen carrier. Scanning electron microscopy (SEM) images of solid reaction products indicated some changes in the surface morphology of a CuO-coal sample after reduction/oxidation reactions at 800 °C. However, significant surface sintering was not observed. The interactions of fly ash with metal oxides were investigated by X-ray diffraction and thermodynamic analysis. Overall, the results indicated that it is feasible to develop CLC with coal by metal oxides as oxygen carriers.
Introduction Fossil fuels supply most of the world’s energy needs. However, the combustion of fossil fuels is one of the major sources of carbon dioxide (CO2), a greenhouse gas. It is necessary to develop technologies that allow use of fossil fuels while reducing greenhouse gas emissions. Commercial CO2capture technologies that exist today are expensive and energyintensive. The main disadvantage of these techniques is the large amount of energy that is required for the separation, reducing the overall efficiency of a power plant. Chemical-looping combustion (CLC) has been suggested as an energetically efficient method for producing high-purity CO2 from combustion of fuel. This is an entirely new combustion technology that involves the use of an oxygen carrier, such as metal oxide, that transports oxygen from the air to the fuel, thereby avoiding direct contact between fuel and air. CLC is similar to oxy-fuel combustion, in that the combustion products with oxy-fuel are just CO2 and water vapor. However, unlike oxy-fuel, no direct supply of oxygen is needed, avoiding the cost and energy penalty of supplying oxygen to the power system. Instead, in CLC, the oxygen comes from a metal oxide, as explained next. The CLC system is composed of two reactors: an air reactor and a fuel reactor. In the fuel reactor, the fuel in the gaseous form reacts with the metal oxide. * To whom correspondence should be addressed. Telephone: +01-304285-4513. Fax: +01-304-285-0903. E-mail: ranjani.siriwardane@ netl.doe.gov. † United States Department of Energy. ‡ Parsons.
fuel (CO, H2) + metal oxide f CO2 + H2O + metal or reduced form of metal oxide
(1)
The metal or a reduced form of metal oxide is oxidized in the air reactor to form metal oxide. metal or reduced form of metal oxide + O2 f metal oxide (2) The regenerated metal oxide is ready to initiate a second cycle. The exit gas from the fuel reactor contains CO2 and H2O and, after condensation of the water, produces a pure stream of CO2, which can be used for sequestration. The significant advantage compared to normal combustion is that a concentrated CO2 stream not diluted by N2 can be obtained without expending any major energy required for separation. The CLC of synthesis gas derived from coal has been reported previously.1-7 However, only a few studies have been conducted on the combustion of solid fuels, such as coal.8-17 A hybrid (1) Diego, L. F.; Labiano, F. G.; Adanez, J.; Gayan, P.; Abad, B.; Corbella, M.; Palaciob, J. M. Fuel 2004, 83, 1749. (2) Cho, P.; Mattisson, T.; Lyngfelt, A. Fuel 2004, 83, 1215–1225. (3) Jin, H.; Ishida, M. Proceedings of TAIES’97 International Conference, Newcastle, New South Wales, Australia, 2001. (4) Jin, H.; Ishida, M. Fuel 2004, 83, 2411–2417. (5) Siriwardane, R. V.; Poston, K.; Chaudhari, A.; Zinn, A. T.; Simonyi, T.; Robinson, C. Energy Fuels 2007, 3, 1582–1591. (6) Copeland, R. J.; Alptekin, G.; Cessario, M.; Gerhanovich, Y. Proceedings of the First National Conference on Carbon Sequestration, Washington, D.C., 2001, DOE/NETL, Pittsburgh, PA; LA-UR-00-1850. (7) Gupta, P.; Velazquez-Vargas, L. G.; Fan, L.-S. Energy Fuels 2007, 21, 2900–2908. (8) Kronberger, B.; Lyngfelt, A.; Lo¨ffler, G.; Hofbauer, H. Ind. Eng. Chem. Res., 2005, 44, 546–556. (9) Wolf, J.; Anheden, M.; Yan, I. International Pittsburgh Coal Conference, Newcastle, New South Wales, Australia, 2001.
10.1021/ef9001605 CCC: $40.75 2009 American Chemical Society Published on Web 07/08/2009
3886
Energy & Fuels, Vol. 23, 2009
combustion-gasification CLC system for production of hydrogen was introduced by Alstom Power, Inc. and uses CaSO4 as the oxygen carrier for gasification of coal. Cao and Pan have introduced a concept of using a transport reactor-bubbling bed combination for combustion of coal with metal oxides.11 Lyon and Cole have conducted some coal gasification experiments in a fluidizing bed with Fe2O3 as the oxygen carrier; it was found that the amount of produced CO and CO2 is consistent with the amount of coal added in the reaction system.13 This preliminary test proved the feasibility of coal CLC with metal oxide as an oxygen carrier. Leion et al. and Berguerand and Lyngfelt reported gasification of petroleum coke with Fe2O3 and steam.14-16 Scott et al. also reported gasification of char with Fe2O3.17 There are many challenges associated with direct CLC of coal using oxygen carriers. The oxygen carrier should have a sufficient combustion rate suitable for various reactor systems, sufficient oxygen release capacity facilitating the coal-oxygen carrier interactions, stable reactivity during multiple cycles, high attrition resistance, and low reactivity with ash and other contaminants. Circulation of a large amount of ash, separation of ash, and designing a suitable reactor system to obtain an efficient process are some of the other challenges. In the present work, various metal oxides, CuO, Fe2O3, NiO, Mn2O3, and Co3O4, were tested as oxygen carriers for direct combustion of coal in the presence of either nitrogen or CO2. Thermogravimetric analysis (TGA) was used to determine the rates of reaction and percentage of combustion/oxidation. Bench-scale flow reactor tests were also conducted to verify the TGA data. Thermodynamic calculations and X-ray diffraction (XRD) analysis were conducted to understand the interaction of ash with the metal oxide oxygen carriers. Multicycle tests were conducted to understand the reaction performance of the oxygen carriers. Experimental Section TGA experiments were conducted in a thermogravimetric analyzer (TA model 2050). The samples were placed in a 5 mm deep and 10 mm diameter crucible. Illinois #6 coal was used in this study with the particle size of 100 µm. Coal was mixed physically with the metal oxide in a Crescent WIG-BUG shaker. The mass ratio (metal oxide mass/coal mass) covered the range from 15 to 22.5, which corresponds to a stoichiometric oxygen supply. For example, for CuO, the stoichiometric reaction for coal with a C/H ratio of 1 is as follows:
2.5CuO + CH f CO2 + 1/2H2O + 2.5Cu Noting the molecular weights of copper oxide (79.5 g/gmol) and the coal (∼13 g/gmol), the mass ratio for copper oxide to dry ashfree coal is 15.3:1. The particle sizes of CuO (99%, Aldrich), NiO (99.98%, Alfa Aesar), Mn2O3 (98%, Alfa Aesar), Fe2O3 (99.98%, Fisher), and Co3O4 (99.7%, Alfa Aesar) were 5, 44, 44, 44, and 70 µm, respectively. CuO with average diameter of 63 µm was also prepared and tested as an oxygen carrier in this study. About 150 mg of the coal-metal oxide mixture was heated in a quartz bowl from ambient to 900-1000 °C at a heating rate of 15 °C/min in (10) Andrus, H. E., Jr.; McCartney, M. S. Hot solids gasifier with CO2 removal and hydrogen production. U.S. Patent 2,083,658 B2, Aug 2006. (11) Cao, Y.; Pan, W. Energy Fuels 2006, 20, 1836–1844. (12) Cao, Y.; Casenas, B.; Pan, W. Energy Fuels 2006, 20, 1845–1854. (13) Lyon, R. K.; Cole, J. A. Combust. Flame 2000, 121, 249–261. (14) Leion, H.; Mattison, T.; Lyngfelt, A. Int. J. Greenhouse Gas Control 2. 2008, 180–193. (15) Leion, H.; Mattison, T.; Lyngfelt, A. Fuel 2007, 86, 1947–1958. (16) Berguerand, N.; Lyngfelt, A. International Journal of Greenhouse Gas Control 2008, 2, 169–179. (17) Scott, S. A.; Dennis, J. S.; Hay Hurst, A. N. AIChE J. 2006, 52, 3325–3328.
Siriwardane et al. either pure nitrogen or pure CO2 at a flow rate of 100 cm3/min. The sample was then kept isothermal at either 900 °C (for CuO) or 1000 °C for about 60 min. Afterward, air was introduced at 40 cm3/min for about 60 min. The fractional conversions (fractional reduction and fractional oxidation) were calculated using the TGA data. The fractional conversion (X) is defined as
fractional reduction (X) ) (Mo - M)/(Mo - Mf) fractional oxidation (X) ) (M - Mf)/(Moxd - Mf) where M is the instantaneous weight of the metal oxide-coal mixture, Mo is the initial weight of the metal oxide-coal mixture, Mf is the weight of the metal oxide-coal mixture after the reaction in either N2 or CO2 (i.e., reduced metal + ash + unreacted coal), and Moxd is the weight of the completely oxidized sample after introducing air. The fractional conversion data as a function of time was fitted to obtain the polynomial regression equation. The global rates of reactions (dX/dt) at different fractional conversions (X) were calculated by differentiating the fifth-order polynomial equation. The percentage of combustion and percentage of oxygen uptake were obtained using the weight change data from the TGA using the following equations:
percent combustion ) (actual weight loss of coal from TGA/ theoretical weight loss based on the carbon content in the coal sample) × 100 percent oxygen consumption ) (experimental oxygen consumption/theoretical capacity of oxygen present in the metal oxide) × 100 Bench-scale fixed-bed flow reactor tests were conducted with a 1 g sample containing CuO and coal prepared similarly to the samples used for the TGA tests. The outlet gas composition (CO2 and CO) from the reactor was measured using a mass spectrometer. Because the mass spectral peaks for nitrogen overlap with those for CO, argon was used as the flow gas; the sample was heated in argon with a flow rate of 100 cm3/min from 25 to 500 °C and kept isothermal at 500 °C for 60 min. Then, the sample was heated up to 710 °C and kept isothermal for 60 min; afterward, the sample was heated up to 800 °C and exposed to air for 90 min. HSC Chemistry 6 thermodynamic software was used to evaluate the possible phases formed when ash and various metal oxides were heated up to 1000 °C. The input parameters used were 1 mol of metal oxide, 1 mol of fly ash (0.6 mol of SiO2 and 0.4 mol of Al2O3), and 0.1 mol of N2. The silica/alumina ratio used in the calculations represents the typical composition of fly ash. XRD analyses were carried out using a Panalytical PW 3040 X-Pert Pro XRD system equipped with a 60 kV PW 3373/00 Cu LFF high-power ceramic tube with a Cu anode and a PW 3011/20 detector. The X-ray wavelength used was Cu KR-1 at 1.540 56 Å. The maximum goniometer resolution was 0.003° (2θ). System calibration was carried out using a polysilicon-pressed disk with the Si〈111〉 referenced to 28.443° (2θ). Sample data were acquired at 40 kV and 45 mA in a line-focus mode using a standard PW3071/ 60 powder diffraction stage. X-ray photoelectron spectroscopy (XPS) spectra were obtained with a Physical Electronics (PHI) model 32-096 X-ray source control and a 22-040 power supply interfaced to a model 04-548 X-ray source with an Omni Focus III spherical capacitance analyzer (SCA). The system is routinely operated within the pressure range from 10-8 to 10-9 Torr (from 1.3 × 10-6 to 1.3 × 10-7 Pa). The system was calibrated in accordance with the procedures of the manufacturer using the photoemission lines Eb of Cu 2p3/2 ) 932.7 eV, Eb of Au 4f7/2 ) 84 eV, and Eb of Ag 3d5/2 ) 368.3 for a magnesium anode. All reported intensities are experimentally determined peak areas divided by the instrumental sensitivity factors. Charge correction was obtained by referencing the adventi-
CLC of Coal with Metal Oxide Oxygen Carriers
Energy & Fuels, Vol. 23, 2009 3887
Table 1. Component Analysis of Illinois #6 Coal weight (%) moisture volatiles ash H C N S O
7.44 37.97 12.45 4.63 67.32 1.3 4.81 9.49
tious C 1s peak to 284.8 eV. Software analysis was conducted using RBD Instruments analysis software.
Results and Discussion The composition of Illinois #6 coal used in this study is listed in Table 1. The reaction mechanisms of coal combustion using an oxygen carrier are not well-defined in the literature. It is reported that the gaseous products from coal pyrolysis may initiate the reaction with metal oxides;12 some literature also suggested that the oxygen released from the metal oxides may initiate the reaction.13-15 In addition, direct combustion of metal oxide and carbon is also thermodynamically favorable under high temperature.12 The major goal of the CLC process is to obtain a concentrated CO2 stream after the combustion process; therefore, the reaction in the fuel reactor has to be conducted either in the presence of CO2 or steam. TGA: Baseline Tests with Coal. Baseline tests were conducted with pure coal by heating it up to 1000 °C in air, nitrogen, and CO2. The weight profile and global reaction rate obtained by differentiating the fractional conversion data for combustion of coal with air in the TGA are shown in Figure 1. A broad peak from 250 to 800 °C with a maximum at 485 °C was observed with air, and 100% coal combustion was obtained. This indicated that the coal volatilization/combustion initiates around 250 °C, and the combustion reaction proceeds up to 800 °C in air. A TGA profile of coal heated up to 1000 °C in nitrogen is shown in Figure 2. A small peak with a peak maximum at 82 °C corresponding to dehydration of coal and a narrow peak (because of coal volatilization) starting at 250 °C with a peak maximum at 435 °C were observed. This indicated that the coal volatilization reaction initiated at 250 °C with the maximum rate at 435 °C. The weight loss because of coal volatilization in nitrogen was about 45%, and an additional weight loss of 55% because of the combustion was observed once air was introduced following nitrogen introduction at 1000
Figure 1. TGA blank test of coal with air.
Figure 2. TGA profile test of coal in N2: (a) weight and reaction temperature versus reaction time and (b) weight and weight loss rate versus reaction temperature in reduction.
°C. When coal was heated in CO2, the weight loss profile and rate data were similar to that with nitrogen. TGA: Tests with Coal and Metal Oxides in Nitrogen and CO2. The weight loss profile and corresponding rate data of the CuO-coal mixture during heating in nitrogen up to 900 °C is shown in Figure 3. There is a continuous weight loss with the coal-CuO mixture during the heating that indicates that CuO contributes to the combustion of coal after coal volatilization. The weight loss corresponds to 100% coal combustion with CuO in the presence of N2. The rate information in Figure 3 indicates the presence of two major peaks at 425 and 708 °C. The major peak at 708 °C is due to coal combustion by oxygen from CuO, and the coal combustion appears to initiate around 500 °C. The outlet CO2 concentration measured by a mass spectrometer during the reaction is also shown in Figure 3. The rate data from TGA measurements and the outlet CO2 profile appear to be very consistent. TGA data during the introduction of air after the reduction reaction with coal are shown in Figure 4. The rate of oxidation is significantly higher than that for the reduction reaction. In addition, the weight gain during the oxidation reaction is consistent with the amount of oxygen present in the original CuO sample, which indicates that the reduced copper could be fully oxidized. When the reaction was performed in the presence of CO2 for a CuO-coal system, the weight loss data and rate information were similar to that with nitrogen, as shown in Figure 5, which
3888
Energy & Fuels, Vol. 23, 2009
Figure 3. (a) TGA data of the combustion segment of coal with CuO in nitrogen. (b) On-line MS signal of CO2 in the outlet of TGA tests during the combustion segment.
Figure 4. Oxidation segment of coal after CLC with CuO in nitrogen.
indicates that the CO2 has no significant effect on the CLC reaction of coal and CuO. Combustion experiments similar to the CuO-coal system were also performed with coal and NiO, Fe2O3, Mn2O3, and Co3O4 in the presence of both N2 and CO2. The results are shown in the Table 2. It is interesting to note the lowest combustion reaction temperature at 700 °C and the highest combustion rate were observed for the CuO-coal system. In addition, the combustion reaction is exothermic with CuO. From a practical
Siriwardane et al.
Figure 5. CLC of coal with CuO under a CO2 environment: (a) combustion segment and (b) oxidation segment.
standpoint, this could be an advantage in using CuO (versus other metal oxides) because both the fuel and air reactor are exothermic, avoiding the need to transfer heat transfer from the oxidizer to the reducer. Full combustion and oxidation were also obtained for coal combusted by CuO with a bigger particle size (63-173 µm) but at a higher reaction temperature (788 °C). The percentage of combustion and oxidation was close to 100% for the CuO, indicating that the complete coal-combustion reaction can be obtained with the CuO and reduced copper can be completely oxidized at 900 °C. For an Fe2O3-coal system, the maximum combustion rate takes place at a higher temperature (973-977 °C) than that with CuO. It is possible to achieve about 95% combustion with Fe2O3. The percentage of oxidation was about 77% when the oxidation-state change of Fe was assumed to be from FeIII to FeII during the CLC reaction. XPS analysis conducted with samples after the combustion reaction also indicated that metallic iron was not present on the surface of the sample. The heat of the reduction reaction for the Fe2O3-coal system is endothermic, and heat transfer from the oxidizer to the reducer will be necessary for a CLC process with Fe2O3. The combustion rate for Fe2O3 appears to be lower than that with the CuO system, but the oxidation rate appears to be higher with Fe2O3. The characteristics of the NiO-coal system were similar to the Fe2O3-coal system, as shown in Table 2. For NiO-coal, the combustion reaction, which is also endothermic, takes place at a higher temperature of 993 °C with a lower reaction rate than that
CLC of Coal with Metal Oxide Oxygen Carriers
Energy & Fuels, Vol. 23, 2009 3889
Table 2. Thermodynamic and Reaction Properties of Coal CLC on Various Metal Oxides combustion sample CuO (5 µm) CuO (63 µm) NiO (44 µm) Fe2O3 (44 µm) Mn2O3 (44 µm) Co3O4 (70 µm)
oxidation
gas media
combustion temperature (°C)
combustion rate (min-1)
% combustion
CO2 N2 N2 CO2 N2 CO2 N2 CO2 N2 CO2 N2
703 708 780 993 1000 973 977 905 978 781 781
0.098 0.083 0.079 0.061 0.017 0.055 0.05 0.011 0.01 0.096 0.096
100 100 100 73.05 68.4 94.9 91.6 76.76 71 83.3 83
with a CuO-coal system. However, the percentage of combustion with NiO-coal was lower than that observed with CuO-coal. For Mn2O3-coal, the combustion reaction took place at 900 °C but showed the lowest combustion rate as compared to the four other metal oxides, but the reaction is slightly exothermic. The percentage of combustion was similar to that with NiO. For Mn2O3, the CLC reaction appears to occur between the oxidation states MnIII and MnII. XPS analysis data also confirmed that the Mn0 is not present in the sample after combustion. The Co3O4-coal system appears to be similar to that of CuO with a low combustion temperature of 781 °C, but the oxidation rate was highest for Co3O4. The heat of reaction was slightly exothermic, yet the percentage of combustion was lower than that with CuO-coal. For Co3O4, the combustion-oxidation reaction was assumed to be between Co3+ and Co2+ oxidation states. XPS analysis verified the presence of Co2+ on the surface of the samples after combustion. From the five metal oxides tested, CuO appeared to have the best combustion characteristics. It had the lowest combustion temperature, the highest combustion rate, and the highest percentage combustion. The CuO and Cu metals have a relatively lower melting point (∼1000 °C) than the other metal oxides. However, the coal combustion reaction with CuO takes place at a lower temperature than the melting point of copper oxide and may be a potential candidate for the direct coal CLC process. Effect of CO2 on the Reaction Performance. One of the goals in the CLC process is to obtain a concentrated sequestration-ready CO2 stream. To obtain a concentrated CO2 from the product stream, the direct coal CLC process must be conducted under either CO2 or H2O. Therefore, TGA experiments were also conducted in the presence of CO2 instead of N2. The summary of the TGA performance data for all metal oxides and coal mixtures in the presence of both CO2 and N2 are summarized in Table 2. The combustion reaction temperatures and the rates were similar in both N2 and CO2 for all of the metal oxides, except for NiO, as shown in Table 2. To further investigate the effect of CO2, pure metal oxides were also heated in both N2 and CO2 in TGA up to 1000 °C. The weight profiles for CuO, Fe2O3, Co3O4, and Mn2O3 were similar in both CO2 and N2; however, a slight weight gain was observed for NiO at 800 °C in the presence of CO2. Similarly, when the Ni metal was heated in CO2, a weight gain was observed in TGA. It has also been reported that the nickel can catalyze the decomposition reaction of CO2.18,19 The significant increase of the CLC reaction performance for NiO in the presence of CO2 is probably due to the enhancement of the decomposition reaction of CO2 to CO in the presence of nickel. (18) Kato, H.; Tsuji, M. K.; Tamura, Y.; Chang, S. G. J. Mater. Sci. 1994, 29, 5689–5692. (19) Osaki, T.; Mori, T. React. Kinet. Catal. Lett. 2005, 87, 149–156.
∆H (kJ/mol)
oxidation rate (min-1)
oxygen uptake (%)
-96.5
0.172 0.175 0.174 0.84 0.82 0.77 0.78 0.42 0.38 1.74 1.74
98.6 99.2 99.2 77.5 71.6 93.7 FeII 90.6 FeII 72.2 MnII 68.3 MnII 78.2 CoII 78.0 CoII
-96.5 75.2 79.2 -36.1 -8.6
∆H (kJ/mol) -156 -156 -327.7 -347.4 -216.4 -243.9
Effect of the Reduction Temperature on the CuO-Coal Reaction. Because CuO showed promising performance, additional experiments were conducted with it. According to the rate information in Figure 3, the CuO-coal combustion reaction appears to initiate around 500 °C. To understand the temperature effect on the reaction rate, TGA experiments were conducted at different final temperatures. The CuO-coal samples were initially held isothermal at 500 °C for 30 min. Then, the samples were heated up to 640, 660, 700, and 720 °C at a ramp rate of 15 °C/ min and held at the final temperature for 60 min. The TGA instantaneous weights were normalized with respect to the weight at 500 °C, at which the combustion reaction initiates. The percentage of combustion calculated from the TGA data are shown in Figure 6a. The data indicated that the combustion reaction can take place as low as 640 °C but the rate is significantly low. The rate appears to be very similar at 700 and 720 °C. To obtain full combustion at an appreciable rate, the reaction temperature of about 700 °C is required for the CuO-coal system. Effect of the Oxidation Temperature on the CuO-Coal Reaction. The influence of the oxidation temperature on the reduced CuO sample after the CuO-coal combustion reaction was investigated at various oxidation temperatures. The TGA profiles are shown in Figure 6b. It was observed that full oxidation can be obtained at the temperature as low as 600 °C and the reaction rate slightly increased with an increasing temperature up to 650 °C. After 650 °C, the oxidation performance was unaffected by increasing temperature. This indicated that, after the coal-combustion reaction with CuO at 700 °C, CuO can also be oxidized with air at 700 °C at an appreciable rate. Because Cu has a low melting point (1100 °C), an operation temperature of 700 °C should be suitable for the CuO-coal CLC process. As discussed in more detail later, this temperature is also well-suited for steam cycle power generation from fluid-bed boilers. Interaction of Metal Oxide with Ash. In a CLC process that involves direct combustion of coal using metal oxides, it is important to understand how the ash in coal would interact with the metal oxides. It is expected to remove ash from the solid mixture after each combustion/oxidation cycle. Because the density of ash is significantly lower than that of the metal oxide, density separation using a cyclone can be used for the separation. However, it is important to understand how the metal oxide interacts with ash because there may be some residual ash left after the separation. Thermodynamic analysis was conducted by HBC Chemistry software to understand the possible products that could be formed by the reaction of ash with metal oxides. The results of the analysis are shown in Figure 7. The thermodynamic data indicated that the interaction of SiO2 and Al2O3 is minimal with Fe2O3 and Co3O4 at temperatures up to 1000 °C. The formation of copper oxide/alumina solid solution is favorable as low as
3890
Energy & Fuels, Vol. 23, 2009
Siriwardane et al.
Figure 8. XRD patterns of metal oxide and coal ash mixture in N2 (40 cm3/min) after heating at 1000 °C for 4 h: (a) fly ash, SiO2, Al2O3, and Fe2O3, (b) CuO-fly ash, (c) NiO-fly ash, (d) Fe2O3-fly ash, and (e) Mn2O3-fly ash.
Figure 6. CLC of coal with CuO at various (a) reduction and (b) oxidation temperatures.
Figure 7. Thermodynamic analysis of the interaction between ash and various metal oxides (HBC software input: 1 mol of metal oxide, 1 mol of fly ash (0.6 mol of SiO2 and 0.4 mol of Al2O3), and 0.1 mol of N2).
200 °C. The formation of silicates and aluminates with NiO is also thermodynamically feasible, and Mn2O3 also appears to
interact slightly with SiO2 even at low temperatures (
