Coal Chemical-Looping with Oxygen Uncoupling (CLOU) Using a Cu-Based Oxygen Carrier Derived from Natural Minerals

: Chemical-looping with oxygen uncoupling (CLOU) is considered a promising technology to burn solid fuels with improved CO 2 capture and has the potential to improve fuel conversion and reaction rates. Cu-based oxygen carriers (Cu-OC) are often used in solid fuel CLOU. This study focused on investigating Cu-OC derived from a natural mineral for solid fuel CLOU because of their potentially lower cost compared to synthetic OCs. Reactivity and recyclability of a natural ore-derived Cu-OC on coal char (Powder River Basin sub-bituminous coal) were studied at 900 ◦ C in Ar and air using TGA-QMS and ﬁxed-bed reactor-QMS for ﬁve cycles. Cu-OC was prepared by simply heating chalcopyrite in air. Chalcopyrite is one of the principle copper sulﬁde ores and one of the primary ores for copper. The prepared Cu-OC had primarily CuO and CuFe 2 O 4 (CuOFe 2 O 3 ) as active compounds based on XRD analysis and an oxygen capacity 3.3% from oxygen uncoupling. The carbon conversion e ﬃ ciency Xc was 0.94 for reduction at a ratio of Cu-OC to char φ = 75 and the product gas was primarily CO 2 with trace O 2 . The reactivities and the rates were similar for ﬁve redox cycles. These results indicate that the natural ore-derived material with low cost has potential as a competitive oxygen carrier in solid fuel CLOU based on its reactivity in this study. ,


Introduction
Chemical-looping with oxygen uncoupling (CLOU) is a promising solid fuel chemical-looping combustion (CLC) process for CO 2 emission control [1,2]. CLOU uses a solid oxygen carrier (OC) that releases gaseous O 2 to combust the solid fuels [3]. It results in a higher fuel conversion and reaction rate than in-situ gasification CLC (iG-CLC), which uses an OC (typically Fe-based OC) to convert syngas generated from fuel gasification [1,4]. Some metal oxides with oxygen uncoupling properties are CuO/Cu 2 O, Mn 2 O 3 /Mn 3 O 4 , and Co 3 O 4 /CoO [3]. Cu-based OCs are often used in solid fuel CLOU and have demonstrated complete fuel conversion with near 100% CO 2 capture [5,6]. Cu-based OCs are also the focus of this study. Synthetic Cu-based oxygen carriers (60 wt.% CuO on a MgAl 2 O 4 support) have successfully been tested in a 1.5 KWth continuously-operating CLOU reactor using bituminous coal [7] and pine wood [8]. Recently, efforts towards development of OCs for CLOU have focused on combined metal oxides consisting of Cu-based and Mn-based materials mixed with Ca, Mg, Cu, Fe, or Si [9][10][11]. For reactivity and kinetic studies of the Cu-based OCs in CLOU, Thermogravimetric analysis (TGA) is typically used along with fixed bed and fluidized bed reactors [12][13][14][15]. The majority

Cu-Based OC (Cu-OC) Preparation and Characterization
The Cu-OC was prepared from a chalcopyrite sample mined in Mexico. The chalcopyrite was analyzed to determine its mineral composition using X-ray diffraction (XRD) (PANalytical X'Pert PRO). The fresh chalcopyrite sample consisted of chalcopyrite (CuFeS 2 , 75 wt.%) with minor quartz (SiO 2 , 20 wt.%), trace kaolinite (a clay mineral, Al 2 Si 2 O 3 (OH) 4 , 3 wt.%), and trace siderite (FeCO 3, 2 wt.%). The chalcopyrite ore composition by weight percentage was estimated by the reference intensity ratio (RIR) method from the XRD data. Table 2 summarizes the XRD analysis results for the various OC samples generated in this study. To prepare the Cu-OC, the chalcopyrite was crushed, ground, and sieved to a particle size range of 106-180 µm. The particles were then heated at 900 • C for two hours in a fixed bed reactor under flowing air to convert the sulfide ore to the oxide form releasing S as sulfur dioxide (SO 2 ). The selected heating temperature of 900 • C to prepare the Cu-OC was based on an ore oxidation behavior test using a TGA (Linsens HS-TGA) and a quadrupole mass spectrometer (QMS) (Pfeiffer Ominstar GSD 301) ( Figure 1).

Cu-Based OC (Cu-OC) Preparation and Characterization
The Cu-OC was prepared from a chalcopyrite sample mined in Mexico. The chalcopyrite was analyzed to determine its mineral composition using X-ray diffraction (XRD) (PANalytical X'Pert PRO). The fresh chalcopyrite sample consisted of chalcopyrite (CuFeS2, 75 wt.%) with minor quartz (SiO2, 20 wt.%), trace kaolinite (a clay mineral, Al2Si2O3(OH)4, 3 wt.%), and trace siderite (FeCO3, 2 wt.%). The chalcopyrite ore composition by weight percentage was estimated by the reference intensity ratio (RIR) method from the XRD data. Table 2 summarizes the XRD analysis results for the various OC samples generated in this study. To prepare the Cu-OC, the chalcopyrite was crushed, ground, and sieved to a particle size range of 106-180 μm. The particles were then heated at 900 °C for two hours in a fixed bed reactor under flowing air to convert the sulfide ore to the oxide form releasing S as sulfur dioxide (SO2). The selected heating temperature of 900 °C to prepare the Cu-OC was based on an ore oxidation behavior test using a TGA (Linsens HS-TGA) and a quadrupole mass spectrometer (QMS) (Pfeiffer Ominstar GSD 301) (Figure 1).   In the ore oxidation test, the ground chalcopyrite (approximately 175 mg) was tested under "simulated air" (22.4 vol.% O 2 /Ar) at 1000 • C using a non-isothermal method. The sample was heated from room temperature to 100 • C at 5 • C/min and held at 100 • C for 1 min to remove moisture. Then, the temperature was increased to 1000 • C at 60 • C/min and held for about 30 min. The product gases were analyzed on-line by the QMS every second during the reaction. The resulting Cu-OC was analyzed to determine the crystalline phase composition by XRD.
During the TGA-QMS test of Cu-ore oxidation, the TGA detected the sample weight loss with a maximum rate at approximately 500 • C. The final weight loss was approximately 16% and was primarily due to decomposition of CuFeS 2 releasing SO 2 with contributions from FeCO 3 releasing CO 2 and Al 2 Si 2 O 3 (OH) 4 releasing H 2 O. All of the expected gases (SO 2 , CO 2 and H 2 O) were detected by QMS during the chalcopyrite oxidation test. Figure 2 shows the SO 2 ion current from the QMS and the sample temperature versus time for the ground chalcopyrite oxidation. Most of SO 2 was released in the temperature range of 436-845 • C with the maximum at about 600 • C. Thus, the temperature of 900 • C was selected to prepare Cu-OC from the chalcopyrite based on this result. In the ore oxidation test, the ground chalcopyrite (approximately 175 mg) was tested under "simulated air" (22.4 vol.% O2/Ar) at 1000 °C using a non-isothermal method. The sample was heated from room temperature to 100 °C at 5 °C /min and held at 100 °C for 1 min to remove moisture. Then, the temperature was increased to 1000 °C at 60 °C/min and held for about 30 min. The product gases were analyzed on-line by the QMS every second during the reaction. The resulting Cu-OC was analyzed to determine the crystalline phase composition by XRD.
During the TGA-QMS test of Cu-ore oxidation, the TGA detected the sample weight loss with a maximum rate at approximately 500 °C. The final weight loss was approximately 16% and was primarily due to decomposition of CuFeS2 releasing SO2 with contributions from FeCO3 releasing CO2 and Al2Si2O3(OH)4 releasing H2O. All of the expected gases (SO2, CO2 and H2O) were detected by QMS during the chalcopyrite oxidation test. Figure 2 shows the SO2 ion current from the QMS and the sample temperature versus time for the ground chalcopyrite oxidation. Most of SO2 was released in the temperature range of 436-845 °C with the maximum at about 600 °C. Thus, the temperature of 900 °C was selected to prepare Cu-OC from the chalcopyrite based on this result.  Table 2). The oxidation is a complex process and may involve several stages of reaction [23]. The overall oxidation of chalcopyrite may be viewed as follows: In the oxidation process, CuFe2O4 may be formed from CuO and Fe2O3. Conversion of the chalcopyrite to an oxygen carrier is a relatively simple process compared to synthetic Cu-Fe OC preparation which could require multiple steps and many chemicals, another benefit of using chalcopyrite as a resource for a Cu-based OC with low cost.

Cu-OC Oxygen Uncoupling
Cu-OC oxygen uncoupling was tested using a TGA-QMS under Ar at 60 mL/min. A sample of the Cu-OC (approximately 30 mg) was heated from room temperature to 100 °C at 30 °C/min and held at 100 °C for 5 min to remove moisture from the OC. Then the temperature was increased to 900 °C at 60 °C/min and held for about 60 min. The QMS quantified the oxygen gas (O2) on-line every second during the reaction. The resulting reduced Cu-OC was analyzed to determine the crystalline phase composition by XRD.  Table 2). The oxidation is a complex process and may involve several stages of reaction [23]. The overall oxidation of chalcopyrite may be viewed as follows: In the oxidation process, CuFe 2 O 4 may be formed from CuO and Fe 2 O 3 . Conversion of the chalcopyrite to an oxygen carrier is a relatively simple process compared to synthetic Cu-Fe OC preparation which could require multiple steps and many chemicals, another benefit of using chalcopyrite as a resource for a Cu-based OC with low cost.

Cu-OC Oxygen Uncoupling
Cu-OC oxygen uncoupling was tested using a TGA-QMS under Ar at 60 mL/min. A sample of the Cu-OC (approximately 30 mg) was heated from room temperature to 100 • C at 30 • C/min and held at 100 • C for 5 min to remove moisture from the OC. Then the temperature was increased to 900 • C at 60 • C/min and held for about 60 min. The QMS quantified the oxygen gas (O 2 ) on-line every second during the reaction. The resulting reduced Cu-OC was analyzed to determine the crystalline phase composition by XRD.
The non-isothermal method used in this study was done to quantify the gaseous oxygen released from the OC. For the isothermal method used for CLOU, a sample was heated to the test temperature in an air atmosphere then tested by switching to inert gas [25]. In the isothermal method, it is difficult to determine the released oxygen from the OC because the gas from the TGA is a mixture of residual oxygen from the air and the released O 2 from the OC with a relatively low concentration (<1% O 2 in this study).

Recyclability of the Cu-OC with Coal Char in a TGA-QMS
The recyclability of the Cu-OC with coal char CLOU was evaluated at 900 • C for five cycles in the TGA-QMS. For one CLOU cycle, Cu-OC reduction with the coal char was first tested under Ar using the same experimental procedure described in Section 2.3.1. A test sample of the Cu-OC and the char was physically mixed with a ratio of Cu-OC to char (φ) by weight of 75. This ratio was calculated based on oxygen. The available gaseous oxygen in the Cu-OC from oxygen uncoupling is the stoichiometric amount of oxygen required to fully combust the char to CO 2 (Reaction R2). After the reduction of the sample, the reduced Cu-OC was immediately reoxidized at 900 • C for about 15 min under an oxygen gas environment. The oxygen gas was 22.4 vol.% O 2 /Ar at 60 mL/min. The QMS quantified the primary reaction gases (O 2 , CO, CO 2 , CH 4 , and H 2 ). The reduced and reoxidized Cu-OC were analyzed to determine the crystalline phase composition by XRD.
After one CLOU cycle, the system was cooled to room temperature. The gas was switched to Ar for 2nd cycle testing. The same amount of coal char was again added to the sample crucible with the oxidized Cu-OC and coal ash from the previous cycle. The 2nd cycle and subsequent cycles were conducted following the same procedure.

Recyclability of the Cu-OC with Coal Char in a Fixed Bed Reactor-QMS System
The same multicycle tests were repeated using a fixed bed reactor-QMS system to generate samples for BET surface area analysis. The purpose of this test was to obtain more sample for the BET measurements because Cu-OCs, like other OCs, have low surface areas. The BET surface areas of the fresh and used Cu-OC were evaluated by N 2 adsorption isotherms performed at 77 K in a Quantachrome Autosorb 1-C.

Data Analysis
For TGA data, the reactivities of the Cu-OC are evaluated based on mass changes in solids. The Cu-OC oxygen uncoupling, reduction, and re-oxidation reactivity with coal char, and recyclability were evaluated by calculating sample weight change and rates of change. m o for oxygen uncoupling, m red for reduction, and m ox for oxidation are defined as follows: where m, m oc and m c are the mass of the test sample at the reaction time, the initial mass of the Cu-OC and the initial mass of the coal char, respectively. The maximum oxygen uncoupling m o is the oxygen transport capacity R oc (R oc = (m ox − m red )/m ox *100, m ox : mass of the fully oxidized oxygen carrier; m red : mass of the reduced oxygen carrier [4]). For QMS data, the reactivities of the Cu-OC are evaluated based on changes in the number of moles of the gases produced. The oxygen uncoupling (X O ) was determined by calculating the moles of oxygen generated by uncoupling and rate of the change is (dXo/dt). Xo is defined as follow: where n is the moles of oxygen generated over the reaction time (t) and n max is the maximum moles of the oxygen generated.
In the coal char CLOU, the char reacted with gaseous oxygen released by heating Cu-OC and generated primarily CO 2 and CO. The reactivity of the coal char was evaluated by calculating the carbon conversion efficiency (Xc) and its rate (dXc/dt) based on the QMS data [5]. Xc is defined as follows: where n c is the moles of carbon in product gases such as CO 2 and CO over the reaction time n o is the moles of the carbon in the initial coal char sample; F is the gas flow rate; and y i is volumetric concentrations of the carbon containing gas (i). A major purpose for developing coal CLOU is for CO 2 emission control, so a primary goal of coal CLOU is to convert carbon to CO 2 . In this study, the coal CLOU process performance was evaluated by calculating the CO 2 conversion efficiency (Sco 2 ) based on the QMS data [5]. Sco 2 is defined as follows: where n co2 and n t are the total moles of accumulated carbon in CO 2 and in the total carbon containing product gases over the reaction time (t), respectively.

Cu-OC Oxygen Uncoupling
When the Cu-OC was heated over 600 • C, gaseous O 2 was initially generated and detected by the QMS in the product gas and simultaneously a sample weight loss was detected by the TGA (Figures 3  and 4). When the temperature was increased over 820 • C, the gaseous O 2 concentration and the oxygen uncoupling rate quickly increased with a maximum at 900 • C (Figures 3 and 4). These oxygen uncoupling behaviors of the Cu-OC were similar to those of pure CuO [5]. Based on the TGA data, the maximum oxygen uncoupling (R oc ) and oxygen uncoupling rate were 3.3% and 0.0032%/s, respectively (Table 3). Table 3 summarizes the reactivities results of OC uncoupling and carbon conversion in reduction for the Cu-OC in this study and pure CuO from reference [5]. The oxygen uncoupling (R oc ) was similar to the 4% oxygen uncoupling of 40 wt.% CuO with ZrO 2 support OC synthesized by mechanical mixing followed by pelletizing by pressure [25].  [3]). XRD indicated that Cu2O was generated and no CuO was detected in the residue of the Cu-OC oxygen uncoupling test. CuFe2O4 decomposes to generate gaseous O2, delafossite (CuFeO2), and Fe2O3 (Reaction R4) [26]. The XRD analysis of this residue also showed that a Cu-Fe-O compound of CuFeO2 (Cu2OFe2O3) formed. The Fe2O3 decreased and the CuFe2O4 (CuOFe2O3) phase changed (tetragonal to cubic) compared to the fresh Cu-OC (Table 2). So, some CuFe2O4 in the Cu-OC decomposed and released O2 because CuFe2O4 was detected by the XRD. Both CuO and CuFe2O4 are active compounds in the Cu-OC. Zhao et al. studied a Cu-ore oxygen carrier with 21% CuO and reported both the CuO and CuFe2O4 released gaseous O2 but the CuO released oxygen faster than the CuFe2O4 [19].   Based on simple stoichiometric calculations, the maximum oxygen uncoupling or oxygen transport capacity of pure CuO from oxygen uncoupling is 10 wt.% [2]. The equilibrium oxygen concentration at 900 °C is about 1.7 vol.% [3]. Comparing of the oxygen transport capacities of the Cu-OC with pure CuO, the Cu-OC was estimated to contain approximately 33% active CuO (releasing of gaseous O2) by weight. The active CuO content in the Cu-OC was lower than the ideal CuO content from the material prepared from chalcopyrite. Based on reaction R1, the ideal CuO content should be ~40 wt.% CuO since 20% of the chalcopyrite consists of SiO2. The reason for the lower active CuO is due to CuFe2O4 formation from the CuO and Fe2O3 reaction. CuFe2O4 in the Based on simple stoichiometric calculations, the maximum oxygen uncoupling or oxygen transport capacity of pure CuO from oxygen uncoupling is 10 wt.% [2]. The equilibrium oxygen concentration at 900 • C is about 1.7 vol.% [3]. Comparing of the oxygen transport capacities of the Cu-OC with pure CuO, the Cu-OC was estimated to contain approximately 33% active CuO (releasing of gaseous O 2 ) by weight. The active CuO content in the Cu-OC was lower than the ideal CuO content from the material prepared from chalcopyrite. Based on reaction R1, the ideal CuO content should be~40 wt.% CuO since 20% of the chalcopyrite consists of SiO 2 . The reason for the lower active CuO is due to CuFe 2 O 4 formation from the CuO and Fe 2 O 3 reaction. CuFe 2 O 4 in the tetragonal phase was detected by XRD in the fresh Cu-OC. CuFe 2 O 4 in the cubic phase was detected in the Cu-OC after the uncoupling based on XRD. Some CuFe 2 O 4 decomposed and released O 2, which also was from the CuO in CuFe 2 O 4 (Reaction R4) [26].
Based on the QMS data, the maximum oxygen concentration and the oxygen uncoupling rate (dXo/dt) were 0.68 vol.% (Figure 4) and 0.0012 s −1 at T max of 900 • C, respectively ( Table 3). The pure CuO oxygen uncoupling rates tested in a fixed bed reactor were 0.0015 s −1 at T max of 850 • C and 0.0077 s −1 at T max of 922 • C for the tests at test temperatures of 850 • C and 950 • C, respectively (Table 3) [5]. The Cu-OC oxygen uncoupling rate at 950 • C was close to the value for pure CuO at 850 • C and lower than pure CuO at 950 • C. Both CuO concentration and temperature affected the oxygen uncoupling rate.

Reduction and Reoxidization of Cu-OC with Coal Char
The mixture of Cu-OC and char with φ = 75 was evaluated through a redox cycle test of reduction in Ar then reoxidation in air at 900 • C. When the Cu-OC and char were heated above 600 • C, CO 2 was generated and detected by the QMS in the product gas and, simultaneously, a sample weight loss was detected by the TGA (Figure 5). Char combustion took place because the Cu-OC began releasing O 2 at~600 • C from the Cu-OC oxygen uncoupling (Figures 3 and 4) (Reaction R3 and R4), which combusted the char (Reaction R2). The combustion had a peak at~838 • C, which was close to the peak temperature (~833 • C) for the reaction of pure CuO and coal in a TGA under N 2 [27]. The maximum combustion rate was 0.028%/s ( Figure 6) and was eight times faster than the Cu-OC oxygen uncoupling rate (without char) because the combustion reaction (Reaction R2) consumed the oxygen from the Cu-OC uncoupling (Reaction R3 and R4) and increased the Cu-OC oxygen uncoupling rate. The final reduction m red was 4.3% which was the same as the theoretical m red 4.3% based on the char being fully combusted to CO 2 (Reaction R2) and gaseous oxygen being released from the Cu-OC. This means that the char was converted to CO 2 in the reduction of Cu-OC with the char. The product gas was mainly CO 2 with a trace amount of O 2 detected by the QMS (Figure 5). The CO 2 conversion efficiency Sco 2 was 1. The final carbon conversion efficiency Xc was 0.94 (theoretical Xc = 1) and the maximum rate dXc/dt was 0.005 s −1 based on the QMS data (Table 3).  (Table 3). The reduced Cu-OC was fully oxidized at 900 °C in air. The final oxidation mox was 3.2% lower than the final mred because of the mox and mred calculations (Equation 1 and 2). In the mred calculation, m included both char and the Cu-OC. But for mox, only Cu-OC was counted in the reoxidation calculation. The final mox was close to the theoretical mox 3.0% based on the oxygen used to combust the char (Reaction R3) and the mox calculation (Equation 2) including char in the sample weight. The slightly higher mox was due to extra O2 released during the reduction and this was also reoxidized. This additional oxidation was not included in the theoretical mox. The maximum oxidation rate was The reduced Cu-OC was fully oxidized at 900 • C in air. The final oxidation m ox was 3.2% lower than the final m red because of the m ox and m red calculations (Equations (1) and (2)). In the m red calculation, m included both char and the Cu-OC. But for m ox , only Cu-OC was counted in the reoxidation calculation. The final m ox was close to the theoretical m ox 3.0% based on the oxygen used to combust the char (Reaction R3) and the m ox calculation (Equation (2)) including char in the sample weight. The slightly higher m ox was due to extra O 2 released during the reduction and this was also reoxidized. This additional oxidation was not included in the theoretical m ox . The maximum oxidation rate was 0.012%/s, less than half the reduction rate of the OC with char (0.028%/s) ( Figure 6). The temperatures during the reoxidation were slightly higher than during the reduction due to exothermal reoxidation (Reaction R3) ( Figure 6). The reduced Cu-OC was fully oxidized at 900 °C in air. The final oxidation mox was 3.2% lower than the final mred because of the mox and mred calculations (Equation 1 and 2). In the mred calculation, m included both char and the Cu-OC. But for mox, only Cu-OC was counted in the reoxidation calculation. The final mox was close to the theoretical mox 3.0% based on the oxygen used to combust the char (Reaction R3) and the mox calculation (Equation 2) including char in the sample weight. The slightly higher mox was due to extra O2 released during the reduction and this was also reoxidized. This additional oxidation was not included in the theoretical mox. The maximum oxidation rate was 0.012%/s, less than half the reduction rate of the OC with char (0.028%/s) ( Figure 6). The temperatures during the reoxidation were slightly higher than during the reduction due to exothermal reoxidation (Reaction R3) ( Figure 6). The Cu-OC reduced with the char primarily consisted of Cu2O and CuFeO2 which were reoxidized to CuO and CuFe2O4 (tetragonal and cubic phases) based on XRD analysis ( Table 2). The experimental results were very close to the theoretical calculations based on active CuO uncoupling. For the Cu-OC, ϕ = 75 was used to fully combust the char to CO2. For pure CuO ϕ = 26 was selected The Cu-OC reduced with the char primarily consisted of Cu 2 O and CuFeO 2 which were reoxidized to CuO and CuFe 2 O 4 (tetragonal and cubic phases) based on XRD analysis ( Table 2). The experimental results were very close to the theoretical calculations based on active CuO uncoupling. For the Cu-OC, φ = 75 was used to fully combust the char to CO 2 . For pure CuO φ = 26 was selected [5]. Based on a comparison of the φ of the Cu-OC with that of pure CuO, the Cu-OC was estimated to contain~35% of active CuO by weight. Figure 7 shows the reactivities of five cycles of a mixture of Cu-OC and coal char in Ar at 900 • C then oxidized in air at 900 • C. Fresh coal char was added to the oxidized OC, which also contained any residual coal ash, after each cycle. The reactivities and the rates for the reduction of Cu-OC with char were similar for all five cycles. For the reoxidation of the Cu-OC with char, the oxidation m ox and the rates for cycles 2, 4, and 5 were similar but were slight lower for cycle 1 and 3 (Figure 7). This is due to slightly lower oxygen concentrations in the oxidation gas (O 2 /Ar) at beginning of the oxidation. Overall, the result indicates that the Cu-OC was stable during the five cycle tests. then oxidized in air at 900 °C. Fresh coal char was added to the oxidized OC, which also contained any residual coal ash, after each cycle. The reactivities and the rates for the reduction of Cu-OC with char were similar for all five cycles. For the reoxidation of the Cu-OC with char, the oxidation mox and the rates for cycles 2, 4, and 5 were similar but were slight lower for cycle 1 and 3 (Figure 7). This is due to slightly lower oxygen concentrations in the oxidation gas (O2/Ar) at beginning of the oxidation. Overall, the result indicates that the Cu-OC was stable during the five cycle tests. The recyclability of the Cu-OC in coal char CLOU also was tested for five cycles in the fixed bed reactor using the same procedure that was used for the TGA tests. Figure 8 shows the carbon conversion efficiency Xc and generated CO2 concentration for the reduction of Cu-OC with coal char in one cycle of the five-cycle test. The CO2 conversion efficiency Sco2 was 1 because the generated gas was mainly CO2 with little O2 and H2O. There were two CO2 peaks, a very small peak around 600 °C and a large sharp peak around 835 °C. This was similar to the reduction test using pure CuO with PRB coal char at 950 °C, which had a small peak at ~670 °C and a large peak at 833 °C [5]. The small peak may be due to the small particle size fraction of the CuO that released oxygen at a lower temperature because a small peak at a lower temperature of 675 °C along with a large sharp were also observed in pure CuO oxygen uncoupling test in Ar. The carbon conversion efficiency Xc was 0.95, close to the theoretical Xc = 1, which means that the carbon in the coal char was close to complete conversion ( Table 3). The carbon conversion rate dXc/dt was 0.006 s −1 ( Table 3). The reaction rates of pure CuO with PRB coal char were 0.006 s −1 and 0.011 s −1 at test temperatures of 850 °C and 950 °C using the same fixed bed reactor-QMS (Table 3) [5]. For 40 wt.% CuO with Fe2O3 and SiO2 OC synthesized by mechanical mixing followed by pelletizing by pressure at ϕ = 67, the carbon conversion efficiency rate at 1000 °C was 0.0077 s −1 [28] and close to the rate of the Cu-OC from the Cu-ore in this study For the five-cycle test, the carbon conversion efficiency and carbon conversion rates were similar for all five cycles. The results from the fixed bed reactor were similar to the TGA results since both are the same type of reactor the carbon conversion Xc and rate dXc/dt measured in TGA-QMS were slight lower than in the fixed-bed reactor (Table 3) because sample in the fixed-bed reactor was heated faster than in the TGA. In the fixed-bed reactor, The fixed-bed reactor was preheated to 400 °C and had heating rate ~65 °C/min [5]. The recyclability of the Cu-OC in coal char CLOU also was tested for five cycles in the fixed bed reactor using the same procedure that was used for the TGA tests. Figure 8 shows the carbon conversion efficiency Xc and generated CO 2 concentration for the reduction of Cu-OC with coal char in one cycle of the five-cycle test. The CO 2 conversion efficiency Sco 2 was 1 because the generated gas was mainly CO 2 with little O 2 and H 2 O. There were two CO 2 peaks, a very small peak around 600 • C and a large sharp peak around 835 • C. This was similar to the reduction test using pure CuO with PRB coal char at 950 • C, which had a small peak at~670 • C and a large peak at 833 • C [5]. The small peak may be due to the small particle size fraction of the CuO that released oxygen at a lower temperature because a small peak at a lower temperature of 675 • C along with a large sharp were also observed in pure CuO oxygen uncoupling test in Ar. The carbon conversion efficiency Xc was 0.95, close to the theoretical Xc = 1, which means that the carbon in the coal char was close to complete conversion ( Table 3). The carbon conversion rate dXc/dt was 0.006 s −1 ( Table 3). The reaction rates of pure CuO with PRB coal char were 0.006 s −1 and 0.011 s −1 at test temperatures of 850 • C and 950 • C using the same fixed bed reactor-QMS (Table 3) [5]. For 40 wt.% CuO with Fe 2 O 3 and SiO 2 OC synthesized by mechanical mixing followed by pelletizing by pressure at φ = 67, the carbon conversion efficiency rate at 1000 • C was 0.0077 s −1 [28] and close to the rate of the Cu-OC from the Cu-ore in this study The BET surface areas of the fresh Cu-OC and the used Cu-OC after five cycles in the fixed bed reactor were 0.36 and 0.2 m 2 /g, respectively. Both the fresh and used Cu-OC had low surface areas. The fresh Cu-OC surface area in this study was close to the reported by Zhao et al. for a fresh OC prepared from natural copper ore (21 wt.% CuO) for CLOU, 0.217 m 2 /g [19]. Their used OC, after 10 h reduction/oxidation CLOU cycles with anthracite coal in a batch fluidized bed reactor, had a BET surface area of 0.115 m 2 /g. They explained the slight decrease in BET surface area of the used OC as being due to slight sintering of the OC particles or blockage of the pores and available surface area For the five-cycle test, the carbon conversion efficiency and carbon conversion rates were similar for all five cycles. The results from the fixed bed reactor were similar to the TGA results since both are the same type of reactor the carbon conversion Xc and rate dXc/dt measured in TGA-QMS were slight lower than in the fixed-bed reactor (Table 3) because sample in the fixed-bed reactor was heated faster than in the TGA. In the fixed-bed reactor, The fixed-bed reactor was preheated to 400 • C and had heating rate~65 • C/min [5].

Recyclability of Cu-OC in Coal Char CLOU Tested in a TGA-QMS and Fixed-Bed Reactor-QMS
The BET surface areas of the fresh Cu-OC and the used Cu-OC after five cycles in the fixed bed reactor were 0.36 and 0.2 m 2 /g, respectively. Both the fresh and used Cu-OC had low surface areas. The fresh Cu-OC surface area in this study was close to the reported by Zhao et al. for a fresh OC prepared from natural copper ore (21 wt.% CuO) for CLOU, 0.217 m 2 /g [19]. Their used OC, after 10 h reduction/oxidation CLOU cycles with anthracite coal in a batch fluidized bed reactor, had a BET surface area of 0.115 m 2 /g. They explained the slight decrease in BET surface area of the used OC as being due to slight sintering of the OC particles or blockage of the pores and available surface area by the ash generated from coal CLOU, but the impact was very limited [19]. In this study, there was no significant agglomeration observed at 900 • C. After a redox cycle, the OC particles were very weakly bound together but easily loosened to powder again by light pressure. Tian et al. studied Cu ore OC with lignite coal in a batch-scale fluidized bed reactor at 950 • C for 10 cycles and observed a sintering problem in the later cycles [20]. In the future, higher temperatures will be tested to evaluate the sintering, agglomeration, or partial melting of the particles.
Natural ore has potential as a competitive oxygen carrier in solid fuel CLOU based on its reactivity as observed in this study. Based on extensive research on Fe-based OC for iG-CLC, the reactivity of mineral OC is typically lower than synthetic OC [4]. Research on the mineral Fe-OC is continuing because the mineral OC is inexpensive compared to the synthetic OC. For CLOU, a low-cost OC is even more important than iG-CLC because CLOU requires more OC to fully covert the fuels to CO 2 due to its low oxygen capacity. Moreover, low cost is preferred for solid fuel CLOU due to ash from fuel combustion that causes loss of OC reactivity and loss of OC with separation of the ash. Mineral Cu-based OC is much lower in cost than synthetic Cu-OC based on Cu material prices. Cu ore (copper concentrate with 20-30% Cu) costs approximately $0.0695/lb [29], much cheaper than Cu metal at about $2.8875/lb [30]. Cu powder is used to produce CuO by roasting it in air among other different CuO processes. Moreover, Cu-ores are abundant. Converting Cu-ore to an OC is a simple process as compared to production of a synthetic Cu-Fe OC which can require multiple steps and use many chemicals. Future investigation to increase the CuO content of mineral Cu-based OC and operating at higher temperatures could increase the reactivity and make it a more attractive OC. For commercial applications, further study of the long-term cyclic stability of the Cu-OCs with coal is required. Increasing the CuO content in the Cu-OC derived from natural copper ore is needed to increase oxygen transfer capacity and lower the ratio of Cu-OC to fuel to lower the operational cost of a future plant.

Conclusions
This study investigated Cu-based OCs derived from natural minerals for solid fuel CLOU to potentially lower the OC cost. Chalcopyrite, which contains primarily CuFeS 2 , has potential as a Cu-OC resource. It is one of the primary ores of copper, so is readily available. The Cu-OC used in this study was prepared by simply heating ground chalcopyrite in air 900 • C to remove the sulfur. The prepared Cu-OC consisted primarily of CuO and CuFe 2 O 4 (CuOFe 2 O 3 ) from the XRD analysis and had oxygen uncoupling properties based on TGA-QMS testing. The oxygen transport capacity of Cu-OC was 3.3% from both CuO and CuFe 2 O 4 oxygen uncoupling. For the reduction of the Cu-OC with coal during CLOU at 900 • C in Ar, the coal char conversion efficiency Xc was 0.95 and the product gas was primarily CO 2 with trace O 2 . The Cu-OC showed high reactivity and cyclic stability in a five-cycle test. Therefore, the natural ore has potential as a competitive oxygen carrier in solid fuel CLOU based on its reactivity in this study. For commercial applications, further study of the long-term cyclic stability of the Cu-OC with coal is certainly required. Increasing the CuO content in the Cu-OC made from natural copper ore and increasing the reaction temperature are needed to improve the oxygen transfer capacity and lower the ratio of Cu-OC to fuel to reduce the operational cost of a CLOU system. Author Contributions: P.W. is the primary author of this article; B.H., N.M., D.S. and D.B. participated in materials preparation and characterizations, test system setup, discussions, and to the writing and editing of the manuscript.
Funding: This research was supported by the Department of Energy, National Energy Technology Laboratory.