Separation of CO2 from flue gas using electrochemical cells
Henry W. Pennline, Evan J. Granite, David R. Luebke, John R. Kitchin, James Landon and Lisa M. Weiland, “Separation of CO2 from flue gas using electrochemical cells”, Fuel 89(6), 1307-1314 (2010). https://doi.org/10.1016/j.fuel.2009.11.036
This summary was written by a language model from the paper and has not been checked line by line. Rely on the published paper for details.
Question
Conventional post-combustion capture by monoethanolamine scrubbing costs a coal plant close to 30% of its thermal efficiency, largely because the water-rich solvent must be heated to regenerate it, and any sorbent or solvent process loses energy moving the absorbent in and out of the flue gas. Electrochemical pumps are a stationary, one-step alternative in which CO2 and O2 cross an electrolyte as carbonate or bicarbonate ions under an applied potential. Earlier work used molten carbonate cells, which run above 400 C (around 650 C in the cited Winnick studies), are corrosive, need low voltages and therefore huge stacks, and are exposed to SO2 that cannot be scrubbed out upstream at such temperatures. Polymer cells with porous membranes soaked in bicarbonate had needed more than 2.5 V, above the 1.23 V where water splitting begins. The paper asks whether a low temperature cell, placed after conventional SCR, wet scrubbing and particulate control, can separate CO2 from flue gas, and proposes a process that uses the resulting CO2/O2 stream (Pennline et al. 2010).
Approach
This is a joint NETL and Carnegie Mellon concept and proof-of-concept paper. The process concept sends cleaned flue gas through an electrochemical cell in which O2 is reduced at the cathode to hydroxide, raising the local pH so CO2 is absorbed as carbonate or bicarbonate; the anions migrate across the electrolyte, and at the anode hydroxide is oxidized back to O2, lowering the pH and releasing CO2. The CO2/O2 product (at most 2:1 for carbonate carriers, 4:1 for bicarbonate) then feeds an ancillary oxy-fired combustor burning more fuel, with recycle of product gas for temperature control, so that condensing the water leaves nearly pure CO2. The paper reviews oxy-combustion as background, including the cost of cryogenic air separation that this concept would avoid.
Experimentally, Pb, Sn, Ni, Pd, Pt, Cu, Ag and Au wires were screened by cyclic voltammetry in carbonate and bicarbonate electrolytes. Ni particles 1 to 4 μm across were made by a polyol reduction of NiCl2 in ethylene glycol at 160 C and painted at 1 mg/cm2 onto Toray carbon paper backed with XC-72R carbon black. CO2 evolution from KHCO3 solutions was quantified volumetrically in a commercial water electrolyzer with Ni electrodes, inferring O2 from the collected H2. Finally, an adapted fuel cell with a Neosepta anion exchange membrane between the gas-diffusion electrodes was fed O2 or a 50:50 CO2/O2 mixture on the cathode, with a mass spectrometer analyzing the Ar-swept anode gas.
Key findings
Cu, Ag, Pb and Sn corroded and Au was inactive; among Ni, Pd and Pt in 0.1 M KHCO3, Ni gave the highest current between 0.8 and 1.0 V vs. SCE, where oxygen evolution occurs, and the synthesized Ni ink outperformed Ni wire with much less metal. Peaks near 0.5 V vs. SCE were assigned to carbonate adsorption and desorption, and the current rise above that to O2 and CO2 evolution from carbonate decomposition. In the electrolyzer, CO2 evolution grew with bicarbonate concentration and current and was much faster from bicarbonate than carbonate (bicarbonate needs one proton to decompose, carbonate two), but the best CO2/O2 ratio was only about 1, at 3 M bicarbonate, well below the theoretical 4; the authors attribute the shortfall to protons crossing to the cathode, an artifact of that device.
In the membrane cell, pure O2 was pumped across at 1.2 V. Within minutes of switching to the CO2/O2 feed, CO2 appeared at the anode, and transport of both gases stopped when the cell voltage was set to 0 V. Transport occurred below 1.23 V, so no water was split, and O2 flux tracked current density as voltage rose; about 0.8 V was needed for measurable transport (1 mA/cm2). Assuming a 4:1 CO2/O2 product, 5 mA/cm2 corresponds to a CO2 flux of 0.069 ml/(cm2 min) and an electrical demand of 116 kJ/mol CO2 at 1.2 V, falling to 77 kJ/mol CO2 at 0.8 V and 1 mA/cm2. The authors stress that the actual product ratio and fluxes were not yet measured.
Significance
The paper establishes that an ambient-temperature anion exchange membrane cell with Ni gas-diffusion electrodes can electrochemically pump CO2 together with O2 at voltages below water splitting, with transport switched on and off by the potential, and it frames this as the separation step of a hybrid electrochemical plus oxy-firing capture scheme compatible with existing flue gas cleanup (Pennline et al. 2010). Its energy estimates are preliminary, and the extracted text omits the figure data, so only the numbers stated in the text are reported here.