Redox-Mediated Separation of Carbon Dioxide from Flue Gas
John D. Watkins, Nicholas S. Siefert, Xu Zhou, Christina R. Myers, John R. Kitchin, David P. Hopkinson, and Hunaid B. Nulwala, “Redox-Mediated Separation of Carbon Dioxide from Flue Gas”, Energy & Fuels 29(11), 7508-7515 (2015). https://doi.org/10.1021/acs.energyfuels.5b01807
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 membranes separate CO2 by size exclusion or solution diffusion, so they depend on a CO2 concentration gradient and have limited selectivity over N2 and O2. An electrochemically driven membrane could instead pump CO2 actively, even against a concentration gradient, and select it chemically. Earlier electrochemical routes had drawbacks: quinones that bind CO2 directly are sensitive to the water and O2 in flue gas, and the O2 reduction based pump of Kitchin and coworkers carries O2 along with CO2. This paper asks whether the proton coupled electron transfer (PCET) of a quinone in aqueous bicarbonate can be used as a proton pump that sets up a pH gradient across a thin liquid membrane, so that CO2 is absorbed at the high pH cathode and released at the low pH anode (Watkins et al. 2015).
Approach
The intended anode chemistry is oxidation of the hydroquinone (QH2 to Q, releasing two protons and two electrons), with each proton converting a bicarbonate ion to CO2 and water, so ideally one CO2 is released per electron; the cathode runs the reverse and generates transient hydroxide that captures CO2. Two couples were studied, hydroquinone/benzoquinone and 2,6-dimethylbenzoquinone/2,6-dimethylhydroquinone (DMBQ/DMHQ). Cyclic voltammetry on glassy carbon (Ag/AgCl reference) examined scan rate, electrolyte identity (Na2SO4 versus NaHCO3, under Ar or CO2), pH, NaHCO3 concentration (0.1 to 1.0 M), and drop-cast Pt, Pd and Ru on carbon black catalysts. A half-cell test on Toray carbon paper looked for gas evolution during hydroquinone oxidation. The device was a membrane electrode assembly: a hydrophilic Celgard 3501 polypropylene separator (about 25 µm) soaked with 10 mM DMBQ/10 mM DMHQ in 1 M NaHCO3, between two 25 cm2 Sigracet carbon paper electrodes coated with 10 wt% metal on carbon, mounted in a fuel cell stack. Humidified feeds were pure CO2, pure N2, 50% CO2/50% N2, or simulated flue gas (16% CO2, 3% O2, 81% N2), with argon as the permeate sweep and a mass spectrometer gas analyzer (Watkins et al. 2015).
Key findings
Both quinones gave a single quasi-reversible two electron wave in CO2 saturated 0.5 M NaHCO3 (pH 7.8), with large peak separations (about 500 mV) and diffusion control. Randles-Sevcik analysis gave diffusion coefficients of 3.54 x 10-5 cm2/s for hydroquinone and 3.33 x 10-5 cm2/s for DMBQ. Extra oxidation peaks seen in weakly buffered media were attributed to local pH drops at the electrode, which change the proton acceptor (water, then sulfate or bicarbonate), consistent with the view that PCET operates throughout and apparent mechanism changes reflect local pH. The oxidation peak shifted positive as pH fell while the reduction peak barely moved, so the peak separation grew from about 206 mV at pH 7.5 to about 499 mV at pH 6.5, suggesting the oxidation depends on proton acceptor identity while the reduction is not proton limited. Raising NaHCO3 to 1.0 M removed the shoulder peaks and narrowed the separation, so 1.0 M was used for the membrane. Pt and Pd greatly reduced peak separation and increased current; Ru did not help the oxidation and showed little water reduction (Watkins et al. 2015).
In the half-cell, gas evolved at +0.5 V only when hydroquinone, NaHCO3 and CO2 were all present, identifying the gas as CO2 released by the locally acidified anode. In the membrane device, plain carbon transported almost no CO2. With Ru or Pd, significant CO2 appeared only at 2.0 to 2.5 V together with O2 from water splitting (Pd also decayed quickly), whereas Pt gave the most CO2 at 1.0 to 1.5 V with a stable plateau and no O2. Feed side CO2 dropped with applied potential while N2 and O2 did not change. With a pure N2 feed, an initial CO2 burst (above 1.0 V) decayed, attributed to start-up acidification of bicarbonate; sustained transport needed both potential and CO2 in the feed. Qualitative transport efficiencies at 1.0 V, relative to one CO2 per electron, were 57% (feed) and 5% (permeate) at pH 8.73, 42% and 16% at pH 7.20, and 15% and 14% at pH 6.30 (pH adjusted with sulfuric acid). True Faradaic efficiency could not be measured (Watkins et al. 2015).
Significance
The work is a qualitative proof of concept that a reversible organic redox mediator can sustain an in situ pH gradient and pump CO2 selectively from a CO2/N2/O2 feed at about 1 V, below water splitting, without the O2 co-transport of the earlier O2 based pump. It identifies buffer capacity, pH and the electrocatalyst (Pt best) as the controlling variables, and shows a trade-off in pH between CO2 uptake at the feed side and proton driven release at the permeate side. The authors suggest amine media such as monoethanolamine as a future route to faster CO2 uptake and higher capacity (Watkins et al. 2015).