Microbes Turn CO2 and Waste Glycerol Into Acetate in Dual-Electrode Reactor

Every year, industries that turn plant oils into biodiesel produce mountains of crude glycerol, a sticky byproduct that is cheap, abundant, and chemically rich in stored energy. Meanwhile, carbon dioxide continues to accumulate in the atmosphere as the defining waste stream of the industrial age. A new laboratory study published in Results in Engineering brings these two streams together in a single microbial electrosynthesis system, in which bacteria at one electrode oxidize glycerol to generate electrical current while bacteria at the opposite electrode consume carbon dioxide and convert it into acetate, a versatile industrial chemical. The work offers one of the first head-to-head comparisons of how two very different ways of feeding carbon dioxide to the microbial catalyst shape the output of such a dual-biocatalyzed reactor.

The research, led by Khurram Tahir with Americo G. Duarte, Inês A.C. Pereira, and Felipe Conzuelo, addressed a stubborn bottleneck that has long limited microbial electrosynthesis, or MES. In conventional MES reactors, electroactive microorganisms sit on a cathode submerged in liquid medium and reduce dissolved carbon dioxide into volatile fatty acids. The trouble is that carbon dioxide dissolves poorly in water, so microbes at a submerged electrode are often starved of carbon no matter how much gas is bubbled through the broth. The team attacked this problem with a gas-diffusion electrode, a porous, PTFE-treated carbon structure that creates a three-phase interface where gas, liquid, and microbial biofilm meet. Carbon dioxide diffuses only a short distance from the gas phase to the hungry biofilm, bypassing the slow dissolve-and-drift route through bulk liquid.

On the other side of the reactor, the researchers swapped out the usual abiotic anode, where water is oxidized to oxygen at a high energetic cost, for a biological one. A graphite felt bioanode was fed with glycerol at 65 millimolar, and the microbial community there harvested electrons from the oxidation of this reduced compound. Because glycerol oxidation is thermodynamically more favorable than water splitting, this arrangement has the potential to lower the overall energy demand of the full cell, while simultaneously converting a low-value biodiesel byproduct into useful chemicals. The two chambers, each holding 100 milliliters, were separated by a Nafion 117 proton-exchange membrane, and the biocathode was held at a controlled potential of minus 0.800 volts versus an Ag/AgCl reference electrode.

The central experiment compared two cathodic carbon-feeding strategies running in parallel reactors. One biocathode received a continuous stream of gaseous carbon dioxide, supplied as an 80:20 mixture with nitrogen at 1.0 milliliter per minute through the gas-diffusion electrode. The other received a single batch dose of 42 millimolar sodium bicarbonate at the start of each run. Both systems were operated through three sequential 96-hour batch cycles, with the electrolyte fully replaced between cycles, allowing the team to track how performance evolved as the biofilms matured. Inocula came from mixed anaerobic sludge collected at the Rio Tranção near Lisbon, enriched separately for six weeks on bicarbonate or glycerol before being introduced into their respective chambers.

The results favored the gas-fed configuration across every measure of cathodic performance. In the first cycle, the carbon dioxide-fed system averaged a cathodic current density of minus 927 milliamperes per square meter, compared with minus 736 for the bicarbonate-fed reactor. By the third cycle, the gap had widened dramatically: the gas-fed system reached minus 1,703 milliamperes per square meter, an 84 percent increase over its starting point, while the bicarbonate system climbed 46 percent to minus 1,075. The estimated charge transferred through the carbon dioxide-fed cathode rose from 288 to 530 coulombs over the same span. Current density peaked at 48 hours in the final cycle, precisely when acetate production in that chamber hit its maximum, a telling coincidence that ties electron flow to product formation.

Acetate was the star product at the biocathode under both conditions, but the gas-fed system outproduced its rival consistently. It reached a maximum acetate concentration of 98 plus or minus 4 millimolar, equivalent to 5.89 grams per liter, in both the first and third cycles, translating to an apparent productivity of 2.94 grams per liter per day. The bicarbonate-fed system peaked at 79 plus or minus 9 millimolar, or 4.74 grams per liter, with a best productivity of 2.28 grams per liter per day. When benchmarked against earlier studies, the carbon dioxide-fed configuration sits comfortably within the reported range of acetate titers for gaseous CO2 systems, which span roughly 1.3 to 7.8 grams per liter, while its current density of minus 1.703 amperes per square meter is comparable to the best carbon-felt systems operating in potentiostatic mode.

The bioanode told its own chemical story. During the first cycle, acetate dominated the volatile fatty acid profile, reaching 84 plus or minus 9 millimolar in the carbon dioxide-fed system and 74 plus or minus 4 in the bicarbonate system. But as the cycles progressed, the product spectrum shifted, and by the third cycle propionate had overtaken acetate as the major product, peaking at 58 plus or minus 10 millimolar in the carbon dioxide-fed reactor. This acetate-to-propionate transition is consistent with known glycerol fermentation routes, including the acrylate and methylmalonyl-CoA pathways, whose activity depends on the redox balance of the enriched community. Butyrate remained negligible throughout, and the researchers caution that residual glycerol was not quantified, so glycerol conversion and anodic carbon recovery could not be calculated.

Electrochemical fingerprints and genetic sequences reinforced the picture. Cyclic voltammetry at the end of the third cycle showed that the carbon dioxide-fed biocathode delivered a maximum cathodic current response roughly 1.5 times higher than the bicarbonate-fed one, minus 15.1 milliamperes versus minus 9.9 at a scan rate of 100 millivolts per second. Sequencing of the 16S rRNA gene revealed strikingly different endpoint communities. The bioanodes were dominated by Firmicutes, accounting for around 60 percent of reads, including the genera Clostridium sensu stricto, Lachnoclostridium, and Proteiniphilum, taxa frequently associated with organic acid production and extracellular electron transfer. The biocathodes instead hosted Proteobacteria, Firmicutes, and Bacteroidota, with the gas-fed cathode notably enriched in Chloroflexi at 16 percent relative abundance, compared with barely 1 percent under bicarbonate feeding, along with elevated Anaerolineae and Bacteroidia.

The authors are careful to frame these findings as an operational comparison rather than a mechanistic proof. The two feeding modes were not matched in cumulative carbon loading or flux, so the performance gap cannot be isolated as a pure gas-diffusion mass-transfer effect. The cathodic chambers contained multiple carbon pools, including citrate from ferric citrate in the medium, yeast extract, and biomass-associated carbon, and no carbon-13 isotope tracing was performed to attribute the acetate to a single source. Each feeding condition was represented by a single reactor, so the sequential cycles serve as analytical rather than biological replicates, and full-cell voltage, glycerol consumption, coulombic efficiency, and energy demand were not monitored. Current densities also remain at least two orders of magnitude below those of conventional carbon dioxide electrolysis, underscoring how far MES still has to travel toward industrial relevance.

Even with those caveats, the study sketches a compelling blueprint for a new class of carbon-recycling reactors. Pairing a gas-diffusion biocathode that drinks gaseous carbon dioxide with a glycerol-fed bioanode that turns a waste stream into both electrons and volatile fatty acids is an elegant piece of integrated engineering, one that attacks carbon supply, anodic overpotential, and waste valorization in a single design. The researchers point toward the next hurdles with unusual candor: gas-diffusion electrodes are notoriously vulnerable to pore flooding, salt accumulation, and loss of hydrophobicity over long runs, and sustained productivity may demand continuous product recovery to prevent inhibition. Future work with isotope tracing, matched carbon loading, independent reactor replicates, and full energy accounting will determine whether this dual-biocatalyzed platform can graduate from a 100-milliliter laboratory curiosity to a genuine technology for turning emissions and biodiesel waste into chemical feedstocks.

Subject of Research: Comparative evaluation of carbon dioxide feeding strategies in dual-biocatalyzed microbial electrosynthesis with a glycerol-fed bioanode

Article Title: Comparative evaluation of CO₂ feeding strategies in a dual-biocatalyzed microbial electrosynthesis system with a glycerol-fed bioanode

Article References: Tahir, K., Duarte, A. G., Pereira, I. A., & Conzuelo, F. (2026). Comparative evaluation of CO₂ feeding strategies in a dual-biocatalyzed microbial electrosynthesis system with a glycerol-fed bioanode. Results in Engineering, 32, Article 113341. https://doi.org/10.1016/j.rineng.2026.113341

Image Credits: AI Generated

DOI: 10.1016/j.rineng.2026.113341

Keywords: microbial electrosynthesis, carbon dioxide, gas-diffusion electrode, glycerol, acetate, bioanode, biocathode, volatile fatty acids, propionate, bioelectrochemistry, microbiome, biodiesel byproduct