The team, led by Eric M. Garcia at the Federal University of São João del-Rei, focused on the oxygen evolution reaction, or OER, the half of water electrolysis that has long been the technology’s Achilles heel. While hydrogen gas bubbles off the cathode with relative ease, the anode must tear four electrons away from two water molecules to release one molecule of oxygen, and that multi-step process is kinetically sluggish. Electrolyzers must therefore push the anode several hundred millivolts beyond the thermodynamic requirement, an excess called overpotential that is paid for directly in electricity. Because roughly a million tons of lithium-ion batteries are expected to reach end-of-life by 2030, the researchers saw an opportunity to attack two problems at once: recovering cobalt from a growing mountain of electronic waste and using it to lower the cost of green hydrogen production.

The recycling chemistry begins with a battery dismantled by hand, its cathode material gently scraped from the aluminum foil current collector. After a thermal treatment at 400 degrees Celsius and a thorough washing to remove residual electrolyte, 5.3 grams of recovered active material were leached in 3.0 molar sulfuric acid at 100 degrees Celsius in the presence of 1.0 molar hydrogen peroxide. The peroxide plays a crucial chemical role: cobalt in lithium cobalt oxide sits in the +3 oxidation state, which dissolves poorly, but the reducing agent converts it to the far more soluble Co2+ species. Atomic absorption spectroscopy of the resulting leachate showed a cobalt concentration of 0.054 moles per liter, corresponding to about 3.18 grams of cobalt recovered from a theoretical content of roughly 3.19 grams, an apparent leaching efficiency of approximately 99.7 percent.

With the cobalt in solution, the researchers then raised the pH to 10 by slowly adding potassium hydroxide, precipitating cobalt hydroxide with a gravimetrically determined yield of about 98 percent. X-ray diffraction identified the product as predominantly the hexagonal brucite-like beta phase of Co(OH)2, in which charge-neutral layers of octahedrally coordinated Co2+ ions are stacked through hydrogen bonding. Applying the Scherrer equation to five diffraction peaks gave apparent crystallite sizes between roughly 5 and 10 nanometers, averaging 7.5 nanometers, a small domain size that typically favors catalytic activity by maximizing exposed surface. Infrared spectroscopy added nuance: broad hydroxyl stretching bands and the absence of the sharp 3645 per centimeter feature characteristic of well-ordered beta-phase material pointed to local structural disorder and hydration, likely influenced by residual sulfate species carried over from the acid leaching step, which energy-dispersive X-ray spectroscopy confirmed as trace sulfur alongside potassium from the precipitant.

Rather than coating this powder onto a conventional substrate, the team pressed it directly into the body of an electrode. Graphite powder was mixed with the recycled cobalt hydroxide at loadings ranging from 5 to 80 percent by weight, dispersed in ethanol, ultrasonicated for two hours and dried. The powder was then blended into molten paraffin wax at 80 degrees Celsius and compacted under ten tons of uniaxial pressure into a self-supported cylindrical pellet of two square centimeters. The paraffin serves as a hydrophobic binder that gives the pellet mechanical cohesion, while the graphite provides electrical percolation throughout the bulk. The approach deliberately avoids polymer binders such as Nafion and skips film-deposition steps entirely, meaning the catalyst is distributed through the electrode volume rather than confined to a thin surface layer.

Cyclic voltammetry screening revealed a clear optimum at 40 percent cobalt hydroxide, where the current density at 0.6 volts versus Ag/AgCl peaked before declining at higher loadings. The researchers attribute this maximum to a trade-off between catalytic content and conductivity: too much insulating hydroxide disrupts the graphite network through which electrons must travel, and the hydrophobic paraffin matrix already partially restricts electrolyte penetration to particles buried deep in the composite. The voltammograms also displayed a distinctive anodic process beginning near 0.2 volts, matched by a cathodic peak on the reverse scan, corresponding to the reversible oxidation of Co(OH)2 to cobalt oxyhydroxide, CoOOH. This transformation is widely regarded as the activation step in cobalt-based oxygen evolution catalysis, because the oxyhydroxide phase is more conductive and hosts the active sites where hydroxide ions are adsorbed and oxidized.

Quantitatively, the optimized electrode delivered a current density of 10 milliamperes per square centimeter at an overpotential of approximately 310 millivolts in 1.0 molar potassium hydroxide, with a Tafel slope of 75 millivolts per decade. Both figures compare favorably with literature benchmarks: the overpotential is lower than values reported for nickel phosphide on glassy carbon at 390 millivolts, carbon-supported NiCo nanospheres at 330 millivolts, and cobalt hydroxide on glassy carbon electrodes at 414 millivolts, though it does not match the 236 millivolts achieved by a sophisticated NiFe alloy nanotube catalyst. For an electrode whose active ingredient was pulled from a dead mobile phone and bound in candle wax, the performance is striking. The researchers also measured a reaction order of 0.87 with respect to hydroxide ions by varying the KOH concentration from 0.1 to 6.0 molar, a near-unity value indicating that hydroxide adsorption participates directly in the rate-determining step, consistent with the classical Bockris-Otagawa mechanism in which an adsorbed hydroxyl intermediate reacts with a second hydroxide ion to form a peroxide-like species before oxygen release.

Interface diagnostics reinforced the picture. Double-layer capacitance, a proxy for the electrochemically accessible surface area, rose from 1.69 to 14.2 microfarads per square centimeter upon incorporating the recycled hydroxide, an 8.4-fold increase. Polarization resistance, measured near the equilibrium potential, fell from about 11.4 to 3.15 kiloohm square centimeters, a 72 percent reduction. Most importantly for any practical device, the electrode held a current density close to 10 milliamperes per square centimeter for 180 hours of continuous operation at the 310 millivolt overpotential, with a slight upward drift in current that the authors interpret as progressive electrochemical activation of the cobalt surface under anodic polarization.

The authors are careful about the limits of their evidence. They did not measure Faradaic efficiency for oxygen, perform post-mortem microscopy or spectroscopy, or run electrochemical impedance spectroscopy, so the long-term test demonstrates operational stability of the current rather than proving that the electrode’s structure survived intact. They likewise note that definitive identification of the CoOOH-like active phase would require operando spectroscopy, since cobalt surfaces are known to reconstruct substantially under oxygen-evolving conditions, changing oxidation state and coordination as they work. Trace impurities below the detection limits of their analytical methods also cannot be excluded.

Even with those caveats, the study offers a compelling template for a circular hydrogen economy. It demonstrates that a battery destined for the shredder can pass through a single acid leach and a single precipitation step and emerge as a competitive water-oxidation anode, without phase engineering, nanostructure synthesis or noble metals. As electrolyzer deployments scale alongside renewable power, and as the first great wave of lithium-ion batteries reaches retirement, routes like this one suggest that yesterday’s phones could help supply the electrodes that split tomorrow’s water.

Subject of Research: Recycling cobalt from spent lithium-ion battery cathodes into low-cost electrodes for the oxygen evolution reaction in alkaline water electrolysis

Article Title: Recycled Co(OH)₂ from spent LiCoO₂ cathodes as a low-cost graphite-based composite electrode for the oxygen evolution reaction

Article References: Garcia, E. M., Taroco, H. A., Melo, J. O. F., & Taroco, C. G. (2026). Recycled Co(OH)₂ from spent LiCoO₂ cathodes as a low-cost graphite-based composite electrode for the oxygen evolution reaction. Ionics. https://doi.org/10.1007/s11581-026-07540-w

Image Credits: AI Generated

DOI: 10.1007/s11581-026-07540-w

Keywords: lithium-ion battery recycling, cobalt hydroxide, oxygen evolution reaction, electrocatalysis, water electrolysis, green hydrogen, LiCoO2 cathode, graphite composite electrode, cobalt recovery, Tafel analysis, alkaline medium, circular economy

The starting materials belong to a family of cyanometallate coordination polymers with the general framework T[Fe(CN)₅NO], where T is a transition-metal ion. In the study, T was cobalt, nickel, or copper. Pyridine molecules bind to the outer metal centers and change the architecture from a three-dimensional bulk structure into layered two-dimensional sheets. This transformation is important because the parent nitroprussides have limited surface area and disperse poorly, whereas thin layers can be distributed across a conductive carbon support. Pyridine is therefore not introduced primarily as an oxygen-evolution promoter. Its principal role is structural: by selectively replacing axial metal–cyanide connections, it enables the layered morphology and helps stabilize neighboring sheets through pi–pi and dipole–dipole interactions. The resulting materials were combined with reduced graphene oxide in a one-to-one weight ratio. The graphene network provides pathways for electron transport and helps expose the coordination-polymer domains to the alkaline electrolyte.

Several complementary measurements confirmed that the metal identity altered both the structure and the electronic environment of the composites. X-ray diffraction showed layered materials with characteristic basal-plane reflections for the cobalt and nickel systems, while the copper composite displayed different stacking behavior. Electron microscopy revealed average particle sizes of approximately 133 nanometers for the cobalt material, 47 nanometers for the nickel material, and 214 nanometers for the copper material. Infrared and Raman spectroscopy detected the cyanide, nitrosyl, pyridine, and graphene-related signatures expected from the composite design. The cyanide stretching frequency shifted systematically as the metal changed from cobalt to copper, reflecting differences in the cations’ polarizing power. Raman measurements also indicated a defect-rich reduced graphene oxide support. The disorder can be useful in electrocatalysis because defects may create additional sites for oxygen-related reactions, although the measured surface area alone did not determine which composite performed best.

Electrochemical testing revealed a clear division of labor among the three materials. For OER in alkaline solution, the nickel composite showed the lowest onset potential and the strongest current response, reaching 10 milliamperes per square centimeter at an overpotential of about 398 millivolts. The cobalt and copper versions required approximately 437 and 518 millivolts, respectively, under the same benchmark. Nickel also exhibited the lowest charge-transfer resistance across the tested potential range. Cobalt, however, produced the lowest fitted OER Tafel slope, 61 millivolts per decade, compared with 74 for nickel and 156 for copper. The researchers caution that a lower Tafel slope by itself does not establish the fastest overall reaction; onset potential, overpotential, resistance, and operating conditions must be considered together. For ORR, the pattern reversed. The copper composite delivered the lowest onset potential and the highest limiting current, indicating the most favorable oxygen-reduction behavior among the tested materials.

Rotating-disk experiments provided further information about the oxygen-reduction pathway. Koutecky–Levich analysis indicated that all three composites predominantly followed a four-electron route, which is desirable in zinc-air batteries because it converts oxygen more directly and limits peroxide formation. The estimated electron-transfer numbers were approximately 3.95 for cobalt, 3.63 for nickel, and 3.97 for copper. The nickel value suggests that a mixture of four-electron and two-electron pathways may occur, potentially producing some hydrogen peroxide. The authors note that rotating-ring-disk measurements would be needed to investigate that possibility directly. Together, the half-cell results show why optimizing a rechargeable zinc-air battery is difficult: nickel favors the charging reaction, copper favors the discharging reaction, and neither is automatically the best complete-device material. The measurements instead point toward a compromise in which both reactions remain sufficiently active.

That compromise emerged when the composites were assembled into working zinc-air batteries. The cobalt-based cathode produced the highest peak power density, 54 milliwatts per square centimeter, narrowly exceeding the nickel system at 53 and clearly surpassing the copper system at 39. It also maintained higher discharge-voltage plateaus, particularly as the current density increased. During galvanostatic cycling at 10 milliamperes per square centimeter, the cobalt battery began with a voltage gap of 0.99 volts between charge and discharge and reached 1.18 volts after 24 hours. The corresponding gaps for copper were 1.05 and 1.29 volts, while nickel showed 1.43 and 1.36 volts. Initial round-trip energy efficiencies were 52 percent for cobalt, 50 percent for copper, and 34 percent for nickel. After the cycling period, they were 46, 42, and 33 percent, respectively. The cobalt device also retained a relatively stable impedance and delivered an average energy above 8.5 watt-hours per square meter per cycle.

The study’s central chemical finding is that the synthesized nitroprussides do not remain unchanged during battery operation. Post-cycling analysis of the best-performing cobalt electrode showed that the characteristic cyanide signal disappeared, while hydroxyl and metal–oxygen signatures emerged. X-ray diffraction and Raman spectroscopy likewise indicated conversion into a cobalt hydroxide or oxyhydroxide phase. The original material is therefore a precursor rather than the final catalyst. Importantly, X-ray photoelectron spectroscopy detected a pyridinic nitrogen signal after cycling, suggesting that the pyridine ligand remained associated with the reconstructed active phase instead of being completely lost. The researchers propose that the retained molecules act as structural pillars and electronic modifiers, helping prevent the newly formed oxyhydroxide sheets from agglomerating. The same alkaline environment that drives the battery reactions promotes this reconstruction. Similar hydroxide or oxyhydroxide phases are considered likely for the nickel and copper composites, although direct post-cycling structural confirmation was performed most extensively for cobalt.

Magnetic measurements offered an unusual window into the reconstructed electrodes, indicating that the active hydroxides are locally disordered rather than simple, perfectly ordered crystals. Measurements of magnetic susceptibility showed metal-dependent interactions, including antiferromagnetic correlations, weak ferromagnetic-like behavior, metamagnetic transitions, and magnetic relaxation. The cobalt and nickel hydroxide phases displayed low-temperature features associated with competing or complex interactions, while the copper material showed behavior consistent with antiferromagnetic correlations and a distorted local environment. These observations matter because local disorder can change the electronic states available for binding oxygen intermediates. Valence-band X-ray photoelectron spectra placed the approximate d-band centers in the order nickel below cobalt below copper. Within the commonly used d-band framework, a higher d-band center can strengthen adsorption of oxygen intermediates, helping ORR but potentially making OER less favorable. The results fit that qualitative picture: copper was strongest for ORR, nickel for OER, and cobalt, with an intermediate electronic position, offered the best balance. The authors emphasize that the d-band values are approximate descriptors, not proof of a single causal mechanism.

The work presents coordination polymers as tunable molecular precursors for practical oxygen electrocatalysis, while also highlighting the limits of laboratory battery demonstrations. The cobalt composite approached a zinc-specific capacity of 809.2 milliampere-hours per gram at 5 milliamperes per square centimeter, close to the theoretical value of 820, but the researchers caution that this result reflects a relatively low current density, a high-purity zinc anode, and a particular concentrated alkaline electrolyte. Further studies will be needed to assess operation at higher rates, longer lifetimes, different electrolyte compositions, and larger electrode areas. In situ spectroscopy and density-functional calculations could also test how the retained pyridine, disordered hydroxide structure, and d-band position cooperate during cycling. Even with those questions unresolved, the comparison demonstrates a useful design principle: the best bifunctional catalyst is not necessarily the material that wins either half-reaction independently. In this series, deliberately balancing electronic structure, conductive architecture, and electrochemical reconstruction allowed the cobalt-based pre-electrocatalyst to deliver the most practical zinc-air battery performance.

The precursor strategy also addresses a practical materials challenge beyond catalytic activity. Conventional oxygen electrocatalysts often rely on platinum for reduction or iridium- and ruthenium-based oxides for evolution, but the scarcity and cost of these elements complicate deployment at scale. Nitroprusside coordination polymers offer a different platform because their metal sites, cyanometallate framework, and organic coordination environment can be varied systematically. The cobalt, nickel, and copper comparison therefore functions as a controlled probe of how first-row transition-metal chemistry affects a common molecular architecture, rather than as a simple search for a single universally optimal composition.

That architecture may be especially valuable for studying catalyst reconstruction under realistic operation. Because the electrochemically generated hydroxide is the working phase, evaluating only the pristine powder could obscure the properties that govern a battery’s behavior. The reported combination of spectroscopy, diffraction, microscopy, and magnetic susceptibility illustrates why identifying active materials requires measurements before and after polarization. It also suggests a broader design opportunity: coordination ligands can serve as temporary structure-directing components while influencing the local environment retained after conversion. Determining how much pyridine survives, where it is located, and whether its effect persists over extended cycling will be important for distinguishing a genuine electronic contribution from a short-lived synthesis advantage. Such questions are relevant to translating these composites from proof-of-concept electrodes into durable, scalable air-cathode materials.

Subject of Research: Transition-metal nitroprusside and reduced graphene oxide pre-electrocatalysts for rechargeable zinc-air batteries

Article Title: Pyridine-based nitroprussides coupled to reduced graphene oxide as bifunctional pre-electrocatalysts for zinc-air batteries

Article References: Quintanilla-Serrano, E. A., Acevedo-Peña, P., Ávila, Y., González, M., & Reguera, E. (2026). Pyridine-based nitroprussides coupled to reduced graphene oxide as bifunctional pre-electrocatalysts for zinc-air batteries. Discover Electrochemistry, 3(1), Article 77. https://doi.org/10.1007/s44373-026-00164-9

Image Credits: AI Generated

DOI: 10.1007/s44373-026-00164-9

Keywords: zinc-air batteries, electrocatalysis, reduced graphene oxide, nitroprussides, oxygen reduction, oxygen evolution, cobalt hydroxide, d-band center, Pyridine-based, coupled, reduced, graphene