Electrochemical Hydrogen Pumping and Separation with Proton-Conducting Ceramics
Electrochemical hydrogen pumps use a proton-conducting solid to selectively transport hydrogen from one gas stream to another. Unlike a conventional pressure-driven membrane, the device converts molecular hydrogen into protonic defects at one electrode, transports those defects through a dense ceramic, and reforms molecular hydrogen at the opposite electrode.
Proton-conducting ceramic membranes can also support hydrogen permeation without an external electrical circuit when the membrane has both protonic and electronic conductivity. These two operating concepts are related but should not be confused: an electrochemical pump uses an external circuit to close the electron pathway, whereas a mixed proton–electron conducting membrane can sustain coupled transport internally.
This article explains the mechanisms, material requirements, cell configurations, and measurements needed to evaluate ceramic hydrogen separation.
1. Why separate hydrogen electrochemically?
Hydrogen-containing streams may also contain nitrogen, carbon monoxide, carbon dioxide, methane, steam, or other gases. The required hydrogen purity depends on the downstream process, such as fuel-cell operation, chemical synthesis, or gas recycling.
A dense proton-conducting ceramic can selectively transport hydrogen because the hydrogen molecule is dissociated at an electrode and transferred through the solid as protonic defects. Other gas molecules do not normally cross the dense electrolyte through the same proton-transport mechanism.
Potential advantages include: - Selective hydrogen transport without relying on a porous molecular-sieving layer. - Electrochemical control of hydrogen transfer rate. - Operation at elevated temperatures compatible with selected process streams. - Possibility of hydrogen recovery from dilute mixtures. - Integration with catalytic reactors or other high-temperature processes.
These advantages do not imply that ceramic pumps are automatically more efficient or more economical than polymer membranes, palladium-alloy membranes, or pressure-swing adsorption. The comparison depends on gas composition, temperature, hydrogen partial pressure, electricity use, membrane area, sealing, and lifetime.
2. Fundamental proton transport
In many proton-conducting perovskites, acceptor doping creates oxygen vacancies that can incorporate hydrogen from water vapour or hydrogen-containing atmospheres.
Hydration is represented by:
\[ H_2O(g)+V_O^{\bullet\bullet}+O_O^x \rightleftharpoons 2OH_O^{\bullet} \]
The proton is stabilized as a hydroxyl defect associated with a lattice oxygen. It moves through the oxide by local reorientation and transfer between neighbouring oxygen sites.
For hydrogen pumping, the electrode reaction can be written in a simplified form as:
\[ H_2(g)\rightleftharpoons 2H^{\bullet}_{(electrolyte)}+2e' \]
The protonic defects move through the ceramic electrolyte, while electrons move through an external electronic circuit. At the receiving electrode, the reverse reaction forms hydrogen:
\[ 2H^{\bullet}_{(electrolyte)}+2e' \rightarrow H_2(g) \]
The net result is hydrogen transfer from the feed side to the permeate side. The electrolyte must conduct protons while suppressing electronic leakage; otherwise, part of the applied current may bypass the intended hydrogen-transfer process.
3. Two distinct membrane concepts
3.1 Electrochemical hydrogen pump using a proton conductor
In a conventional electrochemical pump, a proton-conducting electrolyte is placed between two electronically conducting electrodes. An external voltage drives hydrogen oxidation at the feed electrode and hydrogen evolution at the permeate electrode.
The electrons travel through the external circuit. The protons travel through the ceramic. The two pathways complete the electrochemical circuit.
This configuration is useful when hydrogen transfer must be actively controlled or driven against a hydrogen chemical-potential gradient.
3.2 Spontaneous permeation through a mixed proton–electron conductor
A dense mixed protonic–electronic conducting (MPEC) membrane can transport hydrogen when a hydrogen chemical-potential gradient exists between its two surfaces.
At the feed surface, hydrogen dissociates into protonic and electronic defects. Both species migrate through the membrane in a coupled manner and recombine at the permeate surface.
No external electron circuit is required for the idealized spontaneous-permeation mode because the membrane itself supports both transport pathways. The overall flux is limited by the coupled protonic and electronic transport, as well as surface exchange.
A material that conducts protons but has negligible electronic conductivity cannot sustain this same steady-state spontaneous-permeation mechanism without an external electronic pathway.
3.3 External short-circuit and assisted membrane operation
Some mixed-conducting membrane designs use an external short circuit or an applied electrical potential to modify electron transport and increase hydrogen permeation. This is an electrochemically assisted membrane configuration and should be distinguished from purely spontaneous permeation.
The experimental description should state whether the membrane is operated open-circuit, short-circuit, or under a controlled applied voltage. These modes have different charge-balance conditions and energy requirements.
4. Cell architecture and materials
A laboratory electrochemical hydrogen pump commonly contains:
- A dense proton-conducting ceramic membrane.
- A porous hydrogen-oxidation electrode on the feed side.
- A porous hydrogen-evolution electrode on the permeate side.
- Gas chambers and manifolds that keep the two streams separate.
- Current collectors and electrical leads.
- A high-temperature sealing system.
4.1 Electrolyte materials
Candidate electrolytes include acceptor-doped cerate-, zirconate-, and cerate–zirconate-based perovskites. Examples include doped barium cerates, barium zirconates, and mixed compositions.
Material selection requires balancing: - Proton conductivity. - Chemical stability in CO\(_2\)- and H\(_2\)O-containing streams. - Electronic leakage. - Sinterability and densification. - Thermal expansion and mechanical strength. - Compatibility with electrodes and seals.
A high proton conductivity measured in humid air is not sufficient evidence of good hydrogen-pump performance. The relevant conductivity must be evaluated under the actual hydrogen, steam, and temperature conditions used in the device.
4.2 Electrode requirements
Both electrodes must catalyse hydrogen dissociation or formation and provide electronic conduction. Porous noble-metal electrodes are often used in fundamental studies, while ceramic–metal composites and other catalytic electrodes are investigated for improved cost and compatibility.
Electrode performance depends on catalyst dispersion, porosity, gas access, contact with the electrolyte, and stability during thermal cycling. Electrode polarization can dominate the cell voltage even when the electrolyte resistance is low.
4.3 Membrane thickness
Reducing electrolyte thickness can lower ohmic resistance, but very thin membranes are more vulnerable to pinholes, cracks, handling damage, and gas leakage. A practical design balances ionic resistance with mechanical integrity and manufacturability.
For a planar membrane, the idealized area-specific electrolyte resistance scales as:
\[ ASR_{elyte}=\frac{L}{\sigma_H} \]
where \(L\) is electrolyte thickness and \(\sigma_H\) is the protonic conductivity. This expression neglects grain-boundary, contact, and interfacial contributions.
5. Thermodynamic driving force and applied voltage
Hydrogen transfer is governed by the difference in hydrogen chemical potential between the feed and permeate sides. For ideal gases:
\[ \mu_{H_2}=\mu_{H_2}^{\circ} +RT\ln\left(\frac{p_{H_2}}{p^{\circ}}\right) \]
The reversible voltage associated with a hydrogen partial-pressure ratio follows from the Nernst relation. Its magnitude is:
\[ |E_{rev}|=\frac{RT}{2F} \left|\ln\left( \frac{p_{H_2,permeate}}{p_{H_2,feed}} \right)\right| \]
The sign depends on the chosen cell-voltage convention and transfer direction. For non-ideal mixtures, hydrogen fugacity should replace partial pressure.
The applied voltage must account for the reversible requirement and the additional losses associated with electrolyte resistance, electrode polarization, contacts, and gas transport. Consequently, increasing the applied voltage does not necessarily increase hydrogen production proportionally; parasitic current and material degradation may become more important.
6. Hydrogen flux and Faradaic efficiency
For an ideal electrochemical pump, two electrons are required per hydrogen molecule transported. The theoretical hydrogen production rate is:
\[ \dot n_{H_2,theoretical}=\frac{I}{2F} \]
where \(I\) is current in amperes and \(F\) is the Faraday constant.
If the measured hydrogen production rate is \(\dot n_{H_2,measured}\), the Faradaic efficiency is:
\[ FE=\frac{2F\dot n_{H_2,measured}}{I}\times100\% \]
The equation assumes that the measured hydrogen is generated by the intended electrochemical reaction and that current integration and gas quantification are accurate.
The hydrogen flux normalized by membrane area is:
\[ J_{H_2}=\frac{\dot n_{H_2}}{A} \]
where \(A\) is the active membrane area. Report the units explicitly, for example mol m\(^{-2}\) s\(^{-1}\) or mmol cm\(^{-2}\) h\(^{-1}\).
A high current density is not equivalent to a high hydrogen flux if Faradaic efficiency is low. Product analysis is therefore essential.
7. What limits hydrogen-pump performance?
7.1 Electrolyte ohmic resistance
The electrolyte resistance depends on proton conductivity, thickness, grain-boundary contributions, and temperature. A low-conductivity or poorly densified electrolyte requires a larger voltage to sustain a given current.
7.2 Electrode polarization
Hydrogen dissociation, adsorption, charge transfer, and recombination may be slow. Electrode polarization can be investigated using impedance spectroscopy and symmetric-cell measurements.
7.3 Electronic leakage
If the electrolyte has appreciable electronic conductivity, part of the applied current may pass through the ceramic without transporting the expected amount of hydrogen. This reduces current efficiency.
The electronic contribution can vary with oxygen chemical potential, hydrogen activity, temperature, and material composition. It should be measured or bounded under the actual operating atmosphere.
7.4 Gas transport and concentration polarization
The hydrogen partial pressure at the active electrode surface may differ from the inlet value because of gas conversion, stagnant boundary layers, or insufficient flow. This changes the local Nernst potential and can reduce the observed flux.
Report feed composition, flow rate, pressure, utilization, and the method used to determine the permeate-side hydrogen rate.
7.5 Sealing and gas crossover
High-temperature seals are a major experimental challenge. Leakage can cause an apparent loss of hydrogen recovery, contaminate the permeate, or create misleading current-efficiency values.
Leak tests should be conducted at operating temperature where possible. A dense membrane should also be checked for cracks and pinholes after testing.
8. Electrochemical characterization workflow
A reliable hydrogen-pump study should combine electrical measurements with direct gas analysis.
- Characterize the ceramic. Confirm phase composition, density, dimensions, and microstructure.
- Verify gas tightness. Test the membrane and seals before applying current.
- Establish the atmosphere. Record feed and permeate gas composition, humidity, flow rate, pressure, and temperature.
- Measure open-circuit behaviour. Check baseline voltage and gas crossover before pumping.
- Perform impedance spectroscopy. Determine the total resistance and evaluate electrode polarization over a suitable frequency range.
- Record current–voltage curves. Use controlled voltage steps and allow the gas and cell response to stabilize.
- Quantify hydrogen production. Use a calibrated gas chromatograph, mass spectrometer, or suitable flow measurement method.
- Calculate Faradaic efficiency. Compare measured hydrogen production with the charge passed.
- Repeat at different operating conditions. Vary temperature, hydrogen partial pressure, flow rate, and current density systematically.
- Assess durability. Monitor voltage, current efficiency, gas purity, and impedance during extended operation.
- Post-test characterization. Inspect the membrane, electrodes, and interfaces for cracks, reactions, and microstructural changes.
The time required to reach steady state should be established experimentally. A stable current does not guarantee a stable permeate composition.
9. Distinguishing hydrogen pumping from spontaneous permeation
The following comparison helps prevent confusion between two related technologies.
| Feature | Electrochemical pump | Spontaneous MPEC membrane |
|---|---|---|
| Main transport pathways | Protonic ceramic plus external electronic circuit | Protonic and electronic conduction within membrane |
| Driving force | Applied electrical potential | Hydrogen chemical-potential gradient |
| Electron pathway | External circuit | Internal electronic transport |
| Control variable | Applied voltage or current | Feed/permeate composition, temperature, membrane properties |
| Key performance metric | Hydrogen flux and Faradaic efficiency | Hydrogen permeation flux and selectivity |
| Main losses | Ohmic, electrode polarization, electronic leakage, parasitic current | Coupled transport resistance and surface-exchange limitations |
The table is a conceptual comparison. Assisted MPEC designs can combine an applied electrical bias with internal mixed conduction, so the exact operating mode must be described.
10. Practical data analysis
A useful report should include:
- Membrane composition and phase purity.
- Relative density, thickness, and active area.
- Electrode composition, loading, and microstructure.
- Operating temperature and thermal history.
- Feed and permeate gas composition.
- Hydrogen partial pressure and total pressure on both sides.
- Gas flow rate and flow-meter calibration.
- Applied voltage, current, and measurement duration.
- Hydrogen production rate and analytical calibration.
- Faradaic efficiency and uncertainty.
- Leak-test results and post-test characterization.
When comparing different membranes, normalize hydrogen production by active area and state whether the reported value is measured at the same temperature, hydrogen partial-pressure difference, and applied voltage. A flux measured at a larger driving force should not be presented as an intrinsic material improvement.
11. Common experimental errors
- Treating total ceramic conductivity as proton conductivity.
- Assuming a proton conductor can support spontaneous steady hydrogen permeation without an electronic pathway.
- Confusing an externally biased hydrogen pump with open-circuit MPEC permeation.
- Calculating Faradaic efficiency from current without directly measuring hydrogen production.
- Ignoring gas leakage and crossover.
- Reporting hydrogen flux without active membrane area or gas conditions.
- Comparing data collected at different hydrogen partial-pressure ratios as if the driving force were identical.
- Attributing all voltage losses to the electrolyte without impedance or electrode controls.
- Assuming a stable current means the permeate gas composition has reached steady state.
- Ignoring the effects of water vapour and oxygen chemical potential on electronic leakage.
12. Research opportunities
Important opportunities include: - Proton conductors with higher conductivity and improved tolerance to CO\(_2\) and steam. - Electrode catalysts with faster hydrogen oxidation and evolution kinetics. - Thin, gas-tight membranes with reliable high-temperature seals. - Architectures that reduce concentration polarization and improve gas utilization. - Mixed proton–electron conductors with balanced transport properties. - Operando measurements that separate bulk transport, surface exchange, and leakage. - Long-duration demonstrations using realistic dilute hydrogen feeds.
A practical hydrogen pump must be evaluated as a complete electrochemical system. High proton conductivity is necessary, but it is not sufficient without efficient electrodes, reliable seals, low electronic leakage, and verified hydrogen recovery.
13. Key takeaways
- Electrochemical hydrogen pumps transport hydrogen through a proton-conducting ceramic using an external electronic circuit.
- Spontaneous hydrogen permeation requires a membrane that supports both protonic and electronic transport.
- Hydrogen production is linked to charge by Faraday’s law, but actual performance must be verified by direct gas analysis.
- Membrane thickness, electrode kinetics, electronic leakage, gas transport, and sealing all affect performance.
- Faradaic efficiency and hydrogen flux should be reported together with operating conditions and uncertainty.
- Long-term stability and gas purity are essential metrics alongside initial current density.
References
- S. Cheng, X. Li, X. Huang, Y. Ling, S. Liu, and T. Li, “Hydrogen separation via proton conducting ceramic membranes: A review,” International Journal of Hydrogen Energy 70 (2024), 654–665. https://doi.org/10.1016/j.ijhydene.2024.05.181
- Z. Tao, L. Yan, J. Qiao, B. Wang, L. Zhang, and J. Zhang, “A review of advanced proton-conducting materials for hydrogen separation,” Progress in Materials Science 74 (2015), 1–50. https://doi.org/10.1016/j.pmatsci.2015.04.002
- “Mixed protonic–electronic conducting ceramic membranes for hydrogen separation: A review on recent advances in structure optimization and preparation,” Separation and Purification Technology 355 (2025), 129582. https://doi.org/10.1016/j.seppur.2024.129582
- J. Tong, P. Zhang, F. Zhuang, Y. Zheng, B. Liu, X. Qiao, and X. Zhu, “Mixed-conducting ceramic membrane reactors for hydrogen production,” Reaction Chemistry & Engineering 9 (2024), 3072–3099. https://doi.org/10.1039/D4RE00372A
- M. G. Buonomenna, “Proton-Conducting Ceramic Membranes for the Production of Hydrogen via Decarbonized Heat: Overview and Prospects,” Hydrogen 4 (2023), 807–830. https://doi.org/10.3390/hydrogen4040050