Research
To become fossil fuel-free, society needs to increase the use of renewable energy. This requires hydrogen as an energy carrier and large energy storage systems. Urgently, performance and cost of the related devices (e.g. fuel cells, electrolyzers, flow batteries) need to be improved.
To this end, we develop ion conducting polymers for use as membranes or electrode binder in electrochemical devices.
The activities range from polymer synthesis over membrane formation to the testing of membranes in the application.
1. Alkaline Stability of AEMs (Anion Exchange Membranes) - a problem which can be solved
Anion exchange membrane water electrolysis (AEMWE) combines the advantages of alkaline water electrolysis (cheap non-noble metal catalysts) with those of PEM water electrolysis (low gas crossover, excellent performance).
However, widespread use of AEMWE is impeded by the low alkaline stability of the separators, quaternary ammonium-based anion exchange membranes.
Ion solvating membranes (ISM) are discussed as separator in highly alkaline water electrolysis systems, but so far showed extremely low conductivity in the KOH concentration used in AEMWE, e.g. ≤ 1M KOH.
We found that an ion solvating membrane based on sulfonated para-PBI has a conductivity in 1M KOH which exceeds that of most AEMs.
The absence of quaternary ammonium groups boost the alkaline stability: In a 6 month stability test (1M KOH, 80 °C), conductivity and mechanical properties hardly changed.
This potentially solves the stability issues of AEMWE, and we expect that "AEMWE" with an ISM separator will have high market potential, and will strongly promote production of green affordable hydrogen.
Advanced Energy Materials (2023), https://doi.org/10.1002/aenm.202302966
2. Flow Batteries - VRFB with a world-class energy efficiency of >90% at 80 mA/cm2 achieved
■ An overview over flow battery technology can be found here: www.dx.doi.org/10.1115/1.4037248
■ Polybenzimidazole membranes very effectively block transport of vanadium ions. As expected, the voltage efficiency increases, when the thickness is reduced from 35 to 15 µm. Most interestingly, the coulomb efficiency remains similar or even slightly increases as well. This seems to be related to the lower charging voltage for thin membranes, which should reduce side reactions and reduces the number of vanadium ions which enter the membrane by migration in the electric field.
http://pubs.acs.org/doi/abs/10.1021/acsami.7b10598
■ Polybenzimidazole (PBI) can be produced as a highly conductive gel and as a highly selective dense film. We combined these two forms of PBI into a membrane assembly (gel/film/gel), in which the gel PBI protects the thin PBI film. EE reached 90.5% at 100 mA cm-2.
https://onlinelibrary.wiley.com/doi/abs/10.1002/smll.202206284
■ The Figure below compares different membrane types developed by us.
Membrane type 1: PBI pre-swollen in 10 M sulfuric acid (Chem. Eng. J. 435 (2022) 134902)
Membrane type 2: Stack of 3 membranes: soft protective gel PBI, selective dense PBI, soft protective gel PBI (Small 18 (2022) 2206284)
Membrane type 3: sulfonated polystyrene coated with selective PBI layer (Adv. Energy Mater. 2024)
Membrane type 4: unpublished results
3. Alkaline water electrolysis: World-level 1.8 A cm-2 at 1.8 V achieved
■ This performance appears to be the highest reported for alkaline water electrolysis in >25wt% KOH feed solution. It was made possible by casting a gel PBI membrane and reinforcing it with a porous Teflon support. While PBI membranes fail within 300 hours, our membrane was tested successfully for 1000 hours by project partner DLR.
https://pubs.rsc.org/en/content/articlehtml/2022/ee/d2ee01922a
4. Water electrolysis in low alkaline solutions using Anion exchange membranes
■ We developed a simple method for testing the true-hydroxide conductivity in water, which can be easily adopted by most labs. This solves the issue of CO2 absorption from air, which results in erroneously low conductivity values.
https://www.mdpi.com/2077-0375/12/10/989
■ Industrial scale alkaline water electrolysers can have an active area of up to 3 m2. Looking ahead, large scale AEM water electrolysers may reach similar dimensions, and low dimensional changes during cell assembly and start of electrolyser operation will become essential. We suggest to pre-swell the membranes with a high-boiling solvent like ethylene glycol to minimize dimensional changes.
https://www.sciencedirect.com/science/article/pii/S0376738822010894
■ Reinforcing the membranes may be necessary to guarantee high mechanical strength and low swelling. The different swelling of porous supports and ion conducting matrices results in delamination and increased gas crossover. We approach this by engineering a strong interphase and covalently connecting a PBI nanofiber mat (support) with the AEM matrix.
https://www.sciencedirect.com/science/article/pii/S0376738821007766
5. HT PEMFC - 2000 hours stable operation at 800 mA/cm2 achieved
HT PEMFC membranes are usually based on phosphoric acid (PA) doped polybenzimidazole. We develop new crosslinking methods (avoding potentially unstable C-N linkages), but are also interested in other chemistries, and develop membranes with pyridine or tetrazole groups in the side chain. Because of the low basicity of the tetrazole group, it cannot be easily doped with PA and is not attractive for HT PEMFC. However, we found a way to increase the basicity of tetrazole groups. In detail, we found that 5-(2,6-dioxyphenyl) tetrazole is forced into a coplanar arrangement, allowing resonance stabilization of the positive charge in protonated tetrazolium ions.
Tetrazole: http://pubs.rsc.org/is/content/articlehtml/2015/ta/c5ta01936b
Pyridine: http://www.sciencedirect.com/science/article/pii/S0013468616326342
Crosslinking via aromatic sulfone bonds:
http://www.sciencedirect.com/science/article/pii/S0376738817326273
http://pubs.rsc.org/-/content/articlehtml/2017/ta/c6ta07653j
Recently, we achieved a stable performance at very challenging 800 mA/cm2 for over 2000 hours.
Performance of our membrane in the HT PEM Fuel Cell
6. Membranes for Actuators - exceptionally fast response and large mechanical power density
Ionic polymer actuators are materials which move when a potential is applied. The resulting force can be used to propel underwater vehicles or to move endoscopic surgery tools. By blending a sulfonated polymer with a sulfonated phthalocyanine, we prepared membranes which showed the fastest response to the applied voltage and the largest mechanical power density reported so far. This work was in collaboration with the group of Dr. Chongmin Koo.
7. Make membranes resist the mechanical stresses they experience in the applications
Continuous swelling and shrinking of the membranes due to changes in temperature and/or humidity lead to cracks in the electrode and membrane, membrane-electrode delamination, and formation of cracks along the edges of the active area.
We developed two technologies which help to stabilize the membranes:
a) Membranes with a shape memory effect (SME)
By drying membranes in a fixed geometry (e.g. clamped on a board, or held in shape by rolls), the membranes can only shrink in the thickness. When humidified again, the membranes mainly swell in the thickness, regaining their original shape. In the fuel cell the membranes are clamped between the bipolar plates, and the SME is re-established in every drying cycle. Therefore, the membranes always keep a smooth surface, no wrinkles and folds are observed.
http://onlinelibrary.wiley.com/doi/10.1002/macp.201500063/full
http://pubs.rsc.org/is/content/articlehtml/2014/ta/c4ta01467g
b) Selective strengthening of the most sensitive membrane areas
Gas inlet/outlet areas and the edge of the active area experience most mechanical stress. With a metal stamp these areas can be selectively heated to temperatures at which the membranes start to crosslink and/or lose ion conducting groups. The process is fast (just a few minutes), and leaves the active area intact.
By using this process, high IEC membranes (which easily fail in the fuel cell due to excessive swelling) can be used without failure, while untreated membranes brake along the edge of the active area.
http://www.sciencedirect.com/science/article/pii/S0376738813009733