Have you ever attempted to replicate findings from the energy storage literature? Have you ever tested an electrochemical cell and struggled to achieve the same results? You are not alone.
In the last two years, my colleagues and I at Queen’s University Belfast (QUB), UK, along with researchers from the Massachusetts Institute of Technology (MIT), US, have fostered a community of flow battery researchers who have consistently encountered this specific issue. Each of them is eager to contribute to a solution.
It all began at the UK Flow Battery Network Annual Symposium in 2024, where an international assembly of researchers attended a presentation by MIT’s [Fikile Brushett](https://cheme.mit.edu/profile/fikile-r-brushett/). Speaking openly, he shared the routine hurdles his labs faced in correlating performance metrics from peer-reviewed articles. His forthrightness, coupled with his genuine interest in the experiences of attendees, sparked an engaging conversation.
This dialogue underscored a collective understanding: advancing flow battery technology necessitates enhanced collaboration throughout the research community. To boost its contribution to renewable energy storage and to make the field more inviting for new researchers, those in attendance reached a consensus on the necessity of establishing more effective, consistent, and accessible testing methodologies.
Among those in attendance was [Hugh O’Connor](https://pure.qub.ac.uk/en/persons/hugh-oconnor-2/), who was a PhD student at the time and is now my colleague. During his PhD, O’Connor created a low-cost, 3D-printable flow battery cell and personally experienced the challenges of achieving reproducible experimental outcomes. Capitalizing on the momentum from the symposium, QUB and MIT collaborated with researchers from Queen Mary University of London, Eindhoven University of Technology, Harvard University, the University of Cambridge, and the University of Manchester to conduct an interlaboratory study. Utilizing O’Connor’s cell design as a mutual foundation and a methods-centered experimental protocol, the study aimed to investigate how much variability exists in widely used electrochemical testing techniques across different laboratories.
We supplied all necessary cell components, along with an assembly manual, and asked for three experiments: charge–discharge cycling, polarization curve analysis, and electrochemical impedance spectroscopy. Parameters were specified where feasible, while some were left to the discretion of the participants. These were gathered, along with the data, in a post-experiment questionnaire for analysis.
Expectations vs. Reality
We anticipated that all participants would conduct the same tests, with some variation likely due to differences in auxiliary equipment, such as tanks or pumps. However, we discovered that particularly regarding polarization testing, the methodologies employed by different groups varied significantly. In fact, each of the eight research teams followed a slightly different protocol. This revealed that a definitive ‘best practice’ protocol does not exist within the flow battery community.
Considering that our experimental directive matched the level of detail found in a standard flow battery publication, caution is warranted when comparing polarization curves from literature, unless the specific protocol has been adequately reported. Indeed, we advocate for a standardized methodology for polarization curve analysis in flow batteries, similar to those established for [fuel cells](https://onlinelibrary.wiley.com/doi/book/10.1002/9781119191766) and electrolysers.
In the case of charge–discharge cycling, researchers mainly executed their experiments in a similar fashion. Nonetheless, the study revealed considerable variability in key metrics, such as electrolyte utilization (±9.2%) and capacity decay (±2.5%/day). The impedance data also exhibited a wide range of values for area-specific resistances related to ohmic, charge transport, and mass transport losses, even while employing the same equivalent circuit model. Our findings illustrate that identical chemistry and identical cells do not guarantee replicable performance; other sources of error are evidently affecting the results.
Despite gathering data from eight distinct research groups, we couldn’t definitively link the experimental variations among participants to the quantifiable changes in their data. Consequently, we are now embarking on even larger interlaboratory studies, involving nearly 40 researchers worldwide. Nevertheless, metadata collected from our post-experiment surveys suggests potential areas for error reduction, including electrical connections and electrolyte tank design. This prompted the two leading institutions to investigate the effects of two- and four-point probe connections and cable length, as well as the impact of tube positioning and stirring within the vessels containing the electrolyte. We demonstrated that a significant portion of the variability in ohmic losses in the system may originate from inadequate electrical connections, and non-uniform electrolyte utilization may be influenced by tubing arrangements that promote fluid bypass, especially when the electrolyte is unstirred.
So, what steps can you take to help alleviate this issue in your own laboratory? Here are five recommendations: