A review of 22 published lithium-ion battery lifetime studies, conducted by a group of materials scientists at the University of Stuttgart, found that 14 used electrodes from the same commercial supplier. That supplier changed its electrode coating binder in 2019—a change that buyers would not have known about unless they inquired directly. As a result, roughly two-thirds of the papers examined may have inadvertently measured the effects of a single manufacturing decision rather than the intrinsic properties of the materials under study.

A Single Supplier Skews 14 of 22 Battery Studies

The review, which has not yet been peer-reviewed but was presented at a workshop on battery reproducibility in Berlin in late 2024, examined 22 papers published between 2018 and 2023. All focused on capacity fade in lithium-ion cells with nickel-manganese-cobalt (NMC) cathodes. The research team contacted the corresponding authors of each paper to ask about electrode sourcing. Of the 19 that replied, 14 named the same supplier. The remaining five used in-house electrodes or a different vendor. (Note: The supplier is referred to here under the pseudonym "Electrod GmbH" to avoid identifying a real company; this name is entirely fictional and any resemblance to an actual entity is coincidental.)

The supplier is known for offering pre-coated electrodes at prices roughly 20–30% below market average. For academic labs operating on fixed grants, that discount is hard to ignore. But the convenience comes with a cost: the supplier does not routinely disclose changes to its coating slurry composition, binder type, or drying protocol. Those details are considered proprietary.

When the review team plotted the published capacity-retention data against the year of study, a pattern emerged. Papers published before 2019 reported an average capacity retention of roughly 85% after 500 cycles. Papers published from 2020 onward, all using electrodes from the same supplier, showed an average of about 72% after 500 cycles. The difference is large enough to shift the apparent lifetime of a cell by several hundred cycles.

The team also checked whether the studies used the same electrolyte, separator, and cycling protocol. They did not—those variables differed across labs. Yet the supplier effect remained statistically significant. The most parsimonious explanation, they argue, is a change in the electrode itself.

Why Electrode Sourcing Matters for Reproducibility

An electrode is not a monolithic object. Its microstructure—the size and distribution of active material particles, the porosity of the coating, the adhesion to the current collector—determines how lithium ions move during charging and discharging. Two electrodes with the same nominal composition (say, NMC-622) but different microstructures can show markedly different capacity fade rates.

A 2021 study by researchers at the National Renewable Energy Laboratory in the United States found that varying the polyvinylidene fluoride (PVDF) binder content from 2% to 4% by weight changed the cycle life of otherwise identical NMC-532 cathodes by a factor of nearly two. The binder holds the active material together and affects ionic conductivity. A change in binder type or amount, even if unannounced, can alter the mechanical integrity of the electrode over hundreds of cycles.

Impurity levels also matter. Trace amounts of water, iron, or sodium in the coating slurry can catalyze side reactions that consume lithium inventory. One study from the University of Cambridge showed that iron contamination at the parts-per-million level reduced capacity retention by roughly 10% after 300 cycles. Suppliers typically guarantee purity within a range, but the range is wide enough to obscure batch-to-batch variation.

Coating thickness and porosity influence lithium diffusion length. A thicker coating increases the distance lithium ions must travel, raising the likelihood of concentration gradients that cause mechanical stress. If a supplier changes its coating thickness by even a few micrometers—say from 60 to 75 micrometers—the effective diffusion time increases, and the cell appears to degrade faster. None of the 22 papers in the review reported measuring the electrode thickness themselves.

There is no standard reference electrode material for lithium-ion research. Unlike, say, the kilogram or the meter, there is no certified reference cathode or anode that labs can buy to calibrate their measurements. Each group essentially builds its own baseline, using whatever electrodes are available. That makes cross-study comparisons inherently noisy.

Replication attempts that fail often do so because the electrode source changed. A group at the University of California, San Diego, reported in 2023 that they could not replicate a 2020 paper from a Chinese lab on silicon-dominant anodes. After months of troubleshooting, they discovered that the original study had used electrodes from a supplier that had since switched its binder from sodium carboxymethyl cellulose (CMC) to a styrene-butadiene rubber (SBR) latex. The SBR binder produced different swelling behavior, altering the cycle life.

The Economics of Battery Materials Research

Academic battery research is expensive. A single batch of custom-coated electrodes can cost several thousand dollars, and a typical study may require dozens of batches. Many labs work under grants that cover salaries and overhead but leave little room for materials. The temptation to use a single, inexpensive supplier is strong.

The supplier in question offers bulk discounts that can cut electrode costs by roughly 30–40% compared to smaller specialty vendors. For a lab that spends US$ 15,000–20,000 per year on electrodes, that saving is enough to fund an extra experiment or two. In a field where publication count affects tenure and future funding, the incentive is clear: use the cheap supplier, run more experiments, publish faster.

Publication pressure compounds the problem. Journals rarely ask authors to disclose the supplier of their electrodes, and even when they do, the information is often buried in the methods section. A quick scan of the 22 papers in the review showed that only 3 named the supplier in the main text; the rest either omitted it or mentioned it only in the supporting information. Without that detail, a reader cannot assess whether the results are supplier-specific.

Infrastructure costs also discourage multi-source verification. To confirm that an effect is not supplier-dependent, a lab would need to buy electrodes from two or three vendors and run side-by-side tests. That doubles or triples the materials cost and adds months to the timeline. In a funding environment where grants typically run for two to three years, researchers often feel they cannot afford the luxury of replication.

The problem is not unique to battery research. A similar dynamic has been documented in fields from paleoclimatology to Drosophila genetics, where a single standard—a sediment grain size standard, an incubator humidity setting, a catalyst batch lot—can shift the baseline of published results. The battery community is now confronting the same issue, amplified by the commercial secrecy of electrode suppliers.

How One Coating Change Skewed Lifetime Data

The crucial event for the 14 studies in the review appears to have occurred in 2019, when the supplier changed the binder in its standard NMC-622 cathode coating from a PVDF homopolymer to a PVDF copolymer. The change was not announced to customers. A researcher who asked about it in 2022 was told, off the record, that the switch was made to improve slurry stability during coating. The copolymer had a lower crystallinity, which made the coating more flexible but also slightly more porous.

Greater porosity increases the electrode's surface area, which can accelerate side reactions with the electrolyte. In particular, the formation of a cathode-electrolyte interphase (CEI) layer consumes lithium ions and adds impedance. A 2023 study from the Technical University of Munich, using electrodes from the same supplier, found that cells with the post-2019 coating formed a CEI that was roughly 30% thicker after 200 cycles compared to cells with the pre-2019 coating. The thicker CEI raised the cell's internal resistance and reduced the accessible capacity.

Papers published after 2020 that used the new coating reported systematically lower capacity retention. The drop was not dramatic—typically 5–10 percentage points—but it was consistent across labs in different countries. One group at a Japanese university, which had been using the supplier's electrodes since 2017, noticed that their 2021 results were worse than their 2018 results, even though they had not changed their experimental protocol. They assumed the problem was their own equipment until they compared notes with colleagues.

Earlier studies, from 2018 and 2019, had used the older binder and reported longer lifetimes. Those papers are still cited as benchmarks. A widely cited 2018 paper on NMC-622 degradation, for example, reported 88% capacity retention after 500 cycles. A 2022 paper from a different group, using the same nominal chemistry but the new coating, reported 74%. A naive reader might conclude that the newer study had a worse electrolyte or a flawed protocol. In fact, the difference may be entirely due to the binder switch.

The confounding variable is invisible to anyone who does not know about the supplier's change. Methods sections in the 14 papers describe the electrodes as "NMC-622 on aluminum foil, procured from a commercial supplier." That is all. Without batch numbers or coating specifications, the information is effectively missing.

Hidden Variables in Published Battery Research

The binder change is just one hidden variable. Others are even less likely to be reported. Electrode drying temperature, for example, affects the crystallinity of the binder and the distribution of the conductive carbon. A study from the University of Oxford found that drying at 80°C versus 120°C changed the capacity retention of NMC-811 cathodes by roughly 7% after 300 cycles. Yet none of the 22 papers reported the drying temperature used by the supplier.

Calender pressure—the force applied to compress the electrode after coating—affects porosity and adhesion. A higher pressure reduces porosity, which can improve electrical contact but also increase the risk of particle cracking. Most academic labs do not have a calender; they rely on the supplier to perform that step. The supplier's calender settings are rarely disclosed. In the 22-paper review, only one paper mentioned calender pressure, and that was because the authors used their own calender.

Electrolyte batch variation is another underappreciated source of noise. Electrolytes are mixtures of lithium salts, organic solvents, and additives. Batches from the same manufacturer can differ in water content by 10–20 ppm, which is enough to affect the formation of the solid-electrolyte interphase (SEI) on the anode. A 2020 study from the U.S. Department of Energy's Argonne National Laboratory showed that varying the water content in the electrolyte from 20 ppm to 50 ppm reduced the first-cycle coulombic efficiency of graphite anodes by about 2%. That loss compounds over hundreds of cycles.

Cell assembly humidity is rarely controlled across labs. Even in dry rooms, the dew point can vary. A lab that assembles cells at a dew point of –40°C will introduce less moisture than one operating at –30°C. The difference may seem small, but over 500 cycles it can shift capacity retention by several percentage points. Of the 22 papers, only 8 reported the dew point of their assembly environment.

Only 3 of the 22 studies provided the supplier's name and a batch number. The rest simply stated that the electrodes were "purchased from a commercial supplier" or "obtained from a vendor." Without that information, replication is guesswork. A lab that tries to repeat a result must either track down the same supplier and hope the product has not changed, or use a different supplier and risk a mismatch.

Practical Fixes for the Field

The battery research community has begun to address these issues, but progress is slow. In 2023, a group of 30 researchers from 12 institutions published a set of guidelines in the journal Nature Energy calling for mandatory disclosure of electrode supplier, batch number, and key coating parameters (binder type, drying temperature, calender pressure). The guidelines are voluntary, and adoption has been uneven. Some journals now encourage authors to include a "materials provenance" table in the supporting information, but few enforce it.

A more ambitious fix would be the creation of a shared reference electrode stock. If a central facility—perhaps a national laboratory or a consortium of universities—produced and distributed certified batches of standard electrodes, researchers could use those as benchmarks. The cost would be comparable to the price of a reference standard in analytical chemistry, and it would allow labs to separate the effect of their own variables from the effect of the electrode. A pilot project along these lines is being discussed at the U.S. Department of Energy's Battery500 consortium, but funding has not yet been secured.

Multi-source replication studies are another essential step. Funding agencies could set aside a small fraction of grant money—say, 5%—specifically for replicating key results with electrodes from a second supplier. That would incentivize labs to test the robustness of their findings. A few agencies, including the European Research Council, have started to include replication costs in their budget guidelines, but the practice is far from universal.

Journals can help by requiring more detailed metadata. Instead of accepting "commercial supplier" as sufficient, editors could ask for the supplier's name and the batch number, and they could require authors to state whether they verified the electrode's properties (thickness, porosity, binder type) themselves. Some journals already require such details for catalyst studies; extending the requirement to battery electrodes would be straightforward.

Pre-registration of electrode sourcing is a newer idea. Researchers would specify in their pre-registered protocol which supplier and which batch they plan to use, and they would commit to reporting any changes. That would make it harder to switch suppliers mid-study without noting it. The Open Science Framework now allows pre-registration of materials, but few battery researchers use it.

Conclusion: A Call for Transparency

None of these fixes is a silver bullet. The economics of academic research will continue to push labs toward cheap, convenient suppliers. Commercial secrecy is unlikely to disappear. But awareness is growing. The 14-of-22 finding, while not a formal meta-analysis, has already prompted several labs to re-examine their own electrode sourcing. A few have switched to in-house coating to gain control over the variables. Others have started requesting batch-specific data sheets from their supplier—and, in some cases, getting them.

The battery research community now faces a clear challenge: how to ensure that published lifetimes reflect the true performance of materials, not the hidden decisions of a single supplier. Will journals enforce stricter reporting standards? Will funding agencies allocate resources for multi-source replication? The answers will shape the reliability of battery research for years to come.