Chemistry Under Pressure: How Publish or Perish Is Compromising the Quality of Materials and Catalysis Research

In May 2025, ACS Catalysis retracted a paper on photoelectrochemical synthesis that had been published just six months earlier. The retraction notice stated that NMR spectra reported in the Supporting Information and raw data files had been edited to obscure impurities in analysed samples. A correction had been published first, but following the correction, additional concerns were raised about post-collection editing of spectra — and since the compounds whose spectral data were altered were central to the conclusions of the work, the article was retracted in full.

This is not an isolated case. It is a representative one. Across chemistry, materials science, and catalysis, the pattern of NMR manipulation, figure duplication, data fabrication, and result misrepresentation has become sufficiently common that journals in these fields have begun implementing post-publication image integrity screening as a standard editorial step — not as an exceptional response to specific concerns, but as a routine part of the publication process.

The question worth asking is not why individual researchers make these choices. It is why the system produces them so reliably, and what the consequences are for anyone whose work depends on the reliability of the published chemistry literature.

The Scale of the Problem in Chemistry and Materials Science

The annual count of retractions has surged dramatically in recent years, with 2023 setting a record of over 14,000 retraction notices issued, nearly 12,000 articles published in 2022 subsequently retracted, and 2024 witnessing more than 9,000 additional retractions. As of August 2025, the total had already surpassed 5,000 for the year.

Chemistry and materials science sit at the centre of this wave rather than at its periphery. An analysis of chemistry manuscripts retracted between 2001 and 2021 found that 58.5% were retracted due to misconduct — of which 40.5% were attributed to self-plagiarism and 36% to fraud — while errors and concerns unrelated to misconduct constituted 26% of all retractions. Retracted manuscripts had a median retraction time of 1.7 years and a peer-review time of 71 days, but fraudulent manuscripts were peer-reviewed in only 43 days on average.

That last figure deserves attention. The papers most likely to contain fabricated data were the ones that moved through peer review fastest — suggesting either that their apparent coherence made them seem uncontroversial, or that the review was cursory precisely because the results seemed clean. Both interpretations are consistent with a system under pressure to process volume rapidly.

A 2026 bibliometric study specifically examining materials science research integrity confirmed an ongoing pattern of retracted articles concentrated in particular research areas, with data problems and image integrity issues as the leading causes. The study was cited alongside new work examining whether ISO/IEC 17025 laboratory accreditation standards — designed for industrial analytical laboratories — might offer a framework for preventing scientific fraud in chemical research environments, which gives some sense of how seriously the community is now taking the integrity question.

What Chemistry's Retraction Patterns Look Like in Practice

The specific mechanics of research misconduct in chemistry and materials science differ in important ways from those in biomedical research, and understanding those differences matters for anyone reading or building on this literature.

1. NMR data manipulation

   ACS Catalysis Retracted May 2025 Photoelectrochemical synthesis

   NMR spectroscopy is the primary method by which chemists confirm the identity and purity of newly synthesised compounds — it is, in essence, the fingerprint of a molecule. When NMR spectra are edited to obscure impurities, the compounds whose spectral data are altered are typically central to the paper's conclusions. The whole point of the manipulation is to make an impure or incorrectly characterised compound appear to be something it is not. A catalyst reported to achieve a particular yield or selectivity, characterised by manipulated NMR data, is not the catalyst that was used in the reaction. The reported results are therefore not the results that were obtained. Everything built on that foundation — subsequent synthesis work, industrial interest, investment in scale-up — is built on something that did not happen.

2. Figure duplication and image manipulation

   Nanomaterials retraction 2025 Carbon nanotube electrocatalysts

   In materials science, figure duplication typically involves electron microscopy images, XRD patterns, or characterisation spectra being reused across papers with different reported compositions or conditions. A 2025 retraction of a nanomaterials paper on carbon nanotube electrocatalysts identified inappropriate editing and duplication between figures across two papers by the same corresponding author, published simultaneously under evaluation at the same time — a pattern consistent with paper mill production methods applied to materials science outputs rather than biomedical ones.

3. Self-plagiarism and duplicate submission

   Frontiers in Chemistry Retracted 2024 Most common misconduct in chemistry

   Submitting substantially the same work to multiple journals simultaneously or sequentially is the most common form of misconduct in chemistry overall. A 2024 retraction in Frontiers in Chemistry identified an unacceptable level of similarity between a published article and a Green Chemistry paper by the same authors, where the Green Chemistry paper had been published while the Frontiers submission was still under review. The pressure to accumulate publication counts produces exactly this behaviour: the same data appearing in multiple venues, each counted as a separate publication for career evaluation purposes.

4. Cross-coupling catalysis and contamination fraud

   Nature Catalysis Retracted after 324 days Suzuki–Miyaura, palladium contamination

   Cross-coupling catalysis — Suzuki, Heck, and related reactions — has become a particular locus of reproducibility problems, partly because it is a high-activity field with intense competition, and partly because palladium contamination from reagents can produce apparent catalytic activity in reactions that were not designed to be palladium-catalysed. A paper reporting palladium-free polymer synthesis using an amine-catalysed Suzuki–Miyaura coupling was later shown to be catalysed by residual palladium impurity in the amine reagent. Both the original Nature Catalysis paper and a follow-up Chemical Science paper were retracted, with retraction times of 324 and 51 days. The first had been published in one of the most prestigious journals in the field and cited extensively before the error was identified.

Why Publish or Perish Hits Chemistry Particularly Hard

The publish-or-perish dynamic affects all research fields, but it interacts with the specific culture and economy of chemistry in ways that make the field particularly vulnerable to the integrity problems now visible in its retraction record.

Chemistry is a field where experimental results are, in principle, highly verifiable — the compounds either exist and have the properties claimed, or they do not. NMR spectra, crystallographic data, and reaction yields are concrete, measurable, reproducible claims. This verifiability should make chemistry resistant to fraud. In practice, it makes fraud particularly damaging when it occurs, because the downstream consequences of building on a false chemical claim are concrete and often expensive. A pharmaceutical company that pursues a synthetic route based on a fabricated catalyst result does not merely waste publication budget — it wastes laboratory time, reagent costs, and potentially years of development work.

The publish-or-perish culture grew seriously after World War II as universities began judging faculty by their academic publishing record, and journals proliferated as a result — including irresponsible ones open to paper mills. In chemistry specifically, this proliferation created a tiered landscape of journals with widely varying standards, where the same data could be published in a low-impact journal with minimal peer review almost as easily as it could be submitted to a high-impact one with stringent editorial processes. The incentive to accumulate publications — in any venue, at any tier — has consistently outrun the incentive to produce fewer, more reproducible results.

The Investor and Commercialisation Dimension

For a materials science or catalysis startup — a company building on published chemistry to develop a catalyst, a functional material, a battery electrode, or a synthetic process — the retraction crisis in these fields is not an academic concern. It is a due diligence problem.

A startup whose core technology was developed in a research group that has subsequently retracted related publications is not necessarily working on flawed science. But it faces a credibility challenge that investors will identify and that regulatory bodies will scrutinise. The connection between a retracted paper and a commercial claim built on adjacent work is not always straightforward, but the association is damaging — and in a funding environment where investors are conducting more rigorous technical due diligence on early-stage deep-tech ventures than at any previous point, the provenance of the underlying science is under genuine scrutiny.

More practically: a company whose technology depends on reproducing a published synthetic route or catalytic process that was itself fabricated or contaminated faces the possibility of discovering this at an advanced and expensive stage of development. Independent replication of the foundational chemistry — conducted before significant investment is committed, not after — is the most straightforward mitigation. This is not standard practice in the industry. It should be.

What Researchers Can Do

The systemic pressures producing chemistry's retraction crisis will not be resolved by individual behaviour changes alone. But there are specific practices that researchers in materials science and catalysis can adopt to protect both the integrity of their work and the credibility of the commercial claims built on it.

The Bottom Line

The chemistry literature is, in aggregate, one of the most reliable bodies of knowledge in science — the overwhelming majority of published results are what they claim to be. But the fraction that is not is growing, concentrated in specific high-pressure subfields, and consequential in proportion to how much commercial and research activity is built on it.

In materials science and catalysis, where the gap between a published result and an investable technology is shorter than in almost any other field, that fraction matters more than the headline retraction rates might suggest. The integrity of the literature is not an abstract concern. It is the foundation on which commercial chemistry stands — and when the foundation is compromised, the cost falls not on the authors of the retracted paper, but on everyone who built on it before the retraction notice arrived.