Is Teflon Really Inert? What PTFE Particles Do to Gut Cells

Polytetrafluoroethylene — PTFE, better known by the brand name Teflon — has been sold on a single promise: it is inert. It resists heat, repels almost everything, and, we’ve been told, passes through the body untouched if a flake ever ends up in your food. A 2025 study in the Journal of Hazardous Materials puts that assumption under a microscope, literally, and finds it wanting. [Abass et al., 2025]

Researchers exposed human intestinal cell models to PTFE micro- and nanoplastics and watched the particles slip inside the cells, press against nuclear membranes, and lodge inside mitochondria — all while the cells kept living. [Abass et al., 2025] The particles didn’t kill the cells. They quietly damaged them.

That distinction — harm without death — is the whole story here, and it’s why “chemically inert” is no longer a satisfying answer.

Why PTFE ended up on the research bench

Non-stick coatings shed. Recent work cited by the study’s authors shows PTFE-coated cookware can release thousands to millions of particles during routine use — brief cooking sessions or minor scratches are enough — and many of those particles transfer to food or linger on the pan’s surface, creating a chronic ingestion risk. [Abass et al., 2025] Despite that, PTFE has received little toxicological attention compared with more-studied plastics like polystyrene. [Abass et al., 2025]

The gut is the logical place to look first. Ingestion is the primary route by which humans take in microplastics, and PTFE particles have already been detected in human organs and bodily fluids alongside other microplastics. [Abass et al., 2025] The intestine is where swallowed particles make first contact with living tissue.

How the study was built

To make the findings physiologically relevant, the team used two human intestinal cell setups. The first was simple undifferentiated monocultures of Caco-2 and HT29-MTX cells. The second was a differentiated Caco-2/HT29-MTX co-culture grown for 21 days into a functioning intestinal barrier — complete with polarized cells, tight junctions, and a mucus-secreting layer that mimics the real gut lining. [Abass et al., 2025]

They tested two sizes of PTFE particles: a nanoscale type averaging about 250 nm and a microscale type averaging roughly 2 µm. Cells were exposed at 50, 100, and 200 µg/mL for 24 and 48 hours. [Abass et al., 2025] That two-size, two-timepoint, multi-dose design is what let the researchers separate the effects of size, dose, and duration.

The particles get in — and the small ones get everywhere

Imaging confirmed that both particle sizes were internalized, but not equally. At 200 µg/mL, the nanoscale particles showed up throughout the cell interior, clustering around the nucleus, pressing against the nuclear membrane, and in some cases sitting inside mitochondria. [Abass et al., 2025] The microscale particles got in too, but to a lesser extent, staying mostly near the nucleus or loose in the cytoplasm. [Abass et al., 2025]

This is a clean size-dependent uptake pattern: the smaller the particle, the deeper it penetrates. Nanoscale plastics have a large surface-area-to-volume ratio, which makes them both easier to internalize and more reactive once inside. [Abass et al., 2025]

Damage without death

The most striking result is what didn’t happen. Across every concentration and both exposure times, in both the monocultures and the barrier model, PTFE particles did not significantly reduce cell viability. [Abass et al., 2025]

0

A standard toxicity screen that only checks whether cells are alive would have waved these particles through. The problem is that the cells were alive and damaged. Under that surface of normal viability, the researchers documented a cascade of subtler harm.

Mitochondrial damage. Electron microscopy revealed mitochondria that were swollen, with disrupted cristae and damaged membranes — the structural signature of dysfunction. [Abass et al., 2025] Measurements of mitochondrial membrane potential showed significant depolarization after 48 hours at the highest dose. In the barrier model, only the nanoscale particles produced that effect. [Abass et al., 2025] The team even captured images of nanoparticles sitting inside mitochondria, forming large central vacuoles. [Abass et al., 2025]

Oxidative stress. Reactive oxygen species rose in a concentration- and time-dependent way, most clearly at 48 hours, for both particle sizes. [Abass et al., 2025] Oxidative stress is one of the most common responses to nanosized materials and a frequent starting point for downstream cellular harm. [Abass et al., 2025]

Inflammation. The researchers measured IL-8, an inflammatory signaling molecule. After 24 hours there was no significant change, but by 48 hours the barrier cells showed a modest but statistically significant increase of roughly 10–15%, with the nanoscale particles producing a clear concentration-dependent response. [Abass et al., 2025]

The genotoxicity finding

Perhaps the most consequential result concerns DNA. Using the comet assay, the team found that DNA strand breaks increased significantly in all cell types at both 24 and 48 hours, most notably in Caco-2 cells. [Abass et al., 2025] The nanoscale particles induced greater damage than the microscale ones — the same size-dependent pattern seen throughout the study — and damage scaled with dose and time. [Abass et al., 2025]

A modified version of the assay, using the FPG enzyme, distinguished direct DNA breaks from oxidative DNA lesions. Oxidative DNA damage in the barrier model climbed from 15–20% at 24 hours to 22–30% at 48 hours for both particle types. [Abass et al., 2025]

What this study is not

It’s worth being precise about the limits, because honest limits make the finding more useful, not less.

This is an in vitro study — human cells and engineered barriers in a dish, not people who cook with non-stick pans. The particles used were commercial PTFE micro- and nanoplastics, not particles scraped from a specific pan. And while the study demonstrates what PTFE particles can do to intestinal cells at the tested doses, it does not establish how much a person actually absorbs during ordinary cooking or eating.

The authors are explicit that this is preliminary mechanistic work and that the next step is chronic, low-dose exposure studies in living systems to map the specific pathways involved. [Abass et al., 2025] They also point out that some earlier work found PTFE to have low toxicity — dietary PTFE in rats produced no noticeable changes in clinical signs, organ pathology, or body weight, and a marine gastropod showed no adverse effects on exposure. [Abass et al., 2025] This study doesn’t erase those results; it argues that viability-and-pathology endpoints alone can miss the subtler, sub-lethal damage happening at the cellular level.

The bottom line

The headline claim that PTFE is “biologically inert” was built on its chemical stability — and the researchers confirmed, by FTIR, that both particle types are chemically stable PTFE. [Abass et al., 2025] But chemical inertness and biological inertness are not the same thing. The harm here appears to come from physical interactions — particles physically entering cells and organelles — as much as from any chemistry. [Abass et al., 2025]

For a reader, the practical takeaway is modest and evidence-bound. There’s no need to panic about a single meal cooked in a non-stick pan; this study can’t tell you what a real-world dose does over a lifetime. But the reasoning that made scratched non-stick cookware feel automatically safe — “the particles just pass through” — no longer holds up unexamined. The authors argue plainly that PTFE safety in food-contact applications should be reassessed and that safer alternatives deserve development. [Abass et al., 2025]

If you want a low-effort hedge while the science matures: retire non-stick pans once the coating is visibly scratched or flaking, since damage is exactly what releases particles. [Abass et al., 2025] It’s a small change against an uncertain risk — and it costs you nothing if the risk turns out to be minor.