Correcting the record
The 2024 study that made people bin their spatulas was corrected twice. The chemicals it found are real and still undisputed. The exposure figure that made it frightening was wrong by roughly forty-fold once both corrections landed.
In October 2024 a study went round the world with a simple message: your black plastic spatula is leaching flame retardants into your food at levels approaching a federal safety limit. A lot of people threw their utensils out.
The study contained a maths error. It was corrected twice — once in December 2024 and again, more substantially, in July 2025. Almost none of the coverage that made people throw things away has been updated.
Here is what was found, what was wrong, and what an honest reading leaves you with. It is not “never mind”, and it is not the original headline either.
This site exists to check what companies publish. The same standard has to apply to research we would otherwise be happy to quote — especially when the corrected version is less alarming. A finding that gets weaker after checking is exactly the kind we are most tempted to leave alone.
The paper, by Liu and colleagues, was published in Chemosphere on 11 September 2024. The researchers screened 203 black plastic consumer products sold in the US, 109 of them kitchenware, for bromine. The 20 highest-bromine samples then went for full chemical analysis.
They found brominated or organophosphate flame retardants in 85% of the bromine-positive samples. The most common were TBBPA, BDE-209 (decabromodiphenyl ether) and 2,4,6-tribromophenol, each in at least 70% of tested samples. One sample carried 22,800 mg/kg of total flame retardants. One sushi tray measured 18,600 ppm bromine.
None of this was corrected. None of it is in dispute. There are restricted flame retardants, including one classified as a probable human carcinogen, in a large share of the black plastic kitchen items tested — almost certainly because the plastic was recycled from electronic waste, where flame retardants belong. The Food Packaging Forum’s original write-up covers this well.
In its discussion, the paper estimated that a black plastic kitchen utensil could transfer 34,700 nanograms per day of BDE-209 into food, and compared that against the EPA reference dose for a 60 kg adult — concluding it approached the safety limit.
The reference dose was written as 42,000 ng/day. The correct figure is 420,000 ng/day. A digit had been dropped converting 7,000 ng/kg of body weight per day for a 60 kg adult. Corrected, the estimate was about an order of magnitude below the reference dose rather than close to it. Chemosphere published the correction on 15 December 2024.
This one got far less attention. In July 2025, a further error surfaced — the exposure formula had been misapplied, so the 34,700 ng/day daily-transfer figure was itself too high. The corrected estimate is 7,900 ng/day, roughly a further four-fold reduction. Retraction Watch broke it; Popular Science and Scientific American followed.
Put the two together and the estimated daily exposure went from “approaching the reference dose” to roughly 2% of it.
Toxic-Free Future’s position is that the error does not affect the paper’s conclusion, because the exposure estimate was a discussion item rather than part of the detection work. The core finding — that restricted flame retardants are turning up in food-contact items that have no fire-safety reason to contain them — is unaffected.
That is a fair defence of the chemistry. The industrial chemist Mark Jones argued publicly that it is not a fair defence of the conclusion, since the paper’s claim of significant contamination carried an implication of meaningful health risk that the corrected numbers no longer support. His critique and the authors’ rebuttal were published alongside the second correction.
We think both are partly right, and the distinction is one worth internalising because it comes up constantly in this subject:
On the corrected numbers, we would not tell you to. The estimated exposure sits well under the EPA reference dose, and the reference dose already includes safety margins.
What we would say is this: there is no good reason for a spatula to contain a restricted flame retardant, you cannot tell from the outside which items do, and there is no labelling that would help you. When one wears out, a replacement that is not black recycled plastic costs a few dollars and closes the question permanently. That is a sensible thing to do at the normal replacement moment, and a disproportionate thing to do tonight.
There is no marking that tells you whether a black plastic item came from recycled electronics. The study had to buy products and test them, because no manufacturer publishes it. That, rather than the dose arithmetic, is the durable problem here — and it is the same problem we score across the rest of the site.
We are not currently in any affiliate programme, so nothing on this page is a paid link. When that changes, the disclosure page says so first.
New scores, brands that finally answered, and any rule change that affects what you can buy. The list is not open yet — email us and we will add you when it is.