Blending Ice for 30 Seconds Released Hundreds of Millions of Plastic Particles

The short answer

Almost every microplastic story about kitchen appliances is really a story about heat — hot water through a plastic pod, a sous vide bag at 60 °C, a kettle with a plastic reservoir. The blender is different, and that is what makes it interesting. Nothing about a smoothie is hot. The jar is doing something no other container in your kitchen does: it is holding a hard abrasive that is being flung against it, thousands of times a minute, on purpose.

A plastic surface under that treatment loses material. When researchers actually looked for it, they found a great deal of it — and importantly, they found it in the nanoplastic range as well as the microplastic range, which is the size class that is hardest to measure and most able to move around the body.

What the study actually did

The paper is Detection of microplastics and nanoplastics released from a kitchen blender using Raman imaging, published in the Journal of Hazardous Materials in 2023 by a team at the Global Centre for Environmental Remediation, University of Newcastle, Australia. The framing in the abstract is disarmingly plain: they tested a blender they use in their own kitchen to make juice.

The method was ice blocks in Milli-Q water — laboratory-pure water, so that anything found came from the blender rather than the ingredients — blended for thirty seconds using the pause function, roughly a second of blending followed by a second of rest. They filtered what came out and analysed it with Raman imaging, building a scanning spectrum matrix of 900 spectra in a 30 × 30 grid, then reconstructing images from it to pick out particles too small for conventional Raman resolution.

Their estimate: approximately 0.36–0.78 × 10⁹ micro- and nanoplastic particles released from the plastic container within thirty seconds. That is 360 to 780 million.

Why ice is the worst case

The mechanism the authors describe is straightforward. The blade generates a high-speed vortex; the vortex throws the ice blocks outward; the ice scratches the inner wall of the container; the wall sheds debris into the liquid. Ice is hard, angular, and does not deform. It is, functionally, a tumbling abrasive.

This is worth holding onto because it sets the boundaries of the finding. Blending a ripe banana into oat milk is not this experiment. Blending a fistful of ice, frozen mango and almonds at maximum speed for ninety seconds is more than this experiment. The study establishes that the mechanism is real and that it operates at surprising scale; it does not tell you what your particular Tuesday smoothie produces.

What the number does and does not mean

A figure in the hundreds of millions is arresting, and it deserves three pieces of context:

• Most of those particles are nanoplastics, which are tiny. A count is not a mass. Hundreds of millions of nanometre-scale fragments can add up to a vanishingly small weight of plastic. Counts and masses tell different stories and the study reports a count.
• It is one blender. The authors did not name the model or identify the jar polymer, and they did not compare brands. Nobody has run the equivalent test across a shelf of blenders, which is the study the field actually needs.
• No health outcome was measured. This is a measurement paper about analytical method as much as about blenders. It tells you particles are released. It does not tell you what they do.

How this compares with the rest of your kitchen

The closest well-characterised comparison is the plastic chopping board. In a 2023 study in Environmental Science & Technology, researchers estimated per-person annual exposure of 7.4–50.7 g of microplastics from a polyethylene board and around 49.5 g from a polypropylene one, corresponding to roughly 14.5–71.9 million polyethylene particles a year, or about 79.4 million polypropylene particles. Chopping carrots released more than chopping nothing, which is the same abrasion story.

Set those side by side and the point lands: a year of chopping on a plastic board is in the tens of millions of particles, while thirty seconds of blending ice was estimated in the hundreds of millions. The two figures come from different methods and different size ranges, so this is an illustration of scale rather than a clean ratio — but no reasonable reading of them puts the blender low on the list.

The same study also found that the released particles did not significantly change the viability of mouse cells in lab testing, which is a genuine and often-omitted part of that result. Detection is not the same as harm, and we would rather say so than let the number carry an implication the paper does not.

Why “BPA-free” and “Tritan” miss the point

Blender marketing is almost entirely about material chemistry. Vitamix states its containers are made from BPA-free Eastman Tritan copolyester. Blendtec says the same of its jars, adding NSF food-safety certification. NutriBullet uses Tritan too, and has moved its blending cups to Tritan Renew, made with 50% certified recycled content.

Every one of those claims is true, and none of them addresses what the study measured. BPA is a chemical that migrates out of certain polymers. A microplastic is a fragment of the polymer itself. A jar can be entirely free of the first and still produce the second — and the second is what a vortex full of ice generates.

We have seen this exact pattern before. When researchers counted particles from reusable Tritan bottles, the source was not the wall of the bottle but the screw cap: 131 particles per litre after a single opening, rising to 242 after eleven. The bottle body was fine. The place where two plastic surfaces ground against each other was not. Tritan is a genuinely good drinkware plastic — tough, clear, no bisphenol in its structure — and it is still a plastic being abraded.

What your blender jar is actually made of

Manufacturers publish this, so it is worth checking rather than assuming. All of the below is taken from the brands’ own sites:

Two things stand out. The first is that the premium end of the market is uniformly Tritan, so paying more does not move you off plastic — it moves you to a better-engineered plastic that is still the abraded surface. The second is that the two straightforward non-plastic routes are a Vitamix stainless container, which fits an appliance many people already own, and an inexpensive Oster with a glass jar. Neither is a premium purchase.

Five things that change how much your blender sheds

Every one of these acts on abrasion, because abrasion is the mechanism. In rough order of effect:

• How frozen and how hard the ingredients are. Ice is the worst case and it is the one that was tested. Letting frozen fruit sit out for five or ten minutes, using chilled fruit instead of frozen, or adding the liquid first so the solids are cushioned all reduce the direct ice-on-wall contact the study describes.
• How long you run it, and how fast. The result came from thirty seconds. Ninety seconds on maximum is three times the exposure at higher vortex energy. Most smoothies are blended considerably longer than they need to be.
• How scratched the jar already is. A cloudy, scuffed, chalky-looking jar has been abrading for years. A damaged surface has more exposed area and shears more readily than a smooth one. If you can see the wear, the jar is past its useful life for this purpose — and it is the cheapest single part to replace.
• Whether you dry-grind in the same jar. Grains, nuts and coffee beans with no liquid to cushion them are the harshest thing you can do to a plastic container. Vitamix sells a separate dry grains container precisely because the job is different. If you grind, do not do it in the jar you make smoothies in.
• Whether you use the hot soup program. Vitamix states that blade friction alone brings soup to 170 °F in about five minutes on High. That is five minutes of maximum-speed abrasion applied to a jar wall that is simultaneously being heated toward the top of what the polymer is comfortable with. Nobody has measured particle release under that specific combination, which is exactly why it is worth flagging: it stacks both known stressors at once, and blending soup on the stove first and warm — not boiling — afterwards avoids the question.

What to buy if you are replacing one

• A stainless container, if your machine takes one. Vitamix’s 48 oz stainless steel container is the cleanest upgrade for an existing owner: no shedding surface, and you keep the motor you already paid for. The trade-off is that you cannot see what is happening inside.
• A glass jar. Oster’s Classic Series models ship with a 6-cup Boroclass glass jar as standard and cost a fraction of a high-performance blender. Glass is heavy and can chip, but it does not abrade into your food.
• An immersion blender used in a metal pot. For soups especially, a stick blender with a stainless shaft and bell, used directly in a saucepan, removes the plastic vessel from the equation altogether. Check the bell — some are plastic — and this is a genuine replacement only for soups and purées, not for frozen smoothies.
• Nothing at all, if your blender works. Replacing a functioning appliance on the strength of one study on one unit is not a proportionate response. Changing how you use it is free and acts on the same variable.

What is not worth doing

Straining a smoothie will not help — a mesh that catches nanoplastics would not pass liquid. Buying a “BPA-free” replacement jar does not address the mechanism, as above. And microplastic “detox” supplements have no evidence behind them regardless of where the particles came from.

What we still do not know

• No brand comparison exists. One unnamed blender has been tested. Whether jar polymer, wall thickness, blade geometry or motor speed dominates is unknown, and it is the obvious next experiment.
• Ordinary smoothies have not been measured. The tested condition was ice in pure water. Soft fruit, yoghurt, protein powder and oat milk behave differently and nobody has quantified how differently.
• Counts are not doses. The paper reports particle counts, not mass, and translating either into an exposure estimate would require assumptions the study does not provide.
• The heat-plus-abrasion case is unstudied despite being a marketed feature on high-performance machines.

Related reading: is Tritan a microplastic, plastic cutting boards, electric kettles, air fryers, protein powder, and which kitchen materials are actually plastic.