Sort Plastics by Float Test Like a Recycler Does

Cut a square from a milk jug and a square from a drinks bottle and drop both in a glass of water. One floats, one sinks, and your child has just performed the first step of a process real recycling plants use at industrial scale. This plastic density sorting experiment turns unlabelled recycling into identified resin types using nothing but water, salt and rubbing alcohol. The best moment comes near the end, when a rigid polystyrene cup sinks and a lump of polystyrene foam floats.

Key Takeaways

• You’ll sort plastic samples through three baths of different density and match them to their resin identification codes.
• It teaches density as an identifying property, and how industrial sink-float separation actually works.
• Best for ages 9 to 15. An adult does the cutting; everything else is the child’s.
• The usual failure is trapped air and surface tension making things float that should sink. One drop of detergent in the water fixes it.
• The recycling triangle is an identification code, not a promise that the item is recyclable.

What You’ll Need

Item	How much	Where to get it	If you don’t have it
PET sample, code 1	A 2 cm square from a clear drinks bottle	Recycling bin
HDPE sample, code 2	A 2 cm square from a milk jug or shampoo bottle	Recycling bin
PVC sample, code 3	A 2 cm square of rigid pipe offcut or blister packaging	Garage, medicine packaging	The hardest one to find. Skip it if you must
LDPE sample, code 4	A 2 cm square from a carrier bag or squeezy bottle	Kitchen drawer
PP sample, code 5	A 2 cm square from a yoghurt pot or a bottle cap	Recycling bin
PS samples, code 6	A rigid cup or CD case and a lump of white packing foam	Recycling bin, parcel packaging	Both are needed. The comparison is the point
Table salt	400 g	Supermarket
Isopropyl alcohol, 70%	250 mL	Pharmacy, as rubbing alcohol	Surgical spirit. Flammable, so keep it away from any flame
Three clear glasses or jars	3, about 400 mL	Kitchen
Tweezers or a fork	1	Drawer
Scissors, adult only	1	Drawer	A craft knife is worse, not better
Kitchen scale and a syringe or measuring cylinder	1 each	Kitchen, pharmacy	For the optional density calculation

Nothing here is food and nothing is tasted.

Time: prep 20 min, experiment 35 min, cleanup 5 min Cost: a bottle of rubbing alcohol; everything else is free from the bin Mess level: low, but the alcohol needs ventilation

Do the Experiment: A Plastic Density Sorting Experiment

Step 1 - Prepare clean, labelled samples

An adult cuts a 1 to 2 cm square from each item. Note the number inside the recycling triangle on the original item and write it on a scrap of paper beside each square. Scrub off labels, adhesive and any grease with washing-up liquid and dry them, because glue and oil change what floats.

What you should see: six or seven bare plastic squares. Keep the paper labels, because the whole exercise is about confirming or contradicting them later.

Step 2 - Bath one: plain water with a drop of detergent

Fill the first glass with tap water and add one drop of washing-up liquid, stirring gently without making foam. Water has a density of 1.00 g/cm³ at room temperature. Drop each sample in, push it under with tweezers to release trapped air, and record float or sink.

What you should see: PP, LDPE, HDPE and the foam float. PET, PVC and the rigid polystyrene sink. The detergent matters enormously: without it, surface tension holds thin squares on top of the water and you get false floats that ruin the whole experiment.

Step 3 - Bath two: alcohol and water to split the floaters

Mix roughly 130 mL of 70% isopropyl alcohol with 120 mL of water. That gives a bath of roughly 0.93 g/cm³, lighter than water. Put only the samples that floated in Step 2 into this bath, again pushing them under.

What you should see: PP floating, HDPE sinking, LDPE in between and sometimes hovering. You have just separated two plastics that feel almost identical, using a 0.03 g/cm³ difference. Adjust with a splash of water to make the bath denser, or alcohol to lighten it, until they separate cleanly.

Step 4 - Bath three: salt water to split the sinkers

Stir salt into 300 mL of warm water until no more dissolves, which takes about 110 g, then let it cool. A saturated salt solution reaches roughly 1.20 g/cm³. Put the samples that sank in Step 2 into this bath.

What you should see: rigid polystyrene now floats, while PET and PVC still sink. Three baths have sorted your pile into four groups, and the only information you used was density.

Step 5 - The foam-versus-cup comparison

Put the rigid polystyrene cup and the polystyrene packing foam side by side in plain water. Both are the same polymer, code 6.

What you should see: the cup sinks and the foam floats high. Ask for an explanation before you give one. The foam is mostly trapped air, with a bulk density near 0.02 to 0.05 g/cm³, while the polymer itself is near 1.05. Density is a property of the object, not only of the material.

What’s Actually Going On

Density is mass divided by volume, and for a pure polymer it is set by how tightly the chains pack. Polyethylene and polypropylene are simple hydrocarbons with no heavy atoms, so they come in under 1.00 g/cm³ and float on water. Add chlorine, as in PVC, or a benzene ring and ester linkages, as in PET, and the density climbs above water. That is the whole basis of sink-float sorting.

The published values worth knowing are these. PET, code 1, sits near 1.38 g/cm³. HDPE, code 2, is 0.94 to 0.96. PVC, code 3, runs from about 1.16 for flexible grades to 1.45 for rigid filled ones. LDPE, code 4, is 0.91 to 0.93. PP, code 5, is the lightest common plastic at 0.90 to 0.91. PS, code 6, is 1.04 to 1.06 as solid polymer, while expanded polystyrene foam is 0.01 to 0.05 because it is mostly gas. Code 7 is a catch-all, and includes polycarbonate near 1.20 and the bioplastic PLA near 1.24.

Industrially this works, with caveats. Bauer and colleagues reported on sink-float density separation of post-consumer plastics in the Journal of Material Cycles and Waste Management in 2018, and the technique remains a workhorse because it is cheap and needs no instruments. What it cannot do is distinguish polymers whose densities overlap, which is why real plants pair it with near-infrared spectroscopy.

Three practical complications are worth telling your child about, because they will meet all three.

First, fillers and pigments shift density. Talc-filled polypropylene, common in black food trays and automotive parts, can exceed 1.00 g/cm³ and sink even though pure PP floats. Carbon black is also why black plastics are hard for optical sorters to read. So a sinking sample is not proof of identity; it is evidence.

Second, trapped air and surface tension produce false positives. A thin square of PET will sit on an undisturbed water surface indefinitely, held by surface tension, which is exactly why Step 2 specifies a drop of detergent and pushing each sample under. Industrial lines use agitation and wetting agents for the same reason.

Third, and this is the correction worth making loudly: the number inside the recycling triangle is not a recyclability rating. It is a resin identification code, standardised as ASTM D7611, and its purpose is to tell a sorter what the polymer is. A code 5 yoghurt pot may or may not be collected where you live, and a code 1 bottle and a code 1 tray are often handled completely differently because the tray is multilayer. The EPA’s figures on plastics in municipal waste show how much of what carries a triangle is not in fact recovered. Children learn the triangle as a promise, and the float test lets you replace that with something true: the triangle is a label saying what it is made of, and whether it gets recycled depends on your local facility and the economics.

How to Explain It to Your Kid

Ages 5–8: heavy for its size

Keep it to bath one. The idea to land is that floating isn’t about being heavy, it’s about being heavy for its size. A big foam block floats and a small bottle square sinks. Let them do the pushing-under with a fork and the sorting into two piles. Then show them the foam and the cup together and let them be puzzled. Puzzled is a good place for a six-year-old to finish.

Ages 9–12: a table and a prediction

Introduce density as mass divided by volume, and run all three baths with a proper results table: sample, code, bath one result, bath two result, bath three result, then the published density range from the article. Have them predict each result before the sample goes in. The satisfying part is when their table matches the published values for five or six samples out of seven, and the mismatch is usually a filled or printed plastic.

Ages 13+: measure a density and explain the overlaps

Have them measure an actual density. Weigh a sample on a kitchen scale, then find its volume by displacement: fill a syringe or measuring cylinder with water, read the level, push the sample under, read again. Density is mass over the volume difference. Then the real question: given that PVC spans 1.16 to 1.45 and PET sits at 1.38, can a single sink-float bath ever separate them reliably? No, and working out why leads straight to why near-infrared sorting exists. Finish by asking what the density of a multilayer food tray even means.

The question to ask: “If I gave you a liquid with a density of exactly 1.00 and a sample that hovered in the middle without rising or sinking, what would you know about it, and what would you still not know?”

Why Teaching This Early Pays Off for Materials Identification

This experiment does something most density activities don’t: it uses density to identify an unknown rather than to explain a known. A child who has sorted seven squares into four groups has used a physical property as a diagnostic tool, which is how analytical chemistry actually works and is a genuinely different intellectual move from “oil floats on water because it’s less dense.”

The curriculum connections are solid. Density, mass and volume are primary and lower-secondary science, usually taught with regular solids and a ruler, and displacement measurement is the standard technique for irregular objects and is rarely practised. And the recycling content, which usually arrives as exhortation, becomes a technical problem with real constraints.

The misconception work is the other half. Smith, diSessa and Roschelle argued in 1994 that misconceptions are productive raw material rather than errors to be stamped out, and “the triangle means it gets recycled” is a perfect example: the right response is not to tell a child they are wrong but to hand them a test that shows what the symbol is actually for. The National Research Council’s Framework for K-12 Science Education lists “analysing and interpreting data” among its practices, and comparing your own results against published density ranges is that.

Before You Start: What Parents Should Know

Best age	9 to 15. Bath one alone works from about 6; the three-bath sort and the density calculation need 9 and up
Supervision	An adult does all the cutting and handles the alcohol. The baths and the sorting belong to the child
Real hazards	Three things. Scissors, so an adult cuts every sample and no child cuts rigid plastic, which snaps and flicks. Isopropyl alcohol, which PubChem records as a flammable liquid: keep it away from any flame or hob, work with a window open, do not let anyone drink it, and wash hands after handling. And small plastic squares, which are a choking hazard, so no children under 3 in the room and all squares counted back into a bag at the end
Don’t	Don’t substitute acetone, nail polish remover or white spirit for the isopropyl alcohol. Acetone dissolves polystyrene outright, which destroys your sample and produces a sticky mess, and the others are more hazardous without being any denser or lighter. If you cannot get isopropyl alcohol, run baths one and three only and say so in your results

When It Doesn’t Work

• Everything floats. Surface tension and trapped air. Add a drop of washing-up liquid to the water, push every sample fully under with tweezers, and tap the side of the glass to dislodge bubbles. This single fix accounts for most failed attempts.
• PP and HDPE won’t separate in bath two. Your mixture is the wrong density. Add water a teaspoon at a time to make it denser until HDPE starts to sink, or alcohol to make it lighter if PP sinks. The window between 0.91 and 0.94 is narrow and worth hunting for.
• A sample sank that should have floated. Almost certainly filled or pigmented. Black trays, glittery plastics and anything with a mineral filler can be denser than the pure polymer. Note it as a result rather than an error; it is what industrial sorters struggle with too.
• The salt water isn’t dense enough. You have not saturated it. Keep adding salt to warm water until crystals sit on the bottom, then cool it to room temperature before testing, because warm brine reads denser than it will be.
• You cannot find a PVC sample. It is genuinely scarce in household recycling now. Rigid pipe offcuts, blister packs and old vinyl flooring are the usual sources, and the experiment works fine without it.

Take It Further: More Plastic Density Experiments

Build a density ladder. Make five salt solutions of increasing concentration, measure the density of each by weighing a known volume, and find the bath in which each plastic exactly hovers. That is neutral buoyancy, and the bath’s density at that point is the sample’s density. It is how density-gradient methods work in materials labs.

Test a bioplastic against the codes. If you have made a film from our cornstarch and glycerin bioplastic, run it through all three baths and see where it lands. Starch films are denser than water and will sink in bath one, which puts them in the same group as PET for sorting purposes and raises a genuinely unsolved problem in recycling.

Audit a week of household plastic. Record the code on every plastic item that leaves your house for a week, then check what your local council or municipality actually accepts. The gap between the two is the most useful thing here, and it connects to our look at microplastics and child development.

What to Watch For Over the Next 3 Months

• Week 4: they start reading resin codes on everything, usually out loud in supermarkets, and noticing that identical tubs carry different numbers.
• Month 2: the phrase “heavy for its size” or the word “density” turns up about something unrelated, often a boat, a helium balloon or a heavy book.
• Month 3: worth encouraging: how a sorting plant tells black plastics apart if optical sensors can’t see them. The answer involves infrared, electrostatics and sometimes giving up, which is an honest introduction to engineering limits.

Frequently Asked Questions

Does the number in the recycling triangle mean it can be recycled?

No. It is a resin identification code standardised as ASTM D7611, and it tells a sorter what polymer the item is made from. Whether it gets recycled depends on your local collection, the item’s shape, whether it is multilayer, and the market price for that resin.

Why does polystyrene foam float when polystyrene sinks?

Because density is a property of the object. Expanded polystyrene is roughly 95 to 98 percent trapped air, giving a bulk density around 0.02 to 0.05 g/cm³, while the polymer itself is about 1.05 and sinks. Same chemistry, opposite result, and it is the single best demonstration in this experiment.

Can we use salt water for everything instead of alcohol?

Salt water can only make a bath denser than water, so it separates the sinkers but cannot split polypropylene from polyethylene, which both float. For that you need a liquid lighter than water, which is what the alcohol mixture provides. If you would rather not use alcohol, run baths one and three and report three groups instead of four.

How accurate is this compared to what a recycling plant does?

Surprisingly good for clean, unfilled samples and genuinely unreliable for filled, printed or multilayer ones. That is true industrially too. Your results agreeing with published densities on five of seven samples is a realistic and respectable outcome.

---

About the author

Ricky Flores is the founder of HiWave Makers and an electrical engineer with 15+ years of experience building consumer technology at Apple, Samsung, and Texas Instruments. He writes about how kids learn to build, think, and create in a tech-saturated world. Read more at hiwavemakers.com.

---

Sources

1. Bauer, M., Lehner, M., Schwabl, D., Flachberger, H., Kranzinger, L., Pomberger, R., & Hofer, W. (2018). “Sink–float density separation of post-consumer plastics for feedstock recycling.” Journal of Material Cycles and Waste Management, 20(2), 1781–1791. https://doi.org/10.1007/s10163-018-0748-z
2. ASTM International. (2021). “D7611/D7611M-21: Standard Practice for Coding Plastic Manufactured Articles for Resin Identification.” https://www.astm.org/d7611_d7611m-21.html
3. US Environmental Protection Agency. “Plastics: Material-Specific Data.” Facts and Figures about Materials, Waste and Recycling. https://www.epa.gov/facts-and-figures-about-materials-waste-and-recycling/plastics-material-specific-data
4. National Center for Biotechnology Information. “Isopropanol, CID 3776: properties and hazards.” PubChem. https://pubchem.ncbi.nlm.nih.gov/compound/3776
5. Smith, J. P. III, diSessa, A. A., & Roschelle, J. (1994). “Misconceptions Reconceived: A Constructivist Analysis of Knowledge in Transition.” Journal of the Learning Sciences, 3(2), 115–163. https://doi.org/10.1207/s15327809jls0302_1
6. National Research Council. (2012). A Framework for K-12 Science Education: Practices, Crosscutting Concepts, and Core Ideas. National Academies Press. https://doi.org/10.17226/13165