The short answer is any substance with a density below about 1.0 g/cm³ (fresh water at 4 °C) that does not dissolve. The everyday example is a solid hydrocarbon: candle wax at roughly 0.9 g/cm³ floats on water indefinitely without dissolving. Ice floats too, and so do polystyrene foam, cork, and most solid plastics. Insolubility and flotation are separate properties, though. A substance floats when it is less dense than the liquid. It stays undissolved when its molecules are held together by forces stronger than their attraction to water. A few substances, sugar included, dissolve without floating; a few float without dissolving; almost none do neither.
The mechanism: why insolubility and buoyancy are different tests
Water is a strongly polar molecule with hydrogen bonding. A substance dissolves when its own intermolecular forces are comparable to water's. Polar molecules and ionic salts give water charged or partially charged groups to grab onto, so dissolving releases energy and increases entropy. Nonpolar molecules, including hydrocarbons and fluorocarbons, cannot form hydrogen bonds. Water molecules prefer to bond to each other, so they push nonpolar material aside. The result is insolubility: the forces holding the nonpolar molecules together exceed anything water can offer.
Flotation is pure Archimedes. An object displaces its own weight of water if it can sink far enough. A solid floats whenever its density is below roughly 1.00 g/cm³ at room temperature. That threshold drops to about 0.996 g/cm³ at 25 °C and to about 0.92 g/cm³ at boiling point. So the same object can sink in hot water and float in cold water if its density lies between those values.
The two properties combine because many nonpolar solids happen to be light. Solid alkanes pack loosely; their carbon chains leave empty space relative to the dense water lattice. Densities around 0.8 to 0.95 g/cm³ are typical, which places them firmly in float territory. A plain-language version: water is crowded with sticky polar molecules, and light waxy molecules can neither squeeze in nor dissolve, so they sit on top.
Quantified context: densities, sizes, and force
Fresh water sits at 1.000 g/cm³ at 4 °C. Seawater at 3.5 percent salinity rises to about 1.025 g/cm³. That small jump matters. A plastic at 1.01 g/cm³ sinks in a lake but floats in the ocean; the density gap is only about 1 percent. Ethanol, though miscible with water, has a density of 0.789 g/cm³, which is why a sealed blob of it would float if it could stay separate. Olive oil at 0.91 to 0.92 g/cm³ floats and does not dissolve.
Particle size and shape change the surface result. A steel needle at about 7.8 g/cm³ floats on water when placed gently because surface tension supports it; break the surface and it sinks. A wax crystal smaller than 1 µm can aggregate and settle even though its own density is low, because buoyancy scales with volume while drag and interfacial effects scale with surface area. Surface tension is roughly 72 mN/m at 20 °C, enough to support a few milligrams of light particles if placed flat.
Comparison table: candidate substances
| Substance | Approx. density (g/cm³) | Soluble in water? | Floats? | Notes |
|---|---|---|---|---|
| Paraffin wax | 0.90–0.93 | No | Yes | Standard float and insoluble example |
| Ice | 0.917 | No (very slightly) | Yes | Rare case of a solid floating on its own liquid |
| Polystyrene foam | 0.02–0.05 (foam), 1.04 (solid) | No | Yes (foam), No (solid) | Density depends on form, not chemistry |
| Cork | 0.24 | No | Yes | Air-filled cells dominate the result |
| HDPE plastic | 0.94–0.97 | No | Usually yes | Floats in fresh water, borderline |
| PET plastic | 1.33–1.38 | No | No | Insoluble but denser than water |
| Granite | 2.6–2.8 | No | No | Insoluble and sinks |
The table's real lesson is the PET and granite rows: insolubility alone never guarantees flotation.
Edge cases and counterintuitive details
Temperature can reverse the answer for ice and for some salt hydrates. Ice at 0 °C has a density of 0.917 g/cm³, so it floats. Liquid water at 0 °C is 0.9998 g/cm³. This is a rare case where a solid floats on its own melt. Ice stays solid because hydrogen bonds fix a hydrogen-bonded lattice that is more open than liquid water. That open lattice exists because hydrogen bonds are directional, forcing a rigid geometry. A wax crystal forms a different lattice, which happens to be light for a different reason: long hydrocarbon chains arrange loosely and trap voids.
A second edge case involves foam and air. Polystyrene foam floats because trapped air, not the polymer, supplies most of the buoyancy. Solid polystyrene at about 1.04 g/cm³ sinks. Form can therefore flip a classification. The same logic applies to pumice: a volcanic glass at 2.4 g/cm³ that floats when gas bubbles make its bulk density below 1.
A third case runs the other way. Heavy insoluble salts such as barium sulfate, 4.5 g/cm³, sink and don't dissolve; they behave as inert solids in water. So insolubility tells you nothing negative about density.
Timeline: from a light pellet to a floating outcome
| Time after placement | What happens |
|---|---|
| 0 s | Object contacts surface; surface tension may hold dense objects briefly |
| 0–1 s | Water wets the surface; microparticles may stay pinned at the interface |
| 1–10 s | Buoyancy and gravity resolve; a pellet below ~1 g/cm³ rises to the top |
| Minutes | Wax or foam establishes a stable floating layer |
| Hours | No dissolution; nonpolar substances may slowly spread as a thin film |
| Days | Temperature changes alter the density threshold; a border case object may sink or rise |
| Weeks | Photodegradation or biological action may change the object; density is not the only factor |
Misconceptions: oil, salt water, and surface films
A common error is thinking solubility predicts flotation. It doesn't. Sugar dissolves without floating. PET does neither. A square-cube effect also confuses people: a large chunk of a borderline plastic can float or sink depending on trapped air, since volume grows faster than surface area. Smaller particles sink more readily because interfacial and drag forces dominate, and because air is less likely to be trapped.
Another error is the "salt water floats everything" claim. Seawater at 1.025 g/cm³ sharpens the boundary but does not move it far. Most plastics remain predicted correctly; only borderline members change category. A serious professional caveat: float tests are weak for identifying materials. A materials engineer reading a flotation result will ask about density, air content, and shape before concluding anything. The float test tells you about bulk density, not about what the substance is. Biomass, for example, floats (wood is roughly 0.5 g/cm³ dry) but is not one substance; it is a composite.
A frequent red flag is mistaking the interface for insolubility. A polar liquid that does not mix with water still can dissolve slowly. Surface tension can suspend a dense object and fake buoyancy and insolubility. A floating metal needle is the simplest false positive and demonstrates the limit of any one-second test.
Practical actions: testing and using this knowledge
For material identification, measure density rather than eyeballing. Cut a regular block, weigh it, and find its volume from dimensions, then divide. Use the correct liquid: fresh water holds 1.00 g/cm³, seawater holds 1.025 g/cm³, and heavy liquid additives can raise the limit for difficult cases. Test temperature representative of the application. A plastic at 1.00 g/cm³ can change category between 4 °C and 60 °C. To lift an object, add buoyant volume. Foam and trapped air reduce bulk density and can float a heavier item. For spill cleanup, wax and oil float while dense insoluble particles such as barium sulfate settle; containment strategy follows from where the material ends up.
The thresholds above reconcile simply: use 1.000 g/cm³ for cold fresh water, 1.025 g/cm³ for seawater, and about 0.99 g/cm³ near room temperature. Anything below the relevant number floats. Insolubility is the separate test; both properties are needed to answer the question as framed.
Limitations
The clean 1.0 g/cm³ rule describes water at 4 °C. Most room-temperature work should use about 0.997 g/cm³, and small errors of 1 to 2 percent will misclassify borderline materials. The rule also assumes the substance is a solid homogeneous object, which excludes foams, composites, and particles large or small enough for surface forces to matter. Finally, no single property identifies a material. Which substance is insoluble and floats in water is answered by the combination of low density and negligible water affinity, not by either test alone.