Why Your Car Doesn’t Freeze Solid at −20°F

So what are colligative properties doing in your radiator?

Colligative properties are one of those ideas that sound abstract in class and then turn out to be hiding in every winter parking lot. The short version: when you dissolve a non-volatile solute in a solvent like water, you change the solvent’s physical behavior. Not by changing its identity, but by messing with the number of particles per unit of solution.

In antifreeze, the solvent is mainly water. The solute is usually ethylene glycol (HO–CH₂–CH₂–OH) or, in some newer formulations, propylene glycol. These molecules dissolve nicely in water and don’t evaporate easily.

The four main colligative properties are:

• Lowering of vapor pressure
• Boiling point elevation
• Freezing point depression
• Osmotic pressure

Under your hood, the stars of the show are freezing point depression and boiling point elevation. Vapor pressure and osmotic pressure are still there in the background, but they’re more like supporting actors.

Why adding glycol makes water harder to freeze

Let’s start with the part everyone cares about in January: freezing point depression.

When pure water freezes, water molecules organize into a crystal lattice—ice. That structure is surprisingly picky. It wants mostly water, arranged in a very specific pattern. When you dissolve ethylene glycol in water, those glycol molecules get in the way. They disrupt the ability of water molecules to line up and lock into the solid structure.

So the solution has to be cooled to a lower temperature before enough water molecules can escape the liquid phase and form solid ice. That temperature drop is what we call freezing point depression.

In ideal cases, we describe it with this formula:

\[ \Delta T_f = i \cdot K_f \cdot m \]

Where:

• \(\Delta T_f\) = magnitude of freezing point depression (°C)
• \(i\) = van ’t Hoff factor (number of particles per formula unit)
• \(K_f\) = cryoscopic constant of the solvent (for water, about 1.86 °C·kg/mol)
• \(m\) = molality of the solution (mol solute per kg solvent)

For ethylene glycol in water, \(i \approx 1\) because it doesn’t dissociate into ions like salt does. So the effect depends mostly on how many glycol molecules you cram into the water.

A quick, real-world style calculation

Imagine a coolant mixture that’s about 50% ethylene glycol by volume, which is very common in cars.

Ethylene glycol has:

• Molar mass ≈ 62.1 g/mol
• Density ≈ 1.11 g/mL

If you mix 1 liter of solution at roughly 50/50 by volume, you have about:

• 500 mL ethylene glycol → about 555 g → about 8.9 mol
• 500 mL water → about 500 g → 0.5 kg

Molality \(m\) ≈ 8.9 mol / 0.5 kg = 17.8 m

Now plug into the formula:

\[ \Delta T_f \approx 1 \cdot 1.86\,\frac{^\circ C\cdot kg}{mol} \cdot 17.8\,\frac{mol}{kg} \approx 33\,^\circ C \]

Pure water freezes at 0 °C, so the ideal predicted freezing point would be:

\[ T_f \approx 0 - 33 = -33\,^\circ C \]

That’s about −27 °F.

In the real world, commercial 50/50 antifreeze-water mixes typically protect down to around −34 °F (about −37 °C), because the solution isn’t perfectly ideal and the formulation includes other additives. But the order of magnitude lines up nicely: pack in a lot of solute particles, and the freezing point drops hard.

The strange part: more antifreeze isn’t always better

Here’s where it gets counterintuitive and, frankly, where a lot of backyard mechanics mess up. You might think, “If 50% antifreeze is good, 80% must be amazing, right?”

Not quite.

Water is actually the better “freezer” in this relationship. You need enough water to form an ice-like structure as the temperature drops. If you push the mixture toward pure glycol, you start losing the very solvent whose freezing point you were trying to manipulate.

In practice:

• Around 50–60% ethylene glycol by volume gives you the lowest freezing point.
• Above that, the freezing point starts to rise again.

So there’s a sweet spot. Too little antifreeze and the coolant freezes. Too much and the protection gets worse again. You also start to hurt heat transfer, because water carries heat more effectively than ethylene glycol.

You can see this in manufacturer charts: as concentration rises from 0% to about 60%, the freezing point steadily drops. Past that, it turns around and climbs.

Why coolant also protects you in August

It’s easy to think of antifreeze as winter gear only, but the same colligative logic helps in summer too. Boiling point elevation is the flip side of the same coin.

The formula looks similar:

\[ \Delta T_b = i \cdot K_b \cdot m \]

Where \(K_b\) is the ebullioscopic constant for water (about 0.512 °C·kg/mol).

Again, \(i \approx 1\) for ethylene glycol. The same dense population of solute particles that pushes the freezing point down also pushes the boiling point up. And then the radiator cap adds pressure, which raises the boiling point even further.

Put it together and a typical pressurized cooling system with a 50/50 mix can have a boiling point well above 220 °F, often in the 230–260 °F range, depending on system pressure. That margin is what keeps your coolant from flashing into steam around hot engine parts.

Steam is a terrible heat transfer medium for this job. Once parts of the coolant start boiling locally, you get vapor pockets, poor heat transfer, and the temperature can spike fast. By lifting the boiling point, the antifreeze mixture buys you stability.

Vapor pressure: the quiet background player

Lowering vapor pressure is the underlying reason both freezing point and boiling point shift. When you dissolve glycol in water, you reduce the fraction of surface molecules that are water. Fewer water molecules at the surface means fewer can escape into the vapor phase.

Lower vapor pressure means:

• You need a higher temperature to reach the external pressure and boil.
• You shift the balance between liquid and solid phases, which shows up as a lower freezing point.

You don’t see “vapor pressure” on the label of an antifreeze jug, but the concept is baked into the behavior of the solution.

A quick detour: why salt on roads feels so similar

If all of this is giving you flashbacks to salted winter roads, you’re not wrong. Road salt is basically the same physics, just a different solute.

With sodium chloride (NaCl) or calcium chloride (CaCl₂), the van ’t Hoff factor \(i\) is bigger than 1 because the salt dissociates into ions. One NaCl unit becomes Na⁺ and Cl⁻, so \(i\) is roughly 2. More particles per mole means a stronger colligative punch per unit of solute.

Ethylene glycol doesn’t split into ions, so you need more of it (in terms of moles) to get the same freezing point depression as a smaller amount of salt. But glycol has advantages: it’s a liquid, mixes smoothly with water, doesn’t corrode metal as aggressively when properly inhibited, and plays nicely with engine materials.

When the chemistry fails: two very different mornings

Think of Alex, who lives in Minnesota and bought an older car from a warmer state. The previous owner had topped off the radiator with plain water—repeatedly. The green color was still there, so Alex assumed the coolant was fine.

Then came a −10 °F night.

By morning, the “coolant” was a slushy mess. Ice expanded inside the engine block, stressing metal, squeezing gaskets, and threatening to crack the block or the radiator. All because the solution had drifted too close to pure water and lost the colligative protection.

On the flip side, there’s Maya, who decided to be extra safe. She drained her system and refilled with almost pure concentrated antifreeze, barely any water. She figured more glycol meant more protection.

Her car didn’t freeze, but it started running hotter. The heater performance dropped, and the temperature gauge crept up on long highway climbs. Why? She’d moved away from the ideal concentration for freezing point depression and seriously reduced the solution’s heat capacity and thermal conductivity. In trying to “improve” the chemistry, she sabotaged the cooling system.

Both stories are the same lesson: the physics cares about particle counts and mixture properties, not about human intuition.

Toxicity, safety, and why propylene glycol is getting attention

Ethylene glycol is chemically convenient but biologically nasty. It’s sweet-tasting and highly toxic if ingested. That’s why you’ll see warnings all over antifreeze containers and why accidental poisonings—especially in children and pets—show up in toxicology reports.

Propylene glycol is often marketed as a “safer” alternative. It’s less toxic and used in some food and pharmaceutical applications. Chemically, it behaves similarly in water as a colligative solute, though its physical properties differ slightly.

From a colligative standpoint, both:

• Dissolve well in water
• Stay in the liquid phase over a wide temperature range
• Contribute one particle per molecule (\(i \approx 1\))

So the same freezing point depression and boiling point elevation logic applies. The choice between them is more about toxicity, cost, and compatibility with engine materials and additives.

For health and toxicity background on ethylene glycol and related compounds, resources like the National Institutes of Health and the Agency for Toxic Substances and Disease Registry are worth exploring:

• https://www.ncbi.nlm.nih.gov
• https://www.atsdr.cdc.gov

Why chemists love molality for this job

You might have noticed the formulas use molality, not molarity. That’s not a random choice.

Molality is defined as moles of solute per kilogram of solvent. It doesn’t change with temperature because mass doesn’t expand or contract the way volume does.

In a cooling system, temperature swings are the whole point. If you built your equations on molarity (moles per liter of solution), the value would shift as the solution expanded and contracted. Molality stays put, which makes it much nicer for predicting colligative effects.

In the lab, this matters for accurate calculations. Under the hood, manufacturers handle the heavy lifting and give you volume-based mixing instructions that approximate the right molality in practice.

Where the “simple” theory starts to bend

Now, if you’re thinking, “Hang on, those calculations looked a bit too clean,” you’re right. Real antifreeze solutions are not perfectly ideal.

At high concentrations:

• Molecules interact more strongly.
• Activity coefficients deviate from 1.
• The linear relationship between \(\Delta T\) and molality starts to curve.

That’s why manufacturers rely on empirical data—actual measured freezing and boiling points for specific mixtures—rather than just plugging numbers into the ideal equation. The colligative formulas give the conceptual framework and a decent first approximation, but the real world adds corrections.

If you dive into engineering data sheets or automotive manuals, you’ll see tables and graphs instead of pure equations. Those charts quietly encode the non-ideal behavior.

Frequently asked questions

Why can’t I just use pure water as coolant if I live in a warm climate?

You could, for a while, but it’s a bad idea. Even in warm climates, nighttime temperatures can dip low enough to cause partial freezing. More importantly, pure water boils at 212 °F at atmospheric pressure and offers no corrosion protection. Modern engines run hot, and you want the higher boiling point, corrosion inhibitors, and lubricants that come with real coolant.

Does mixing different antifreeze brands mess up the colligative properties?

The basic colligative behavior (freezing and boiling points) mostly depends on the overall glycol concentration, so different brands with the same base (like ethylene glycol) will behave similarly in that narrow sense. The real risk is additive package incompatibility—corrosion inhibitors and dyes that don’t play well together. That can lead to sludge, deposits, or reduced protection, even if the freezing point is technically okay.

Is propylene glycol–based antifreeze as effective as ethylene glycol?

In terms of pure colligative action—lowering freezing point and raising boiling point—propylene glycol can perform comparably at similar concentrations. However, it often has slightly different physical properties (like viscosity and heat capacity), and formulations are tuned accordingly. It’s generally favored where lower toxicity is more important than maximum thermal performance, such as in some food-processing or HVAC systems.

Why do I need a specific water-to-antifreeze ratio instead of just filling with concentrate?

Because the mixture’s behavior isn’t linear. There’s a concentration range where you get the lowest freezing point and a nice boost in boiling point while still keeping good heat transfer. Go too concentrated and you start losing water’s excellent thermal properties and even see the freezing point creep back up. The recommended ratios are based on that balance, not on a “more is better” mindset.

Does pressure in the cooling system change the colligative properties themselves?

Pressure doesn’t change the colligative effect from the solute, but it does shift the boiling point of the whole solution. The colligative part raises the boiling point above that of pure water at a given pressure. Then the radiator cap increases system pressure, which raises the boiling point further. They stack together. The freezing point, on the other hand, is only weakly affected by the modest pressures in a car’s cooling system.

Want to dig deeper?

If you’re curious about the underlying thermodynamics and solution behavior beyond the simple equations, chemistry and engineering departments at major universities often publish accessible material. For example:

• https://chem.libretexts.org (open chemistry textbooks and explanations)
• https://web.mit.edu (search for solution thermodynamics or colligative properties)

Antifreeze may look like a mundane maintenance item, but it’s actually a nice little case study in how particle numbers, not just chemical identities, shape the physical world. Next time you see that neon-green or orange liquid, you’ll know there’s a quiet, very quantitative argument with winter going on in there.