Beaker & Batter
❄️

Cold & Freezing

Everything Heat & Transformation covers has a mirror image on the cold side -- and ice cream is where it shows up most directly.

The kitchen answer

Salt on ice doesn't make things colder by magic -- it's the same colligative-property trick as sugar raising a syrup's boiling point (see Sugar Work), just running in reverse to push water's freezing point down instead of up.

What's actually happening

Dissolving something into water changes its freezing point the same way dissolved sugar changes its boiling point -- except freezing point moves down instead of up. Salt dissolved into the ice-water mixture around an old-fashioned ice cream churn can push that mixture's freezing point down to roughly −18 to −21°C, cold enough to freeze the actual ice cream base (whose own freezing point is already depressed by its sugar content) even though plain ice alone sits right at 0°C.

How ice cream freezes matters as much as how cold it gets. Constant churning during freezing mechanically breaks up ice crystals as they try to form, keeping them microscopically small -- that's what gives ice cream its smooth texture instead of a crunchy, icy one. Fat and stabilizers do a second job afterward: they coat the tiny crystals that do form and slow a process called Ostwald ripening, where larger crystals slowly grow at the expense of smaller ones during storage.

0°CPlain ice · freezing point of pure water
−18 to −21°CSalt + ice mixture · depressed freezing point
ChurningMechanically limits ice crystal size
Fat & stabilizersSlow ice crystal regrowth in storage

What does that mean for your recipe

"No-churn" ice cream recipes don't skip the physics, they route around it -- heavy cream whipped to trap air (mechanical aeration, the same idea as creaming in Fats & Dairy) limits how much of the mixture is liquid water able to form large crystals in the first place. That's why fat content matters even more in a no-churn recipe than a churned one -- it's doing double duty.

What can go wrong

Try it yourself

Same base, three freezing treatments:

  1. Freeze one batch with continuous churning.
  2. Freeze a second batch completely undisturbed, no churning at all.
  3. Churn a third batch properly, then deliberately let it partially thaw and refreeze once.

Watch for: the churned batch smooth, the undisturbed batch icy and crystalline from the start, and the refrozen batch grainy despite having started out smooth -- Ostwald ripening demonstrated directly.

Go deeper

Salt water has a true eutectic point -- a specific concentration (around 23% salt by weight) where the freezing point bottoms out at its lowest possible value, roughly −21°C. Past that concentration, adding more salt doesn't push the freezing point any lower; it just leaves undissolved salt sitting in the mixture.

Part of a bigger thread: freezing point depression is the cold-side mirror of boiling point elevation in Sugar Work -- the same colligative-property family, running in opposite directions -- and churning is a third way to get air into a mixture, alongside creaming and whipped egg foams.

Try the tool: Salt & Ice Temperature Estimator

Estimated from the real NaCl-water freezing-point curve, capped at the eutectic minimum around −21°C -- adding salt past roughly 23% of the total mix doesn't get you any colder, it just leaves undissolved salt sitting in the brine.

⚖️

More on the way

Next for this section: overrun (how much air by volume is actually in commercial ice cream, and why it matters for mouthfeel), and stabilizer chemistry (guar gum, locust bean gum) in more depth.