How Salt Makes Ice Cream

It started because there wasn't any ice cream in the freezer.

Isla had checked three times. She also checked the drawer where we keep frozen peas, behind the frozen peas, and, for some reason, the fruit bowl. She was seven, it was a heavy Saturday afternoon in August, and she had the specific kind of outrage that seven-year-olds reserve for weather they feel has personally wronged them.

"We could make some," I said.

She looked at me the way she'd have looked at someone trying to sell her a bridge. "You can't make ice cream. You need a machine."

"You need cold. We have cold."

I could see her weighing it. Cold suspicion against the distant possibility of ice cream.

"Show me."

The Setup

We got two freezer bags. The big zip-lock sort. One small, one large.

Into the small one went double cream, a spoonful of sugar, and a splash of vanilla. I sealed it very carefully, because Isla was watching, and I knew that if it leaked she would take it as a personal failing of the entire method.

Then the large bag. Half full of ice from the freezer. On top of the ice, a surprising amount of salt. She was suspicious about the salt.

"Are we eating this?"

"Not the salty bit. The salty bit stays outside."

I dropped the small bag into the large bag, pushed the air out, and sealed it. Then I wrapped the whole thing in a tea towel, which I had learned the hard way was non-negotiable, and handed it to her.

"Now you shake it."

"For how long?"

"About ten minutes."

"Ten."

"Yes."

"Minutes."

"Yes."

She set her jaw and started shaking.

The Shaking

At two minutes, she complained her hands were cold through the tea towel.

At four minutes, she tried to give the bag to me. I said no. She said this was "not really an experiment, this is just work."

At six minutes, she sat on the kitchen floor and shook it between her knees like she was churning butter. I told her the word "churning" and she seemed slightly mollified by having a technical term for what she was doing.

At seven minutes, she said, "I think it's lying to us."

At nine minutes, I told her to stop and opened the outer bag.

I lifted the small bag out. It was heavier than it should have been, stiff in the corners, and when I pressed it, nothing sloshed. It had gone solid. Not rock-solid. Soft-solid. Ice cream.

Isla stared at it.

"Oh."

The Science Bit (Which Happened Naturally, I Promise)

"Why does it work?"

"When ice melts, it gets cold by stealing warm."

"Stealing warm?"

"Ice can't just melt on its own. It has to pull warmth out of whatever's touching it. That's why holding an ice cube makes your hand feel freezing. The ice is taking the warm out of your skin."

"So it took the warm out of the cream."

"Exactly. And when you pull enough warm out of cream, the cream goes hard. That's what ice cream is."

"Then why the salt?"

"Because plain ice only eats warm down to a certain point, then it stops. It gets to zero degrees and says, right, I'm full. But cream needs to be a tiny bit colder than that before it freezes. So plain ice gets close, but never quite there."

"And salt makes it hungrier?"

"Salt makes ice really hungry. Salty ice doesn't stop at zero. It keeps eating warm all the way down past zero. Minus ten, minus fifteen if you pour in loads. So it eats enough warm out of the cream to freeze it."

"So why did I have to shake it?"

"Two reasons. The cream touching the bag freezes first. If you don't shake, that frozen bit just sits there and stops the rest from getting cold. Shaking keeps moving warm cream out to the edge."

She shook the bag. "Plain ice is polite. Salty ice is greedy."

"Now you've got it."

The Testing Phase

She wanted to know what happened if we used less salt. (We tried. The next bag, half an hour later, came out as cold flavoured cream but not ice cream. She declared it "disappointing in a scientific way.")

She wanted to know if rock salt worked better than table salt. (It worked about the same, but rock salt felt more correct to her, which I now understand to be a common scientific bias.)

She wanted to know if she could make strawberry. (Yes. A smashed strawberry and a slightly smaller amount of sugar. It was fine. The colour was more off-pink than pink, but she wasn't fussed.)

She wanted to know if it would work with orange juice. (No, not really. Orange juice without cream just freezes into a block you can't eat with a spoon. I did not have the heart to tell her this is called a sorbet.)

By the end of the afternoon, the kitchen smelled like vanilla and brine and there was a small puddle of salty meltwater under the table that nobody had confessed to. I counted that as a win.

The Next Day

She told her dad at breakfast that she had made ice cream "by making the ice greedy."

He said, "Sorry, what?"

She said, "When you put salt on ice, the ice keeps drinking heat out of everything next to it, and that's how a fridge works too, probably."

He looked at me.

"Thermodynamics," I said.

"Right," he said. "Of course."

Your Turn

You need: a small zip-lock bag, a larger zip-lock bag, about half a mug of double cream, a tablespoon of sugar, a splash of vanilla, three or four handfuls of ice, and a proper amount of salt. Not a pinch. About sixty grams of it. More if the bag is big. You also need a tea towel, because without one, the shaker's hands will stop after ninety seconds.

Put cream, sugar, and vanilla in the small bag. Seal it well. Put ice and salt in the big bag. Put the small bag inside the big bag. Seal. Wrap. Shake.

If it doesn't work, the most likely reason is not enough salt. The second is not enough shaking. The third is a small-bag leak, in which case you now have salty cream, and I'm sorry. If it falls flat in a way you can't place, the usual reasons a kitchen science experiment goes wrong are worth a read.

A note on the shaking: most children will try to hand the bag back at about the five-minute mark. That is more or less exactly when a motivated kid will give up on anything. If you can keep them going another three or four minutes, the bag rewards them for it. It genuinely does.

Try chocolate next. A spoonful of cocoa powder in with the sugar. Try it with a frozen berry already in the small bag. Try it without salt, once, so your child can watch it not work and understand that the salt was the whole experiment. If they get curious about what else salt does to water, you can grow a salt crystal at home over three days and let them name it.

That last one is the most important. The salt is the whole experiment. The cream is just along for the ride. And if the kitchen suddenly looks more like a laboratory than it did this morning, it is worth knowing that there is quiet maths hiding in the dinner, the snow, and the bugs outside as well.