Ice Cream in a Bag: Why Salt Makes Ice Colder

Plain ice cannot freeze ice cream. Put a bag of sweetened cream in a bag of ice cubes, shake for ten minutes, and you get cold cream. Nothing more. Add a third of a cup of salt and the same cream turns to soft-serve in about seven minutes. The ice cream in a bag experiment earns its place because of the gap between those two attempts, and because nearly every explanation a kid hears for it is wrong. Salt does not make ice colder by adding cold. It changes the temperature at which ice and liquid water can sit together, and a salted bath stabilises near minus 15 degrees Celsius while plain ice stays stubbornly at zero.

Key Takeaways

• You will make about 200 mL of real soft-serve in a zip bag surrounded by salted ice, in roughly seven minutes of hard shaking.
• The concept is freezing point depression. For sodium chloride and water the floor is minus 21.1 degrees Celsius at 23.3% salt by mass.
• Best for ages 5 to 12. Younger kids shake. Older kids put a thermometer in the ice and collect numbers.
• The thing that ruins it: brine leaking into the inner bag. Double-bag the cream and press the air out.
• This one is genuinely safe to eat. Every ingredient is food and the sealed inner bag never meets the salt.

What You’ll Need

Item	How much	Where to get it	If you don’t have it
Half-and-half or light cream (contains milk)	240 mL	Supermarket dairy case	Whole milk freezes icy; heavy cream works but churns toward butter
Granulated sugar	25 g	Pantry	Honey, 20 g, though it dominates the flavour
Vanilla extract	2.5 mL	Pantry	Skip it, or 1 tbsp cocoa powder plus an extra tsp sugar
Ice cubes	about 900 g	Freezer	Bagged ice from a petrol station is cheapest
Coarse salt (rock, kosher or coarse sea)	90 to 140 g	Supermarket, or hardware shop ice-melt	Table salt, same mass, but the cold fades sooner
Small freezer-grade zip bags	2, 500 mL size	Supermarket	A jam jar with a screw lid, rolled inside a towel
Large zip bag	1, 4 L size, freezer grade	Supermarket	A 2 L tub with a tight lid
Dish towel or waterproof gloves	1	Kitchen drawer	Oven mitts work surprisingly well
Instant-read thermometer	optional	Kitchen drawer	Without one, skip the data version

Time: prep 5 min, experiment 10 min, cleanup 5 min Cost: everything is probably in your kitchen except the rock salt, which costs about what a bag of chips does Mess level: medium. Salt water on tile is slippery, so work at the sink or over a tray.

Do the Experiment

Step 1 — Mix the base in the inner bag

Pour 240 mL of half-and-half into one small zip bag. Add the sugar and vanilla. Seal it with as much air squeezed out as you can manage, then massage it until the sugar stops feeling grainy through the plastic. An inflated bag touches less ice and freezes slower.

Step 2 — Double-bag it

Put the sealed bag inside the second small bag and seal that too. This feels redundant and it is the most important line in the article. Zip seals fail under hard shaking, and a teaspoon of brine ruins the batch.

Step 3 — Build the ice bath

Fill the large bag with roughly 900 g of ice. Add the salt on top, then nestle the double-bagged cream into the middle so ice surrounds it on all sides. Seal the large bag with most of the air pressed out.

Step 4 — Read the starting temperature

Slide a thermometer probe into the ice beside the inner bag. Plain ice and water reads about 0 degrees Celsius. Within a minute of the salt going in it should be heading down through minus 5.

Step 5 — Shake for seven minutes

Wrap the big bag in a dish towel or put gloves on. Then shake, squeeze and knead continuously, not gently. The cream has to keep moving so fresh liquid touches the cold plastic. Pass the bag between people every minute, because hands get uncomfortable fast.

Step 6 — Rinse the bag, then open it

Check the texture at five minutes and again at seven. When it holds a spoon shape, rinse the outside of the double bag under cold tap water and dry it before unzipping. Salt crystals clinging to the outside otherwise fall straight into the ice cream.

What you should see: The bath drops to between minus 10 and minus 18 degrees Celsius within two minutes of salting, and the outside of the bag frosts over as room moisture freezes onto it. The cream thickens in stages rather than all at once: thin, then like a melted milkshake at three minutes, then scoopable between six and eight. Expect about 200 mL, closer to soft-serve than to a hard-frozen tub. Shake for 15 minutes and it turns firmer and slightly grainy.

What’s Actually Going On in the Ice Cream in a Bag Experiment

Freezing point depression is the whole mechanism. Any solute dissolved in water lowers the temperature at which that water freezes, and raises the temperature at which ice melts into it. Not because the salt is cold. Because dissolved sodium and chloride ions get in the way of water molecules trying to lock into the ordered lattice of ice. The liquid becomes a more comfortable place for a water molecule to sit, so ice has to be pushed colder before it can hold onto its own molecules.

The sequence inside the bag runs like this. Ice carries a microscopically thin film of liquid water on every surface, even a couple of degrees below zero. Salt dissolves into that film and makes brine. The ice is now touching brine whose freezing point is well below zero, so the ice melts. Melting costs energy: 334 joules per gram converted from ice to water. The system takes that energy from the nearest available heat, which is the ice, the brine, the bag and the cream. Temperature falls until melting and freezing balance again at a new, lower equilibrium.

How low? For sodium chloride and water the limit is the eutectic point, minus 21.1 degrees Celsius at 23.3% salt by mass. Robert Bodnar’s 1993 revision of the freezing point depression equation for water and sodium chloride gives the curve between those extremes, and Potter, Clynne and Brown had tabulated it in 1978. A kitchen bag, with warm hands around it and cubes rather than crushed ice, realistically reaches minus 10 to minus 18. That is plenty, because the base does not freeze at zero either. Dissolved sugar depresses its freezing point to roughly minus 2 or minus 3. Which is exactly why plain ice fails: zero is not cold enough to freeze sweetened cream.

The popular version says salt makes ice melt faster, and melting is cold. Half right, in a way that blocks understanding. Faster melting would not help if the equilibrium temperature stayed at zero. The trick is that the equilibrium temperature moves. A good test of whether a kid has it: ask why a salted road at minus 5 stays wet, and why salting a road at minus 25 wastes salt.

The shaking does separate work. Goff and Hartel, in the standard reference text on ice cream manufacture, describe commercial freezing as simultaneous freezing and whipping: a chilled barrel wall, a blade scraping thin frozen layers off it, a dasher beating air in. Your bag is a crude version of the same machine. Small crystals plus incorporated air is what “creamy” physically means, and both come from agitation. Leave the bag still and you get a grainy puck.

How to Explain It to Your Kid

Ages 5-8: Ice has a rule, and salt breaks it

Tell them ice follows one rule: it turns to water at zero degrees, and it will not get colder than that while it is still melting. Salt breaks the rule. Run the plain-ice version first for two minutes so they feel that nothing happens, then add salt and let them touch the bag ninety seconds later. The lesson lands through their hands, not through the word “depression.”

Ages 9-12: Freezing point depression, with a number

Give them the term and the number: salt water can stay liquid down to minus 21 degrees Celsius. Then ask them to work out the consequence. If plain ice sits at 0 and ice cream mix freezes at minus 2, how could plain ice ever freeze it? It cannot. That is a real deduction, and the kind of reasoning that makes chemistry feel solvable rather than memorised. Let them weigh the salt so 23.3% means something physical.

Ages 13+: Colligative properties and why only the count matters

Freezing point depression is a colligative property: it depends on how many dissolved particles are present, not what they are. Sugar depresses the freezing point too. Sodium chloride is efficient because each formula unit dissociates into two ions, so one mole of salt delivers roughly two moles of particles. Ask them to predict whether calcium chloride, which splits into three ions, does better. It does, which is why airports use it. Then measure the bath at 0%, 10%, 20% and 30% salt by mass and plot it.

The question to ask: “If I put a cup of pure water in the freezer next to a cup of salt water, which freezes first, and what happens if I leave them both overnight?”

Why Teaching This Early Pays Off

Freezing point depression arrives in school chemistry under colligative properties, usually around ages 14 to 16, and students find it arbitrary because it shows up as a formula with a constant in it. A kid who has already felt salted ice get cold enough to ache has something physical to attach that formula to. It also pre-loads a larger idea: dissolved things change the properties of the solvent. That is the doorway to osmosis, to boiling point elevation, and to why seawater does not freeze when a lake does.

The honest version of the learning claim is narrower than the enthusiasm usually suggests. One afternoon of ice cream does not change a child’s trajectory in science. What the evidence supports is specific: Alfieri and colleagues, reviewing 164 studies in 2011, found that unassisted discovery learning performs no better than direct instruction, while guided discovery, where an adult explains alongside the activity, beats both. That is the difference between handing a kid a bag and walking away, and running the plain-ice control together and asking why it failed.

The concrete payoff: a kid who has done the control will never again accept “salt makes ice colder” as an explanation. That instinct to ask how travels, and it is the same one that makes the difference between a chemical and a physical change worth understanding rather than memorising.

Before You Start: What Parents Should Know

Best age	5 to 12. Under 5, hands get too cold to shake long enough and the result disappoints. Over 12, run the thermometer version or it is just a snack.
Supervision	Light. An adult measures the salt and rinses the bag before opening. Kids do the shaking.
Real hazards	Prolonged skin contact with a minus 15 degree brine bag can cause frostnip, so use a towel or gloves and swap hands every minute. Spilled brine on tile is slippery. Contains milk, an allergen.
Don’t	Never substitute dry ice for the salted ice. It sits at minus 78 degrees, burns skin on contact, and releases carbon dioxide that must not be trapped in a sealed bag.

On eating it: yes, this one is safe to eat, which is unusual for a science experiment. Everything in the inner bag is food, the bag stays sealed, and salt never contacts the cream if you double-bag and rinse. Use pasteurised cream, which is what supermarkets sell. If the inner bag leaks, bin the batch rather than tasting it to check.

When It Doesn’t Work

• It tastes salty. The inner bag leaked, or salt fell in when you opened it. Double-bag, press the air out before sealing, and rinse the outside under the tap. There is no rescue for a leaked batch.
• It never thickened past milkshake. Not enough salt, or not enough ice. Aim for 10 to 15% salt by mass relative to the ice: 90 to 140 g per 900 g. Check that ice surrounds the inner bag rather than sitting under it.
• It froze into a grainy block. You shook too long, or stopped partway and let it sit. Agitation keeps crystals small. Knead the bag for 30 seconds and it will loosen.
• It turned into butter. Heavy cream at 36% milkfat or more, shaken for 12 minutes, coalesces. Switch to half-and-half, shorten the shake, or lean into it and read our guide to churning cream into butter.
• The ice melted in three minutes. Crushed ice plus fine table salt in a warm kitchen. Coarse salt and cubes hold the cold far longer.

Take It Further

Vary the salt fraction and plot it. Build four baths with 900 g of ice each and 0 g, 90 g, 180 g and 270 g of salt. Record the bath temperature at 30, 60 and 120 seconds with the same thermometer, then plot temperature against salt percentage by mass. The curve flattens near the eutectic, which is a real feature of the phase diagram rather than sloppy measurement.

Change the solute, not the amount. Compare 90 g of table salt against 90 g of sugar against 90 g of calcium chloride ice-melt pellets, each in 900 g of ice. Record the lowest temperature reached. The ranking tells you about particle count per gram.

Measure freezing time against fat content. Run three inner bags with whole milk, half-and-half and heavy cream, identical sugar, identical shaking. Record the minute each holds a spoon mark, then rank them for texture blind. Fat slows freezing slightly and improves mouthfeel a lot; separating those two effects pairs well with the science of why whipped cream holds its shape.

What to Watch For Over the Next 3 Months

• Week 4: They mention salt on roads or sidewalks in winter, unprompted, and get the direction of the effect right.
• Month 2: A phrase like “it lowers the freezing point” appears in an unrelated conversation, about why the sea does not freeze or why you salt pasta water. The pasta case is a useful trap: salt shifts the boiling point by a fraction of a degree, not meaningfully.
• Month 3: They propose their own variable without being asked. Sugar instead of salt, two bags at once, a stopwatch. Designing a comparison is harder than following a procedure, so that is the signal worth encouraging.

Frequently Asked Questions

Can we use table salt instead of rock salt?

Yes, by mass rather than volume. Table salt packs denser in a measuring cup, so a third of a cup of it is considerably more salt than the same volume of rock salt. Weigh out 90 to 140 g. Table salt dissolves faster, so the bath cools quickly and then warms sooner.

Why does my ice cream come out icy instead of creamy?

Two causes. Not enough fat, if you used skimmed or whole milk, which leaves more free water to form large crystals. Or not enough continuous agitation, which lets crystals grow. Half-and-half plus nonstop shaking is the reliable combination.

Is it safe for my kid to eat?

Yes. All the ingredients are food, the inner bag stays sealed, and the salted ice never touches the cream. Use pasteurised cream, rinse the outside of the bag before opening, and discard the batch if the seal failed.

Should I worry about frostbite from the cold bag?

A good bath runs minus 10 to minus 18 degrees Celsius, cold enough to make bare hands ache within a couple of minutes and to cause frostnip over longer contact. A dish towel or gloves solves it completely. Swap hands every minute.

My kid wants to watch a video of this instead. Is watching enough?

Watching teaches less, and the research here is counterintuitive. Deslauriers and colleagues found in 2019 that students in actively engaged classrooms learned more while feeling they had learned less than students who watched a polished presentation. Feeling unproductive is not evidence of being unproductive.

Can we use a jar instead of bags?

Yes, and it leaks less. Put the cream in a small screw-top jar, set it in a tub of salted ice, and roll it rather than shake it. Glass slows heat transfer, so allow a few extra minutes. It can be rolled along the floor, which younger kids much prefer.

---

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. Bodnar, R. J. (1993). “Revised equation and table for determining the freezing point depression of H2O-NaCl solutions.” Geochimica et Cosmochimica Acta, 57(3), 683-684. https://doi.org/10.1016/0016-7037(93)90378-A
2. Potter, R. W., Clynne, M. A., & Brown, D. L. (1978). “Freezing point depression of aqueous sodium chloride solutions.” Economic Geology, 73(2), 284-285. https://doi.org/10.2113/gsecongeo.73.2.284
3. Goff, H. D., & Hartel, R. W. (2013). Ice Cream, 7th edition. Springer. https://doi.org/10.1007/978-1-4614-6096-1
4. University of Guelph Department of Food Science. “Ice Cream Manufacture.” Dairy Science and Technology. https://www.uoguelph.ca/foodscience/book-page/ice-cream-manufacture
5. U.S. Food and Drug Administration. 21 CFR 131.150, “Heavy cream.” Code of Federal Regulations. https://www.ecfr.gov/current/title-21/chapter-I/subchapter-B/part-131/subpart-B/section-131.150
6. Alfieri, L., Brooks, P. J., Aldrich, N. J., & Tenenbaum, H. R. (2011). “Does discovery-based instruction enhance learning?” Journal of Educational Psychology, 103(1), 1-18. https://doi.org/10.1037/a0021017
7. Deslauriers, L., McCarty, L. S., Miller, K., Callaghan, K., & Kestin, G. (2019). “Measuring actual learning versus feeling of learning in response to being actively engaged in the classroom.” PNAS, 116(39), 19251-19257. https://doi.org/10.1073/pnas.1821936116