from Hacker News

Why do we salt the ice when making ice cream?

by timrice on 9/21/22, 6:23 PM with 106 comments

• by ortusdux on 9/21/22, 8:51 PM

  Testing for melting point depression is a common diagnostic method used in chemistry to check for impurities. Pure compounds melt at known temperatures, and typically do so within a narrow range (+/- 0.5 °C). Impurities almost always lower the melting point and widen this band. I spent a lot of my undergrad chem courses packing my products into capillary sized test-tubes and watching them slowly melt.

  https://www.mt.com/us/en/home/applications/Application_Brows...

  Some companies leverage this effect to make non-reversible temperature indicators that change color at specific temperatures.

  https://www.mcmaster.com/temperature-indicating-stickers/

• by Taniwha on 9/21/22, 9:09 PM

  Surely part of the issue is that the ice at < 0C while the liquid portion is at 0C (because of the equilibrium thing) - but it's the liquid portion, not the ice, that's most physically connected to the inner container you're trying to freeze (this is the important point).

  If you add ice you reduce the equilibrium temp and as a result the < 0C ice temp can be passed to the liquid phase and as a result on to the inner con tain er where you're making the ice cream

• by CliffStoll on 9/21/22, 9:52 PM

  Now check out Eutectic mixtures ... old-timers may remember soldering with 63-37 tin/lead solder.

  The reason? With any other mixture of lead/tin, the liquid solder freezes over a temperature range, often resulting in what very-old-timers called a "cold solder joint". For example, 50-50 tin/lead mixture starts melting at 183C and is fully melted at 214C.

  Using Eutectic Solder, the phase transition happens at exacctly 183 C ... the lump is solid at 182C and liquid at 184C.

  Geologists take advantage of this: when non-eutectic mixtures of lava freeze (say, a basalt flow in Hawaii or on the moon), different minerals will be found in the rocks. Analyzing the minerals, and assuming equilibrium, you can understand temperatures and pressures in the origination magma.

  (ps - yep, new ROHS rules have largely eliminated lead based solder)

• by lend000 on 9/21/22, 8:19 PM

  > It turns out, yes! What happens is that when the salt is added some of the ice melts – pulling heat from the system – until the temperature has reached the new, lower equilibrium point.

  Correction, or addendum here: the actual dissolution of the salt is an endothermic process, so even if there was no ice, the temperature of water decreases when salt is dissolved.

• by majikandy on 9/21/22, 8:04 PM

  Always heard about this but never tried it. Sounds like fun. Great description, very clear and much better than just saying it lowers the temperature! Nice writing.

• by lesuorac on 9/21/22, 8:22 PM

  I thought it was going to about adding salt to the ice cream but was not :/

  I have a compressor so I have no use of a salted ice bath but I find that using salt in the mixture will make the ice cream not as hard when left overnight or longer in the freezer.

• by refurb on 9/22/22, 3:25 AM

  It's easier to think of a closed system and what the temperature of water vs. ice would be with just a phase change.

  1 kg of ice turning into 1 kg of water requires 333,550 J.

  1 kg of water require 4184 J to warm up 1 C.

  So ignoring all the physical constraints, if you were to turn 1kg of ice "magically" into liquid water, keeping the total energy of the system the same, you'd end up with 1 kg of water at -80C (yes, I know I'm ignoring entropy).

• by 0xbadcafebee on 9/21/22, 11:19 PM

  Or put the simpler way: salt makes the icewater colder than 0C, and cream needs to be about 0F to freeze.

  The ions from the salt get in the way of water molecules aligning to crystallize into ice. When salted ice melts, the water can't refreeze as readily because the saline isn't pure water anymore and because the freezing point is colder. As more ice melts, more heat is absorbed, bringing the temperature down even lower. (https://www.thoughtco.com/how-cold-does-ice-get-with-salt-40...)

• by ricardobeat on 9/21/22, 11:34 PM

  Adding salt to an ice and water mix in a cooler is a known trick in hotter countries. Makes your beer cool down faster, and can get it down to -2C, the sweet spot for light Pilsner.

• by steve_john on 9/21/22, 10:25 PM

  Creating a saltwater slush and packing this around our ice cream base allows us to cool the base enough so that it starts to thicken and freeze before the ice melts completely.

• by nvr219 on 9/21/22, 8:31 PM

  Great post about ice from Rice.

• by Shadowed_ on 9/21/22, 10:43 PM

  I heard explanation few times before but yours is the most clear and simple of them all.

• by fortran77 on 9/22/22, 4:28 AM

  I always wondered why not other things with lower freezing points, like alcohol/vodka? (Not to add in the ice-cream, but to immerse the container that the ice-cream mix is in.)

• by bouncycastle on 9/22/22, 2:37 AM

  Spoiler: "Salt added to an ice / water slurry in an ice cream machine lowers the temperature of the mixture beyond the typical freezing point of water."

• by nielsbot on 9/21/22, 10:53 PM

  Would this work better with calcium chloride instead?

• by quijoteuniv on 9/21/22, 8:19 PM

  I think I finally almost understood it! :)

• by greenbit on 9/22/22, 11:35 AM

  Because we can't get hold of dry ice to do a proper job of it? This seems to vary by state.

• by shultays on 9/22/22, 8:47 AM

  We add some salt to most deserts as well, to make them more tasty (I am no expert so I don't actually know how it makes them "more tasty"). I highly doubt the salt in ice cream is for melting point, and probably it is also for taste reasons. OP is probably overthinking this

• by OJFord on 9/21/22, 10:12 PM

  Alternatively (edit - in the product itself): because it's delicious.

  If you like 'salted caramel' ice cream, try sprinkling some salt on vanilla ice cream. (I bet you'll find it's the 'salted' you like more than the 'caramel'.)