30 Jul 2008

Can dry ice be food?

During our summer holiday in the United States this summer, I found over-the-counter dry ice in the supermarket. A truly marvelous thing! This opens up a host of possibilities, both in the home kitchen and as a science teacher.

Dry ice is in fact frozen, solid carbon dioxide (CO2), the «carbonated» part of carbonated drinks. One fascinating thing about CO2 is that it sublimes rather than melts – solid CO2 does not melt to give liquid CO2, but goes directly over to become gas. Hence, the argument of the producer that dry ice can be used as coolant without getting everything wet. Further, the freezing point of CO2 is -78 °C (-108 °F), and since this is far colder than that of water/ice the cooling power is larger, but you might get frost burns, so be careful (see precautions below).

The producer gives a number of suggestions for how to use dry ice, such as:
  • Cooling: no wet ice – no water, but note that the low temperature results in water condensing on the outside of the container, especially in humid weather. Stays cold longer than water-based ice
  • Freezing: ordinary freezing or flash-freezing (very low temperature gives rapid freezing. Stays frozen longer than water-based ice
  • Make carbonated beverages: add dry ice into still beverages to get a fizzy drink (they have a recipe for home made root beer on their web site)
  • Special effects – fog or witches brew-like effects
The product
The dry ice from this producer comes in solid blocks packed in plain plastic bags. You need to bring some insulated box or bag. Size will obviously vary since the dry ice is stored in insulated containers (using freezer is not of any great help since the ice is far colder than the freezer). Hence, the bag is weighed upon purchase, and the older the dry ice, the greater loss from the bag. This is no problem, as expiry date is no question (as long as it is kept pure and away from other products). As it is meant for use in carbonated drinks, it must obviously be pure enough to ingest (not in solid form however, see below).

My experiences
  • Dry ice in blocks is quite hard, but brittle. Use a short and hard knife (I used an oyster knife), a screwdriver or similar to chop off pieces. It should also be possible to shave off to get a more powdery dry ice.
  • Dry ice curling: chop off a flat piece and leave it on a table or the floor. After a short while it will start floating with almost no friction against the surface, and you can play around with it. Use gloves or touch is only very briefly

Carbonated drinks (the dry ice must be completely dissolved before drinking)
  • Using dry ice for carbonating drinks results in a quite different effect from using SodaStream-type carbonated drinks. One tablespoon of dry ice in a large glass gave a slight tingling sensation on the tongue. Using too much, results in a somewhat bitter taste. Using shavings is better than large chunks, as the latter take a long time to dissolve. The dry ice was either added and stirred with a spoon (i.e. juice) or added after the liquids in the blender for icy drinks, smoothies etc.
  • I tried making icy drinks (slush) using larger amounts of dry ice, but did not succeed. My approach was to crush dry ice together with ordinary ice. This resulted in a layer of rock-hard ice in the bottom of the blender. In addition, I guess this might damage the plastic container of the blender due to over-cooling.
  • Dry ice might very well be combined with ordinary ice. In crushed-ice drinks, add the dry ice after the liquid contents. When using whole ice cubes, mix dry ice with the liquid and add ice cubes. In all cases, stir or blend well after adding the dry ice.
  • Previously, I've tried using a CO2-charged dessert whip for making carbonated apple juice. Squirting the juice out of the nozzle gave a carbonated juice quite similar to the dry ice version (much of the CO2 bubbles out in the process). Martin’s carbonated fruit seem to be quite different, but that may simply be due to the longer waiting time. Obviously, the dessert whip is not meant to make carbonated drinks as there are dedicated soda siphons on the market. Note that Martin also have a few posts that mention the use of dry ice.
  • The cooling effect of dry ice is low, slow and very local when added to a liquid. In fact, water ice seems much more effective for cooling drinks. This is probably due to the escaping CO2 gas whereas water ice adds cold water to the mixture as it melts.
Tested recipes:
  • Dry iced ice tea: works ok, but the bitterness in the tea might be enhanced by the dry ice
  • Carbonated juices (pina colada mix, cherry mix): works ok
  • Smoothies: I tried a strawberry/banana/coconut drink («Coconut Dream») below and it was good with a slight tingling sensation on the tongue

What might be learned (or rather, taught)
There are lots and lots that might be learned in connection with dry ice. A few topics from the top of my head:
  • Gases, liquids and solids
  • Freezing, melting and evaporation/boiling
  • Friction and force (dry ice curling)
  • Acid-base equilibria (see also "naked egg"-resource which is relevant, CO2 is produced)
  • Solubility of gases in liquids
  • Questions about CO2 itself
See also the Q/A section below.


Main point: dry ice is harmless if you avoid ingesting it in frozen state and if you’re careful with skin/eye contact.

Ingestion of solid dry ice may result in severe frost damage due to the low temperature of the dry ice. The same might happen from direct skin contact, but it may be handled if small pieces are thrown back and forth between the hands, minimising the contact time. Textile gloves work well. Upon sublimation the
CO2 increases its volume by 540 times. Ingesting a tea spoon of dry ice would then produce 2.7 l gas in the stomach, which might result in internal injury. Dry ice might damage certain plastic items, varnished material etc., due to the low temperature.


In my vocabulary dry ice might very well fall into the category of food, but this depends on how it is used. Use it as a cooland or reagent in a lab, and it is a chemical. Use it for a somewhat more tingling smoothie, and it’s food.

Finally, I have only explored a few possibilities. I expect that many others are more proficient in finding even more tempting culinary applications than these simple icy drinks and smoothies (as have already been suggested in Martin's and other blogs, follow links and comments at his relevant entries mentioned above).


A few questions and answers
Q: What is the «smoke» coming from dry ice, either pure or in drinks?
A: The fog is not dry ice, but condensed or frozen water from the air (or the drink). The air above the dry ice is cooled to an extent that water in the air condenses or freezes; miniature snow crystals.

Q: Why is expiry date not
a relevant issue?
CO2 is among the most stable compounds around. Hence, it will not deteriorate apart from sublime («disappear»). However, if other products that might give off odour etc. are in the vicinity, these might condense on the surface and the dry ice is increasingly contaminated.

Q: Why do pieces of dry ice float freely on hard surfaces?
A: In contact with a warmer surface, gaseous CO2 is liberated creating a «gas cushion» between the table and dry ice. The same effect being used in hydrofoil boats and what you see if you spill drops of water onto a hot cooking range. As long as gas is produced in the interface, this «gas cushion» results in almost no friction.

Q: Does dry ice float or sink?
A: In water (or other beverages), dry ice will float due to the porous, and thus gas-filled, structure.
CO2, however, is more dense than air and sinks in air (the CO2 will escape only slowly in a glass filled with CO2 gas if it is not stirred)

Q: Is it possible for private persons to get dry ice in Norway?
A: Yes, from producers AGA (retailers, choose «industry dealer» when prompted) and Yara Praxair (retailers) through their network of retailers. However, it seems to be quite expensive, at least compared to the $0.99/lb in US supermarkets.


  1. can you put dry ice into hot soups to cool down,before putting in refrigerator?

  2. In theory, yes. And it's a very good idea because you wont dilute the soup. However, you might end with a somewhat fizzy soup afterwards. It might also render the soup slightly more acidic, but that depends on what sort of soup you make. I guess it wouldn't make much of a difference in e.g. tomato soup, whereas milk based soups might curdle due to the increased acidity (but that probably needs to be tested).

    Secondly, it might be a good idea to use crushed/powdered dry ice and not bits/pieces. Putting bits of dry ice into water mixtures often results in lumps of frozen soup around pieces of solid dry ice. Hence, it might take some time before it is all melted and homogeneous again.


Please add your comments