How To Make Dry Ice

The Short Version

Dry ice is solid carbon dioxide at -78.5°C. You can't really make it from scratch unless you have industrial equipment, but you can compress it from CO2 gas if you have the right setup. The method below assumes you're starting with a pressurized CO2 source, which is what most people actually mean when they ask how to make dry ice. You need three things: a source of pure CO2 gas (liquid CO2 in a tank is ideal), a dry ice sintering machine or a properly set up CO2 snow gun, and insulated containers for storage. I used to run a small deli that needed 200-400 pounds a week, so I built a modest setup from parts and learned the hard way what doesn't work. Here's the actual process. Connect your CO2 tank to the sintering machine or snow gun through a pressure regulator. The tank should be liquid CO2, not compressed gas from a hardware store. Open the valve slowly and let the CO2 flow through the machine. As the liquid expands through the orifice, it flashes into solid snow. The machine compresses that snow into blocks or pellets under hydraulic or screw pressure. What comes out the other end is your dry ice. That's it, mechanically speaking.

The temperature inside the forming chamber drops to around -110°F during normal operation. You'll see frost on everything within a few minutes if you don't pre-cool the equipment. I learned that one the first time I fired up my used snow gun. Took me twenty minutes just to get the first consistent stream of pellets because the steel was warm and absorbed all the initial cooling. I ran water over the housing to speed it up, which was a mistake—the thermal shock cracked a seal. Now I just let it idle for ten minutes before loading it up. Storage is where most people mess up. Dry ice sublimates at about 5-10 pounds per day in a standard cooler. If you want it to last longer, use a styrofoam chest with a tight seal and don't open it. I keep mine in a 50-pound capacity Styrofoam container and it lasts roughly four days before it's gone. Nothing works better than that, and any claim about plastic coolers holding dry ice for a week is just wrong. Polystyrene is about three times better at insulation than polyethylene, and the difference shows immediately. There's a common misunderstanding about making dry ice at home with a fire extinguisher. You can actually get CO2 snow from a standard ABC or BC fire extinguisher if you aim it upside down and trigger it briefly. The discharge tube gets dangerously cold almost instantly. I've done this a few times in a pinch when I needed a small amount and didn't have a tank. The problem is you only get loose snow, not compressed blocks, and you're mixing in trace amounts of extinguishing agent powder that gets in the snow. It's fine for cooling food short-term, but it won't compress properly and it contaminates anything you need clean.

Another thing people don't figure out until it's too late: the moisture content of your CO2 matters. Cheap refill tanks sometimes have significant water vapor mixed in. When that water freezes inside your machine, it forms ice plugs that block the flow. I had a tank once that was barely producing pellets because ice was building up in the nozzle. I had to disassemble the discharge assembly and let it warm up and drain. After that I started checking the tank pressure and weight more carefully before hooking it up. A full liquid CO2 tank weighs about 20 pounds more than an empty one, and if it's significantly underweight it might be old stock with degraded purity. If you're trying to make dry ice on a regular basis and don't want to buy a commercial sintering machine, there's a DIY approach using a CO2 cylinder and a modified fire extinguisher horn. You connect the cylinder to the horn's discharge opening with a valve assembly and trigger rapid releases. The expansion through the horn creates snow that you can scoop up and compress by hand using a hydraulic press or even a heavy bench vise with a custom die. It's slow and labor-intensive, maybe 10-15 pounds per hour at best, but it works for small-scale needs. One detail that costs people money if they don't know about it: when you're filling the snow gun, keep the outlet at least partially open. If you close it completely while the valve is still open, pressure builds in the chamber and can damage the seals or worst case rupture the line. I've seen it happen. The machine didn't explode or anything dramatic, but the seals blew and I was looking at a two-day downtime while I waited for replacement parts. Always have an exhaust path.

Pure CO2 for this process needs to be food grade or at least 99% purity minimum. Lower grade CO2 from welding supply houses often contains trace hydrocarbons and other contaminants that deposit inside your equipment over time. It doesn't affect the dry ice quality much, but it gums up the mechanism and you'll be cleaning it more often than you should have to. I switched to beverage-grade CO2 exclusively and my maintenance intervals went from weekly to monthly. The yield from a standard 50-pound liquid CO2 tank depends on how efficient your setup is. A decent sintering machine will convert about 85-90% of the liquid CO2 into solid dry ice. The rest escapes as gas during the expansion and compression cycle. I usually get about 42-45 pounds of dry ice out of a 50-pound tank. Some machines claim higher yields but the difference is usually marginal. If you just need small amounts occasionally, buying it is almost always cheaper and easier than making it. A 5-pound block runs about $3-5 at most grocery stores or ice suppliers. The equipment cost for a proper setup starts around $2,000 for a basic machine, and that's before you factor in the CO2 tank and any maintenance. Making it only makes sense if you're going to use more than 50 pounds a week consistently.

The main safety concern is straightforward: CO2 displaces oxygen. Never operate a dry ice making setup in an unventilated space. Even with a small tank, the amount of gas released during normal operation can lower oxygen levels in a closed room. I keep a small CO2 detector near my setup and ventilate the area continuously. Anything above 5,000 ppm is a warning, and above 10,000 ppm you need to stop and get out until it clears. Also wear gloves. Not oven mitts, actual cryogenic gloves or thick leather work gloves. Dry ice sticks to skin and will cause frostbite on contact. I've had it happen, even briefly. It stings for a second, then goes numb, and you don't realize how long you've been in contact with it until you're peeling it off. The fingers are the usual culprit when you're grabbing blocks directly. Don't store dry ice in an airtight container. The sublimation pressure will build up and the container can rupture. Use a standard cooler with the lid loosely closed or a dedicated dry ice storage box with a pressure relief vent. I've cracked lids before because I forgot to leave the seal slightly open. The sound is distinctive and you won't forget it.

There's no real shortcut around the physics here. The phase change from liquid CO2 to solid requires that rapid expansion and pressure drop, and getting consistent results depends on maintaining steady temperature and flow throughout the process. Once you have the equipment dialed in and running, it's just a matter of keeping the tank full and the machine cold. The first few cycles are always the rough ones. After that it becomes routine.