Updated: 5/25/26 to correct an error in a quote attribution.
This is part 1 of a series on “clear ice” that I’ll be rolling out over the next few months, with posts covering my way to make it at home, the global ice trade (a.k.a., the “frozen water trade”, which happened to ship mostly clear ice), and a look at one of the most famous ice vendors of the 19th century.
The Discovery
One of the most useful kitchen skills that I learned in recent years was how to make “clear ice” at home. Beautifully clear ice elevates your cold drinks, melts more slowly than standard ice1, makes a lovely sound when it taps the side of a glass, can make your drink easier to consume2, and is nice to look at.
But the default state of ice at home is cloudy: cloudy ice is produced by automatic ice makers (50-75% of the volume is cloudy), standard ice cube trays, and the spherical forms you might have seen at kitchen or bar supply shops3. Fortunately, it’s not too hard to make crystal clear ice at home.
My introduction to clear ice was a lucky one. In the span of a few weeks, I saw two articles about the topic: How To Make Clear Ice on Epicurious (sign-up or login might be required), and Clear Ice Cubes in the San Francisco Chronicle (subscription required). Both articles mentioned a blog and book by San Francisco-based Camper English (The Ice Book: Cool Cubes, Clear Spheres, and Other Chill Cocktail Crafts), a former bartender who had become obsessed with making clear ice. Updated (originally the quote was attributed to English, but one of the comments below informed me that the quote was from Jim Blakely of ClearlyFrozen. Clearly a mistake in my note taking), 5/25/26: English’s work inspired other innovators, like Jim Blakely, who is quoted in The Epicurious article:
“For six or eight months, I was fiddling around with various structures of things and making my wife crazy because there was water all over the kitchen and I was taking up a ton of space in the freezer.” The experimentation led him to build a Gorilla Glued gadget out of insulation board from Home Depot, a plastic lid from a grocery store cake, and a silicone ice cube tray with a series of holes punched in it. By December 2017, he had worked with a factory to build his gadget, ordered 2,000 pieces, and started selling them under the brand ClearlyFrozen.
The Way to Clear Ice
Dissolved gasses in water make ice cloudy. As the ice freezes, dissolved gases become too large to fit into the solid water matrix, and some nucleate into bubbles that are trapped between the frozen water crystals.
There’s a lot of folklore about clear ice floating around. English tried them all — boil the water first (which theoretically expels dissolved gases), use distilled water (no impurities), use water from melted ice, and so on4 None of them worked (see English, pp. 3-6).
His many experiments led to a breakthrough: English noticed the even in the most cloudy blocks of ice, the outer layer of was crystal clear. So he thought that perhaps there’s something special about the outer surface that keeps it clear. This led to him filling a hard-sided cooler with water and putting that into his freezer. The resulting blocks were clear on top, cloudy on the bottom. He called it “directional freezing,” where freezing water at the top of the from the top gases are pushed down into the liquid reservoir below.
To illustrate this, I made some simple cross-sectional diagrams of cloudy and clear ice. In the cloudy ice diagram, the center of the ice is cloudy, while there is a thin area of clear ice around the edge that forms as freezing water pushes dissolved gases into the large reservoir of liquid water in the center of the block.

The second diagram shows clear ice made with directional freezing where five of the six sides of the container (the sides and bottom) are insulated. Contact between cold air and the water is limited to the top, so as the water freezes, the dissolved gases are pushed down, ending up in the liquid center5. Once removed from the freezer, you need to drain the liquid and clean up the clear ice block.

My Experiments at Home
Building an Ice Chamber
I didn’t have room in my freezer for a cooler, so I needed some smaller clear ice making devices. I had some plastic containers in the cupboard, about 3″ x 3″ at the base, 6″ tall (7.6 cm x 7.6 cm x 15.2 cm). I also had some soft packing foam (2″ (5 cm) thick), so I hand cut it into five pieces to create an insulation jacket. I connected them by inserting long nails in strategic locations. When assembled, I would press the container into its insulating jacket, which provided insulation on five of the six sides. This design created clear ice, but was it wasn’t easy to press the container into its jacket, and the material was fragile. I also suspect that the foam wasn’t a very efficient insulator, as it was designed to protect goods during shipping, not to reduce heat transfer (perhaps a situation of partially open cells vs. fully closed cells).
I pondered this problem for a while and decided that foam building insulation would be a better material. Not wanting to buy a 8 foot by 4 foot (2.4 m x 1.2 m), I was pleased to find that some hardware stores sell relatively small pieces of insulation (like Foamular Project Panels from Owens Corning). I cut the piece to form the sides and base to cover the plastic container. Instead of nails, I used large rubber bands to keep them in place. The next set of illustrations show 1) how it looks when assembled, and 2) a scale drawing of the five pieces with dimensions in the caption (note that the foam is 1 inch (25.4 mm) thick).



Some Results
For the container that fits the device described above, I typically let the water freeze for 12 to 18 hours, which allows the clear section to take up about 1/3 to 1/2 of the container. It’s important to not let the water completely freeze, as water expands when it freezes and that can break your container (I have a few containers with cracks on their base from a complete freeze).
When the ice is ready, I remove the device from the freezer and move the ice block to a shallow dish and then I crack the thin walls to let the water flow out (into the shallow dish, not on the counter!). Finally, on a cutting board, I use a serrated knife to trim away the thin edges and end up with a rough rectangle. I store the blocks in plastic bags in the freezer.
The photos below show one of my ice blocks on top of a piece of paper with “frozen water” written in several languages. It’s not perfect, as there is a crack through the middle (the result of the “polishing” process, which will be described below), and some roughness on the surfaces reduces the transparency (another polishing issue).


Polishing the Ice
After I trim the ice blocks, I store them in plastic bags in the freezer. Over time, a frosty surface builds up, and although the inside is still clear, it wouldn’t deliver the complete “clear ice experience” if you put it in your drink. But don’t worry, it’s easy to fix this with a process that English calls “polishing.”
Polishing ice doesn’t require abrasives or special equipment — all you need is a clean metal pan (or even a ceramic plate or glass pie pan). Place the ice cube in the pan and move it around. As the block surface warms, the frosty part and other irregularities melt away. Continue until all sides are clear. With care you could also square the sides to make a perfect cube or rectangle, or — with a lot of patience — even shape it into a gem-like shape or sphere.

Problems and Mysteries
“Clear ice at home” is still a work in progress. The freezing part is fine, though I need to start timing the freezes so they don’t go too far and break my container. Draining the water and carving off the scraps is also working well. But I have yet to find the best way(s) to reliably break the large piece of clear ice into the sizes I want. Sometimes the ice fractures neatly at a right angle, sometimes it comes off in jagged shapes. Is this related to the way the ice froze, or where and how deep I am scoring the ice block? Would a better method be to drill holes into the block, much like stone is removed from a quarry? Would a three-prong old-style ice pick work? The “polishing” stage also brings the risk of cracking. Most of the time that I put the block on a metal surface, the foggy outer surface melts away. Other times it melts for a few seconds, then I hear a *crack*, and a crack forms within the block or causes a chunk to fall off. Is this about the piece of ice (e.g., air bubbles causing interior weakness) or the temperature (should I let it sit out for a bit to “temper” before polishing?) or something else?
A mystery: over the summer I purchased two sets of silicone molds that are designed to spherical ice cubes. The upper and lower pieces each have half of the sphere, with the upper piece having a hole in the top where you pour the water. But because there is no place for dissolved gases to go, the spherical ice is cloudy. Could these molds be incorporated into a larger gadget to create clear spheres? My first idea is to cut holes in the bottom of the mold, and then suspend it in a bath of water (which can absorb the migrating dissolved gases). I have my doubts about this: the mold will be an insulator that prevents freezing, and the molds will be locked into a large block of ice.
Have you tried to make clear ice at home? Have you tried any of the commercially available clear ice devices?
Notes
- I haven’t done careful tests, but in my experience the clear ice melts slower. This makes sense as a large cube has less surface area than several small cubes with the same overall volume. Additionally, the cloudy part of a cloudy cube is a jagged air-water complex, with much more surface area, so the outside liquid has much more contact with the frozen water.
- Another thing I like: something I don’t like about the semi-circular ice cubes that my freezer makes is that the cubes naturally align with the curve of the glass and prevent the liquid from getting out of the glass. That doesn’t happen with one large ice cube.
- Something interesting I noticed during a multi-hotel trip I took in August: the two hotel ice makers (Hoshizaki and Manitowoc brand) both dispensed clear ice. To be sure, the ice pieces were irregular in shape and smaller than a cube from a standard ice cube tray, but quite clear. I wonder how they make the ice clear…
- Or “hot filtered water”, which is in the Food & Wine Cocktails 2010 and Food & Wine Cocktails 2012 (ISBN 978-1-932624-43-4) books from Food & Wine magazine: “To make clear ice cubes: fill ice cube trays with hot filtered water.”
- I have a feeling that if you measured the gas content of this liquid section, it would be higher than you would have found in the water before freezing begun.

3 comments
Hi:
This is a great piece!
Two things:
To answer your questions – (a) foam is enough of an insulator to induce directional freezing because the process is actually just creating slightly cooler water at the top that descends in a continuous process, the same way that a lake freezes, which doesn’t require a huge temperature differential and (b) the silicone shell of a sphere mold is not enough of an insulator to keep the water inside from freezing – the arrangement you describe does work.
Camper English is brilliant and is THE PERSON who introduced the world to “directional freezing,” which is the method now in use in EVERY clear ice tray. His excellent book about ice – The Ice Book: Cool Cubes, Clear Spheres, and Other Chill Cocktail Crafts – is available on Amazon. However, the Epicurious quote in your piece is actually from Jim Blakey, the person behind ClearlyFrozen, just one of those many trays that all came about only due to Camper’s pioneering discovery.
Thank you for the comment and correction. I rechecked the Epicurious piece and see that I made a mistake in attribution. It’s been corrected in the post.
Very interested info :) I have big problem to get clear ice without bubbles inside. I need one clear block for kakigori and I have tried 10 different way and always The same result with bubbles inside in the middle …