Appliance Science: In search of the perfect ice cube

Last week I showed you how ice makers work, using a neat mechanism to endlessly create ice for your summer drinks. I touched on the importance of the quality of the water, how clean, clear water produces the best ice. But what effect does this really have? I decided to do some more experiments to find out, to see how I might make the perfect ice at home.

As any mixologist will tell you, creating the perfect cocktail is like creating a work of art, and putting a nasty ice cube in it is like painting a Miley Cyrus t-shirt on the Mona Lisa: an insult to all concerned. This is why fancy bars go to incredible lengths to get hold of the best ice, spending thousands of dollars on delivery ice or high-end ice makers that use all sorts of fun tricks to produce clear, dense ice. Most of us don't have thousands of dollars to spend on delivery ice, so we have to make the best of what we have. That's what I decided to do, to find out what I could do to produce the nicest, cleanest ice in my home.

Firstly, I did a series of tests using water filtered in various ways. First up was standard New England tap water, freshly drawn. I have fairly decent tap water, but it is by no means completely clean. I measured the amount of grunge coming from my tap with a Total Dissolved Solids (TDS) meter, which measures the conductivity of the water. The more electricity the water conducts, the more dissolved stuff there is in there.* This measured about 107 parts per million (ppm) of dissolved stuff, which is pretty normal for unfiltered tap water. I froze this water in a plastic tray in the freezer compartment of my top-freezer refrigerator, letting it sit in the closed refrigerator for 24 hours.

The results, pictured above, didn't look so good. The ice is cloudy and filled with gas bubbles, and has an uneven, bumpy surface. It just doesn't look nice.

Next, I tried filtering the water using a Brita pitcher filter. After this filtration, the water measured about 67 ppm on the TDS meter, a significant reduction. So would this cleaner water produce more attractive ice cubes?

Indeed -- that made quite a bit of difference. The ice is definitely clearer, but it is still cloudy and lacks the glass-like quality that we are looking for. On closer examination, you can see that the ice is cloudy because it is full of small bubbles, formed by gas being forced out of the water as it freezes.

So, onto the next test: using distilled water. I keep axolotls, so I have a reverse osmosis water filter, which uses some fancy chemistry to clean the water that I put in their tank and removes most of the chemicals that can make my fluffy-gilled friends ill. I measured this water as having about 15 ppm of dissolved solids, another huge reduction.

There is a definite improvement there, but I am still not getting the kind of clear, glass-like ice that I want. The ice is much clearer, but there is still that annoying streaky bubbling in the middle that looks unattractive. So, having water with less gunk in it helps, but it won't, by itself, produce the clean ice I am looking for.

After pondering this for a bit, I realized that the problem wasn't the water, but the way the ice freezes. When I put the tray straight into the freezer compartment, the ice freezes from the outside in, meaning that any gasses or minerals in the water are trapped. As the ice freezes, they get pushed toward the center of the cube, creating the familiar star-pattern in the middle. How do you get around this? You rethink the process and change the way the ice freezes, so that the gas bubbles aren't trapped.

Rather than use the ice cube trays, I found a different technique demonstrated in this video below by Better Cocktails at Home, where they produce a large chunk of ice. Here, the ice is made in a plastic container that is shielded on most sides with insulating foil. This means that the ice freezes from the top, but only part of the water is frozen, leaving a reservoir of liquid water for the gas and other impurities to escape to at the bottom.

This changes how the water in the plastic container freezes: because the bottom and sides are insulated, the ice in the trays freezes from the top down, pushing the impurities down as the freezing progresses. To test this out myself, I filled the tub above with water from the reverse osmosis filter. After about 24 hours, I removed the whole thing from the freezer to find that the top half was frozen, while the bottom half was still liquid. After draining the water and removing the ice, I chopped the large ice chunk up into smaller cubes, which looked great: they had the clean, almost glass-like quality I was hoping for.

Success! We have produced a large chunk of clearer, cleaner ice using a bit of ingenuity and some basic science. And we did it in a cheap refrigerator, using simple, inexpensive tools. This technique produces a big lump of ice that you have to chop up to form ice cubes, but that's easily done because the clear ice is nice and brittle, and can be chopped and broken into smaller cubes with some scoring using a bread knife and a chisel.

So, it seems there are two factors to producing clean, clear ice cubes. Water quality is one factor: the cleaner the water, the less likely it is that you will get cloudiness in the finished ice. How you freeze it is the other factor: by controlling the direction in which the ice freezes, you can create clearer ice if you freeze the water in one direction, pushing the remaining impurities out of the way as it freezes.

Science side note no. 1: On a few of the ice cubes I made, I got small ice spikes, a strange phenomenon where spikes poke out of the ice as it freezes. The mechanism for how these are made isn't fully understood yet, but it is thought to be caused by the pressure building up inside the ice as it freezes: sometimes, the liquid water forces its way out and freezes into a spike.

Science side note no. 2: Pure water is a very poor conductor of electricity. Absolutely pure water has a very high resistance to electricity (about 18 megaohms per centimeter) because there are very few ions in it to conduct the electricity. The problem is that water is very seldom absolutely pure: it is such a good solvent that it absorbs gases from the atmosphere, and even sometimes chemicals from whatever you store it in. This is the stuff that makes it conductive, because the ionized dissolved chemicals provide a more ready path for the electrons to flow.