Category: ice

Josh Colquitt [facebook page] contacted me on Facebook to share some clear ice work he's been doing. He has been filling a punch cooler (aka "beverage dispenser") with water and freezing with the top off inside a big top-loading freezer. 

The cooler freezes from top to bottom via directional freezing, leaving him with a big round cylinder of ice to use. He then carves stuff out of it using power tools. 

He gave me permission to share pictures of his work, including a huge punch/fruit bowl and other objects. I'd never seen anyone use that container before, great stuff. 

All of the ice projects on Alcademics can be found on this page. 

Tony Gonzales is a bartender and fellow ice nerd who had clear ice cube trays made. The brand is called Ghost Ice. 

Well, not really trays, they're inserts that fit into insulated coolers to take advantage of the directional freezing. If you're familiar with the method of poking holes in the bottom of silicone ice cube trays and placing them on the bottom of the cooler, it's the same thing except these trays rest on the top. 

This is a better system because you can adjust the size of the cubes you make in it. You place the tray on top of your cooler and fill it with water. Fill it so that the ice cubes are only two inches high, or all the way up to three. 

Gonzalez originally built the trays to fit the Coleman Party Stacker cooler – a longer cooler meant for holding a lot of beer cans. Some bars have been using this cooler in their bars (if they have walk-in freezers) to make ice via directional freezing and cutting it up with a band saw like pro ice companies do. With the Ghost Ice trays, the insert makes 48 big cubes in about 48 hours. 

However, due to popular demand, Gonzalez cut some trays down for the Ghost Ice For Home model so that they fit into the Igloo Island Breeze that everyone (me first in 2009) uses to make ice blocks at home. These make 20 cubes in about 2 days. 

He sent me a tray to try. I put it in my freezer which at its highest setting is about -9C/15F. I let it freeze for 48 hours and it was just right – the ice had frozen in the trays and just barely underneath the tray. It made a groovy pattern in the ice below it when I pulled out the tray. 

The ice popped out super easily, better than any other tray I've tried. I should also mention that the silicone these trays is made from is super thick and seems like it will last years. 

The thing is that if you use this size cooler to make ice anyway, this insert actually maximizes your ice harvest compared with freezing a block of ice and cutting it up yourself. These trays go down pretty far into the cooler so that you're not wasting a lot of clear ice beneath them, plus when you carve up an ice block you lose a lot of the ice to chips and shards. With the Ghost Ice insert you can use every bit of the ice that comes out of this tray. I wasn't expecting that; a nice bonus. 

Oh and just because I know a lot of y'all are perfectionists and may not like the slightly 'pointed' tips on these cubes, I wanted to let you know you can always flatten ice even. I used a thaw plate (remember the Thaw Master from 80s TV commercials anyone?), but you can use a cookie sheet or any conductive metal. 

The price on these is quite steep – $300 for the large size and $150 for the small, but this is a one-person company and a custom-made product. For a bar with a walk-in freezer or other large freezer space, and a relatively small volume for cocktails to serve each night, it's a great deal over a short time – if you're paying .50 a cube for two-inch cubes from a specialty ice provider (which is low), that's only 12.5 times (25 days) you have to make ice cubes in the Ghost Tray system (plus the cost of the coolers which is low) to make your money back. 

As a home ice nerd like me, well let's not pretend the small model isn't a big splurge. The cubes are awesome though. 

Check out the videos on the Ghost Ice website and Instagram page for more info and demos.  

The index of all the clear ice experiments on Alcademics is here. 

Alcademics reader Andy L commented on the post about how to Make Clear Ice Balls Using a Thermos with an idea: 

Fill the ice ball with water and set the thermos upside-down on top of it like a dunce cap. This way the ice ball mold would still be insulated on the top half: Would directional freezing allow it to freeze from the bottom-up? 

Alas it seems not. I attempted to do this two ways:

1. The ice ball on bottom, empty thermos upside-down on top. 
2. The ice ball on bottom, partially-filled thermos upside-down on top. 

The theory with the second set-up is that the water inside the thermos would provide additional insulation in case the air in the thermos wasn't doing it. 

The set-up:

The reveal. You can see that in the partially water-filled one on the left that the water remaining inside the thermos is frozen. On the right you can see where the water was pushed out the top hole as it froze. 

The results: 

In both set-ups, the cloudy part of the ice was a tornado-shaped column right up the center facing the hole (the sphere on the right is rotated sideways in the picture). So the center was the last part to freeze.

Most likely when the first water froze it floated to the top and plugged the hole.

Alas, it would have been great. We'll add this to the list of "experiments that didn't work" on the Index of Ice Experiments page.  

Today's post comes from the work of reader Richard Newell. We'll call it the "canned water" technique for improving ice clarity. 

We know that the cloudiness in ice comes mostly from trapped air and impurities, which usually are pushed to the center of an ice cube when it freezes. The technique of directional freezing just relocates the air to one end of an ice cube/block rather than eliminates it. [see all the ice posts on Alcademics here]

Many people have attempted degassing the water to remove the air in various ways – mostly by boiling it, but others have tried putting it under a vacuum. As far as I know, nobody has been successful at fully degassing water so that the ice made with it comes out perfectly clear. 

Today's post unfortunately isn't a fully-degassed ice system either, but it does seem to point to a fairly large visible improvement by using water that has been "canned" – using a canning technique but canning water alone appears to improve the clarity of ice made with that water. Newell used an Instant Pot to do the canning, making it super easy. Traditionally-canned water also works. 

Let's start with Newell's conclusions:

• Vacuum is a good (if not essential) method of home degassing of water [in an Instant Pot]
• "Canning" is a convenient way to "pull a vacuum" on water with simple equipment
• A pressure cooker is not required to get good results
• Distilled water makes little or no difference (assuming your starting water is not too bad)
• Boiling (without a vacuum) does not do an adequate job of degassing [without sealing].
• All of my results also depend upon top-down freezing.

The below image is of "canned water" that was cooled and then set into aluminum thermoses to freeze with directional freezing. Note that there's just a tiny amount of cloudy water at the bottom. 

This is compared with non-degassed water in the same thermoses:

We went back and forth in email so this is an edited version of a series of emails and later experiments. Newell says:

I have had success in removing the gas in the water using a combination of boiling and vacuum with various set-ups, some more complicated and labor intensive than others. Then, I discovered this simple and nearly foolproof technique: Use traditional canning techniques and "can" the water in a pressure cooker. Especially with an instant pot type cooker this is dirt simple and doesn't require much labor or monitoring.

Then, I use the top-down freezing technique to make blocks of ice that I then mold into spheres. [The image above is "canned water" frozen inside insulated thermos, so directional freezing from top-down.]

I use filtered water, mainly since the local tap water tastes so bad. I fill some mason jars with the water with a 1 to 1.5 inch head-space, put on the lids and rings "finger tight" only and set them in a water bath in the cooker.

I put them in the electric pressure cooker for at least the 8 minute automatic cycle, and walk away until it completes. When it is done, I release any remaining pressure, remove the jars, and fully tighten the bands.

As you can see, there are no bubbles, and the top-down freezing left almost no minerals in the top section of the ice. The bottom-most part, is of course, cloudy. After a lot of experiments, too numerous to detail, this method is the most simple and foolproof I have found.

I confirmed that Newell was using glass jars, not fully filled with water and not submerged completely in water, and that there is air headspace in the jar when he does it. This is important because my assumption would have been that any headspace in the jar would contain air that would reabsorb into the water. 

The head-space fills with steam during the pressure-cooking, and when the jars cool the steam condenses and creates a vacuum, sealing the lid. So, the head-space is essential to pull the vacuum used to seal the jars during cooling. The lid acts as a one-way valve: during heating the lid lets steam escape (which is why you don't tighten the bands completely before cooking; the jars might explode), and then during cooling the lid gets sucked down to the jar and seals. The drop in pressure actually causes some more of the hot water in the jar to boil as the headspace cools.

Thus, there really is not that much air in the jar to reabsorb… the head-space is filled mostly with water vapor. I think the vacuum may help remove any remaining dissolved air in the water. I typically leave the jars sealed overnight, and freeze the water from them the next day, though I don't know for sure that this delay is necessary. But, it easily fits into my work flow and has become my habit.

I asked why tighten the bands (the lids) at all, and just pour it into the molds at this point? 

That might work, but I haven't tried it. I always let the jars seal and pull an internal vacuum, as I described above. The jars usually seal themselves without tightening the bands, but just to make sure, I usually still tighten them during cooling.

Back when I was experimenting with simple boiling, I tried going direct from the stove-top to the tumbler, but it didn't really help. I suspect that the water, even when boiled for a long time, was never degassed enough. I believe that reduced pressure is an essential step, as the only experiments I have done that succeeded have all used a partial vacuum as part of the process.

I tried boiling followed by a vacuum, and this did work well. One way I did this was with a vacuum sealer (like a Food Saver brand bag sealer) that has a vacuum pump. You can buy an attachment for most of these type devices that fits over a mason jar to pull a vacuum in it without having to heat/boil it (to seal a jar of nuts, coffee, or flour, for example). But, this was a big nuisance and a bit dangerous to use with hot water. I also experimented with valves and other complications.

Then, I hit upon the pressure canning scheme, which boils the water (actually, at a higher temperature, due to the pressure), and pulls a vacuum inside the sealed jars during cooling, and is dirt simple.

Then Newell did some additional experiments: 

• Using distilled water in the same set-up
• Rather than canning under pressure in an Instant Pot, doing the traditional canning technique that would be used to sterilize Mason jars 

I "canned" some water using low pressure (i.e., atmospheric-only, or "no extra pressure") where the boiling point of water is about 212 F. (Note, I am near sea level). This gave essentially the same results as "high pressure" canning using a conventional pressure cooker, or equivalently, an electric pressure cooker like an Instant Pot, that adds about 15 psi and hence water has a higher boiling point (about 250 F.).

Note that I have a grate at the bottom of the pot to lift the jars a bit off of the bottom of the pan. As anyone who has done much canning knows, if you put the jars directly on the bottom of the pan over the heat, the violence of the boiling can break the jars.

I boiled this for roughly 20 minutes, to make sure the water in the jars reached the full boiling point. Then, I removed the jars and fully tightened the bands, and let them cool. 

The results after top-down freezing look essentially identical to when I used the high pressure cooker (see below). Nice and clear with no bubbles except the artifacts at the bottom. 

Traditional Canning Set-up and Resulting Ice:

One more experiment, using the "canned water" technique with distilled water. 

Finally, I also ran an experiment with distilled water. I took a quart of my usual filtered water and distilled a pint of "extra pure" water from it. This I ran through my usual pressure-canning de-gassing process, and then froze it. Here is the result (below). The distilling made no noticeable difference. Still clear in the middle (due to the de-gassing), but still having the same artifacts at the bottom after top-down freezing.

So, you don't need a pressure cooker to de-gas the water. A simple pot and a mason jar will do the trick. Distilling is a waste of effort.

Another conclusion 

My belief is that as long as the jars seal and form an internal vacuum during cooling, the results will be satisfactory.

In conclusion, it appears in this set-up is effective at reducing cloudiness in ice.

Analysis

If the pressure in the canning step in the Instant Pot is not crucial, then perhaps the main factor is that after boiling, the mason jars are sealed. Many of us ice nerds have tried boiling the water and then freezing it (with some improvement in clarity but in my estimation not worth the added effort)

So it appears that the sealed water doesn't reabsorb nearly as much air when kept under a vacuum as it would when cooling from boiling unsealed. I had always figured that water would naturally reabsorb most of the air when it was cooled to come to equilibrium (after being sealed or not), but maybe not!

This makes me want to do a bunch of experiments (or, you know, inspire you to do some experiments and share your results):

1. Repeat this "canned water" method just to verify it works for others
   1. The ideal set-up would be a double-walled aluminum or other deformable container that could be canned and then put into the freezer while still sealed. 
   2. Or I suppose we could just sacrifice a mason jar – can the water and then make an insulated sleeve for the mason jar to encourage directional freezing.
2. Repeat this "canned water" method but put the water in an Igloo cooler set-up like we do for traditional directional freezing to see if it improves ice clarity.
   1. This should be compared with boiled, uncanned water.
   2. It's possible that the large surface area of the cooler would allow for more reabsorption of air into the water as it freezes.
      1. If this experiment fails, it would be interesting to try putting some sort of lid or just a layer of plastic wrap on top of the water to see if it improves clarity. Actually it would be interesting to do this with just regular water/boiled water! 
3. Buy some commercial canned water like this or this,  though I wouldn't really expect them to have been canned at high temperatures. 
   1. And see what happens when your stick it in the freezer. Likely the can would burst as water expands when it turns into ice, but then cut open the can to see the cloudiness of the ice inside. 
   2. Or get it super cold first, then open the can and transfer it to a directional freezing system.

So… long post and lots of homework, but it's been a while since anyone seems to have come up with an improved technique so let's get to work. Thanks much to Richard Newell for doing the work and sharing it with us. 

The index of ice experiments page on Alcademics is here.