Crystal Clear Ice

[Kohai]

I wanted to share an ice-making technique I stumbled upon this weekend. I may not be the first person treading this ground; I'd love to hear other people's thoughts on this.

At the restaurant where I currently work, we are cursed with a shell ice machine that turns out the wettest, crumbliest slush imaginable. The vast majority of our clientele doesn't really know the difference, but for those that do, we keep block ice in an ice cream freezer behind the bar and chip it by hand.

I have been trying to figure out an easy way to make clear block ice - free from bubble clouds or stress/compression fractures that usually result when freezing large blocks of water. I think I may have figured out a way. Furthermore, it's stupidly simple.

Remembering how Milk and Honey double-freezes their ice, I filled a pan with shell ice and poured near-boiling water from the coffee machine over that. Most of the ice melted; there was a thin layer of cubes floating on the surface but I don't think it lasted long. I put that pan into the freezer overnight

When I took that pan out of the freezer, the resulting block of ice was not only nearly crystal clear, it was also free of the deep cracks that result from the pressure of expansion caused by freezing. And it took less than half the time required to freeze, thaw and then refreeze ice.

I thought I'd share this technique and see if others are able to duplicate the results. Anyone care to try this and report back?

[slkinsey]

Interesting. How deep is the pan?

[Kohai]

It's a deep third pan. Perhaps 6 or 7 inches deep, maybe 12 across, maybe 6 wide. Volume is about a gallon and a half. It's still too early to tell if this actually works consistently but it seems to be working with the last few batches I've made.

[Chris Amirault]

Is there a core of nonclear ice in the center?

[Kohai]

OK, so I'm trying to figure this out. I'll put pictures up whenever I remember to get a camera in to work.

Chris, to answer your question, there is still a small cloudy core at the center of the block - sometimes. But not always. I have successfully made at least one perfectly clear block of ice using this technique. I have a suspicion that this is a factor of the ice : hot water ratio, which I've been varying to see what gets me the best results. (My guess is that using more ice creates blocks without the cloudy core - but it might have to sit at room temperature for a little while so most of the ice melts.)

Even when there is a cloudy core, it's much smaller than what I get using tap water or boiled water. You can pretty much see through to the bottom of the pan at most points in the block.

I have tried this method something like six times now and what I can reiterate with certainty is that there are no fracture lines from the pressure of expansion. This is good because those fracture lines affect the way the ice breaks, making clean hand-chipping more difficult.

(I have been told that commercial ice in Japan has a "grain" to it, like invisible fault lines that run through the ice. It causes the ice to split evenly and cleanly. This helps the carver create near-perfect cubes. I'm not sure how that part of the ice production is accomplished.)

[Chris Amirault]

I'm going to give this a try tonight. I don't have crappy shell ice, though, only Tavolo cubes. I'll break 'em up a bit and see what happens.

[Chris Amirault]

So it kinda happened. I used too much ice, so the cubes only melted by about 30-40% or so, and they remained atop the block in a layer. However, below that layer was a pretty damned clear section of ice. I'm going to try it with less ice and see what happens. Very interesting....

[MaxH]

Kohai, this is enlightening, thanks for posting. (I can't help a side thought, wondering if the particular lack of fracture lines you are getting comes at least partly from the particular pan being well suited, thermomechanically, to freezing water.)

But my main thought is that you are, essentially, experimenting with large pure crystal growth. That subject is comercially important with other kinds of crystals, and your experiments recalled something technical that might even be related to what you're seeing, FWIW. The machine you're reading this text on works via electronic logic created on small dice cut from a large, very thin slice of pure silicon crystal. That technology demands extremely low levels of impurities or crystal defects, but otherwise, the situation is the same: wanting a large clean crystal. A classic process to remove defects is annealing (heating, nearly hot enough to melt, lets atoms in crystals re-arrange themselves into a more stable, which tends to be a more regular, or defect-free, pattern). A clever variation is zone refining: a long cylindrical crystal is slowly moved past an induction heater that heats up a "zone," a short part of the crystal's length, to melting or near-melting. As the crystal moves, the locally melted part moves down the crystal's length. It not only anneals defects, but also gathers and "pushes" contaminents out, leaving a huge ultra-pure, defect-free crystal. Food for thought.

[paulraphael]

I used to live next door to a insane person who also happened to be an ice sculptor. His setup for creating clear ice blocks used some kind of circulator that kept the water in motion while it froze i the molds. Somehow this kept air out of the ice and prevented imperfections from forming. His blocks were at least a couple of cubic feet and were clear as glass.

[Badiane]

When I was in catering, we created crystal clear ice by pouring boiling hot water into a metal hotel pan and putting it in the freezer. So perhaps your success has something to do with the fact that you are starting with hot water from the coffee machine?

[Kohai]

Max, thanks for the scientific insight. Those are some very interesting topics for future reading.

paul, Circulators and/or agitation are definitely ways to do it; I've been trying to find a way that doesn't involve special equipment.

Badiane, I'm sure it's not due (solely) to the hot water because I've already tried freezing boiling water and I've never had satisfactory results. If someone can tell me what I'm doing wrong there I'd be happy to hear but I just can't get any significant difference in terms of clarity and fracture lines than with tap water.

Incidentally, in the past few weeks of using this method I've noticed that it is still not perfect. There are occasionally cloudy cores, but they are usually at the bottom of the pan - and they are much smaller. I'm not quite sure what to make of that. It almost seems like if I could just figure out what I was doing wrong, I could eliminate that cloudy core entirely.

OK, so here's a crazy idea. Mold an identically-shaped pan out of silicone - effectively a huge, gallon-and-a-half ice mold (the walls might have to be a bit stronger than usual, not sure). Then try making ice in that. The walls would stretch as the ice expanded. This would avoid fractures in the ice... but would it result in less dense ice? o_O

Edited August 14, 2010 by Kohai (log)

[dougal]

Boiling the water will drive off dissolved gas.

That'd be very good for clarity when you freeze it.

Starting with 'soft' water should mean that you don't get stuff precipitated when you boil. Any such 'dust' would act as multiple nucleation sites, giving a polycrystalline (lots of little crystals) structure, which won't be as clear.

But there's not going to be a benefit to putting hot water direct in to freeze. Or, other than speed of chilling, adding old ice to it.

To grow 'big' (like ice-cube-sized), defect-free crystals, you want to minimise (or eliminate) random nucleation sites. So, no dust or precipitates.

And you want to grow the crystals SLOWLY. So freeze it at a 'high' temperature. Only slightly below freezing point. (Unfortunately, this is not good for production!)

A larger mass of water will cool slower than a small one - hence the success from large pans.

It should also cool more steadily. With a freezer compressor firing up occasionally and then being off for several minutes, the 'coolth' is applied in bursts. I wonder if a "terrine in a bain marie in the oven" approach to steadying the temperature might help. Put your ice tray into a big water (ice) bath. (Thinking that on the second batch, you'd have a block of ice at the right sort of temperature, with a hole the right size for your tray ...)

And that should also minimise temperature upset on opening the freezer door, etc.

You need stable conditions to grow pretty crystals!

If you have expansion-stress problems, maybe try using a stretchy (rubber, silicone ...) mould.

[dcarch]

Pure water without impurities may not freeze at normal freezing temperature. A condition known as "super cooling".

Super cooled water will freeze instantly if disturbed, which may not give you chrystal clear transparency.

dcarch

[Edward J]

I had the opportunity to visit a pro ice carver--this guy had CNC machines to do a lot of his work, and he did a lot by hand as well. He had this one room that looked like something out of the movie "Invasion of the body snatchers". Filtered water was circulated in huge molds, but the freezing only occured in the top 1/3 of the mold. Basically the ice block was frozen from the top down, and any impurities would natually settle to the bottom. When finished, he would bandsaw the bottom 2-3 inches off that contained any cloudiness I think he said it took something like 48 hrs to freeze a block, and he had 6 freezers for the blocks and a huge walk-in for storage.

He would also do "Custom blocks" with stuff frozen in the center of the block, as well as "Bartenders show-offs" where he would freeze a length of plastic hose in the block and then later carve the block into some shape with a funnel at the top, bartender would pour the drink through the block, making a big show.

[eje]

Can't believe no one has linked to Camper English's Experiments in Clear Ice.

Igloo Cooler Ice

Long story short, the most consistent results he has gotten come from something which could be termed "directional freezing".

That is, he places an open insulated cooler in the freezer, forcing the ice to freeze from the top down. The unclear part is isolated to the bottom percentage, which can then easily be cut from the block.

Short of buying giant sculpture grade ice blocks from a vendor, this is the closest anyone has come to consistently, and easily, creating perfectly clear ice at home.

[MaxH]

My past concern with ice clarity was limited to molded cubes, and all I noticed at the time was the importance of water impurities (dissolved minerals -- "hardness" being one form of this). Processing the water first with activated-charcoal water-purifier filters that reduced the local minerals made a noticeable difference. But for the big blocks I think eje caught it:

Camper English's Experiments in Clear Ice. ... the most consistent results he has gotten come from something which could be termed "directional freezing".

That's consistent with some of the separate practical experiences reported here.

Directional freezing may be a practical low-tech form of zone refining for ice crystals. Similar insight comes from a process called something like freeze distillation, which separates water from impurities by partially freezing; the ice tends to be purer water, and the impurities to remain in the part still liquid. In particular, this happens in some German Eiswein where ice crystals removed from frozen grapes concentrate the sugars and flavor elements left behind. (I happen to know from various experiences of the results that this works, as Clive Coates would say, very well, very well indeed.)

[slkinsey]

It seems to me that the best way of doing this would be to have some kind of apparatus that would vibrate the ice tray.

[eje]

From what I've heard, aside from the difficulty of finding vibrators with AC adapters, vibration only works significantly well in combination with the type of layered freezing and/or water circulation done by commercial ice producers.

The beauty of Camper's method is that it requires nothing more than a freezer and an insulated container. Well, and a saw.

[dcarch]

The way fiber glass (fiber-optics)for tele-communication is made with extraordinary clarity is by first subject the molten glass to vacuum to draw out most of the air bubbles, then a very high pressure is applied to compress any remaining bubbles to very tiny size.

dcarch

[bmdaniel]

The way fiber glass (fiber-optics)for tele-communication is made with extraordinary clarity is by first subject the molten glass to vacuum to draw out most of the air bubbles, then a very high pressure is applied to compress any remaining bubbles to very tiny size.

dcarch

I wonder what would happen if you sealed a bag of water in a chamber vac and then froze?

[Chris Amirault]

So I took a crack at a modification of Kohai & Camper's method, only because I found a mini Igloo cooler at Savers the other day. A layer of Tavolo cubes in a bath of hot water, with a closed, not open, top:

Turns out 72 hours doesn't quite freeze everything:

So that's a hole with about 3c of unfrozen water, with a nice casing around it and, especially, below it. It's great trick if you're trying to create a pretty container for for a nice punch, or a big dose of Everclear and Red Bull. Not my plan today, so I broke it up. The "bubbles" are small imperfections on the surface, not in the ice itself:

Cleaved out a rough rock and poured some Talisker over it:

So that's a pretty easy-peasy way to make big chunks of clear ice. Not quite a block, mind you. Perhaps next time I can let it sit for five or six days, see what happens. But in the meanwhile, I'm pretty happy.

[Kohai]

Wow. Wonderful pictoral evidence, sir.

ETA: I just realized - there's no cloudy core! Is it possible that the impurities in the water (as they were called above) are all contained in that liquid center... and could that perhaps mean that the liquid center would become the cloudy core if it were allowed to freeze... meaning that you can avoid cloudiness just by half-freezing a block and draining off the unfrozen water which contains all the impurities and particulate debris?

ETA one last thing: Why did you close the top instead of leaving it open?

Edited August 18, 2010 by Kohai (log)

[Chris Amirault]

Thanks!

ETA: I just realized - there's no cloudy core! Is it possible that the impurities in the water (as they were called above) are all contained in that liquid center... and could that perhaps mean that the liquid center would become the cloudy core if it were allowed to freeze... meaning that you can avoid cloudiness just by half-freezing a block and draining off the unfrozen water which contains all the impurities and particulate debris?

I was wondering the same thing and tasted the water, which seemed unusually mediocre. Should have tried freezing it. However, given that speed of freezing seems to have quite a bit to do with crystal formation, I'm not sure whether that would be conclusive or not....

ETA one last thing: Why did you close the top instead of leaving it open?

I had other stuff in the freezer that I didn't want to warm up from the hot water.

[dcarch]

Wow. Wonderful pictoral evidence, sir.

ETA: I just realized - there's no cloudy core! Is it possible that the impurities in the water (as they were called above) are all contained in that liquid center... and could that perhaps mean that the liquid center would become the cloudy core if it were allowed to freeze... meaning that you can avoid cloudiness just by half-freezing a block and draining off the unfrozen water which contains all the impurities and particulate debris?--------

I am not sure that freezing water can selectively concentrate impurities. Also, it seems to me that the actions of "zone refining" in chip making is different than water freezing. Zone refining uses progressive thawing and refreezing to move impurities in a chrystaline structure.

The cloudy center of an ice block could be:

Water expands as it freezes. Once ice has formed outside a cube and the inside is still liquid, any dissolved air in the water will have no way of escaping. when the interior of the cube is expending, the remaining air bubbles will get bigger and bigger.

May be this will explain why icicles do not have a cloudy center?

Just a unproven theory.

dcarch

Edited August 18, 2010 by dcarch (log)

[Chris Amirault]

Water expands as it freezes. Once ice has formed outside a cube and the inside is still liquid, any dissolved air in the water will have no way of escaping. when the interior of the cube is expending, the remaining air bubbles will get bigger and bigger.

So maybe I should make with the ice pick a couple of days into the process, pour off a little water, and see what happens?