So would this be ChemistryFilter or PhysicsFilter?

June 2, 2005 6:28 AM

What would happen if you take a cube of ice, at a specific size, and fit it in an airthight glass box with the exact same dimensions, and then let it melt?

Since water expands when it's frozen, I'm interested in knowing what would happen if I did this. Would the water fill the entire box, or would there now be a layer of air above it? If it's the first, I take it the glass box would implode if it was thin enough?

posted by cheerleaders_to_your_funeral to science & nature (29 comments total)

You would be left with a cube filled with water and no bubbles if the ices had no bubbles in it. If the ice cube had air bubbles than you would be left with a glass box and the water with the air bubbles that were in the ice cube.
posted by askmatrix at 6:35 AM on June 2, 2005

Assuming there are no gases dissolved in the ice, you'd get a layer of water vapor above the water.

A boiling point is a function of pressure as well as temperature. In a vacuum, water at room temperature will boil, just as a pot of water at atmospheric pressure will boil at 100 degrees Celsius. In both cases, the boiling continues until the system reaches equilibrium -- in your case, until enough water evaporates to bring the pressure of the empty space to the point where the liquid stops boiling.

(Incidentally, water can go directly from solid to gas; this process is called sublimation... but no matter whether there's sublimation or evaporation or both, at room temperature you'd eventually wind up with a cube filled with an equilibrium mixture of water and water vapor.)
posted by cgs06 at 6:35 AM on June 2, 2005

Water and water-ice take up the same volume. That's why the level in a cold drink doesn't change as any ice in it melts.
posted by veedubya at 6:35 AM on June 2, 2005

Water and water-ice take up the same volume. That's why the level in a cold drink doesn't change as any ice in it melts.

No, it's because ice floats (being less dense than water) and therefore displaces the same volume of water that the ice melts down to.
posted by redfoxtail at 6:47 AM on June 2, 2005

Actually, no — water ice at 0 C is less dense than water 0 C. The reason the water level doesn't change when the ice melts is that objects always float at a level such that the weight of the object is equal to the weight of the displaced fluid. As the ice melts, it turns into "displaced fluid", and the net result is that the volume displaced by the ice cube plus the volume of melted water is constant.

If water & water ice had the same density, ice cubes wouldn't float at all — they'd be neutrally buoyant & just hang out wherever they felt like in your glass, top, bottom, or middle.

On preview, redfoxtail beat me to it. Ah well.
posted by Johnny Assay at 6:50 AM on June 2, 2005

Man, you guys are making me really miss Orbitz.
posted by Caviar at 8:33 AM on June 2, 2005

I think you'd just get some water with some air above it. After all, air is what fills in the gaps in the structure of solid ice... right?
posted by sbutler at 8:37 AM on June 2, 2005

sbutler--Not if it's an airtight box, as the questioner specified.
posted by adamrice at 8:49 AM on June 2, 2005

Water with a bit of air and water vapor above it under a slight vacuum.
posted by caddis at 8:50 AM on June 2, 2005

No, I think the gaps in the ice are actually a vacuum. There are no air bubbles in the water when you typically freeze it, the gaps appear because water molecules are polar, they have a positive and negative end. The like charges repel each other as the water freezes and cause gaps to appear. The gaps are a vacuum, not air (I think), and are the reason ice is less dense than water.

I agree with the folks who said water/water vapor.
posted by knave at 9:05 AM on June 2, 2005

When you buy those nice crystal clear ice cube you are buying ice which was agitated slightly during freezing to encourage air bubbles to move up and out of the freezing slush. When you freeze water in your freezer the ice cube looks whitish due to all of the air bubbles tapped inside during freezing. Ice is not less dense to void therein, it is a physical property of water and the way the molecules align themselves upon freezing.

The glass would implode, or at least crack, if it was thin enough.
posted by caddis at 9:14 AM on June 2, 2005

knave, those bubbles are air pockets. Thaw an opaque ice cube in a glass of water and watch the fizzing.
posted by substrate at 9:23 AM on June 2, 2005

Oh. Ok. :)
posted by knave at 9:38 AM on June 2, 2005

air is what fills in the gaps in the structure of solid ice... right?

Aw geez. Your high school chemistry teacher should be ashamed.

As the ice melts, it takes up less space, creating a partial vacuum, which causes some of the liquid water to vaporize to fill in that vacuum and bring it back to equilibrium.

(Basically all those little tiny chunks of empty space in the ice lattice, which were individually too small for any molecules to fit into, merge together as the ice melts -- forming a larger vacuum into which water molecules can now break free as a gas.)

The vapor pressure inside the box is going to at least partially balance out the atmospheric pressure outside it, so I don't think the glass would be too likely to break. But (other than that it depends on the temperature) I don't remember how to calculate this, so I guess my high school chemistry teacher should be ashamed, too.
posted by ook at 9:46 AM on June 2, 2005

The vapor pressure inside the box is going to at least partially balance out the atmospheric pressure outside it, so I don't think the glass would be too likely to break.

Only a little--the vapor pressure of water at room temperature is about 2% of atmospheric pressure.
posted by DevilsAdvocate at 10:06 AM on June 2, 2005

What's cooler is if you try it with dry ice. As it heats up, more of it sublimates into gas, building up pressure... until eventually the gas is too dense to be gas, and poof! You've got a cube full of liquid CO2.
posted by squidlarkin at 10:15 AM on June 2, 2005

H2O has maximum density at 4 degrees centigrade.

"Water is the only substance on this planet where the maximum density of its mass does not occur when it becomes solidified."

I do believe if you designed the box properly, the ice would stay frozen due to the pressure exerted on it. Or it would explode the box.
posted by shepd at 10:32 AM on June 2, 2005

I do believe if you designed the box properly, the ice would stay frozen due to the pressure exerted on it. Or it would explode the box.

No, I believe if you exert enough pressure on ice, it melts. Also, if you remove pressure enough from a closed container of liquid water, it boils (this is why you need to cook things longer at higher altitudes--the water boils at a lower temperature due to the lower atmospheric pressure). So the cube would melt, producing some water with an "airspace" above that was completely filled with water vapor. Four degrees centigrade is above freezing (0 degrees centigrade), meaning that water has maximum density as a cold liquid. On the other hand, if you filled the glass cube with liquid water and then froze it, then yes the box would explode.

Water with a bit of air and water vapor above it under a slight vacuum.

The glass would implode, or at least crack, if it was thin enough.

Um...no.
posted by LionIndex at 11:15 AM on June 2, 2005

I do believe if you designed the box properly, the ice would stay frozen due to the pressure exerted on it.

No way; not if we're talking about melting the ice cube at room temperature. If you look at the phase diagram of water, at 298 K (20 deg C), it's not going to be ice I. It'll be vapor, liquid, or, at very high pressures, some exotic form of ice. Since we're not exerting any pressure on the water, all pressure will come from vaporization. At 20 degC, the vapor pressure of water is 17.5 mmHg, or 0.023 atm (2.3 kPa). This puts us right on the vapor/liquid coexistence line of the phase diagram, of course.

So it'll be liquid with a gap on top filled with low-pressure water vapor.

The pressure on the container is 1.000 - 0.023 = 0.977 atm. The container would need to be pretty fragile not to stand up to that.

No, I believe if you exert enough pressure on ice, it melts

This is true at most pressures and temperatures we experience in daily life. It's interesting to note, however, that it's not generally true. Look at the second phase diagram here. If -20 degC < T < 0 degC and you put pressure greater than atmospheric pressure on the ice, it'll melt (at around 200 MPa, if T = -10 degC). If you continue putting pressure on this water, however, it'll turn into ice V, which has a different packing arrangement and is therefore denser. If it's colder that about -20 degC, you'll never see liquid water: the ice I will turn directly into a denser form of ice. And yes, if it's cold enough, you'll get ice IX.
posted by mr_roboto at 11:38 AM on June 2, 2005

Pressure will melt ice that is near 0°C--or cause it to remain liquid at slightly below 0°C.

I've seen it happen: I left an unopened bottle of water overnight in my car on a cold night. I got to my car the next morning and found that the water was still liquid--until I opened the bottle, releasing the excess pressure (probably a combination of the bottle being slightly pressurized by the bottler to begin with, and pressure from the water trying to expand as it falls below 4°C) whereupon the whole thing instantly froze. Pretty neat to see--at least neat enough to make up for the disappointment that I didn't have a bottle of liquid water to drink!
posted by DevilsAdvocate at 12:04 PM on June 2, 2005

I've seen it happen: I left an unopened bottle of water overnight in my car on a cold night.

It's also possible that the water in your bottle was supercooled, particularly if the bottle was smooth and the water was very pure. Your explanation is probably more likely, however.
posted by mr_roboto at 12:15 PM on June 2, 2005

I think you can assume from the original question that the ice in the box is originally at atmospheric pressure. Given that the questioner knew that ice has a greater volume than liquid water I think than you can also assume that the ice would be brought to that temperature of maximum density for maximum effect, but in any event the density increases when changing phase from solid to liquid at zero degrees C. Since it takes up less space than before and is in a confined volume pressure will decrease to create a slight vacuum relative to atmosphere. Air is trapped in most ice and will be released; the remainder of the space will be water vapor. If no air was trapped in the ice then water vapor only sounds right. Theoretically the pressure of the surrounding air could overcome the pressure inside the glass cube to cause its failure. This might be tricky to actually achieve unless you designed a cube with strong walls and base to retain the water pressure and a very weak upper surface to fail under the pressure of atmosphere. So, um yeah LionIndex.
posted by caddis at 1:46 PM on June 2, 2005

Theoretically the pressure of the surrounding air could overcome the pressure inside the glass cube to cause its failure.

You're talking about plain old atmospheric pressure being strong enough to break glass that contains a gas at slightly less than atmospheric pressure. I think not. Your standard light bulb is a more of a vacuum than our theoretical airspace here, and you don't see many of those imploding. Now, if you're talking about a glass container that's thinner than, say, saran wrap, then yeah you might be correct.

As far as air being trapped in the ice and then there being an air water/vapor mixture above the melted ice, you're correct. But having a vacuum on top, and then the air/water vapor mixture below that? No--the air/water vapor would just expand to fill the vacuum.
posted by LionIndex at 2:49 PM on June 2, 2005

Vacuum typically denotes less than atmospheric pressure not absolute vacuum, at least with the scientists with whom I communicate over this issue. Absolute vacuum is only a theoretical possibility.

The pressure in the gas space above the liquid would be less than atmospheric therefore under slight vacuum. Plain old atmospheric pressure is 14.7 pounds per square inch, over 2,000 pounds per square foot. Depending upon the vacuum in a space this is more than sufficient to break even quite thick and strong glass. You are not going to get that much of a pressure differential, especially given the flexibility of glass which will allow the space to shrink to help balance the pressure.
posted by caddis at 3:36 PM on June 2, 2005

Your standard light bulb is a more of a vacuum than our theoretical airspace here

Incandescent light bulbs are filled with an inert gas (argon or nitrogen) to reduce sublimation of the tungsten filament.

Also, In this case, the pressure inside the cube is not "slightly less than atmospheric." It is 0.023 atmospheres, as calculated by mr_roboto. The faces of the cube need to withstand a pressure of 0.977 atmospheres, and it's entirely possible that they would be too thin to do so. 1 atmosphere is roughly the same as the pressure exerted by a 10 meter high column of water, so if the glass cube can't withstand that pressure, it won't withstand the pressure difference between the atmosphere and water vapor at 20 C.
posted by yarmond at 3:41 PM on June 2, 2005

Plain old atmospheric pressure is 14.7 pounds per square inch

For an ice cube, all we're really talking about is about a square inch. This square inch would take a gradually increasing, evenly distributed load as the ice melted away from beneath it. Glass is brittle and breaks under sharply applied shear forces, but is otherwise a surprisingly strong material. 14 pounds over a one-inch span isn't that much. Saran wrap was overstating it, but this would have to be a seriously fragile piece of glass to have to worry about breaking under atmospheric pressure. Fragile enough that worrying about it isn't really dealing with the original question.

caddis, this is your original statement: Water with a bit of air and water vapor above it under a slight vacuum.

I took it to mean that you're saying there would be a layer of vacuum above the layer of air/water vapor. Did you basically mean "at a pressure of less than one atmosphere"? If so, then we basically agree on this point.

Incandescent light bulbs are filled with an inert gas (argon or nitrogen) to reduce sublimation of the tungsten filament.

This is correct. Whoops. The idea in my head was of the early bulbs, where the main problem in making the thing work was achieving an adequate vacuum.
posted by LionIndex at 5:19 PM on June 2, 2005

There is so much pseudo-science and so little real science as to make this thread worthless.
posted by caddis at 6:04 PM on June 2, 2005

However, my last comment is not directed at mr_roboto who I think is correct in calculating the pressure, assuming a perfectly rigid pressure vessel.
posted by caddis at 7:01 PM on June 2, 2005

Reading this was painful.

A lot of you who answered need to take a look at ook's answer. If your answer diverged from his significantly, either in readability or in content, you need to seriously re-evaluate whether you should be answering any questions here at all.
posted by ikkyu2 at 10:12 PM on June 2, 2005