Exploring the Liquid-to-Solid Transition of Water

[mikemense]

I am new at this and apologize if this is an inappropriate question per the rules. I hope it is not but I imagine it might be. In all cases, maybe this goes without saying, but in all cases I am looking not only for the answer to my question but also the basis of the answer.

I am interested in the very specific event of liquid water turning to ice.

Is there any scale at which this is an instantaneous event?
Is there a transitional state between liquid water and ice?
I believe that (at least one of) the current understandings of nuclear physics includes certain discontinuities. Is that true? If so, is that understanding of discontinuity related in any way to what might be a discontinuity that might exist (over time I guess) between water in a liquid state and water in a solid state?

Is there something of a critical mass? Do there need to be a certain number of molecules prepared to make the transition before the transition will occur? I guess one molecule of water can't freeze because the act of freezing is a process of entering into a fixed crystalline arrangement with other molecules?

How long does it take for water to freeze? Once it has arrived at the appropriate temperature, is there a formula that tells us how long it will take to convert liquid to solid, given the rate of removal of energy?

 

[Mapes]

Hi mikemense, welcome to PF. These are some great questions.

Is there any scale at which this is an instantaneous event?

Is there something of a critical mass? Do there need to be a certain number of molecules prepared to make the transition before the transition will occur? I guess one molecule of water can't freeze because the act of freezing is a process of entering into a fixed crystalline arrangement with other molecules?

Your intuition is correct about a critical mass; it is usually called a nucleate or a cluster, and the process is called nucleation. There is actually another reason why ice doesn't form two or three atoms at a time: there is an energy penalty for creating the surface between ice and water. But there is an energy benefit to freezing at temperatures less than 0°C. And above a certain nucleate size, the volumetric energy benefit to freezing overcomes the surface energy penalty, and ice forms. (The energy benefit wins out because the volume increases in an expanding sphere faster than the surface area.)

The critical mass happens randomly and occasionally as the water molecules move around due to thermal motion. If the temperature is above 0°C, the molecules disperse again; if the temperature is less than 0°C and the cluster is large enough, the molecules arrange themselves into crystalline ice.

This is a kinetic process, so freezing is never perfectly instantaneous. However, the time delay decreases with decreasing temperature, and there is a temperature at which freezing is practically instantaneous. It is called the homogeneous nucleation temperature.

 

[mikemense]

Thank You
Mapes
I like some of this very much.
I like the role of geometry, the increasing ratio of volume to surface.
But I have more questions.

What does kinetic mean in this context?

Is that homogeneous nucleation temperature -42c?

Can you explain more about the energy price of creating a boundary? Where is it spent?
How does energy maintain a boundary?

Let me put this in my own words. The liquid state molecules (all more or less at 32f) are bouncing around (brownian movement?) and at some point they find themselves in pretty much the same position that they would be in if they were ice and they all decide to give it up and relax and stop bouncing around and form a piece of ice. In the process of doing that they give up a bunch of energy (in the form of infrared waves?). Does that energy somehow create the boundary? Or does that energy heat the adjacent water.

Having read Gould (I think it is) I am familiar with the "one side boundary" or whatever it is, so I understand that over time, all of the water will find itself randomly in the right position and all will turn to ice. But when ice forms next to ice, why doesn't it melt the adjacent ice with the big energy release?

That's enough for now.
Thanks again.

 

[Mapes]

What does kinetic mean in this context?

This means that a rate is involved, that everything doesn't happen at once. Kinetics is the study of rates. For a crystal like ice to grow, a water molecule has to randomly move to the right position and attach in the right position, and the energy given off by the phase change has to diffuse away. All this takes time, and it's the reason why ice cubes don't form instantaneously in the freezer (and why the freezing scenes in the movie The Day After Tomorrow were absurd).

But when ice forms next to ice, why doesn't it melt the adjacent ice with the big energy release?

Again, good intuition. That thermal energy must be removed for freezing to continue if you're right at the melting temperature. But if the water is supercooled, waiting for a cluster to assemble, then the ice-water boundary can move quite fast because the water is cold enough for the energy release to have little impact. See http://youtube.com/watch?v=bDwZqBq-rLQ".

Is that homogeneous nucleation temperature -42c?

Plus or minus around 10°C. By convention, this temperature is the point where nucleation happens quickly in your sample--but the size of your sample and the definition of "quickly" is up to you.

Can you explain more about the energy price of creating a boundary? Where is it spent?
How does energy maintain a boundary?

The idea here is that the inside of a crystal is in a low-energy state. But surfaces are unavoidable defects that contain huge numbers of higher-energy bonds because the crystalline arrangement stops. So it's not like there's a special energy there, more like the unsatisfied bonds are at a higher energy level than fully satisfied bonds. We have to account for this difference when we consider nucleation.

As a crystal grows, though, the surface energy penalty becomes negligible compared to the volumetric energy benefit. Think of a cube of 27 atoms; 26 are surface atoms that don't get the full benefit of being nicely packed in a crystal. But in a cube of 1,000,000,000 atoms, there are only about 6,000,000 surface atoms, a much smaller fraction. The surface energy is only competitive when the first few atoms or molecules happen to gather together into a cluster.

 

[mikemense]

This means that a rate is involved, that everything doesn't happen at once. Kinetics is the study of rates.

If it's necessary to have a term for something that has a rate involved, what are things called that don't have a rate involved? Please give an example of something that is not kinetic (and therefore happens instantaneously?).

I am understanding then that the rate of freezing is a statistically predictable phenomenon.
But when a specific molecule will freeze is not predictable? So the anti-commonsensical characteristics of quantum theory have moved up at least to the scale of molecules? Or is this a sophomoric analogy? (I ask this because i have noticed in my own life that even at the scale of humans, things are only statistically predictable. At the scale of planets and beyond, it seems specific events can be predicted, but then they are really just an aggregate of statistically predictable events, I suppose.)


Mapes said See http://youtube.com/watch?v=bDwZqBq-rLQ".

The videos are great (mostly). Thanks! I assume the reason we get slush is because of the released phase change energy is still adequate to keep some of the water from freezing?

Why can you roll the supercooled water around but a sharp rap starts the crystallization?
I think its a little weird that the difference in the characteristic of the disturbance that we (humans) understand seems also to "have meaning" for the molecules. Or is this statistical again -- sometimes the rolling around will start the process because randomly the molecules will find themselves in the right place -- but the rap creates a much greater set of motions and so statistically it is much more likely that the molecules will find themselves in the right place. It is possible to rap the bottle and have nothing happen, I bet.

Mapes said, The idea here is that the inside of a crystal is in a low-energy state. But surfaces are unavoidable defects that contain huge numbers of higher-energy bonds because the crystalline arrangement stops.

I am understanding this to mean that it just takes more energy to maintain the boundary.
Well, duh! (I am sure it is appropriate for me to apologize for my inability to avoid making wildly inappropriate analogies and metaphors across scales.)

 

[mikemense]

H two O singles start club

This occurrence of water molecules randomly finding themselves in the right places to form an ice crystal (is that a careful use of the term? Is "pure ice", whatever that might mean, really crystalline? -- but that's a side issue, please let's not get distracted) with a surrounding environment that is very ready to accept the released energy -- this all makes great sense to me. But I want to shorten the interval, if that's possible.

Has it been possible to watch this process microscopically? Do two molecules bond and then others join, one at a time? Or does this group that has a good enough ratio of volume to surface area -- do they all jump at once? Have we seen this? Can we slow down the video? (again, I apologize for my analogies) -- But if it happens as a group, do they all really do it simultaneously. Is it even appropriate any more to ask about the temporal duration of the transition. If the transition has duration, then there is something between "water" and "ice". What is that? Now, I am not talking about the big environment, the H[2]O in the bottle, no, I am instead talking about either the couple that bonds or the group that bonds. If there is no duration, then don't we have an instantaneous discontinuity?

I hope somebody is with me on this.

 

[DaveC426913]

Conceptually, a molecule doesn't "stop moving and then give up its energy", conceptually, a molecule "has its energy stolen from it through collision with another molecule, and, once robbed of that energy, it is moving slower, and able to be captured by the bonds of other molecules".

If you play with billiard balls for a while, you will discover that a fast moving ball will richochet off another ball, and in doing so, transfer most of its energy. The original fast-moving ball is often stopped dead in its tracks, while the slower one is often propelled away at high speed.

Now, what if, everytime that fast-moving ball hit a rail, rather than letting it continue to careen around the table, you stopped it and put it back on the table moving very slowly? Eventually, the table would be covered in near motionless balls.

Now think of the balls as atoms of water and the pool table rails as the surrounding air and/or ice cube tray. They bleed the energy out of the system by stealing the molecules' energy.

The balls can careen around in the centre of the table as long as they want; they will lose energy only at the rails. And they do so one ball at a time.

 

[mikemense]

dave, thanks
Let me see if I can straighten this out in my own mind. The billiard ball model makes perfectly good sense to me as a way to describe the homogenization of the temperature of a quantity of water. In every collision, the slower moving molecule relieves the faster moving molecule of energy (I think we should keep the moral aspects of stealing out of this). OK
But we are talking about a phase change occurrence. And you are suggesting, I think, that the metaphorical description (the suggested physical analogy) of this event is that all of the energy goes to the slow moving ball (molecule) because the fast moving ball just stops in place (held in place apparently by some sort of bond that develops between it and the adjacent also frozen water molecules). And, plausibly to me at least, that previous slow moving now energy overloaded molecule immdiately starts relieving itself with all the nearby still free floating molecules.

Why does it happen sometimes that heat transfer interactions occur and sometimes that phase change interactions occur?

Thanks, and hope to hear from you again.

 

[mikemense]

heat transfer media

In my work, we talk about three forms of heat transfer, conduction, convection and radiation. I've always understood that, scientifically, this is a messy list. Convection doesn't really belong on this list, it's just a larger scale example of the other two.

But I have continued to believe that conduction and radiation are two different things. Now, I am wondering. Heat, at the scale of molecules it seems to me, is just motion. Is that just a useful metaphor or is it understood to be true in some Platonic sense? Conduction of heat is the exchange of momentum? If that's true, in this context, what are infrared waves (or rays, are rays and waves the same thing?)? Does radiation, distinct from conduction, occur between molecules?

I think the following is another way of asking the same question.
One could almost argue that phase changes are only a contingent apparency (I hope that's a word) resulting from the specifics of sensory processes (applying apparently to all forms of life on earth). Except for those huge energy spikes that seem to be required for phase change.

 

[DaveC426913]

And you are suggesting, I think, that the metaphorical description (the suggested physical analogy) of this event is that all of the energy goes to the slow moving ball (molecule) because the fast moving ball just stops in place

The fast-moving balls throughout the water will slow or stop or reverse or whatever, depending on initial conditions. This all averages out in the middle of the water sample, since no heat will leave or enter the system. But: the key is that, at the interface between the water and its container, you are preferentially removing energy from the system by letting the water molecules transfer their energy to molecules of air or metal, which are cooler (thus, slower) than the water molecules. This energy is not returned to the system. It is transferred to the coolant in the refrigerator and exhausted elsewhere.

(held in place apparently by some sort of bond that develops between it and the adjacent also frozen water molecules). And, plausibly to me at least, that previous slow moving now energy overloaded molecule immdiately starts relieving itself with all the nearby still free floating molecules.

The bond between water molecules that causes it form form a lattice is weak; only slow-moving molecules will succumb to it. Anything faster will not be held.

Why does it happen sometimes that heat transfer interactions occur and sometimes that phase change interactions occur?

Indiividual Water molecules in any sample are always slowing and bonding into a lattice or speeding up and flying off (evaporation).

The phase change occurs when all the molecules are slowed down through energy loss. The reason it occurs at a fairly discrete temeprature is becasue water is a good thermal conductor. Cool one part and the entire sample of water acts like a heat reservoir, meaning that one part will easily steal energy from the entire sample rather than freezing. Thus, the entire sample will reach freezing (or boiling for that matter) at nearly the same time.

 

[mikemense]

The fast-moving balls throughout the water will slow or stop or reverse or whatever, depending on initial conditions. This all averages out in the middle of the water sample, since no heat will leave or enter the system. But: the key is that, at the interface between the water and its container, you are preferentially removing energy from the system by letting the water molecules transfer their energy to molecules of air or metal, which are cooler (thus, slower) than the water molecules. This energy is not returned to the system. It is transferred to the coolant in the refrigerator and exhausted elsewhere.



The bond between water molecules that causes it form form a lattice is weak; only slow-moving molecules will succumb to it. Anything faster will not be held.

What is the nature of this lattice?

Indiividual Water molecules in any sample are always slowing and bonding into a lattice or speeding up and flying off (evaporation).

Are you saying that in any sample of water at, say, room temperature, statistically ignorable numbers of molecules are forming an unstable quantity of ice, which does not survive because the environment is not supportive, and at the same time, another statistically ignorable number of molecules are evaporating? I think this is a pretty cool idea. But its surprising to me. Phase change requires a large amount of energy, at least as I understand things, so it is surprising to me that this sort of energy transfer is happening and unhappening (if you will) randomly and pretty much constantly. Under this theory, if that is what this is, ice is happening all the time, but it only happens for us when the temperatures are suitable. As an alaskan, I am tempted to point out that this is analogous to how glaciers form. Glaciers are formed pretty much everywhere in alaska every year, but 99% of them don't survive to the second winter, so we don't call them glaciers.

The phase change occurs when all the molecules are slowed down through energy loss. The reason it occurs at a fairly discrete temeprature is becasue water is a good thermal conductor. Cool one part and the entire sample of water acts like a heat reservoir, meaning that one part will easily steal energy from the entire sample rather than freezing. Thus, the entire sample will reach freezing (or boiling for that matter) at nearly the same time.

What does "all" mean here? I know that all the water in the glass or lake does not freeze all at once. I know that all the water in the pot does not all boil at once. So what do you mean by "all"?

 

[nucleus]

Mikemense
I don’t know if this will help or not, but here is a link to lots of information on water:
http://www.lsbu.ac.uk/water/anmlies.html
There are several other parts to the site, such as:
Hot water may freeze faster than cold water; the Mpemba effect

 

[DaveC426913]

What does "all" mean here? I know that all the water in the glass or lake does not freeze all at once.

Once ice on a lake starts freezing the process is dramatically changed. The icce floats which actually insulates the rest of the water from losing more heat. In the billiards analogy, imagine the slowing-moving balls all collect along the rails, preventing the fast moving balls in the middle from losing their energy to the rails.

I know that all the water in the pot does not all boil at once. So what do you mean by "all"?

Well, it all reaches pretty much the same boiling temperature, though it can't physcially all turn to vapour at once since there's the weight of the water on top of it to hinder it (both temp AND pressure are factors in liquids turning to vapour).

 

[mikemense]

Dave,
Thanks for your efforts. I think I am still looking for specific answers to my specific questions.

 

[mikemense]

thanks, this looks good!

 

[DaveC426913]

Dave,
Thanks for your efforts. I think I am still looking for specific answers to my specific questions.

What questions do you still have?