[gsmollin (OP)]

This is a continuation of the "ice cube" thread, but I started
a new thread because the old one does not belong to me.

I decided to do the hot/cold water freezing experiment myself.
The water samples were from my tap. GH = 187 ppm. KH = 68 ppm.
The cold sample was prepared first. 232 g of tap water was put
into a 1 pint Corningware dish. The hot water was boiled, then
232 g was poured into another 1 pint Corningware dish. Both
samples were carried to a chest freezer at -24 C, and placed
on a wire rack in the freezer. The first temperature reading
was taken at this time, which would have been about 5 minutes
after the boiling water was poured from the kettle.

Temperature data was taken after 20 minutes in the freezer,
with a type K thermocouple. After the surface ice became too
thick to poke the thermocouple wire into, I stopped measuring
that sample.

Temperature Data

• TIME __HOT T COLD T


• 00 min  66.0 C  16.8 C


• 20 min  23.1 C  06.1 C


• 30 min  14.7 C  03.4 C


• 40 min  09.3 C  01.0 C


• 50 min  04.5 C  00.7 C


• 60 min  01.8 C  00.7 C


• 72 min  00.8 C  00.8 C


• 82 min  00.8 C  thick ice

Freezing Observations

The cold sample had a thin layer of ice form at 40 minutes.
The hot sample had a thin layer of ice form at 60 minutes.

Post-freezing Mass

The cold sample contained 232 g of water.
The hot sample contained 215 g of water.

Conclusions

Hot water does not freeze faster than cold water, but it is
apparent why one could come to the opposite conclusion. First,
the volume of hot water needed to add 232 g to the Corningware
dish was noticably larger than the cold water. If one were using
a volume measurement, one would start with less hot water. Then
the hot water lost most of its heat in the first minutes of the
experiment, even before I could transfer the containers to the
freezer. Most of this heat loss was carried by evaporation, and
that showed up in the mass measurement of the frozen water. If
one were performing this experiment using a shallow pan, such
as an ice cube tray, that would cause a large amount of mass to be
lost to evaporation. The remaining hot water could freeze before
the more massive cold sample.

The observed freezing temperature of the cold sample was 0.8 C,
and the hot sample at 1.8 C. It is tempting to say that was caused
by loss of GH due to boiling in the hot sample, but my observation
intervals are too large to be sure.

[neilep]

*sigh*......someone give this individual a medal.....a big gsmollin....nicely done !

Men are the same as women.... just inside out !!

[chimera]

Ah, but how many times must you go through all this to be sure it's not just a statistical fluke?

[]

[gsmollin (OP)]

What random variables are present that would invalidate the results? I suppose I would get somewhat different temperatures in another run, but the time difference between freezing of the cold sample, and the hot sample was 33% of the total time in the freezer. It was not a close race. The time would have been even further apart had I covered the dishes, and blocked the evaporation of the hot sample.

I think this is a closed issue. The evaporation of the hot water from a shallow ice cube tray is almost certainly the cause of the discrepancies.

[Ultima]

http://math.ucr.edu/home/baez/physics/General/hot_water.html I remember hearing about this ages ago...

wOw the world spins?

[gsmollin (OP)]

I read that some time ago. It is a good example of an article with a lot of anecdotes, but no real information. They suggest that water has states containg memory of its past, i.e., water knows it's been boiled and freezes faster afterwards. The only state in water that meets this criterion is its mineral content. The general hardness can be reduced by boiling, so that is the one possibility. Some water samples may be highly mineralized, and boiling it raises its freezing temperature. I may have seen that in my experiment; the boiled water began freezing at 1.8 C, and the unboiled was down to 1.0 C.

The other effect is the evaporation. Between them I think the "Mpemba" effect is explained.

[chris]

Hallelujah

Perhaps we should send the thread to scientific american ?!

"I never forget a face, but in your case I'll make an exception"
 - Groucho Marx

[chimera]

Opponents might say you forget two things:

- freezing is crystallisation, and by poking at intervals with a (metal? recalibrated?) thermometer and possibly shaking the whole setup (freezer) you influence that process negatively.

- secondly, and I admit it's a possibility, the minerals in the water, which are basically metallic, might play a role in the 'memory' department, much as some metals have a shape-memory, and that you do not take into account things like magnetic fields inside the freezer, with its pump kicking in at times etc. and they might even say that a metal container could give different results from using plastic or porcelain containers...

[Seriously: I really applaud the experiment and the solid way you go about things... but science is part nitpicking, too, so I'll play devil's advocate to the bitter end here]

[neilep]

As a firm believer in empirial study, I too shall now conduct the same experiment, Apparatus involved: One Ice tray, One Freezer, One kettle and water......I'll report later.

Men are the same as women.... just inside out !!

[gsmollin (OP)]

quote:Originally posted by chimera

Opponents might say you forget two things:

- freezing is crystallisation, and by poking at intervals with a (metal? recalibrated?) thermometer and possibly shaking the whole setup (freezer) you influence that process negatively.

- secondly, and I admit it's a possibility, the minerals in the water, which are basically metallic, might play a role in the 'memory' department, much as some metals have a shape-memory, and that you do not take into account things like magnetic fields inside the freezer, with its pump kicking in at times etc. and they might even say that a metal container could give different results from using plastic or porcelain containers...

[Seriously: I really applaud the experiment and the solid way you go about things... but science is part nitpicking, too, so I'll play devil's advocate to the bitter end here]

-Crystallization: I poked with a type K thermocouple. When there was a skin of ice on the surface of the water, I poked through the ice to measure the temperature of the liquid water beneath. This was admittedly crude, and I was able to poke through repeatedly, so I know I disturbed the crystalliztion. However- by the time an ice skin had formed on the cold water sample, the race was over, and cold water had won. Disturbing the ice formation only slowed down the freezing of the cold water. The hot water was still liquid. Poking a thermocouple into it did stir it, so that would cause the hot water to lose heat and freeze sooner. Yes, this does disturb the detailed results, and causes small errors, but it does not change the result that cold water freezes faster than hot water.

-secondly: The minerals involved are calcium and magnesium. These are diamgnetic and are not affected by small magnetic fields. The freezer has a steel skin, so it excludes external magnetic fields. In any event, both the hot and cold water see the same environment in the freezer, magnetic fields and all. It would also be easy to repeat the experiment with distilled water so minerals are not part of the unknown variables. I predict this would widen the gap between cold and hot.

A metal container could give different results, especially aluminum (aluminium for you Brits). I did not use metal because it can release ions into the water, and would do that to the hot water more than the cold. The higher conductivity of metal and ceramic helps cold water to freeze faster than hot, because heat can be lost in the water to conduction through the walls of the container. In a plastic ice cube tray, less heat is lost through the container walls than in ceramic, so more heat is lost through evaporation. This reduces the mass of the hot water, and speeds the freezing of the remaining water. As I said in the previous post, this is what causes the "Mpemba effect". I can believe that an experiment that filled plastic ice cube trays with hot water and cold water would show the hot water freezing first, but weighing the ice would also show a large loss to evaporation in the hot water. Removing this variable would require adding more hot water to the tray, so the frozen masses are equal. I think the results will then be different.

[neilep]

well...after my amateurish attempt...even I can report that the colder water froze, before the hot water. Incidentally, I used cold water from the tap, warm water, hot water and freshly boiled, they all froze respectively in the order one would logically expect.

Men are the same as women.... just inside out !!

[DrPhil]

2 identical air-tight glass containers
117g distilled water in each container (warm water was not boiled)
Digital thermometers with remote sensors
Freezer at -16.6 Celsius
Temperature recorded by remote sensors without disturbing the containers or opening freezer.
Freezing defined as zero degrees, at which time the freezer was entered and containers retrieved.

1st run:
Initial temperature: Warm water 46.9 , Cold water 12.0
No contest. Cold reached zero first. (halted before Warm allowed to reach zero)

2nd run:
Initial temperature: Warm water 33.4, Cold water 12.3
Cold beat warm by 1 min 13 sec.

3rd run:
Initial temperature: Warm water 23.8, Cold water 11.5
Warm beat cold by 1 min 47 sec.

[neilep]

quote:Originally posted by DrPhil

3rd run:
Initial temperature: Warm water 23.8, Cold water 11.5
Warm beat cold by 1 min 47 sec.

Fascinating !!

Men are the same as women.... just inside out !!

[chimera]

quote:Originally posted by DrPhil
3rd run:
Initial temperature: Warm water 23.8, Cold water 11.5
Warm beat cold by 1 min 47 sec.

Now THAT is funny, indeed. Sounds like a nice setup, too. How did you heat the 'specimens'? Microwave?

[DrPhil]

I heated the water on an electric hotplate in a separate container (not the container I put in the freezer), stirring gently with a alcohol filled glass thermometer until approx. desired temp was reached.

Just for grins I repeated run #2, but this time I pre-chilled both containers in the freezer before adding the water. Previously I had the container at the same temp as its water. This time Warm beat Cold by 23 sec.

[chimera]

Wow. What gets me is the fact you used demineralised water - that means that's out as an influencing factor.

I know that in boiling water they have observed hexagonal cells (named after a Frenchman, starts with a B, but forgot), which reminds me of the fact that snowflakes also tend to go 6-sided.

Now this guy has a theory, which I think is largely correct, but I do not see how it can help explain 'your'/'our' phenomenon though.

Good read, anyway - quote:

"Finally when one has set up the model as described, another curious phenomenon is explained. Going from ice to liquid, one hits a cataclysmic event at melting. That is, at 0.0000oC ice and water co-exist but at 0.0001 oC the ice completely disappears. However In the reverse direction no cataclysmic event occurs. The liquid can exist without any ice at temperatures well below OoC. This dissymmetry in direction is accounted for in the proposed model."

http://www.morenos.com/sec~1.html

[gsmollin (OP)]

quote:Originally posted by DrPhil
...1st run:
Initial temperature: Warm water 46.9 , Cold water 12.0
No contest. Cold reached zero first. (halted before Warm allowed to reach zero)

2nd run:
Initial temperature: Warm water 33.4, Cold water 12.3
Cold beat warm by 1 min 13 sec.

3rd run:
Initial temperature: Warm water 23.8, Cold water 11.5
Warm beat cold by 1 min 47 sec.

You have been busy freezing water. This topic has generated a lot of interest.

Some questions: How did you bring the cold sample to 11-12 C? What is the temperature data, especially around 0 C. This is a very flat region in the temperature data because of the latent heat of fusion in the water, and I am frankly suspicious of temperature data at 0 C. That was why I inspected visually for ice formation. The time differences between hot and cold samples are 1-2 minutes. How long were the samples in the freezer?

[chimera]

quote:Originally posted by gsmollin

What is the temperature data, especially around 0 C. This is a very flat region in the temperature data because of the latent heat of fusion in the water, and I am frankly suspicious of temperature data at 0 C. That was why I inspected visually for ice formation.

With the phenomenon described in my previous post, that may be very wise, although you could be well below zero before any ice is formed at all (even to -40 C).

It is a curious region: to raise water temperature takes 4.2*10^3J per kg for each degree of temperature, to convert ice to water without any temperature rise, 3.3*10^5J per kg...

[DrPhil]

I did not record temperatures as I was originally only concerned about which one reached zero first (is my ignorance of the process being investigated showing yet?)

I used a countdown timer to remind me to check the thermometers every 5 minutes. As soon as one of them came within a few degrees of zero I watched continuously. Total time was between 15 to 20 minutes.

The bulk bottle of distilled water was stored in a cooler with the temp set to approx. 7C. It was removed from the cooler during sample preparation. By the time the Cold sample was placed in the freezer its temp had risen to approx. 11 to 12C.

Sources of error:

Grams: Plus or minus half a gram

The thermometers were calibrated to zero using ice/water bath of city well water (chlorinated and fluoridated.) However, this seemed to have worked out reasonably well since, amazingly, the temperatures of the samples bottomed out and remained at exactly 0.0 C as they began to freeze.

Digital readout to only one-tenth of a degree.

The thermometers' digital readouts were not continuous. They take the temperature and updated the readout once every 10 seconds. The internal clocks of the two thermometers can not be synchronized. This was a big concern for me since this was, after all, a timed race, but I had no choice. Timing of a very close race would have been a problem.

Placement of the temperature probes was done by eye. Care was taken to ensure that the sensors were completely submerged and not touching the walls of the container, but the exact depth was not measured. The depth of the water was approx. 25mm and the probes were place about halfway down. On reflection, I guess the results could have been different if the probes were placed just below the surface or near the bottom.

timing: Quartz clock with analog display. The kind where the second had "jumps" every second, so measuring fractions of a second was not possible. This didn't seem to be a major concern considering the possibility of a 10 sec error due to the thermometers limitations.

~~~~

I didn't really think that this would be the definitive test. I merely report the facts and let you interpret. I'm sure that one of you with a bit more lab experience could do better.

[DrPhil]

ps:  With the exception of the 1st experiment, which was halted early, I waited for both samples to reach zero degrees, so there was ice present in both containers when they were finally removed from the freezer. I cannot say exactly when that ice formed. But I can say that there was both liquid and solid present and the liquid did not go below zero degrees. Based on my limited knowledge of how water freezes and the inability of my thermometer to register anything less than one-tenth degree increments I suspect I would not see negative temperatures until shortly after the samples froze solid. But the samples were removed long before they froze solid.

I'm sure this is full of holes, so go ahead and poke away. I did the best with what I had on hand.