Department of Energy Argonne National Laboratory Office of Science NEWTON's Homepage NEWTON's Homepage
NEWTON, Ask A Scientist!
NEWTON Home Page NEWTON Teachers Visit Our Archives Ask A Question How To Ask A Question Question of the Week Our Expert Scientists Volunteer at NEWTON! Frequently Asked Questions Referencing NEWTON About NEWTON About Ask A Scientist Education At Argonne Frozen and Unfrozen Bottle
Name: Burr
Status: scientist
Grade: other
Location: OH
Date: November 2007

Hello fellow Ask-a-Scientist volunteers! I am also a volunteer for Ask-a-Scientist, and let me start by thanking you for your time and effort to make this program great. My colleagues and I (professors and PhD candidates in the Chemical Engineering Department at a major research University) have been debating the cause of a mysterious phenomenon I observed. I thought some of you might have some ideas. I am going to provide as much detail as I can -- although some info may be irrelevant. After an evening where the temperature dipped below freezing (I did not record the low, but it was in the 20's), I realized I had left a pair of 2-liter containers of carbonated soft drinks in the trunk of my car. My car is a Honda Civic, 4-door sedan. I had parked my car in the driveway, facing East. The 2-liter containers were in my trunk, laying horizontally, oriented in the same direction, tops facing front/east, double-wrapped in standard (Kroger supermarket brand) plastic grocery sacks. The type of beverage was Diet Ruby-Red Squirt. I had just purchased the bottles, although I do not know when they were manufactured. In the trunk with the bottles were my golf clubs, some miscellaneous clothes (mostly cold-weather golfing gear), and various sporting goods. The trunk was roughly half-full of junk. In the morning, when I went outside to retrieve them, the temperature had risen back above freezing, but there was still frost on the ground. I feared they would have frozen and exploded, but found a most unexpected result: One bottle had frozen solid, and had expanded its container but did not rupture it. The other bottle, immediately adjacent to it, was unfrozen and unaffected in any noticeable way. My question: How do you explain this? Despite considerable discussion with several thermodynamics experts in our department, no consensus was reached. Various theories were proposed (and I have my favorite), but none was indisputably correct. Other information: The two bottles were in the same bag. They were double-bagged (one bag in the other). They were not tied shut, but they had laid closed (e.g. limited convection). The bag was roughly left of center as I faced the trunk, about 6" from the back tailgate and about 18" from the left wheel well. They rested on the floor of the trunk (there is a spare tire beneath them). There were clothes and my golf bag to each side and behind them. The pop was diet -- no high-fructose corn syrup. Grapefruit sodas contain both heavier-than-water and lighter-than water components emulsified and suspended in the beverage. I presume this influences nucleation, but I don't know how. The bottle that had frozen did not rupture or leak, but it did show stress lines in the plastic, and the bottle was irreversibly expanded. The pop from the frozen bottle, after being thawed in my house at room temperature, remained carbonated normally (at least to taste) after opening. After thawing, there was a red sediment in the bottom of the bottle. The soda is red normally, and I could discern a very slight change in color of the soda, presumably due to some red components precipitating out. The same sediment was not observed in the unfrozen bottle. Can anyone explain these observations? Why/how did one freeze, but not the other?

What an interesting question. I will throw out a couple I can think of that might affect it.

Where were the bottles relative to the gas tank or any other item/insulation on the bottom or side of the car that might expose one bottle to more cold air than the other or otherwise prevent heat from being lost at the same rate from that bottle?

Is it possible there is a difference in chemistry of the bottles---are they the same lot numbers? Same expiration dates?

Patricia Rowe

Burr, My suspicion is that the quality control of the manufacturer is not as precise as one might think. A small change in the amount of carbonation or dissolved solutes can noticeably affect the freezing point of the sodas. I suspect that conditions were just right so that the more concentrated soda (or the one with more carbonation and at a higher pressure) was barely above its freezing point while the less concentrated/less carbonated soda was below its freezing point given the temperature in the trunk of your car.

If shaking the unfrozen bottle did not cause immediate freezing, then it is not an issue of nucleation and one could discount kinetic factors. If opening the unfrozen bottle caused immediate freezing then one could say that the bottle was kept from freezing by the pressure of the contents, i.e. it was very close to its freezing point.

Greg (Roberto Gregorius)

This has probably already been suggested to you, but I suspect that part of the puzzle is that the frozen bottle just happened to have the right ice nucleus in the right place to initiate freezing before the other bottle. The heat released by the freezing of the water in the frozen bottle then prevented the other bottle from getting started. Ice nuclei are fairly hard to come by. Once the one bottle got started, the other was doomed. The first had the growing crystals to enable the crystallization, while the other couldn't even get cold enough to begin.

Just my $0.02

Richard Barrans, Ph.D., M.Ed.
Department of Physics and Astronomy
University of Wyoming

There are many possibilities, although my guess is that you have differences in temperature within your trunk. I do not know where the tailpipe of a civic is relative to where the bottles were, but that would be a nonuniform heat source that could give one bottle a head start on cooling (and thus freezing). Likewise, the spare tire bay sometimes has drainage holes, the bottle further away would cool less. If the bay has no holes, the board covering the spare tire bay sits directly on metal in some spots and over open space in others, again leading to temperature differences.

The best way to know would of course be to take temperature measurement at various points in your trunk by whatever means you have available. Obviously as you open your trunk everything gets driven towards ambient air temperature but you might be able to see a residual difference (especially if it is a differential heat source, and you measure right after stopping)

Don Yee


Again let me emphasize that my take on this is that kinetic factors are not in play. In my opinion, the bottles were in the trunk of the car long enough to equilibrate, and that temperature variance in the sealed trunk was minimal at the coldest point of the night. Thus, even if we imagine that one bottle froze sooner than the other and caused the heating of the second bottle, it still begs the question of why the first bottle should freeze before the second -if the temperature was homogeneous and the systems were at equilibrium.

So going on the premise that the system was in equilibrium and homogeneous, then the only explanation is a thermodynamic one and having to do with a difference in concentration or pressure causing a depression of the freezing point of the second bottle.

Moreover, if we allow that concentrations and pressures are not going to be very different between bottles, then the second bottle may have been near its freezing point, hence the agitation and nucleation issue. But, that, I think is secondary.

Greg (Roberto Gregorius)

Eliminating all the details that probably were irrelevant, I suspect that the "unfrozen" bottle of soda supercooled. In the absence of some sort of "seed" with all the "stuff" in a soda it is possible to cool the soda below its freezing point for a considerable time without inducing freezing. Had someone bumped your car the results might have been quite different. There also could have been some undetectable differences in the formulas of the two samples, but I do not think I would want to repeat that "experiment" in the trunk of my car!!

Vince Calder

Click here to return to the General Topics Archives

NEWTON is an electronic community for Science, Math, and Computer Science K-12 Educators, sponsored and operated by Argonne National Laboratory's Educational Programs, Andrew Skipor, Ph.D., Head of Educational Programs.

For assistance with NEWTON contact a System Operator (, or at Argonne's Educational Programs

Educational Programs
Building 360
9700 S. Cass Ave.
Argonne, Illinois
60439-4845, USA
Update: June 2012
Weclome To Newton

Argonne National Laboratory