The Mysterious collapsing Lucozade Bottles..!

A scientific investigation of a curious phenomenon.....

I occasionally drive a Ford Transit van for some friends in a band, and after a long, hard day, sometimes feel the need to partake of a refreshing sugar-rich soft drink to quench my thirst and keep alert for the drive home. The resultant empty bottles tend to get discarded  on the dashboard, and have accumulated in the area between tyhe dash and the windscreen.
Over a number of months I observed that some of these empty bottles have collapsed, as shown above. I was curious as to what was causing this effect, and initiated the following investigation....

The environment

Inside the windscreen of a Ford Transit van (K-Reg, diesel, Blue). The interior can get quite hot in summer, and this may help soften the bottles, so the internal pressure is typically close to atmospheric. Squeezing the bottles appears to confirm this.

The presence of Bill the Monkey (left) is not considered to be a significant environmental factor.

Tests

The displacement volume of the bottles was measured by immersion in a cylindrical water vessel. The results are as follows :

Notes :

1) The non-collapsed bottle samples 1 and 5 were probably the result of the cap not being tightly closed. These were used as the reference for the percentage figures.
2) As no data is available on the length of time that the collapse takes, the sell-by dates are recorded as they may give a very approximate idea of this, however as this is heavily dependent on uncontrolled variables such as shop stock rotation, it should be regarded as being of little or no use.

Further tests

It would be interesting to analyse the composition of the gases remaining in the collapsed bottles, in particular Nitrogen, Oxygen and Carbon Dioxide. If anyone in the UK has access to a suitable gas analyser I'll happily send a bottle for analysis!

Conclusions

If it is assumed that the lowest figures above are the ultimate values, and little or no further shrinkage occurs, then the amount of shrinkage corresponds quite well to the percentage of oxygen in the atmosphere.

The amount of collapse is far to high to be due to barometric pressure variations.

Some process is clearly ocurring in the bottles, which is causing a change in the internal pressure. My best guess is that it is either :
1) A chemical reaction involving the remnants of Lucozade, the plastic, air, and the lid seal materials
2) A biological process, bacterial contamination occurring during the drinking process making use of the very sugar-rich Lucozade remnants to convert the air inside.

Any useful suggestions on the exact explanation are welcome! Any further information will be added to this page.

Update :

The phenomenon has also been seen in Mars Drink and Drinking Yoghurt bottles.

Many people have suggested a mechanism based on the cap acting as a one-way valve with heat fluctuations. I'm not convinced of this as the caps are screwed on tightly.

Others have suggested that it is a heat-shrink effect of the sleeve. This is not the case as the bottles return to their original shape when the cap is removed.

From: Robert J. Furmanak :
I have a possible explanation, derived from my experience with polymers. One thing about polymers, they all pass water, over time a soda bottle will "sweat" all of the moisture out of it. Perhaps this is what is happening to your bottles, a small amount of
moisture remaining has escaped, reducing the pressure in the bottle? I would suggest repeating your experiment with a thoroughly dry bottle.

From: Goncala Gamboa :
With regard to the mysterious shrinking bottles, I may add that I have observed this effect plenty of times in my lab (I am an organic chemist). The effect is most evident in our 2 litre low density polyethylene bottles containing acetone. After using the acetone we often keep some of these bottles for a while as they are convenient for short term storage of solutions. Anyway, most times they are stored with some residues of acetone and after a couple of weeks most of them will have shrunk quite a bit. It is my opinion too that what happens is an oxidative process that consumes most of the oxygen in the bottle. The acetone must undergo some sort of oxidation. In your lucozade bottles the same must be happening. Glucose, one of the components of lucozade, has an aldehyde group that easily gets oxidized to a carboxylic acid, consuming oxygen. That would explain the circa 20% reduction in volume. The same explanation probably could apply to the Mars and Yogurt drinks. Even if they don't contain nearly as much glucose as lucosade, they surely have some other oxidizable substrates.  It would be interesting to check if an oxygen-enriched atmosphere lucosade bottle would shrink more than 20%. If you are really interested, you can prepare some O2 by decomposing hydrogen peroxide (the sort you use to clean wounds, 3-6%) with some catalyst such as MnO2. The MnO2 can be taken from a used 1.5 V normal (not alkaline) battery. Instead of MnO2 you could simply use a little bit of cow liver as it contains plenty of catalase that will catalyze that reaction. In any case, take care handling the chemicals and don't enrich the bottle with too much O2 as you don't want the oxidation to be "too intense" and risk blowing the bottle or something like that...Hope this helps!

From Richard Jordan :  My dad has worked in the plastics industry for all his career, and he was
always been describing the properties of plastics to me as a kid. All plastics are permeable to some degree to the smaller molecules of air. If those lucosade bottles were heated by the sun, the pressure differential created inside would very slowly force the smaller of the air molecules (hydrogen) out of the bottle through the plastic. Upon cooling down, the pressure differential would disappear, forcing the bottle to collapse in on itself. The bottle therefore is acting as a partially permeable membrane. In theory, the force of the bottle trying to reexpand would not be enough to force the molecules back the other way. Anyway, just a theory.

Another contributor : PET (polyester) is used in soft drink bottles because it has a very low permeation rate for CO2. This allows for a fairly long shelf life before the bottles go flat. The permeation rate for CO2 in PET is about 2-3 times that of oxygen (the permeation rate of N2 is even lower). The driving force for permeation of a gas is not the absolute pressure difference between the two sides of the membrane, but is the partial pressure difference for the particular gas. If the inside of the bottle has a high concentration of CO2 in it, this CO2 will move (permeate) from the inside to the outside to try and equalize the partial pressures inside and outside the bottle. The oxygen and nitrogen outside the bottle will try and do the same but at a slower rate. Raising the temperature of the bottle will increase permeation rates. You can verify this by taking one bottle and leaving it on its side opened for a day before you cap it. This will be the air filled control. Take the upright test bottle and put a little chunk of dry ice in the bottom and let it evaporate. Immediately after the dry ice has evaporated, but not before, cap this bottle. This will be the CO2 filled test bottle. Put both bottles in the test chamber (between the dash and windscreen) and watch for results. The peculiar final shape of the bottles is due to the stresses induced by the blow-molding process used in making the bottle. At higher temperatures the PET will try to change shape to reduce these stresses.