Does "light" really mean light? - part II

A follow-up on the last entry, comments and Brian's Diet Coke Floating blog entry

The comments on the last entry on Coke density spurred my curiosity, and I decided to follow up with a few experiments:

1. What's the density of different kinds of Coke? I did a simple experiments, weighing measured amounts of Coke (ca. 100 ml) and got the following result. I did the same measurement for water and corrected this according to standard water density values from a web based water density calculator:
   
   
   
   
   
2. What about water density as function of temperature? The density changes according to the reference values are small, far too small to make a difference (my cans/bottles held between 13 and 21 deg. C), as seen from the density curve as function of temperature (the same density calculator used):
   
   And, of course, the density-temperature difference should not be very different from water to Coke, so at temperatures of high water density, the same should apply to Coke.
   
   
3. Experiments should always be tested for repeatability, so I used two cans of coke. Also, I tested whether the same would happen for Coke bottles (500 ml plastic bottles), and if the same would happen for Coke Zero. The pictures below tell the whole story. The difference between Coke Light and Coke Zero, from the ingredients list, seems to be the sweeteners. Coke Light contains Sucralose and Acesulfame K, while Coke Zero contains Aspartame and Acesulfame K.
   

What seems really strange to me is the measurement of Coke light and Coke Zero having densities lower than water, especially the light variety which is well below any temperature dependent variations. How can it be that a water solution with dissolved matter has lower density than water? Carbon dioxide? I don't think so. Carbon dioxide is still matter dissolved in water and should contribute to a higher density rather than lower (regardless of its density in pure, gaseous form).

Anyway, the safe explanation to the floating Coke light is of course the air pocket (both in cans and bottles), and I think I'll stick to this as the main explanation rather than densities of Coke. Ordinary Coke is a clean cut case, anyway.

Erik