12 Mar 2007

Does "light" really mean light?

In a simple demonstration playing around with light products floating/sinking in water a fascinating contrast emerges.

Try submerging the following two pairs of products: ordinary and light mayonnaise, and diet coke (or other cola vs. cola light).

Light mayonnaise floats lower than ordinary mayonnaise. In the case of the soft drink, cola light floats while ordinary cola sinks! What's going on?

The term in question is density:

The main ingredients in mayonnaise are water and fat/oil:
Ordinary mayo: 80% fat, 16% water
Light mayo: 40% fat, 50% water

Fat floats in water. A larger proportion of fat makes the mixture closer to pure fat and vice versa.

Cola can be considered as water with some dissolved material. Ordinary cola is, as such, a sugar solution with a few other additives (taste, aroma, colour etc.).

Cola light contains the artificial sweeteners Sucralose and Acesulfame K. Both are far sweeter than table sugar, sucrose (650 times and 180-200 times, respectively, ref. Belitz). Thus, far less sweetener is needed. If we assume that all other ingredients are the same, then far less material is dissolved in the light version. The same total volume with less material --> lower density. Thus: pure cola light would float up in ordinary cola taken that they didn't mix.

In both the mayo and cola cases, there is some air (or trapped gas) inside the tube/can. This makes the tube/can float higher than in the case of the pure mayo/cola. However, as long as the volumes are the same, this doesn't make any difference. If it wasn't for the air, both cans would in fact sink, and the experiment wouldn't be.

A question of using and understanding scientific concepts
The concepts relevant to this is not only "light", but also (amongst others) "(chemical/dietary) energy" and "density". If something floats, we usually say that it's "lighter than water". However, two kilograms of wood is heavier than one kilogram of water, but it still floats. To me as an adult, it's probably easy to grasp, but placing this in an educational context makes it important to use the correct terms. So, "cola has higher density than cola light" would be more correct.

It's quite easy to put this to the test: measure both the volume and weight of a can/bottle of cola and compare. The one that sinks (highest density) weighs the most taken the same volume. Another version of this experiment is concealing the labels, letting the students know the content without telling which is which. The task is then to use knowledge and reasoning to deduce which is which.

Using the term "diet" rather than "light" would of course make the whole case less diffuse, but then a fascinating aspect in the experiment and following discussion is lost. This is the reason for using "light" instead of "diet" in the first hand. However, this may be a nice way of introducing the energy concept of (chemical) energy, kcal and kJ, and contrast this against "light" used in different contexts.

Natural sciences are evidently not only concerned with nature itself, but just as much the language describing nature.


- Belitz et al., Food Chemistry 3rd ed., Springer 2004
- A Swedish version of the cola experiment at SkolKemi pages of University of Umeå


  1. I'd like to mention that the coke will also float to the surface of the water, just not as quickly as the diet coke does. I mention because there's apparently an urban legend about Diet Coke floating when Coke does not, and it's not entirely true.

    I wrote something up on Diet Coke Floating on my site.

    Naturally, this is tangential to your main point, but I thought it worth mentioning. A fun post all around.

  2. Brian,

    interesting comment and nice picture on your blog on floating Coke. I've left a comment on your observation in your blog entry.

  3. From my response to your response:

    After reading through your similar entry on the subject, I realized that I completely ignored the air bubble issue. If an aluminum can is completely filled with a water-like liquid, it *should* sink to the bottom. Dissolved solids shouldn't decrease the density of the liquid, but should increase it. Dissolved Carbon Dioxide might decrease the density, but I don't know if it's appreciable.

    I didn't really do anything special to my coke can to get it to float; I actually had to be quick to get the photo with the coke and the diet coke at different levels. It occurs to me that there could be a few different possibilities. First, there could be enough variation in the canning process that different amounts of liquid will be in the can, which seems unlikely. Second, as you are blogging from outside the Norway, it's possible that your Coke is made with real sugar; ours is generally made from Corn Syrup, which may have a different density. Third, there could be variations in the shape of the coke can that allows for the same volume/weight of liquid with additional space for air. I'm specifically thinking of the indentations at the bottom of the can, as that's not something that would be paid much attention. Fourth, Norwegian Coca-Cola might have a different concentration of Carbon Dioxide than US Coca-Cola which, if it's contributes significantly to the density of the liquid (and I have no idea if it does), then that could be it.

    Obviously, more experimentation is called for. The first thing to do would be to measure the volume and weight of our respective cans of Coke and Diet Coke and see if there's any variation. We could also try seeing if there's any variation in can weight and volume within a certain area, to see if the process has any natural variations, or if it might be regional. My guess is that your weight and volume between the two types of Coke will be more consistent than mine.

    I can't see that the Carbon Dioxide would affect anything, though. Because if it did, that would imply that flat Coke would weight more than carbonated coke, which would imply that, if you sat a can of coke on a scale, weighed it, then opened it and let it go flat, that it would weight more after going flat. And that seems silly.

    So, my front-runners for the reason your Coke doesn't float and mine does are: 1) Variations in can volumes; 2) Sugar vs. Corn Syrup. When I can do some more experiments, I'll post to and let everyone know.

  4. Nice pictures - I wasn't aware of the mayonnaise experiment!

    I have repeated the coke/diet coke experiment several times, even with picture proof.

    Brian, the fact that you don't observe a floating diet Coke is probably due to small differences in filling height. The amount of trapped air in the cans is crucial for the experiment to work.

  5. Hi, Martin,

    Actually, I do get the diet coke to float. But I also get the coke to float. In your various experiments, did you get the coke to float? I agree, it's likely an air pocket issue, but I'm curious if it's intentional or accidental, and how reproducible it is.

    I could imagine, for example, that someone in product development decided that it was worthwhile to give the can a little extra volume for an air pocket so that, for example, it would float in an ice bath, thus keeping people from having to submerse their hand in order to get a drink.

  6. Guys,

    I guess the next step is for all of us to measure our respective Cokes. Measuring the weight and volume of the contents as well as the total volume of the can. That way, we can compare the size of the air pocket in the respective cans.

    Anyway, I've observed this with more than one pair of cans bought at different places quite long time apart (months/year). In addition, I guess the industrial reproducibility in bottling/canning is quite high as the producer probably careful about wasting anything.

    I haven't got any high-accuracy scales and volumetric equipment, though. Only ordinary high school labware sort of stuff. I'll give it a go anyway.


Please add your comments