The Mentos geyser: From demo to real science

This article is one of a series of Experiments meant to teach students about how science is done, from generating a hypothesis to designing an experiment to analyzing the results with statistics. You can repeat the steps here and compare your results — or use this as inspiration to design your own experiment. 

If you’re ever up for a fun and messy time, take one two-liter bottle of diet soda outside. Open the cap and place it on the ground. Use a small paper tube to quickly slide in 6 Mentos mint candies. Then, make way!

The resulting geyser is one of the most popular science demonstrations. But it’s not an experiment. Despite its explosive power, it doesn’t answer questions about how the geyser works or what aspects of its ingredients matter most. Luckily, it’s easy to turn this demonstration into your own experiment. I’ll show you how.

What gives the soda geyser its energetic gush? When people talk about this demonstration on the Internet, they always use Mentos. They are very specific about the candy — it has to be Mentos.

I looked up other science sites to find out why Mentos are important. By reading some of the scientific papers on this, I learned that Mentos are covered in tiny little pits. This allows bubbles to form, triggering our fountain of fizz.

Behind those bubbles, I learned, is a process called nucleation. In chemistry, this process causes structures to form by themselves. In the case of the soda spray, it’s the bubbles of carbon dioxide that assemble by themselves. Using high pressure, beverage makers pump this colorless, odorless gas into a drink to provide a pleasant tingle. That gas remains dissolved in the liquid until the soda container is opened. This releases the pressure. At once, carbon-dioxide molecules begin to collect as tiny bubbles.

Usually this bubble formation — nucleation — isn’t very explosive. But that changes when you add a little bit of candy. Mentos have a surface full of tiny craters. The little bumps and dips serve as nucleation sites— spots where tiny carbon-dioxide bubbles can cling. With all those craters, bubbles form so fast they need to explode outward. And in a bottled up soft drink, they have nowhere to go but up.

Knowing how the geyser works gives me a way to form a hypothesis — an idea I can test. If nucleation sites are important to form a geyser, then I might hypothesize that candies with rough surfaces will produce bigger fountains than candies with smooth surfaces.

I decided to test three kinds of sweets. As my control, I’ll again use those Mentos, with their pitted surfaces. Next, I’ll try M&Ms, which have a very smooth candy shell. Finally sugar cubes have pits and craters much larger than the Mentos.

These three sweets also start out as different sizes. So instead of putting in the same numbers of candies, I will put in the same mass — 22 grams (about 0.8 ounces) of each type.

I measured fountain size in two ways. First, I set off all my geysers in front of a wall. I put marks up the wall every 10 centimeters (or 3.9 inches). I then videotaped all of my geysers to identify their maximum height.

But there’s another way to measure geysering: weigh the soda! A soft drink that erupts with a high spout will lose more of its liquid than one that emits a small squirt. By weighing the soda bottles before and after each test, I can determine how much liquid was lost in each test.

I also took care to test each type of candy many times. Each fountain spray will be slightly different, even if each test uses the same kind and amount of candy. For each type of candy — my test variable — I added up all of the geyser sizes and calculated the average. Then I ran statistics, comparing the average geyser size for each type of candy.

I made six fizzing fountains for each type of candy. Then I graphed my results, so that everyone can see my data.

You can see on the right that M&Ms, with the smoothest shell, triggered the smallest eruption, measured by loss of beverage mass. Sugar cubes made a bigger one. Mentos made the biggest of all. The stars and plus signs on my graph are used to indicate a p value —the probability of seeing a difference as big as the one I found if the difference between geysers was due to chance. These stars and plus signs indicate a very small p value of less than 0.001 — or less than one percent — suggesting the result was significant, and not due to chance. 

I also weighed my sodas before and after. In the second graph at right you can see that the sodas with M&Ms lost less weight than sodas with sugar cubes. Mentos geysers spouted out the most soda.

Finally I performed a correlation — or comparison identifying a possible link — between geyser height and the amount of soda lost during a geyser. 

The graph below shows there is a relationship between the height of the geyser and the amount of weight lost from each soda bottle during the test. A taller geyser is associated with a larger loss in soda mass — a positive correlation.

The findings supported my hypothesis. The M&Ms’ smooth candy shell produced a smaller geyser when dunked into a Diet Coke than did the pitted sugar cubes. But the pits in the sugar cubes might be too large to produce the tiny bubbles needed for a really good rushing geyser. For that, it appears, you need the tiny nooks and crannies in a Mentos candy, which produced the best geyser of all. At least, that’s my tentative conclusion.

Have you ever done a soda geyser? Turn a simple fountain into an experiment, and see if you can repeat my results.