Original by Philip Gibbs and Cecil Adams 1996.
There are a few questions about fizzy drinks such as beer and carbonated colas that come up from time to time in the physics newsgroups. On the whole, these tend to be questions about phenomena which turn out to be myths. It is easy to invent theories about why these phenomena might be real but it is equally easy to do simple experiments that show that the effects are at best imperceptible. This provides a good lesson in how to do experiments correctly.
The Straight Dope by Cecil Adams. Copyright 1996 Chicago Reader, Inc. Reprinted with permission:
Dear Cecil: I have always laughed at people who, before opening a carbonated drink which has been shaken, tap the top of the can with their finger so that it doesn't explode upon opening. After lengthy arguments, we even performed a semiscientific experiment by shaking a drink and opening it with and without tapping the top, but with no solid scientific conclusion. We would like to know what you, in your infinite wisdom, think of this. --Benjie Balser, Dallas, Texas Dear Benjie: This is not a problem that requires infinite wisdom. This is a problem that requires an IQ above two, apparently a stretch for some folks these days. First I called the folks at Coke central in Atlanta. I did this in the interest of thoroughness, in case Coke physicists had discovered quantum mechanical aspects of beverage carbonation that had previously eluded the notice of science. However, they didn't return my calls. There are two possible explanations for this: (1) everybody was out in the plant stamping out souvenir Olympic bottles, or (2) Cecil's message was a little too detailed. This is an inherent risk in my business. If you tell some low-level gatekeeper type you have a question about poultry, you may actually get through. Tell them you want to know which end of the egg comes out of the chicken first, on the other hand, and they'll have security trace the call. No matter. First let's consider the matter from a theoretical perspective. Carbonation is produced by forcing carbon dioxide into solution with H2O under pressure. Shake up the can and you create thousands of micro-size bubbles. Each bubble offers a tiny surface where CO2 can rapidly come out of solution, creating the potential for explosive fizzing should you open the can prematurely. Wait a while though, and the bubbles will float to the top of the can and disappear, and eventually all will be as before. But suppose you're the impatient type. You tap the can. What, pray tell, is this supposed to accomplish? Are we going to noodge the tiny bubbles to the surface faster, after the manner of herding cows? Right. Are we going to maybe dislodge a few bubbles that have stuck to the sides of the can? Maybe we are, but the difference is slight. Open that baby and you're still going to get a faceful of froth. We confirmed this to our satisfaction out in the Straight Dope Backyard of Science with a half dozen cans of pop. OK, so I didn't replicate my results 50,000 times. I figure if extraordinary claims demand extraordinary evidence, stupid claims demand. . . well, something a little less rigorous. I should tell you that when I had Little Ed broach this issue recently on the Usenet he heard from a science teacher, among others, who insisted tapping the can really did reduce fizzing and bragged about a classroom demonstration he did to make just this point. We wrote back to the science teacher: had he, like, tried comparing the amount of froth from tapped versus untapped cans--one of your more basic experimental procedures? Rarely, he replied. 'Nuff said. The Usenet debate dragged on for weeks. On the one hand there were various learned personages offered advanced theoretical reasons why tapping should reduce the froth. On the other hand were a hardy band of experimenters who kept trying to find a "tapping effect" and failing. The lack of experimental confirmation failed to deter the believers. Finally one Martin Heinz wrote in exasperation: "I did try some `pseudo-experiments' (thank you very much) which led me to believe that tapping does *not* have any effect. However, the variations are endless. People argued: "1) It depends on *how* you tap. (Top? Side? Flicking a finger? Knocking?) "2) Just *when* is a soda can agitated enough for the hypothesized effect to work? (Shaking? Dropping? How much? How high? Does it count if the can is scooped up by a plane dropped in a forest fire?) "3) It depends on what *kind* of soda you use. I used Diet Coke, then someone said, `Noooo, the NutraSweet throws everything off.' "4) It depends on the temperature of the soda. "5) It depends on the country the experiment is conducted in, because `tapping' is one of the many words in Inuit related to soda and means `sticking a can up your butt.' "You be the judge." You'd think that would have ended the debate. But no debate on the Usenet ever ends. It just sort of peters out. So I decided to appeal to authority. I know, I AM the authority. But I figured it was probably smart to ask for a second opinion. I called up physicist Jearl Walker, who's written about the physics of beverage carbonation in Scientific American. Jearl, you'll remember, is the guy who used to plunge his hand into a vat of molten lead as a classroom demonstration of the Leidenfrost effect. This makes him either certifiably crazy or a genius--in either case somebody you want to listen to with an attitude of respect. Jearl had heard similar claims about the efficacy of tapping and had a similar reaction: these guys are nuts. He said he could only attribute the persistence of the practice to the same suppressed macho ethic that makes people tap the ends of their cigarettes before lighting up. If you want a real solution, try this. It's an implacable fact that a warm can of pop that's all shook up will fizz more than a shaken cold can. If you absolutely must pop the top on that jug of Jolt, stick it in the fridge first. You'll chill the contents and chill the carbonation too, an inevitable consequence of increased gas solubility and Charles's law. Too much trouble? Fine. Just let the can sit awhile to give the microbubbles time to disappear. To pass the time you can even tap the top of the can. Just don't kid yourself the tapping itself makes the bubbles go away. --CECIL ADAMS
Answer: It doesn't!
To be more precise, provided the can is already in equilibrium, shaking will not cause a noticeable change in pressure. If a bottle has been opened and closed it will no longer be in equilibrium. Shaking the bottle will then cause the pressure to increase. If it is left standing it will slowly come back into equilibrium anyway but this could take several hours if it is not shaken.
Beer, colas and other carbonated drinks contain carbon dioxide which is held dissolved in the liquid under pressure. In fact it reacts with water to form a mild acid which improves the taste of the drink. Provided it has been standing at the same temperature for a long time the gas in solution will be in equilibrium with the gas in the head-space of the can. Shaking the can will not cause any change in pressure that can be detected by hand.
It is possible to produce plausible sounding theories that suggest that the pressure might increase. Perhaps momentary areas of low pressure in the can cause bubbles to come out of solution which are not so easily redissolved under high pressure. It is also conceivable that agitation could reverse the weak reaction between the carbon dioxide and water causing gas to come out of solution. To show that these theories are incorrect it suffices to perform a simple experiment, but care is needed to avoid a misleading result.
Here is an experiment anyone can try. Take six cans of fizzy drink. Let them stand for twenty-four hours at room temperature to ensure that they are in equilibrium. If the cans are taken from out of a fridge the experiment will be invalid since they may warm up during the course of the experiment and will then no longer be in equilibrium.
Mark three of the cans underneath and shake them vigorously. The other three cans should be left undisturbed so that they can serve as controls. Now get someone to shuffle the cans around while you are not looking so that you no longer know which ones were shaken. Without looking at the marks underneath, try to pick out the three cans that seem to be at the highest pressure. Finally look underneath to see if you managed to pick out the three that had been shaken. Of course, you had a one in twenty chance of getting it right by luck so it may be necessary to do it several times to improve the statistics.
The reason for doing it blind is to avoid the possibility of misjudging the pressure in accordance with your own preconceptions. feeling the pressure of the cans with your fingers is very subjective and cannot be relied on if you know which cans have been shaken. This kind of experimental mistake is partly to blame for the existence of the myth in the first place.
So why does the can froth so violently if it is opened after it has been shaken? This is caused by the bubbles that are generated by the agitation. These multiply and expand when the can is opened.
But what of that anecdote about the time someone shook a plastic bottle of cola so hard that it burst? Well the pressure inside a bottle of cola can be reasonably high, especially if it is not chilled. Shaking it very hard can strain the bottle near to its limits just because of the force of the liquid as it hits the sides. Add to that the increased pressures caused by the hard squeezing grip that is necessary to hold it as it is shaken, and it is not difficult to account for it bursting.