241 Pizza

What’s going on? Beer, sparkling wine and other fizzy drinks are supersaturated with gas. Although the laws of thermodynamics favour the gas bubbling out of the dissolved state, bubble formation is unlikely since bubbles must start small. Because the pressure of these tiny bubbles can reach about 30 atmospheres in a bubble only 0.1 micrometres in diameter and the solubility of gases increases with increasing pressure (Henry’s law), the gas is forced back into solution as quickly as it comes out.

However, bubbles can form around dust particles, surface irregularities and scratches. These areas, know as nucleation sites, are hydrophobic (they repel water) and allow gas pockets to form without first forming the tiny bubbles. Once the gas pocket reaches a critical size, it bulges out and rounds up into a properly convex bubble, the radius of curvature of which is sufficiently large to prevent the self-collapse described earlier.

Then there is a cascade effect. If the bubbles reach a certain critical number per unit volume, this in itself constitutes a physical disturbance and results in the release of yet more bubbles.

Nucleation may be precipitated by a variety of imperfections. Minute salt crystals, such as calcium sulphate, may be present if the glass has been left to dry by evaporation after being washed in hard water. Or there may be tiny cotton fibres if the glass has been dried with a tea cloth. Dust particles may have settled on the glass if it has been left standing upright on a shelf for any length of time. And tiny scratches will be present on the inside surface of all but brand-new glasses.

Once the inside of the glass is wet, any salt crystals will have dissolved and any cotton fibres will no longer function as centres of nucleation. Most of the dust particles and all of the scratches will, of course, still be there. However, these will have been coated with liquid and the fresh carbonated liquid will reach them very slowly, by diffusion. Bubbles will still be produced, but at a rate that is too slow for the cascade effect to come into play. As a result, the drink will not froth over.

To demonstrate this, take a glass and thoroughly coat the inside with olive oil, which is a more efficient surface-covering agent than water. Then add the wine or beer (if you are feeling frugal, use a carbonated drink such as lemonade). The effervescence will be nil or minimal. Throw in the dust you scraped from behind the wine rack and see how it can affect the behaviour of the liquid. Finally, add a few million centres of nucleation from a large spoonful of granulated sugar and prove to yourself how volcanic the effervescence can be.

PS: Thanks to modern production techniques, today’s glasses are of such good quality that some manufacturers build in deliberate imperfections, especially in beer glasses, in order to generate enough bubbles to maintain the head on the top of your tipple.

There are almost certainly similarities between this experiment and the Mentos and cola effect from the ‘Overreaction’ experiment on page 194, although the jury is still out on why Mentos and cola produce quite such a spectacular effect. If you are going to do either of these indoors, go for the wine or beer version …

For more experiments go to www.newscientist.com/hamster

Explore Books

• Book Extracts

• Book Reviews

• Books Group

• Audio Books

Times Recommends

• The 100 Best Books of the Decade
• The Times picks the 50 best paperbacks of 2009
• The conversation: James Ellroy