Proteins give champagne its distinctive fizz.
WHEN WE HEAR THE UNMISTAKABLE SOUND of the cork popping off a bottle of champagne, we stop talking and look closely at what happens as it is poured into our glass. If the foam subsides slowly, if the frill of bubbles is delicate and persistent, and if the liquid is effervescent, the wine is considered to be of good quality. Conversely, a rapid fall in the level of the fizz, the absence of a collar, and the presence of large bubbles are taken to be signs of inferior quality, even though the taste of the beverage may otherwise be satisfactory. Champagne makers therefore strive to produce wine that has a delicate and stable foam.
Researchers at Moët et Chandon recently obtained European Union funding to study the physical chemistry of the foam of champagne. In particular they wanted to know which molecules are responsible for its stability and which physical mechanisms account for changes in its structure. At the beginning of the project, evaluation was conducted solely by a panel of judges, a labor-intensive and time-consuming procedure. In short order two devices for obtaining reliable measurements were put into operation. In one, known as a Mosalux, a gas was injected into a fritted glass at the base of a cylinder containing the wine in order to measure dilation and the average duration of the foam. The other, a video system equipped with pattern recognition software, was used to track changes in the structure of the foam in actual champagne glasses.
These devices were first used in connection with filtration studies. Wine growers filter their products to give them clarity and to reduce the concentration of colloids before precipitating the tartaric acid because colloids limit the crystallization of the acid—this despite the fact, as they are well aware, that filtering hurts the quality of their wines, causing them to lose their roundness in the mouth.
How does filtering change the foam of champagne? It is generally thought to be harmful because it eliminates proteins, which are tensioactive molecules. These molecules are composed of hydrophilic parts that dissolve easily in water and hydrophobic parts that avoid water, preferring to be in contact with the air inside the bubbles. Proteins help stabilize egg white foams and the bubbles in champagne: By coating the bubbles they impede the formation of new bubbles and prevent existing bubbles from fusing with their neighbors.
Brewers of beer had observed that filtering did not harm its head, but this is because proteins are much more plentiful in beer than in wine. Research has also shown that proteins with a molecular mass higher than 5000 create a stable head.
In 1990, Alain Maujean and his colleagues at the University of Rheims noted a relationship between the concentration of proteins and foaming capacity in thirty-one wines chosen at random. Yet they were unable to determine the conditions for producing a stable foam. Three researchers at Moët et Chandon, Joël Malvy, Bertrand Robillard, and Bruno Duteurtre, used the Mosalux to study this problem. By subjecting still wines to ultrafiltration, they were able to separate out a part rich in macromolecules (notably proteins) and another part poor in such molecules. Then, by mixing them in different proportions with the base wine, they obtained wines containing various protein concentrations.
Foam Measured
On insufflating these wines with carbon dioxide, the researchers observed the same pattern: The foam initially accumulated and rose in the cylinder of the Mosalux, then subsided a bit, reaching an equilibrium for a time before finally dropping as the gas ceased to be injected. Wines whose protein concentration was reduced by 20–100% exhibited similar foaming behavior, but the higher the content of proteins having a molecular mass greater than 10,000, the greater the measured quantity of foam. Although the bubbling action of a wine during the first seconds was independent of its macromolecular concentration, these macromolecules substantially retarded the foam’s decline and the accompanying dilation of its bubbles. As expected, filtering greatly harmed the foam. The disappearance of only a milligram of protein per liter of wine (out of about ten milligrams normally) reduced its ability to foam by half; removal of 2 milligrams of proteins shortened the average duration of the foam by half as well.
These first studies were supplemented by measurements of the foaming power of wines from which colloids and particles had been removed through successive filtrations. The dilation and average duration of the foam diminish by more than half when a wine is filtered by membranes whose pores have a diameter of 0.2 micrometers. With wines that have been aged for more than a year, the consequences of filtering them through pores of the same diameter are still worse, whereas wines filtered through pores 0.45 or 0.65 micrometers in diameter have a more stable foam. This shows that the macromolecules and particles do not act in the same fashion on the bubbles.
How, then, do they act? In the first seconds after injection of the gas, the films separating the bubbles are stabilized by the tensioactive macromolecules. Once the quantity of these macromolecules at the liquid–gas interface exceeds a certain minimum threshold, the foam rapidly coalesces because the bubbles, which contain only carbonic gas, are not in equilibrium with the ambient atmosphere. It may be that other macromolecules in addition to proteins—sugars, for example—act on the bubbles as well. But what is clear—as clear as filtered champagne—is that filtering interferes with foaming and so must be done with care.