Molecular Gastronomy: Exploring the Science of Flavor (42 page)

sees the fusion of molecules that contain three fatty acids and have only single

chemical bonds between carbon atoms.

Instead, physical chemists Frédéric Lavigne, Michel Ollivon, and their col-

leagues in the faculty of pharmacy at Chatenay-Malabry used melted butter to

study the opposite of fusion: crystallization. To separate the various parts they

therefore performed a split crystallization, slowly cooling the liquid and isolat-

ing crystals of the same molecular type that appear at the same temperature.

New Butters

Having thus isolated these similar parts, Lavigne, Ollivon, and their col-

leagues next looked for a way to form mixtures that would be spreadable

286 | a c uisine f or t omor r ow

straight out of the refrigerator. They hit on the idea of mixing high–fusion

temperature triglycerides, which remain solid at room temperature, with a

suitable proportion of low–fusion temperature triglycerides, which are liquid

at room temperature.

In this way one obtains an apparently solid body that, like traditional but-

ter, contains a proportion of molecules in liquid form (even in milk the fatty

droplets are partially solid, the proportion of solid matter reaching 70% at 4°c

[39°f] but only 10% at 30°c [86°f]). Enrichment by low–fusion temperature

molecules makes the mixture easier to spread. The parts that fuse at high tem-

peratures are used to make pastry (still under the name
butter
because the law

permits it), particularly puff pastry.

Why, then, does a solid that contains liquid appear to be solid? Because of

the crystals that increase with cooling and interlock with one another: Scraped

with a knife, butter seems to soften, not because it is heated but because the

crystals are separated.

To have an idea how these discoveries can be used in cooking, try testing

split crystallization yourself. Melt the butter and skim off the solids as they

form, just as the physical chemists did. You will then be able to manufacture

your own butters by mixing proportions of solids and liquids and in this way

obtain the specific texture appropriate to a particular dish.

Butter: A False Solid
| 287

86

Liver Mousse

Its aromatic qualities depend on its texture.

t h e f o o d i n d u s t r y i s f o r e v e r l o o k i n g for ways to make its prod-

ucts lighter by reducing lipid content and increasing water or air content, with-

out the taste suffering as a result. Even liver mousse, renowned for its sturdy

lipidic constitution, has not been spared. Michel Laroche and his colleagues

in the Laboratoire d’Étude des Interactions des Molécules Alimentaires at the

Institut National de la Recherche Agronomique (inra) station in Nantes stud-

ied liver mousses to which a large amount of starch had been added, in order

to determine by how much the fat content of such products can be reduced

without affecting flavor. Physicochemical measurements and sensory analyses

showed that the attractiveness of lighter versions depends principally on their

meltability, that the sensation of fattiness does not depend on the quantity of

lipids substituted for by starch, and that the perception of flavor depends on

their consistency.

Making a low-fat liver mousse is particularly tricky because one wants it to

be easily spread on toast, a property that in classic mousses seems to result

from their high lipidic concentration (as much as 50%). Earlier, in 1985, two

other researchers from the Nantes station, René Goutefongea and Jean-Paul

Semur, showed that this property could be preserved by adding hydrocolloids,

which is to say dispersions of long molecules in water. The recent experiments

tested the partial or complete substitution of lipids by starch paste obtained

288 |

from fava beans, the seeds of a leguminous plant (
Vicia faba
) cultivated in

France, which preliminary studies suggested might be a suitable candidate.

In a liver mousse prepared in the classic fashion, by grinding up pork liv-

ers with egg whites, lactoserum, gelatin, sodium nitrite, salt, pepper, onions,

shallots, and cognac, the Nantes researchers replaced the various quantities of

lipids with fava bean starch paste (15% starch in water) for a first batch of sam-

ples and with a fixed quantity of paste having variable concentrations of starch

for a second batch (so that the result would be 50% lighter but with different

starch concentrations). These preparations were compared with liver mousse

prepared by traditional methods that was purchased at a local grocery store.

Various mechanical measurements were made to characterize these differ-

ent mousses, which were then tasted by ten trained judges. In a room dimly

illuminated by red light, the tasters were instructed to evaluate four sensory

elements: meltability (defined as the ease with which a sample melts between

the tongue and palate), fattiness, granular texture (defined as the perception of

particulate matter), and intensity of flavor. Finally, they gave an ordered rank-

ing of these elements with respect to their relative contribution to the overall

perception of quality.

First, meltability was found to increase with the quantity of starch in the

first batch of substituted products but to diminish in the second. The percep-

tion of granular texture was not changed by the amount of starch unless it was

very large. And the sensation of fattiness, which was independent of the actual

quantity of fat for the batch with variable lipid content, diminished by con-

trast when the starch content was increased in the second batch. An analogous

variation was observed for the perception of flavor, which increased for the first

batch but diminished with the quantity of starch in the second batch.

The four sensory indices turned out to be strongly correlated, but the rela-

tionship between meltability and the perception of flavor is the most interest-

ing: The increase in the quantity of starch in the second batch was associated

with diminished meltability and a diminished perception of flavor. Should it

be concluded, then, that the aromatic compounds were adsorbed by the starch?

Or that they were adsorbed by the water? No; both of these conclusions are

invalidated by the results obtained for the first batch, where an increase in the

quantity of starch went hand in hand with an increase in the water content of

the mousse and a notable increase in meltability.

Liver Mousse
| 289

Because the composition of the samples does not explain the relationship

between meltability and aromatic quality, it seems either that the increase in

meltability, which is associated with a more even distribution of the mousse

in the mouth, improved the perception of flavor, or that the less smooth the

texture, the more trouble the tasters had perceiving the other elements.

Meltability was well correlated with high marks for the other sensory indi-

ces, in order of importance for the overall impression of quality: meltability, fat-

tiness, flavor, and granular texture. These four parameters were in turn strong-

ly correlated with the measurements of hydration and mechanical resistance.

What is the optimal proportion of starch? Using a 15% starch solution, one can

replace two-thirds of the lipids without diminishing the overall quality of the

mousse; at higher concentrations, however, the mousse becomes too soft.

These studies lead us to conclude three things. First, meltability is the chief

factor in determining the overall quality of a liver mousse. Second, the sen-

sation of fattiness is independent of the quantity of lipids that are replaced.

Third, the perception of flavor depends on texture. This final result calls to

mind the finding of Patrick Étiévant and his colleagues at the inra station in

Dijon, in 1990, that the addition of pectins to strawberry jam firmed up the

texture while also reducing its aromatic qualities.

290 | a c uisine f or t omor r ow

87

In Praise of Fats

Whatever else may be said about them, fats are to be welcomed in

cooking.

f a t t y f o o d s a r e a c c u s e d o f b l o c k i n g our arteries and making us

fat. Unsurprisingly, perhaps, one hears calls nowadays for banning fats from

the kitchen. Nonetheless, fats are an indispensable part of the cook’s reper-

toire. Let’s consider the reasons why.

Deep frying, which involves temperatures of 200°c (392°f) or more, gives

French fries and fritters their crispiness. Because water cannot withstand such

temperatures without boiling, the surface of fried food is dried out without

the water inside having time to diffuse outward. The crust that is formed in

this way is what produces the sensation of crispiness. If deep-fried foods were

cooked in a very hot oven, the results would be different. Chemists in Bristol

and Nantes have demonstrated that fats are an essential element of Maillard re-

actions between sugars and amino acids. These reactions differ depending on

the presence or absence of fatty matter; indeed, the good taste of the browned

surface of fried foods is directly attributable to fats. It is for this reason that

quail, for example, is wrapped in fat or bacon before roasting.

It is a mistake to baste meat with the juices that drip into the roasting pan,

by the way; these juices are mostly water, which softens the crispy surface (it-

self the consequence of the water inside the skin evaporating) and so produce

exactly the opposite result of what one hopes to achieve by roasting. Here again

fat is the cook’s friend. Ideally one would use a decanting drip pan to recover

the melted fat while eliminating the water.

| 291

From Roasts to Emulsions

Whether one uses oil or butter, fats are inevitable in mayonnaise, béarnaise,

hollandaise, beurre blanc, and other emulsified sauces formed from a base

thickened with butter or cream. These emulsions consist almost exclusively of

fatty matter. Oil droplets are packed together in water so closely that they no

longer have room to move, with the result that the sauce has trouble flowing.

In the case of some sauces the coagulation of the egg yolk adds solid particles

to the oil droplets dispersed in the water, yielding suspension emulsions rather

than emulsions proper.

Might it nonetheless be possible to increase the proportion of water and

reduce the proportion of fat? One could use a hand mixer rather than a fork to

divide the fatty matter up into smaller and therefore more numerous droplets,

but little would be gained. Alternatively, one could use thickeners and gelati-

nizing agents, but it is difficult to reproduce the fluid behavior of an emulsi-

fied sauce in a suspension (of expanded starch granules, for example) or a

concentrated solution (as when hydrocolloids—molecules surrounded by a lot

of water molecules—are dissolved in water).

This observation may help us finally to transform a small evil into a great

good. Butter, for example, acquires an unpleasant odor in the refrigerator

because many aromatic molecules are soluble in fats. Makers of perfume ex-

ploit this solubility in order to extract fragrances from the most delicate flow-

ers: They place freshly cut blossoms on a neutral fat for a few hours, then

discard them and melt the fat in order to recover the aromatic molecules that

have dissolved in it. This process is known as enfleurage. Butter serves as the

fatty substrate for enfleurage in a refrigerator, as do chocolate (made of cocoa

butter) and cream.

Separating Aromas

This property could be put to more systematic use in cooking. Why not

wrap cheeses in aromatic plants, for example, so that the aromatic molecules

slowly dissolve in the fatty matter of the cheese? We would also do well to recall

the underlying principle of a famous recipe for sage butter, recommended in

Italy as an accompaniment for pasta: When one cooks the leaves of this herb

292 | a c uisine f or t omor r ow

in butter, the heat causes their cells to burst and release aromatic molecules,

which are then dissolved in the melted butter.

Not all aromatic molecules are fat soluble, however. One way to dissolve

them is to use a separating funnel, long familiar to chemists as a useful device

for separating mixtures. Put oil and water in the funnel, and then add chopped

or ground pieces of an aromatic food such as cepe mushrooms. When the

funnel is shaken, the hydrophobic aromatic molecules are dissolved in the oil

while the hydrophilic aromatic molecules are dissolved in the water.

In this way two flavors are created out of one because the aromatic mol-

ecules are different in the two solvents. If you don’t have a separating funnel,

simply use a jar that can be hermetically sealed. Put oil and water in it, add an

aromatic food, and when the aromatic molecules have been dissolved, slowly