FOOD DYES ARE OFTEN disparaged today, but, as Cicero tells us, a man who knows only his own generation remains always a child. The history of cooking shows that coloring agents have been popular since the earliest times. In the Middle Ages, cooks used a variety of substances derived from spices and vegetables, even insects. Green was obligatory in the Christian West, where it symbolized the resurrection of Jesus Christ. Modern cooks continue to use the green pigment of spinach to color sauces, for example. Nothing could be simpler: grind up some spinach leaves in a blender, press out the juice through a cheesecloth-lined sieve, and then simmer over low heat until a very green froth—the dye itself—bubbles up from the brownish liquid.
Achieving mastery over the outward appearance of foods products is an ancient ambition because it has long been known that the faculty of taste can be fooled by the visual aspect of a dish. Chefs today continue to be intrigued by the possibility of using first impressions in order to influence the judgment of diners. Physiologists, for their part, have thoroughly investigated how the brain processes the information it receives about foods once they have been swallowed. The sensations detected by taste and olfactory receptors are encoded as neural signals in a dense and rapid train of electrochemical impulses. Only certain nerve fibers are excited in reaction to the detection of a particular flavor. Accordingly, a mapping of fiber-excitation patterns reflects the coexistence of various flavors. For each of our senses, a cerebral area of a few square centimeters registers a neural image of the sensation that has been transmitted to the brain for interpretation. First, this image is presented to memory, which reports back that it does or does not recognize the image or that the image resembles another previously stored image. Then the brain integrates all the relevant sensory and cognitive data, determines the level of pleasure aroused by the sensation, and commits this new piece of information to memory, where it forms part of the correspondingly enlarged archive that will be consulted in assessing the next sensation. In other words, we are guided in our judgment of a dish by the preliminary estimates made by the brain—in the same way that we are pleased once again to see someone we like because our brain associates seeing this person with a pleasurable response. This fact has a crucial implication, that our first impression is right (or at least not entirely mistaken)—which implies, in turn, that vision is an important aspect of our perception of flavor.
Experimental confirmation was strikingly provided a few years ago by researchers at the Institute of Oenology at the University of Bordeaux. Subjects trained in the use of wine-tasting terms (so-called descriptors that characterize the aroma, or bouquet, of red and white wines) were asked to describe two samples: a simple white wine that was made from a mixture of grapes and displayed no distinctive varietal character (or “typicity”), and the same wine colored red using pigments extracted from red wine. None of this was known to the subjects, who proceeded in all innocence to describe the white wine with the vocabulary of white wine and the faux red wine with the vocabulary of red wine! The tasters’ response is not as surprising as it seems. The eye is preeminent among human sense organs: visual information is the first type of sensory information to be processed since it is the first to reach the brain. Once a neural image has been received, the brain seeks to specify the conceptual category to which it belongs by consulting its archive of gustatory information. If it is fooled in the first place by the color of a substance, it has a hard time finding its way out from the category into which the true perception of this color initially led it.
Cooks are free, of course, to go to the trouble of making green dye from spinach and other natural foods, but why should they deprive themselves of coloring agents that food manufacturers are legally authorized to use? The list of additives formally approved for use in the European Union and Switzerland contains an impressive variety of products bearing E numbers in the 100s. Let’s take a brief tour, noting a few details along the way that may be of interest to cooks wishing to experiment with products they have never tried before.
E100 Curcumin is a yellow dye found in the rhizome of a plant of the ginger (Zingiberaceae) family, Curcuma longa, which contains turmeric acid, the substance that gives curry powders their pungency. Obviously curcumin is not to be confused with the spice itself, turmeric. It has the advantage of imparting color without also imposing the powerful trigeminal sensation of the spice. Should turmeric in its powdered form nevertheless be preferred to E100? There are two reasons for doubting it. First, the powder is obtained by grinding and drying, operations causing chemical degradations that make the powder less pure than the liquid extract. (Just so, it would be a welcome thing in my view if we were to stop calling turmeric powder “turmeric,” for the plant and the powder are very different in their molecular structure.) Second, the powder is much more liable to adulteration than E100, whose manufacture is strictly controlled.
E101 Riboflavin and riboflavin-5’-phosphate are yellow dyes extracted chiefly from yeasts, wheat germ, eggs, and animal livers. Their chemical synthesis is understood as well.
E102 Tartrazine is a synthetic yellow coloring agent. Food manufacturers sometimes mix it with Brilliant Blue (E133) or with Green (E142) to obtain various shades of green, particularly in order to add color to canned baby peas. One hears it said still today that tartrazine causes the testicles to shrivel and reduces sperm count, but as long ago as 1917 it was shown to carry no such risk—another sign of the irrationality to which we are prone in the face of perceived dangers to our health.
E104 Quinoline (or Quinoline Yellow WS, to give it its official name) is a synthetic yellow dye of a subtly different shade than its neighbors.
E110 Sunset Yellow (Orange Yellow S) is a synthetic orange dye.
E120 This code designates a trio of natural red dyes, cochineal, carminic acid, and carmine. They are prepared from aqueous, aqueous-alcoholic, and alcoholic extracts of the dried shell of the female cochineal insect (Dactylopius coccus Costa)—found in Central America, the Canary Islands, North Africa, and southern Spain—from which the color takes its name. The shell encloses eggs and young larvae. About fifteen thousand of these insects are needed to obtain a hundred grams of dye. The principal pigment is carminic acid.
E122 Carmoisine (or azorubine) is a synthetic red dye discovered about a century ago. First used (along with cochineal) by textile manufacturers, it was subsequently adopted by printers and later by the food industry.
Should the commercial migration of dyes from cloth production to edible products make us uneasy? The answer to all such questions depends on how a given compound interacts with the human body. Is it metabolized? If so, how? What chemical products are formed by the metabolic reactions? Do these products have receptors? In what quantities do they have an effect on the organism? Regulatory standards specify approved uses for additives. The use of carmoisine, for example, is approved in Europe (but not in the united States) for pork sausage casings as well as for preserved fruits, candies, ice creams, and beverages. Industrial applications are carefully monitored on the whole, but the artisanal use of such products often goes unnoticed. Titanium dioxide, for example, is a white dye approved for sausage casings, cheese rinds, decorative sugar work, and chewing gums, but I have seen more than one pastry chef use it for other purposes without anyone objecting. Evidence of a double standard?
E123 Amaranthe is a red synthetic dye. Some thirty years ago it became a source of toxicological controversy even though many studies had shown it to be harmless. Without going into the details of this episode, let me say simply that commercially manufactured additives are required to indicate on their labels an acceptable daily intake (ADI), which is based on a calculation known as LD50: the letters l and D stand respectively for “lethal” and “dose”; the number 50 means that this dose is sufficient to kill half the members of a test group of laboratory animals. In determining the ADI, toxicologists take great care to multiply the LD50 value by a considerable factor in order to offset the risks posed by worst-case scenarios.
Note-by-note cooks must be sure to inform themselves about all the precautions needing to be taken whenever they consider using a compound, either by researching the matter themselves or by consulting their supplier. To give you some idea of the proportions that are typically involved, the ADI for carmoisine is 0.8 milligrams per kilogram of body weight (about 0.36 milligrams per pound). That doesn’t sound like much, but it has to be kept in mind that carmoisine in its pure form is an extremely powerful staining agent. It should be used in minimal quantities.
E124 Ponceau 4R (also called Cochineal Red A) is a synthetic red dye that has nothing to do with the natural dye called cochineal that I mentioned earlier.
E127 Erythrosine is a red food coloring. As in the case of some other dyes, commercial use of its sodium, calcium, and potassium salts is also permitted. These salts are designated by the same code, supplemented by the letters a, b, c, and so on (or by the numerals i, ii, iii, and so on) as necessary.
E128 Red 2G (or azogeranine) is, as the name suggests, yet another red coloring agent. (I forgot to mention, by the way, that the dyes discussed here are for the most part powders, granules, or concentrated solutions.)
E129 Allura Red AC. Red, again—only different.
E131 Patent V Blue. A synthetic blue dye.
E132 Indigo carmine (or indigotine). In France it was a pigment of this type—used to make blue jeans (toiles de nîmes, source of the English word denim) blue—that helped bring wealth and prosperity to Toulouse and the neighboring area of Cocagne (between Toulouse, Castres, and Albi).
E133 Brilliant Blue, the blue dye I mentioned earlier in connection with tartrazine.
E140 Chlorophylls and chlorophyllins. Let’s pause here for a moment, for these pigments contribute very largely to the green color we see in algae and plants. There are many chlorophylls, and every vegetable contains several of them (which means that the term
chlorophyll, in the singular, should be avoided because it suggests there is only one kind). But they are not solely responsible for the color of vegetables. Nor are all of them green. The green we see in spinach, for example, is due to a combination of pigments in which several chlorophylls (yellow green, green, green blue) are mixed with carotenoids (yellow, orange, and red). Painters are therefore right to add red to green when they draw plants, for this in fact is how plants make their green colors—something that cooks who wish experiment with colors need to remember.
The additives jointly designated by the code E140 are produced by extraction from herbs, alfalfa, nettles, and other edible vegetable matter. Eliminating the solvent can lead to total or partial separation of the magnesium found in chlorophyll molecules, which changes the color somewhat. All chlorophylls have a plate-shaped molecule, more or less square with a sort of tail attached to one corner and a magnesium atom at the center. When a green vegetable is heated, especially in an acidic medium, the magnesium is expelled, and the chlorophylls are transformed into pheophytins, which have a bluish tint. This phenomenon is well known to cooks, who counteract it by adding baking soda to the cooking water to prevent the loss of magnesium and to retain the vibrant green color of fresh vegetables.
E141 Copper complexes of chlorophylls and chlorophyllins. These green pigments were first isolated as part of an attempt to explain why the characteristically vivid color of vegetables disappeared during prolonged cooking. It had been observed that vegetables kept their color when they were cooked in copper pans. Cooks, not realizing that the copper served to replace the lost magnesium at the center of the chlorophyll molecules, therefore became fond of using what they called “greening pans.” (For a certain time, as I mentioned earlier, it was even customary to add copper sulfate to the cooking water—a procedure that was subsequently found to pose unacceptable risks and was prohibited by law in France in 1902.) In such matters we should not be reluctant to place our confidence in government agencies: if the use of copper complexes of chlorophylls is permitted today, it is because exhaustive toxological testing over many years has shown it to be safe.
E142 Green S. Yet another green dye.
E150a Caramel. That’s right, ordinary caramel, used as a brown coloring. Generally speaking, compounds containing brown and black pigments have an odor of burnt sugar and an agreeable, slightly bitter taste. But how has plain caramel wound up on a list of color additives? Read on.
E150b Caustic sulfite caramel. Plain (let it be noted, not “natural”) caramel is obtained by cooking sucrose over high heat. It is a complex mixture: some of its constituent compounds impart consistency, others taste, others odor, and others still color. Cooks have long been aware that caramels differ depending on the conditions under which they are made. Syrups to which a dash of vinegar has been added, for example, caramelize differently than the same syrups containing no acid. Adding other ingredients yields other results. And so it is that the caramel identified by the code 150b comes by its name.
E150c Ammonia caramel, due to the addition of ammonia.
E150d Sulfite ammonia caramel. Oh, while I’m thinking of it: a few years ago the Italian ministry of health banned the use of coloring additives in restaurant kitchens as part of a demagogic campaign against molecular cooking. The additives in question? Caramel and its derivatives, such as sulfite ammonia caramel. Revolutions have been started for less than that!
E151 Black PN (or Brilliant Black BN). Yes, a black dye.
E153 Vegetable carbon. A black pigment produced by carbonizing vegetable matter such as wood, cellulose residues, peat, coconut husks, and the outer covering of other vegetables. The raw material is charred at high temperatures.
E154 Brown FK, a brown dye.
E155 Brown HT, yet another brown dye.
E160a Alpha-carotene, beta-carotene, gamma-carotene. Now we come to the carotenoids, pigments present in many vegetables and the source of the brilliant yellow, orange, and red colors of many edible fruits (lemons, peaches, apricots, strawberries, cherries, and so on), vegetables (carrots, tomatoes, and so on), mushrooms (chanterelles), and flowers. They are also present in a variety of animal products, including eggs, lobsters, langoustines, and fish. I won’t go into the molecular details except to note that heating may modify their color and that they can react with different compounds. Beta-carotene, abundant in carrots, is perhaps the best known of the three pigments designated by this code.
E160b Annatto, bixin, norbixin. Annatto is a seed that is very commonly used in South America, among other places. Some indigenous peoples in the Amazon, for example, use it to paint their skin red. Bixin is prepared by extracting annatto seeds from their husks with the aid of solvents, and hydrolyzed to norbixin by using an alkaline solution. Annatto extracts may be either aqueous or oily.
E160c Paprika oleoresin (paprika extract), capsanthin, capsorubin. The extract, obtained by the action of a solvent on natural strains of paprika, contains paprika’s principal pigments, capsanthin and capsorubin.
E160d Lycopene. A pigment naturally present in tomatoes and many other vegetables and fruits. It’s fun to make it yourself, but easier to use when it comes already made in a bottle.
E160e Beta-apo-8’-carotenal. Another red pigment.
E160f Ethyl ester of beta-apo-8’-carotenic acid. An orange-red to yellow pigment.
E161 This designation covers ten compounds, molecular neighbors of the carotenoids and known as xanthophylls. They are found in many vegetable substances and include the next two coloring agents.
E161b Lutein. A xanthophyll prepared by solvent extraction from edible fruits and plants as well as from herbs, in particular alfalfa (lucerne).
E161g Canthaxanthin. A cousin to lutein, canthaxanthin assumes the form of beautiful dark-violet crystals.
E162 Beetroot Red. More formally known as betanin, it is obtained from natural strains of red beetroot either by pressing crushed beetroot until it yields a liquid or by aqueous extraction from shredded beetroot, followed by enrichment of the betacyanins that constitute the main coloring principle.
E163 Anthocyanins—a world of their own, for they comprise the coloring substances of fruits and flowers. Take a rose, grind it up with water, and strain the liquid, and you are left with an impure solution of anthocyanins, the pigments of the rose. If you acidify this solution by adding a drop of white vinegar, you see that the color changes. Now add a little baking soda to raise the ph (by lowering the acidity), and you get another color. It is generally true of anthocyanins that their color depends on the acidity of the medium and on the presence of metal ions. If you add a little iron (found in eggs and meat) or a little zinc or a little aluminum, the colors completely change. Indeed, the entire range of shades can be recreated, from yellow to violet.
Most anthocycanins used as food dyes are prepared by extraction from edible vegetable tissues using sulfited water, acidified water, carbon dioxide, methanol, or ethanol. The pigments themselves contain organic acids, sugars, mineral salts, and tannins. Tannins are an especially interesting kind of phenolic compound, by the way, because they are neither sweet nor sour nor bitter nor salty; they have no odor, nor are they cool or pungent. They are, well, tannic—which means that in binding with proteins in the saliva, they suppress the lubricating action of these proteins, causing the soft tissues in the mouth to feel as though they have been constricted. When the astringent effect is strong, it may be disagreeable, but when it is slight, as in the case of certain grapes, the sensation is marvelous. Conveniently enough, the winemaking industry sells a whole host of “oenological” tannins for the note-by-note cook to play with!
Some colorants are authorized for external use only:
E170 calcium carbonate (chalk)
E171 titanium dioxide (mentioned earlier)
E172 iron oxides and iron hydroxides
E173 aluminum
E174 silver
E175 gold
E180 lithol rubin BK (a reddish synthetic powder approved for staining cheese rinds)
I should emphasize that there is no reason why these pigments must be used in their pure form. By mixing them, note-by-note cooks will be able to produce a range of almost unimaginably brilliant colors and shades, sharper and more vivid than the ones with which they are familiar from traditional cooking. Recipes for new dishes will need to be developed, of course, but this will be scarcely more difficult than learning to paint with gouache—child’s play really, a game that takes its place alongside the ones we learned to play earlier involving shape, consistency, taste, and odor.
The idea of using artificial food dyes has a way of crystallizing unfounded fears and strengthening popular opposition to “chemistry.” The most recent charge is that these products are responsible for hyperactivity in children. Even if hyperactivity were to be proved, however, would it not more likely be the result of kids spending too much time in front of television and computer screens? However this may be, fearmongers are the scourge of the modern world. Some of them are honest, others are not; some of them are naive, others are not. All of them are opposed by trained specialists who reject their opinions—alas, to no avail. If people no longer trust experts or politicians or government agencies or the press, can the day be far off when even our academies of science no longer command respect?
For many years now, much of the public has believed that synthetic colorants should be banned because they are “bad,” but that natural colorants are “good,” especially if they have been prepared without an organic solvent. Without an organic solvent? Ethanol, of which we are so fond, is an organic solvent! Our inconsistency on this point is really quite troubling. We want to have colors—without coloring agents. And yet several years ago, when makers of a certain mint syrup stopped using the green dye they had always added (mint syrup is colorless, as it happens), the clamor for its return was deafening. Arguing over whether natural green should be used rather than synthetic green is utterly beside the point: we have seen more than once that what is natural is not always good and what is artificial is not always bad—quite to the contrary.
In making sense of the claim that natural colorants are inherently superior to artificial colorants, it will be instructive to consider a pigment that French scientists at the National Institute for Agricultural Research laboratory in Rennes recently discovered. They succeeded in isolating a new water-soluble yellow dye, POP2 (POP stands for “phloridzyn oxidation product”), that can be used as a natural substitute for tartrazine (first synthesized and approved for commercial use almost a century ago). In effect, POP2 is a precious waste product. Ten million hectoliters (almost 265 million gallons) of apple juice are pressed from 45 million tons of apples in the world each year, which makes a massive quantity of solid-waste material available for productive uses of one kind or another. Until recently the food industry had been content to extract pectins from this waste material, which are widely used, as we have seen, in manufacturing fruit preserves, fibers, and animal feed. But the pulp of apples also contains phenolic compounds, all of whose molecules contain at least one benzene group (six carbon atoms arranged in the form of a hexagon) and one hydroxyl group (an oxygen atom bonded to a hydrogen atom). One of these molecules, phloridzin, specifically associated with the Rosaceae family (and especially apples), happens to be the precursor of POP2.
Phloridzin is chiefly present in seeds, which means that pressing apples concentrates it. The conversion of this precursor into a pigment results from a reaction that, under other circumstances, would be considered harmful. Phenolic compounds are responsible for many foods turning brown once they have been cut up: the knife releases these molecules, along with enzymes known as polyphenol oxidases, which in intact cells are found in separate compartments. In the presence of air, the enzymes transform the phenolic compounds into reactive quinones, which subsequently form dark compounds. This is why apples, avocados, pears, and indeed the majority of fruits and vegetables turn brown.
Cooks and food engineers make a special point of blocking this reaction. Cooks use lemon juice, whose ascorbic acid (vitamin C) chemically lowers the concentration of quinones. Food engineers have a number of methods at their disposal, both physical (centrifugation, which eliminates browned liquids) and chemical (various antioxidants). Now, it is exactly these normally harmful enzymatic oxidation reactions that produce POP2, which turns out to be yellow rather than brown: chemists studying the phenolic compounds found in apples observed that the color of the pulp changed during storage. Using a separation process known as liquid-phase chromatography (in which a mixture is dissolved in a fluid and injected into a column filled with a pulverulent solid so that the various compounds of the mixture will migrate at different speeds), it became clear that three new phenolic compounds are formed. The first one is a derivative in which a hydrogen atom is replaced by a hydroxyl group. This intermediary is then transformed into another colorless compound, POP1, which subsequently is transformed into POP2, an intense yellow pigment.
The reaction that produces POP2 is remarkable for three reasons. First, it is the apple that furnishes both the precursor, phloridzin, and the enzyme. Next, although the reaction occurs more rapidly with increasing temperature, its yield is the same at 10° or 30° or 40°C (50° or 86° or 104°F), so there is no point expending additional energy to heat the material more quickly. Finally, the POP1 that is produced along with POP2 is a good antioxidant (and may come to have a related application one day); POP2, however, is water soluble like many phenolic compounds and in low concentrations produces a saturated color varying from brilliant yellow (between pH 3 and pH 5) to orange for less acidic media. Will POP2 manage to establish itself in the market? A commercial patent has been granted, but the substance itself will nevertheless have to pass toxicological tests, just as synthetic compounds do, and no one knows in advance whether it will turn out to be safe or unsafe. The moral of the story is that in matters of toxicology we are well advised to adopt an impartial attitude—whether the products being tested are natural or synthetic.
Davidson, Alan. Oxford Companion to Food. Oxford: Oxford University Press, 1999. See especially the entries on food additives and coloring.
Food Additives in Europe 2000: Status of Safety Assessments of Food Additives Presently Permitted in the EU. Copenhagen: Nordic Council of Ministers, 2002.
Guernevé, C., P. Sanoner, J. F. Drilleau, and S. Guyot. “New Products Obtained by Enzymatic Oxidation of Phloridzin.” Tetrahedron Letters 45 (2004): 6673–6677.
Guyot, S., S. Serrand, J. M. Le Quéré, P. Sanoner, and C. M. G. C. Renard. “Enzymatic Synthesis and Physiochemical Characterisation of Phloridzin Oxidation Products (POP), a New Water-Soluble Yellow Dye Deriving from Apple.” Innovative Food Science & Emerging Technologies 8 (2007): 443–450.
Milgrom, Lionel R. The Colours of Life: An Introduction to the Chemistry of Porphyrins and Related Compounds. Oxford: Oxford University Press, 1997.
Rossotti, Hazel. Colour: Why the World Isn’t Grey. Princeton, N.J.: Princeton University Press, 1985.
Scotter, M. J. “Methods for the Determination of EU-Permitted Added Natural Colours in Foods: A Review.” Food Additives & Contaminants 28, no. 5 (2011): 527–596.
Serrand,S., S. Bernillon, J. M. Le Quéré, P. Sanoner, and S. Guyot. “The Role of Polyphenoloxidase in the Synthesis of a Yellow Pigment Derived from Apple.” In Proceedings of Enzymes for Food—Symposium européen, Rennes, France (2006), 131–139. Paris: Institut National de la Recherche Agronomique, 2006.
This, Hervé. “Artificiel ou synthétique? Les fruits sont des mines de molécules utiles et encore non exploitées: Le POPj est un nouveau colorant jaune extrait des pommes très intéressant.” Pour la science 348 (October 2006): 4.
