Chapter VII · FLAVORING EXTRACTS
Reading edition
LEMON EXTRACT
The important ingredients of lemon extract are lemon oil and citral, its aromatic constituent. Oil of citronella and oil of lemon grass are sometimes substituted for lemon oil. Methyl alcohol is sometimes used in place of the more expensive spirit alcohol as a solvent for the lemon oil.
Lemon Oil
The presence of lemon oil may be detected by adding a large excess of water to a small amount of the extract in a test tube. If the mixture does not show some cloudiness, it is a strong indication that no lemon oil is present. The degree of cloudiness gives an idea of the amount of oil present.
Citral
This is present in the oil of lemon grass, which is sometimes used as a substitute for lemon oil. Citral may be detected by the following test by Burgess:
Add 20 cc. of sulfuric acid to 85 cc. of water. Dissolve in this mixture 10 grams of mercuric sulfate. Shake 2 cc. of the sample with 5 cc. of this reagent in a test tube. If citral is present, the liquid will be bright red, and will quickly disappear and give place to a whitish compound, which floats on top.
Oil of Citronella
This is often substituted for lemon oil. It may be detected by the same test which was used for citral. Instead of the red color and the white compound, citronella gives a bright yellow color, which does not disappear for some time.
Tartaric or Citric Acid
Precipitate the oil by the addition of an equal volume of water. Filter and add a very little of the filtrate to a test tube nearly full of cold lime water. A precipitate will form if tartaric acid is present and it will dissolve in an excess of ammonium chlorid or acetic acid. Filter, or, in case no precipitate forms, heat the liquid. Citric acid is precipitated in the presence of a large excess of hot lime water.
Methyl Alcohol
Mullikin and Scudder.—Take 2 ft. of No. 12-15 copper wire and bend at right angles about 8 or 10 inches from one end. Grasp this bent end and an ordinary lead pencil side by side in such a way that the bend will be about the middle of the pencil. Wind the wire around the pencil and toward the free end of the short part of the wire until you have a coil 3 or 4 cm. long. Remove the pencil and twist the unwound parts together for a handle for the coil.
Dilute a portion of the sample 3 or 4 times, and oxidize 10 cc. of the diluted liquid (preferably in a test tube) by heating to a red heat the above copper coil in the oxidizing flame of a Bunsen lamp. Thrust the heated coil quickly into the liquid contained in the test tube. In a second withdraw and immerse in water. Continue this operation till the oxid of copper fails to be reduced (usually 4 or 5 times is sufficient). Cool the liquid by immersing the tube in water. Separate into two parts and test each for formaldehyde by the following methods:
Mullikin, S. P.—Place one of these parts in an evaporating dish, and add to it 1 cc. of strong ammonia, boil gently over the free flame till the vapors cease to smell of ammonia. Add 2 or 3 drops of strong hydrochloric acid and heat just to boiling, and cool quickly by dipping the dish in cold water. Make the test for formaldehyde: Add a drop of a solution of resorcin (1 : 200) and pour this mixture slowly down the side of an inclined test tube which contains 3 cc. of strong sulfuric acid, taking care to keep the liquids separate. After 3 minutes give the tube a rotary motion by rolling between the hands for a minute or more but only gradually mixing the water and acid, but mixing only about half of the acid.
Flecks of a rose-red color form if methyl alcohol is present. Bands of color or flecks of other colors, even though they be tinged with red or a rose-red solution without the precipitate, should never be taken as proof that methyl alcohol is present. These conditions, however, are good grounds for repeating the test; 10 per cent or even less may be detected by this test.
Hydrochloric Acid and Ferric Chlorid Test.—Add a few drops of the other part of the above oxidized liquid to about 10 cc. of milk, known to be free from formaldehyde, in a porcelain casserole, and add 10 cc. of commercial hydrochloric acid (sp. gr. 1.2) which contains 1 cc. of 10 per cent ferric chlorid per 500 cc. Heat slowly over the open flame nearly to boiling. Give the liquid a rotary motion to break up the curd. If formaldehyde is present, the liquid will be colored violet. If not, it slowly turns brown. The presence of formaldehyde proves that methyl alcohol was in the original extract.
Coloring Matter
Preliminary Test.—Treat the sample with strong hydrochloric acid, and if tropæolin or methyl orange be present the solution will turn pink; Martius yellow partially decolorizes the solution; dinitrocresols decolorizes the solution. Turmeric or naphthol yellow produces no color change.
Turmeric
Turmeric may be detected by soaking a piece of filter paper in the sample, drying and dipping it in a dilute solution of boric acid or borax which has been slightly acidulated with hydrochlorid acid. Dry again and a cherry-red color forms if turmeric is present. Add a drop of dilute alkali and if turmeric be present the paper will be colored dark olive.
Coal-Tar Colors
Evaporate some of the extract to dryness; take up the residue with water and extract the coal-tar colors if present, and test for them by the method given under canned vegetables.
VANILLA EXTRACTS
The best grades of vanilla extract are made by treating vanilla beans with 50 per cent alcohol. Coumarin, an extract from tonka beans, may be used in making the extract. This of course would make a cheaper product. If less than 50 per cent alcohol is used in making the extract, some alkali must be added to dissolve the resins which will not dissolve in a weaker alcohol. In artificial extracts some such coloring matter as caramel or tannin is used.
Preliminary Test.—To a portion of the extract add a few drops of lead acetate solution. The absence of a bulky flocculent precipitate shows the extract not to be of high quality. Leach recommends that normal acetate of lead be added to the sample, and if a precipitate does not form it is conclusive evidence that it is not a pure extract.
When a precipitate forms with the above reagent, it should settle immediately and leave a clear and almost colorless liquid. When there is a mere cloudiness only, it may be due to caramel, in which case the extract is to be suspected.
Alkali
Shake a portion of the sample with twice its volume of water. If no precipitate forms, an alkali is present. A flocculent reddish-brown precipitate shows no alkali is present. If the solution is milky it indicates the presence of a foreign resin.
Add hydrochloric acid drop by drop to the diluted extract. Nothing more than a mere turbidity should result. Should it be quite turbid and the color fading after a time, it shows that an alkali has been used.
Foreign Resins
Mix a portion of the extract slowly with twice its volume of water, frequently shaking the mixture. When this solution is milky, it indicates a foreign resin.
Hess’ Test.—Dealcoholize 25 cc. of the sample by concentrating on the water-bath, adding water from time to time to retain the original volume. When no alkali is present in the extract, pure vanilla resin will be thrown down as a reddish-brown flocculent precipitate. Collect the resin, whatever its color, on a filter, and wash. Save the filtrate to test for caramel. Place a piece of the paper and resin in a dilute solution of potassium hydroxid. If the resin is that of pure vanilla it will dissolve, giving a deep-red color, and is reprecipitated when the alkali is neutralized with hydrochloric acid. Dissolve another part of the precipitate in alcohol, and to a part of this solution add a few drops of ferric chlorid; and to the other part, hydrochloric acid. There should be no marked coloration in either case if the resin is that of pure vanilla. Foreign resins nearly always produce a coloration.
CARAMEL
Shake the bottle of vanilla, and if the bubbles, which form, are a bright caramel color, keeping the color till all are gone, the presence of caramel is indicated.
Concentrate a portion of the filtrate, which was saved in making the test for foreign resins, at a rather low temperature until it has about the same color as the original extract. Add a few drops of strong hydrochloric acid and heat very gently. If caramel is present, a yellowish-red flocculent precipitate will form. After the liquid cools, filter and wash with water. Should this precipitate contain caramel, it will not dissolve in water, ether, and alcohol, but will dissolve in sodium hydroxid, dilute alcohol, and glacial acetic acid.
Tannin
Test another portion of the filtrate made in testing for foreign resins, with a few drops of a solution of gelatin. A slight precipitate only should form due to the presence of a small amount of tannin normally present in this filtrate. A large excess shows that it has been added to the extract.
Coumarin
Leach’s Test.—Dealcoholize a portion of the sample as above and treat with ammonia, add 3 or 4 volumes of chloroform in a separatory funnel. Evaporate the chloroform extract in an oven, not permitting the temperature to rise above 60° C. To the residue add a few drops of water; warm gently, and add a little of a solution of 1 gram of crystallized potassium iodid in 50 cc. of water, and the solution saturated with iodin. If coumarin is present, a brown, precipitate will form, and if stirred with a rod it will collect in dark green flecks.