TEA AND COFFEE.

579. Special Analysis.—Aside from the examination of teas and coffees for adulterants, the only special determinations which are required in analyses are the estimation of the alkaloid (caffein) and of the tannin contained therein. It is chiefly to the alkaloid that the stimulating effects of the beverages made from tea and coffee are due. The determination of the quantity of tannin contained in tea and coffee is accomplished by the processes described under the chapter devoted to that glucosid.

The general analysis, viz., the estimation of water, ether extract, total nitrogen, fiber, carbohydrates and ash, with the exceptions noted above, is conducted by the methods which have already been given.

For detailed instructions concerning the detection of adulterants of tea and coffee the bulletins of the Chemical Division, Department of Agriculture, may be consulted.[593]

580. Estimation of Caffein (Thein).—The method adopted by Spencer, after a thorough trial of all the usual processes for estimating this alkaloid, is as follows:[594] To three grams of the finely powdered tea or coffee, in a 300 cubic centimeter flask, add about a quarter of a liter of water, slowly heat to the boiling point, using a fragment of tallow to prevent frothing, and boil gently for half an hour. When boiling begins, the flask should be nearly filled with hot water and more added from time to time to compensate for the loss due to evaporation. After cooling, add a strong solution of basic lead acetate until no further precipitation is produced, complete the volume to the mark with water, mix and throw on a filter. Precipitate the lead from the filtrate by hydrogen sulfid and filter. Boil a measured volume of this filtrate to expel the excess of hydrogen sulfid, cool and add sufficient water to compensate for the evaporation. Transfer fifty cubic centimeters of this solution to a separatory funnel and shake seven times with chloroform. Collect the chloroform solution in a tared flask and remove the solvent by gentle distillation. A safety bulb, such as is used in the kjeldahl nitrogen method, should be employed to prevent entrainment of caffein with the chloroform vapors.

The extraction with chloroform is nearly complete after shaking out four times; a delicate test, however, will usually reveal the presence of caffein in the watery residue even after five or six extractions, hence seven extractions are recommended for precautionary reasons. The residual caffein is dried at 75° for two hours and weighed.

The principal objection which has been made to Spencer’s method is that the boiling with water is not continued for a sufficient length of time. For the water extraction, Allen prescribes at least six hours cohobation.[595] In this method six grams of the powdered substance are boiled with half a liter of water for six hours in a flask, with a condenser, the decoction filtered, the volume of the filtrate completed to 600 cubic centimeters with the wash water, heated to boiling, and four cubic centimeters of strong lead acetate solution added, the mixture boiled for ten minutes, filtered and half a liter of the filtrate evaporated to fifty cubic centimeters. The excess of lead is removed with sodium phosphate and the filtrate and washings concentrated to about forty cubic centimeters. The caffein is removed by shaking four times with chloroform. Older but less desirable processes are fully described by Allen.[596]

In France this method is known as the process of Petit and Legrip, and it has been worked out in great detail by Grandval and Lajoux and by Petit and Terbat.[597]

581. Estimation of Caffein by Precipitation with Iodin.—The caffein in this method is extracted, the extract clarified by lead acetate and the excess of lead removed as in Spencer’s process described above. The caffein is determined in the acidified aqueous solution thus prepared, according to the plan proposed by Gomberg, as follows:[598]

Definite volumes of the aqueous solution of the caffein are acidulated with sulfuric and the alkaloid precipitated by an excess of a set solution of iodin in potassium iodid. After filtering, the excess of iodin in an aliquot part of the filtrate is determined by titration with a tenth normal solution of sodium thiosulfate. The filtration of the iodin liquor is accomplished on glass wool or asbestos. The results of the analyses are calculated from the composition of the precipitated caffein periodid; viz., C₈H₁₀N₄O₂.HI.I₄. The weight of the alkaloid is calculated from the amount of iodin required for the precipitation by the equation 4I: C₈H₁₀N₄O₂ = 508: 194. From this equation it is shown that one part of iodin is equivalent to 0.3819 part of caffein, or one cubic centimeter of tenth normal iodin solution is equal to 0.00485 gram of iodin.

In practice, it is recommended to divide the aqueous extract of the alkaloid, prepared as directed above, into two portions, one of which is treated with the iodin reagent without further preparation, and the other after acidulation with sulfuric. After ten minutes, the residual iodin is estimated in each of the solutions as indicated above. The one portion, containing only the acetic acid resulting from the decomposition of the lead acetate, serves to indicate whether the aqueous solution of the caffein contains other bodies than that alkaloid capable of forming a precipitate with the reagent, since the caffein itself is not precipitated even in presence of strong acetic acid.

In the solution acidulated with sulfuric, the caffein, together with the other bodies capable of combining with iodin, is precipitated. The residual iodin is determined in each case, and thus the quantity which is united with the caffein is easily ascertained. The weight of iodin which has entered into the precipitated caffein periodid multiplied by 0.3819 gives the weight of the caffein in the solution.

Gomberg’s method has been subjected to a careful comparative study by Spencer and has been much improved by him in important particulars.[599]

It is especially necessary to secure the complete expulsion of the hydrogen sulfid and to observe certain precautions in the addition of the iodin reagent. The precipitation should be made in a glass-stoppered flask, shaking thoroughly after the addition of the iodin and collecting the precipitate on a gooch. As thus modified, the iodin process gives results comparable with those obtained by Spencer’s method, and it can also be used to advantage in estimating caffein in headache tablets in the presence of acetanilid.

582. Freeing Caffein of Chlorophyll.—Any chlorophyll which may pass into solution and be found in the caffein may be removed by dissolving the caffein in ten per cent sulfuric acid, filtering, neutralizing with ammonia and evaporating to dryness. The residue is taken up with chloroform, the chloroform removed at a low temperature and the pure caffein thus obtained.[600]

583. Proteid Nitrogen.—The proteid nitrogen in tea and coffee may be determined in the residue after extraction of the alkaloid by boiling water as described above. More easily it is secured by determining the total nitrogen in the sample and deducting therefrom the nitrogen present as caffein. The remainder, multiplied by 6.25, will give the quantity of proteid matter.

584. Carbohydrates of the Coffee Bean.—The carbohydrates of the coffee bean include those common to vegetable substances; viz., cellulose, pentosan bodies (xylan, araban), fiber etc., together with certain sugars, of which sucrose is pointed out by Ewell as the chief.[601] In smaller quantities are found a galactose yielding body (galactan), as pointed out by Maxwell, a dextrinoid and a trace of a sugar reducing alkaline copper solution.

The sucrose may be separated from the coffee bean by the following process:[602] The finely ground flour is extracted with seventy per cent alcohol, the extract clarified with lead acetate, filtered, the lead removed from the filtrate with hydrogen sulfid, the excess of the gas removed by boiling, the filtrate evaporated in a partial vacuum to a sirup and the sucrose crystallized from a solution of the sirup in alcohol.

For a quantitive determination, ten grams of the coffee flour are extracted with ether and the residue with seventy-five per cent alcohol. This process, conducted in a continuous extraction apparatus, should be continued for at least twenty-four hours. The alcohol is removed by evaporation, the residue dissolved in water, clarified with basic lead acetate, filtered, the precipitate washed, the lead removed, again filtered, the filtrate washed and wash water and filtrate made to a definite volume. In an aliquot part of this solution the sugars are determined by the alkaline copper method, both before and after inversion. From the data obtained the percentage of sucrose is calculated.

In a coffee examined by Ewell the percentage of sucrose was found to be 6.34. The pentose yielding constituents of the coffee bean amount to from eight to ten per cent.

When coffee meal is extracted with a five per cent solution of sodium carbonate, a gummy substance is obtained, which is precipitable by alcohol. This gum, after washing with hydrochloric acid containing alcohol, gives a gray, translucent, hard mass on drying. On hydrolysis it yielded 75.2 per cent of dextrose, on distillation with hydrochloric acid, thirteen per cent of furfuraldehyd and, on oxidation with nitric acid, 18.7 per cent of mucic acid. This gum, therefore, consists chiefly of a mixture of galactan, xylan and araban.

585. Estimation of Galactan.—From three to five grams of the substance supposed to contain galactan are placed in a beaker with sixty cubic centimeters of nitric acid of 1.15 specific gravity. The mixture is evaporated on a steam bath until it is reduced to one-third of its original volume, allowed to stand for twenty-four hours, ten cubic centimeters of water added, well stirred and again allowed to stand for twenty-four hours, until the mucic acid is separated in a crystalline form. To remove impurities from the mucic acid it is separated by filtration, washed with not to exceed twenty cubic centimeters of water, placed together with the filter in the beaker, from twenty-five to thirty cubic centimeters of ammonium carbonate solution, containing one part of dry ammonium carbonate, nineteen parts of water and one part of ammonium hydroxid, added and heated to near the boiling point. The mucic acid is dissolved by the ammonium carbonate solution and any insoluble impurity separated by filtration, the filtrate being received in a platinum dish, the residue well washed and the entire filtrate and wash water evaporated to dryness on a steam bath acidified with dilute nitric, well stirred and allowed to stand until the mucic acid separates in a crystalline form. The separation is usually accomplished in half an hour, after which time the crystals of mucic acid are collected on a tared filter, or gooch, and washed with not to exceed fifteen cubic centimeters of water followed with sixty cubic centimeters of alcohol, then with ether, dried at 100° and weighed. For computing the amount of galactose, one gram of the mucic acid is equal to 1.333 of galactose and one gram of galactose is equal to nine-tenths gram of galactan. Before the commencement of the operation, the material should be freed of fatty matters in the case of oily seeds and other substances similar thereto.[603]

586. Revised Factors for Pentosans.—The factors given in paragraph [154] have lately been recalculated by Mann, Kruger and Tollens, and as a result of their investigations the following factors are now recommended.[604] The quantity of furfurol is derived from the weight of furfurolhydrazone obtained by the formula:

The pentoses (xylose, arabinose) may be calculated from the pentosans (xylan, araban) by dividing by 0.88.

The method of procedure preferred for the estimation of the pentosans is that described in paragraph [157]. The phloroglucin is dissolved in hydrochloric acid of 1.06 specific gravity before it is added to the furfurol distillate. The latest factor for converting the phloroglucid obtained into furfurol is to divide by 1.82 for small quantities and 1.93 for large quantities. After the furfurol is obtained, the factors given above are applied.

587. Application of Roentgen Rays to Analysis.—The detection of mineral matters in vegetable substances by roentgen photography has been proposed by Ranvez.[605] This process will prove extremely valuable in detecting the lacing of teas with mineral substances. Practically, it has been applied by Ranvez in the detection of mineral substances mixed with saffron with fraudulent intent.

Barium sulfate is often mixed with saffron for the purpose of increasing its weight. Pure saffron and adulterated samples are enclosed in capsules of black paper and exposed on the same sensitive plate for a definite time to the rays emanating from a crookes tube. In this case the pure saffron forms only a very faint shadow in the developed negative, while the parts to which barium sulfate are attached produce strong shadows. The principle involved is applicable to a wide range of analytical research.