NOVEMBER, 1951 Vol. 76, No. 908 THE ANALYST PROCEEDINGS OF THE AND OTHER SOCIETY OF PUBLIC ANALYSTS ANALYTICAL CHEMISTS DEATHS WE regret to record the deaths of Bernard Cracroft Aston (Honorary Member) Charles Wesley Bayley The Titrimetric Determination of Carbon Dioxide Liberated in the Ninhydrin Reaction with Amino-Acids BY A. M. SMITH AND A. H. AGIZA The a-amino nitrogen in single amino-acids or in mixtures can be accurately determined by measuring the carbon dioxide released by ninhydrin. A simple and convenient unit for the purpose and experimental conditions suitable for a quantitative .reaction are described. Recoveries of from 97 to 102 per cent. were attained with amounts corresponding to about 0.15 mg of a-amino nitrogen. ONE of the standard methods of determining amino-acids is based upon the determination of the carbon dioxide liberated in their decarboxylation by ninhydrin.This reaction is practically confined to the carboxyl carbon adjacent to the amino group and under suitable conditions most of the a-amino-acids, including glutamic acid, liberate one molecule of carbon dioxide; proline and hydroxyproline react similarly, but aspartic acid and cystine give two molecules of carbon dioxide, and the carboxyl group of p-alanine is only slightly reactive. The method is, therefore, extremely useful for measuring the total a-amino nitrogen in a mixture of amino-acids and other nitrogenous compounds likely to occur in a protein hydro- lysate. In this respect the method is complementary to that involving the measurement of the colour produced by the ninhydrin, a method that is more sensitive than the determina- tion of the carbon dioxide but is not so specific and can be used only for individual amino-acids whose identity is known (Smith and Agizal).Mason2 and Van Slyke, Dillon, MacFadyen and Hamilton3 have shown how the carbon dioxide may be measured manometrically with great precision, but the equipment is necessarily rather elaborate and simpler techniques for the titrimetric determination of the carbon dioxide have been devised by Christensen et ~ 1 . ~ 1 ~ and Van Slyke, MacFadyen and H a m i l t ~ n . ~ I t is obvious that in the measurement of the small amounts of carbon dioxide corresponding to less than 1 mg of a-amino nitrogen, special care must be taken to avoid contamination with atmospheric carbon dioxide, and the apparatus and procedure described below were designed to provide a convenient means of achieving this throughout the analysis. 619620 SMITH ASD AGIZA: THE TITRIMETRIC [Vol.76 METHOD APPARATUS- The reaction tube, A, and the absorption flask, B, 26 or 50 ml, are connected to each other by capillary tubes X and Y through the stopcock, S, which in turn connects them through capillary tube 2 with either a vacuum pump or a test tube, C, containing sodium hydroxide. The normal precautions were taken over the selection The apparatus is illustrated in Fig. 1. A‘ s. position t A s, position 2 =!&+ 4 s, position 3 -FEY- 1 . \ - _ , Fig. 1. Reaction and absorption apparatus and cleansing of rubber connections. A small burette is fitted to the stopper of the reaction tube and the whole apparatus is made air-tight, the stopcocks being lubricated with a high- vacuum grease.A sheet of asbestos between A and B helps to preserve the temperature gradient during the distillation of the carbon dioxide. A compact and practically continuous form of aspirator is shown in Fig. 2. Two bottles of 1 litre capacity are built into a rigid unit by means of stoppers and glass tubing. The rubber tubing A connecting the two bottles is long enough to allow either bottle to be turned uppermost so that when the lower one has filled with water, the bottles are inverted, the appropriate stopcocks being closed or opened to maintain the ctlrrent of air through the solution of sodium hydroxide in the flask B.REAGENTS- The reagents are selected according to the approximate quantity of amino-acids taken, namely, (a) “macro,” (b) “micro” and (c) “submicro” for 3 to 5, 0.15 to 1.0 and less than 0.15 mg of a-amino nitrogen respectively. Niahydri.n-Solutions containing (a) 150, (b) 100 and (c) 50 mg of ninhydrin in 1 ml of a citrate buffer of pH 5 are freshly prepared each week.Kov., 19511 DETERMISATION OF CARBON DIOXIDE 62 1 Barium hydroxide-(a) Approximately 0.25 N containing 2 per cent. of barium chloride, (b) approximately 0.125 N containing 2 per cent. of barium chloride and (c) approximately 0.016 N containing 10-5 per cent. of barium chloride. Hydrochloric acid-(a) 0.1428 N , (b) 0.07138 N , (c) 0.02855 or 0.01428 N . These con- centrations are convenient since 1 ml is equivalent respectively to 1.0, 0.5, 0.2 or 0.1 mg of a-amino nitrogen.- Fig. 2. Continuous aspirator Im?icators-For (a) and (b), 1 drop of a 1 per cent. solution of phenolphthalein in 95 per cent. ethanol; for (c), 1 drop of a 0-04 per cent. solution of cresol red in water. Bufers-A solid buffer of pH 4.7 is added to the amino-acid solution ; it consists of 17.65 g of tri-sodium citrate (Na,C,H5O,.2H,O) and 8.40 g of citric acid (C,H,O,.H,O) separately ground and then intimately mixed in a mortar. As a standard of pH 8 for the “submicro” titration, 10-3 g of sodium veronal are dissolved in 500 ml of water, and 7 ml of this solution are mixed with 4 ml of 0.07138 N hydrochloric acid. PROCEDURE- Place 2 to 3 ml of the amino-acid solution and about 100 mg of the solid citrate buffer in the reaction tube, insert the stopper and put a solution of ninhydrin in the burette.Remove the carbon dioxide from the tube by placing it in boiling water and connecting it through XZ (Fig. 1) to the suction pump; turn the appropriate stopcocks and allow air, free from carbon dioxide, to enter the tube by way of the sodium hydroxide solution in C. Repeat this procedure a t least three times. Remove carbon dioxide also from the absorption flask by passing a current of air, free from carbon dioxide, from an aspirator and, while the current is still passing, add 1 to 3 ml of barium hydroxide from a protected automatic burette. Attach the flask to tube Y and evacuate the whole system, refill with air, free from carbon dioxide, and again evacuate with the tap S at position 1.Turn S to position 2, add a suitable amount of ninhydrin solution (usually 1 ml) to the tube A and seal the burette with a drop of mercury. Immerse flask B in an ice-bath and the reaction tube in boiling water for 6 or 7 minutes until the reaction is complete and the carbon622 SMITH AND ACIZA: T H E TITRIMETHIC [Vol. 76 dioxide has distilled with most of the water from the tube to the flask. Shake the flask frequently during the process to bring the carbon dioxide into contact with the barium hydroxide. Turn S to position 3 so that any unabsorbed carbon dioxide in Y may be washed into the flask by a current of carbon dioxide-free air from C. Disconnect the flask, add the indicator and titrate the excess of barium hydroxide immediately with acid while a current of air, free from carbon dioxide, is passing through the solution from the aspirator.Risk of contamination with atmospheric carbon dioxide is eliminated by the careful evacuation of the apparatus before the reaction and by filling the absorption flask with carbon dioxide-free air before disconnecting it for the titration. Blank experiments with carbon dioxide-free water in place of amino-acid solution are carried out from time to time, and the amount of amino nitrogen in milligrams is simply (a - b) x c, where a and b are the amounts of hydrochloric acid in millilitres required to neutralise the excess of barium hydroxide in the blank and in the experimental solutions, respectively, and c is the factor for the acid.The presence of the barium chloride facilitates the precipitation of the carbonate, and with “macro” and “micro” quantities the titration is carried out to the first disappearance of the colour of the phenolphthalein. For “submicro” quantities, the end-point of the titration is matched against the colour of an equal volume of the veronal buffer containing 1 drop of cresol red. DISCUSSION OF RESULTS The time required for a quantitative reaction is not the same for all amino-acids and is shorter at pH 4-7 than at pH 2.5 or 1. Van Slyke, Dillon, MacFadyen and Hamilton2 found, however, that prolonged heating of a protein digest at pH 4.7 was liable to give a result higher than the maximum reached under more acid conditions. This was possibly due to a ‘secondary slow formation of carbon dioxide from glutamic acid and lysine. Those authors, therefore, preferred to carry out the reaction at pH 2-5 and considered that a result approaching the theoretical one would be obtained by adhering closely to the prescribed reaction time, on account of a compensation of errors with the few amino-acids that did not give a theoretical value in that time.Their results for individual acids certainly show smaller discrepancies with acids like cystine and lysine at pH 2.5 than at pH 4.7. In the present investigation it was found that a buffer of pH 4-7, 50 mg of ninhydrin per millilitre of amino-acid solution and a reaction time of 6 to 7 minutes gave very uniform results with twenty-two amino-acids. They were examined separately and in a mixture of equal quantities of each in terms of a-amino nitrogen.Determinations were carried out in quintuplicate, the amount of a-amino nitrogen taken being 150 pg in 1 to 2 ml of solution on each occasion. Replicates did not differ from each other by more than 2 per cent., the average recovery varied from 97.3 per cent. for tryptophan to 102.0 per cent. for aspartic acid and the general average was 99-9 per cent. for all the individual acids. For an unknown reason the recovery from the mixture was higher Lit 102.5 per cent. A selection of the results is given in Table I. TABLE I RECOVERY OF AMINO-ACIDS BY DETERMINATION OF CARBON DIOXIDE Acid a-Amino nitrogen I A I Taken, Found, Recovery, tG P6 % Alaninc . . Threonine . . Cystine . . Methionine . . Aspartic acid Arginine .. Lysine . . Pheny lalanine Tryptophan Proline . . Mixture .. .. .. 150 .. .. 150 .. .. 150 .. .. 150 .. .. 150 .. . . 150 . . .. 150 . . . . 150 .. .. 150 .. .. 150 .. .. 150 150 147 150 150 153 148 151 149 146 150 154 150 147 14‘3 150 I52 150 152 149 1+6 150 154 149 147 150 150 154 151 151 149 146 149 154 150 147 150 149 153 151 152 149 146 151 154 7 151 148 161 151 153 150 152 150 146 150 153 100.0 98.1 100.0 100.0 102.0 100.0 101.1 99.8 97.3 100.0 102.5Nov., 19511 DETERMISATIOS OF CARBOS DIOXIDE 623 The method was used to determine the proportion of total nitrogen in a-amino form in a series of hydrolysates of proteins extracted from grassland species. Preliminary tests indicated that the apparatus might be convenient for the determination of small amounts of calcium carbonate. REFERENCES 1. Smith, A. M., and Agiza, A. H., Analyst, 1951, 76, 623. 2. Mason, M. F., Biochem. J., 1938. 32, 719. 3. Van Slyke, D. D., Dillon, R. T., MacFadyen. D. A., and Hamilton, P., J . Biol. Chem., 1941, 141, 4. Van Slyke, D. D., MacFadyen, D. A., and Hamilton, P., Ibid., 1941, 141, 671. 5. West, E. A., Christensen, B. E., and Rinehart, R. E., Ibid., 1940, 132, 671. 6. Christensen, B. E., West, E. S., and Dimick, K. P., Ibid., 1941, 137, 735. 627. EDINBURGH AND EAST OF SCOTLAND COLLEGE OF AGRICULTURE First submitted, January, 1951 Amended, June, 1951