Detection and Estimation of Ammonia

Moderate quantities in the air or in solution are recognised by the smell, by the brown colour imparted to yellow turmeric, by the blue to red or purple litmus, and generally the alkaline colour to indicators of the "weak alkali sensitive" class, such as Congo blue or methyl orange. The latter, however, are rather too sensitive, as they change in time when exposed to ordinary air. The blackening of mercurous nitrate is also a good test. All these indicators are best applied in the form of test papers.

Traces of ammonia and its salts in solution are recognised by means of Nessler's solution. If sulphides are present, they also impart a dark colour to Nessler's solution; but this persists on the addition of acids, whereas the brown due to ammonia is destroyed by acids.

Free ammonia in solution is usually estimated by titration with standard acid.

Since the ammonia is continually being lost in the vapour, the solutions should be diluted and kept in a well-stoppered b@ttle or titrated at once, or treated with an excess of standard acid.

On account of the hydrolysis of solutions of ammonium salts, which therefore react acid, the end point of the titration is found with an "acid-insensitive" or " alkali-sensitive " indicator such as methyl orange, or better, methyl red.

The ammonia in ammonium salts may be determined after distillation with alkali and absorption of standard acid. A weighed amount of the salt is dissolved in boiled water treated with excess of 10 per cent, boiled sodium-hydroxide solution and distilled from a flask provided with a " spray trap " or " splash bulb." Many forms of apparatus are described in the text-books of analytical chemistry. The distillate is taken up, or the evolved gases pass through, a known excess of standard, usually 0-5N acid, and the excess titrated back with standard alkali as described above.

If the ammonia is present as the salt of a volatile acid [e.g. as carbonate), it may be boiled with standard sulphuric acid, and the excess of this titrated as above. In the determination of nitrogen in organic compounds by the Kjeldahl method, the nitrogen is finally obtained as ammonium sulphate dissolved in concentrated sulphuric acid. This is distilled with an excess of alkali, etc., in the manner already described. If the quantities of ammonia are small, they may be condensed with water and determined by the Nessler method.

Ammonium salts may also be estimated quickly, although less accurately, by loss. The salt (e.g. 10 grams) is dissolved in water, treated with an excess of alkali, and made up to a known volume, e.g. of 1 litre. A measured volume of this, e.g. 25 c.c., is titrated directly, using methyl red (a). Another 25 c.c. is diluted, boiled in a flask until the steam is free from ammonia, and then titrated (b). The difference between the titrations (a) - (b) gives the ammonia. A gas volumetric method is also available, which depends upon the decomposition by sodium hypo- bromite, with evolution of nitrogen:

2NH₃ + 3NaOBr = 3H₂O + 3NaBr + N₂.

The alkaline hypobromite is run into the solution of ammonium salt in a nitrometer and the volume of nitrogen measured and corrected to N.T.P. in the usual manner. Since the decomposition is not quite complete (owing to the existence of a little ammonium hypobromite) a blank analysis is carried out on a standard solution of ammonium salt, and a correction factor is thus evaluated.

Ammonia may be estimated gravimetrically after precipitation by chloroplatinic acid. The precipitate may be dried at 130° and weighed as (NH₄)₂PtCl₆, or ignited to platinum.

Ammonia present as salts may be indirectly estimated by Ronchese's method. Formaldehyde converts the ammonia into hexamethylene tetramine (hexamine, urotropin), and the liberated acid is titrated in the presence of phenolphthalein:

4NH₄NO₃ + 6HCHO = C₆H₁₂N₄ + 4HNO₃ + 6H₂O.

Any free acid originally present is first neutralised, the amount of alkali necessary being determined by a preliminary titration in the presence of methyl orange or methyl red. Formalin containing 20 per cent, of formaldehyde, previously neutralised if necessary to phenolphthalein, is added in slight excess. The titration is then carried out in the usual manner with standard acid. The method is not applicable to the ammonium salts of weak acids.

Small quantities of ammonia, free or combined as salts, are estimated colorimetrically with the aid of Nessler's solution, potassium mercuric iodide, K₂HgI₄, with an excess of sodium or potassium hydroxide. The brown precipitate, or brown or yellow colour, is the compound O-Hg₂- NH₂-I, which has great colouring power.

The solution to be tested is made alkaline with sodium hydroxide and carbonate, allowed to stand until any carbonates of calcium magnesium or iron settle, and a measured portion of the clear solution is tested with 1 c.c. or more of the Nessler reagent. If the amounts of ammonia are small, as usually in waters, 100 or 200 c.c. are treated with sodium carbonate or a suspension of magnesium carbonate (both of which must of course be free from ammonia) and distilled from a flask through a Liebig's condenser. Successive portions of the distillate are treated with 2 c.c. of the Nessler reagent, and the yellow or brown colours are matched against that produced by a very dilute standard solution of ammonium chloride. The well-known cylindrical Nessler glasses of colourless glass, holding 50 or 100 c.c., are used for the comparison. 0.01 mgm. of ammonia can easily be determined by this method.