If hydrogen be evolved in presence of arsenical compounds, it combines with the element to form “arseniuretted hydrogen,” or arsine, AsH3, a colourless gas of alliaceous odour, extremely poisonous, giving, when passed into silver nitrate solution, a precipitate of silver and a solution of As2O3; decomposed at a red heat into As and hydrogen, and burning with a livid flame into As2O3 and water. The flame yields, when a cold surface, such as a porcelain crucible-lid or dish, is depressed into it, a steel-grey lustrous stain or ring of metallic arsenic.
To evolve the hydrogen, Marsh originally used zinc and sulphuric acid. As it is so difficult to obtain zinc pure, magnesium has been proposed. But the evolution is then too rapid. Moreover, the reputed “pure” acids of commerce are scarcely ever free from a trace of arsenic. This difficulty affects equally the galvanic method. Hence it is better to employ sodium amalgam (Edmund Davy, Chem. News, xxxiii., 58, and ditto, 94). One part of sodium, scraped free from oxide, is melted under solid paraffin, and gradually added to ten parts of mercury (previously purified by nitric acid) with constant stirring, the paraffin poured off, and the amalgam cleaned by washing with pure dry benzine. The result is a solid crystalline alloy.
A few fragments of this alloy are placed with water in a flask provided with a thistle funnel, and with a delivery tube dipping into a 4 per cent. solution of silver nitrate. The horizontal part of the delivery tube is heated to just below redness by a lamp, meanwhile being supported to prevent its bending. If, after about an hour, no As ring appears in the tube, and if the silver nitrate, after precipitation of the silver by hydrochloric acid and filtration, gives no arsenious sulphide on addition of sulphuretted hydrogen, the amalgam is pure. Now add to the flask the suspected liquid, put in more amalgam, and continue the heating of the tube and passing of the hydrogen till no further evolution of As occurs. The portion of tube containing the deposit of As may be cut off, weighed, the As dissolved off by aqua regia, and the tube washed, dried, and weighed again. The silver nitrate solution contains the remainder of the As dissolved as As2O3: after removal of the silver by hydrochloric acid and filtration, the arsenious acid solution may be divided, a portion titrated by iodine (see Blyth’s Pract. Chem., p. 392), and the rest tested qualitatively by sulphuretted hydrogen, ammonio-silver nitrate, and ammonio-cupric sulphate (see these tests, post).
If the original liquid be rendered strongly alkaline before adding the amalgam, no antimony will pass off with the arsenic. But from acid liquids, arsenic and antimony pass off together. They both give metallic rings in the tube, and stains on a cold surface. The chief distinctions between them are as follows:—
1. Arsenic.—More volatile, hence deposited further from the flame; bounded by a “hair brown” fringe of suboxide; heated in a current of sulphuretted hydrogen gives yellow As2S3, unchanged by passing dry hydrochloric acid gas: heated in air, it gives easily a sublimate of As2O3 in glistening octahedral crystals. It is soluble at once in chloride of lime solution.
2. Antimony.—Less volatile, hence forming close to the flame; no brown fringe; heated in a current of sulphuretted hydrogen, gives orange or black Sb2S3, volatilized by passing dry hydrochloric acid; heated in air, it gives a white oxide, volatile with great difficulty, and not generally crystalline. It is insoluble in chloride of lime solution.
By the above process, Edmund Davy has detected one-millionth of a grain of As; 1/1000 grain gives a very decided effect in a few minutes. It is applicable not only to As2O3, but to all other arsenical compounds in powder, whether soluble or insoluble. Organic matter interferes very little.
It must be observed that hydrogen alone may give a slight reduction and precipitate in solutions of silver nitrate.
To prepare pure sulphuric acid and pure zinc, see Selmi, Chem. Soc. Journal, May, 1881, p. 311.
REINSCH’S TEST.