LATEST DESIGN

OF

July 2, 1915

A Belgian Refugee Engineer, Manager of a

Chemical Works, seeks Employment during the War.-Address, FERN. LEROY, 6, Central Hill, Upper Norwood.

(Made in England).

Munitions of War. London district.-Address "Munitions," CHEMICAL NEWS Office, 16, Newcastle Street, Farringdon Street, London, EC.

Lecture and Laboratory Courses qualifying for Degrees in Chemistry will be forwarded on application to the REGISTRAR. Applications for admission to the Research Laboratories should be made to the DIRECTOR of the Chemical Laboratories.

ARHWARD

the end-point by testing a few drops of the solution under

NITRO COMPOUNDS.

By A. J. BERRY, M.A., and C. K. COLWELL, B.A.

NITRO Compounds are usually determined volumetrically by reduction to amino compounds by excess of a suitable reducing agent, the excess of which is determined by titration in some convenient manner. The reducing agents usually employed for this purpose are stannous chloride or titanous chloride in acid solution. When stannous chloride is employed in presence of hydrochloric acid the reduction proceeds in accordance with the equation

R.NO2+3SnCl2 + 6HCl = R.NH2 +3SnCl4 +2H2O. The excess of stannous chloride is usually determined in an aliquot portion of the solution by titration with iodine after adding excess of sodium bicarbonate and potassium sodium tartrate.

In the course of some experiments on the quantitative reduction of nitro compounds to amino compounds we have applied a method originally due to Weil (CHEMICAL NEWS, 1871, xxiii., 49) for the volumetric determination of copper to determine the excess of stannous halide. This method depends upon the fact that a solution of a cupric salt strongly acidified with hydrochloric acid is reduced quantitatively to the cuprous condition with simultaneous formation of an equivalent amount of stannic salt in accordance with the equation—

2CuCl2+SnCl2 = Cu2Cl2 + SnCl4.

It is clear that this method might be applied in the reciprocal way; that is, it should be possible to employ a standard solution of a cupric salt to determine a solution of stannous halide, and, consequently, to determine a nitro compound by difference. The results which we have obtained show that the method is capable of yielding satisfactory results.

In our earlier experiments the general procedure was as follows:-A definite quantity of the solution of the nitro compound was measured out, acidified strongly with hydrochloric acid, and a known excess of a standard solution of stannous chloride added, and the resulting mixture boiled gently for a quarter of an hour, a slow current of carbon dioxide being passed through fhe solution all the time. The liquid was then titrated at the boiling-point with a standard solution of copper sulphate strongly acidified with hydrochloric acid. The colour of the cupric solution was of course discharged by reduction to the cuprous condition, but it was found necessary to determine

chloride. The end-point was taken as that point at which the solution ceased to give a turbidity with mercuric chloride.

As the employment of an external indicator is a somewhat tedious process we modified the procedure in our later experiments. Etard and Lebeau (CHEMICAL NEWS, 1890, Ixi., 137) have pointed out that a solution of a cupric salt acidified with excess of hydrobromic acid possesses a deep brownish violet colour, which becomes perfectly colourless on reduction with a stannous salt. The colour change is much more striking than with hydrochloric acid, and the employment of an external indicator is quite unnecessary. In our experiments we added a measured quantity of a strong solution of potassium bromide to the solution of the nitro compound which was being reduced by stannous chloride and hydrochloric acid. After boiling the solution for fifteen minutes in an atmosphere of carbon dioxide, we titrated the excess of the stannous salt in solution by means of a strongly acid solution of copper sulphate containing potassium bromide.. At the end-point of the reaction the liquid assumed an orange-yellow colour, and to obtain strictly consistent results the colour of the solution was compared with that of a dilute solution of potassium dichromate. This second method was found to be much more satisfactory to work than the first, and as far as our experience goes, slightly more accurate also. In

both methods the stannous chloride solution was always titrated against the standard copper solution at the time of carrying out a determination.

Examples.

1. Paranitrophenol. 1938 grm. of this substance was weighed out, dissolved in water, and the solution diluted to 250 cc. Aliquot portions of the solution were titrated with the following results:

25 cc. of the solution with 25 cc. of stannous chloride required 31 cc. of copper solution. 25 cc. of stannous chloride solution alone required 17.6 cc. of copper solu. tion. Hence 13'4 cc. of the copper solution correspond to 25 cc. of the solution of paranitrophenol. The copper solution contained 39 2 grms. of copper (metal) per litre. From this it follows that the weight of paranitrophenol in This experiment the 250 cc. of solution is 1913 grm. was carried out with the aid of mercuric chloride for determining the end-point.

2. Picric Acid.-2.985 grms. of this substance were weighed out, dissolved in water, and the resulting solution This determination was carried out diluted to 500 cc. with the addition of 35 cc. of a 50 per cent solution of potassium bromide to every 25 cc. of the solution of

2KIO3 + Na2S2O3.5H2O + 2HCl =

25 cc. of the picric acid with 25 cc. of the stannous solution required 314 cc. of copper solution. 25 cc. of the stannous solution alone required 140 cc. of the copper

solution.

From this it is clear that 17:4 cc. of the copper solution correspond to 25 cc. of the solution of picric acid. The weight of picric acid in the 500 cc. of solution is equal to 2.944 grms. It should be pointed out that our experiments demonstrate the applicability of this modified reduction method to the determination of the nitro groups in picric acid. Altmann (Journ. Prakt. Chem. 1901, İxiii., 380) quoted an experiment in which he failed to obtain satisfactory results with picric acid, and concluded that this substance could not be determined by reduction with stannous chloride. The experiments of Knecht and Miss Hibbert (Ber., 1903, xxxvi., 1549) have, however, shown that picric acid as well as other nitro compounds may be accurately determined by means of titanous chloride.

3. Paranitrophenol.-1'122 grm. of this substance was weighed out, dissolved in water, and the solution diluted to 250 cc. The acid solution of stannous chloride employed for reduction contained potassium bromide in solution, and the same copper solution as that employed for the determination of the picric acid in No. 2 was employed for the titrations. The following results were obtained :

- Na2SO4 + K2SO4 +2IC1 + 6H2O I cc. = 0.002068 grm. of Na2S2O3.5H2O. The titrations were made in glass-stoppered bottles in exactly the same manner as in the determination of sulphurous acid. It is important to titrate as soon as possible after the thiosulphate is brought in contact with the hydrochloric acid, which should not be warmer than 18° C.; also it is best to dissolve the thiosulphate in a little water previous to adding the hydrochloric acid. For convenience several solutions of sodium thiosulphate of various strengths were prepared and standardised by titration with known amounts of pure iodine in the usual manner, observingall precautions. The results obtained are given in Table I. TABLE I.

25 cc. of the solution of paranitrophenol with 25 cc. of stannous halide required 13.65 cc. of copper solution. 25 cc. of the solution of stannous halide alone required 21.05 cc. of copper solution.

0'0000 0'0000

Downing College Laboratory, Cambridge.

THE VOLUMETRIC DETERMINATION OF

12

The titrations given in Table II. were made with a potassium iodate solution which had the value I cc. = 005300 grms. of Na2S2O3.5H2O.

The results in both tables show a satisfactory agreement among themselves and with the actual amounts present.

A volumetric method for the determination of tetrathionic acid has been described by E. Abel (Zeit. Anorg. Chem., Ixxiv., 395), which is based upon adding an excess of standard iodine solution to the solution of the tetrathionate, which has been made strongly alkaline. Then the solution is acidified with hydrochloric acid, and the

By GEORGE S. JAMIESON.

Sheffield Chemical Laboratory, Yale University.

In a recent paper (Am. Journ. Sci., 1914, xxxviii., 166) from this laboratory it was shown that sulphurous acid could be titrated with a solution of potassium iodate in the presence of 15 to 20 per cent of actual hydrochloric acid and a small volume of an immiscible solvent, such as chloroform, according to the general method of L. W. Andrews (Fourn. Am. Chem. Soc., 1903, 25, 756). It has been found that thiosulphuric and tetrathionic acids can be titrated in the same manner, while dithionic acid, on account of its stability, cannot be determined by this method.

In order to test the method for the titration of thiosulphates, a solution containing 3'567 grms. of normal potassium iodate in 1000 cc. was used. (This solution was previously used for the determination of sulphurous

sodium thiosulphate in the usual manner. The iodate method has the advantage in requiring only a single standard solution. In order to test the method sodium tetrathionate was prepared according to the method of F. Kessler (Pogg. Ann., 1848, lxxiv., 255), by grinding sodium thiosulphate with a slight excess of indine in the presence of a few cc. of water. When the sodium thiosulphate was entirely dissolved, two volumes of absolute alcohol were added. After the solution had stood for two hours the sodium tetrathionate crystals were filtered by suction and washed with 98 per cent alcohol until all the iodine and sodium iodide were removed. The sodium tetrathionate was found to contain some sulphate, which was determined by the following method:-Weighed portions of the salt were dissolved in water and acidified with I grm. of tartaric acid. Barium chloride was added to the solution, and when the precipitate had settled, it was filtered and determined in the usual manner. The results obtained are given in Table III.

Two water determinations were made at about 112° C.
I. 10000 grm. sub. gave 11.27 per cent of water.
II. 0'5000 grm. sub. gave 11'37 per cent of water.
Calc. water for Na2S406.2H2O after correction for
3'35 per cent of Na2SO4, 11*35 per cent.

In the analyses given in Table IV., the actual amount of Na2SO406 present for each titration is stated. The first six titrations were made with the potassium iodate solution mentioned above, which contained 3'567 grms. of KIO3 in 1000 cc., and according to the equation of the

3

A DEVICE FOR OBVIATING THE KINKING OF RUBBER TUBING CONNECTIONS OF WATER

JACKETED CONDENSERS.

By A. COTTRELL, M.Sc., and H. WRIGHT, B.Sc.

THE accompanying photograph shows how the rubber tubing attached to condensers can be kept free from the kinks which are so prone to develop at the points of connection.

taken.

used.

found.

(a) Titrations 7 to 14 were made with an iodate solution which had the value I cc.=0'003297 grm. of Na2S406.

These results show a satisfactory agreement with the actual amounts of sodium tetrathionate taken.

Dithionates were found to react so slowly with potassium iodate in the presence of strong hydrochloric acid that it was possible to titrate and determine the amount of sodium thiosulphate in the presence of large quantities of potassium dithionate. Likewise tetrathionates could be estimated in the presence of dithionates. The stability of dithionates has been observed previously by R. H. Ashley (Am. Journ. Sci., xxii., 259). It was found that barium and sodium dithionate were only partially decomposed by potassium iodate after reacting for a day and no iodine was liberated for some time after bringing the constituents together. The following experiments were made:

Brass wire of 15 S.W.G.Jis wound into a spiral of suitable dimensions, and after being fixed in position the spiral is gently bent to a suitable curve. Technical School, Dewsbury.

SEVERAL years ago (fourteen) the writer published under the above title a description of an apparatus for washing out of fertilisers, water-soluble ammoniates, and phosphates (Fourn. Am. Chem. Soc., xxiii., No. 8; republished in CHEMICAL NEWS, 1902, lxxxv., 28). Water is dropped from a reservoir intermittently by means of a syphon on a weighed portion of substance on a filter. The nitrogen (or phosphoric acid) in the residue is determined by one of the usual methods; this subtracted from the total nitrogen (or phosphoric acid) gives, of course, the watersoluble portion.

This apparatus, improved and simplified in construction, is still in constant use in this laboratory. Two forms of it are shown in Figs. 1 and 2. In the latter each element (a 300 cc. percolator with its syphon) is separate and distinct, and the charging of each with water requires a separate and distinct operation. In the former, each