The details of the condensation between ethyl sodiocyanacetate and ethyl ẞ-methylacrylate have already been given (p. 948). It was found that the length of time during which the condensation mixture must be heated to produce the maximum quantity of ethyl hydrogen salt was 6 hours; hence on treating the product after this time with excess of methyl iodide, a corresponding quantity of ethyl a-cyanoaa'-dimethylglutarate, CO,Et C(CN)(CH3)·CH2 CH(CH3) CO2C2H5, was obtained on working up the product. This ester boils at 181-185°

under 30 mm. pressure. On analysis:

0.2214 gave 0.4839 CO, and 0·1531 H2O. C-59.61; H=7·69. C12H1904N requires C=59.75; H=7.88 per cent.

On hydrolysis with an equal weight of potash dissolved in methyl alcohol, the ester is converted into a potassium salt, evidently that of the dibasic cyano-acid, which can be separated by filtration and purified by washing with methyl alcohol. On analysis :

0.5134 gave 0-3611 K,SO. K = 31.53.

C,H,O,NK, requires K = 31.32 per cent.

This salt, when its solution in water was acidified and extracted with ether, yielded a solid substance, which was not further purified but directly treated with concentrated hydrochloric acid. On boiling, a considerable quantity of carbon dioxide was evolved, which ceased after the heating had been continued for 1 hour. The solution was then extracted with ether, and the ethereal solution shaken three times with a dilute solution of sodium carbonate. On evaporating the ether, a solid remained which separated from hot water in long needles and melted at 173°, evidently consisting of aa'-dimethylglutarimide, CH(CH3) CONH (Auwers and Thorpe, Annalen, 1895, 285, CH(CH,)•CO

CH2

315). On analysis:

0.1885 gave 164 c.c. nitrogen at 17° and 749 mm. N = 9.96.

CHON requires N = 9.93 per cent.

The imide on heating for 2 hours with 50 per cent. sulphuric acid is completely hydrolysed, yielding cis-aa'-dimethylglutaric acid, which separated from water in needles and melted at 127-128°. On analysis:

0.2237 gave 0.4295 CO2 and 0·1172 H2O. C=52.38; H = 7·47. CH1204 requires C=52.50; H=7.50 per cent.

On treatment with acetyl chloride, the acid yielded an anhydride which separated from light petroleum or ethyl acetate in plates and melted at 94-95°. On analysis :

0.1950 gave 0.4212 CO2 and 0·1223 H2O. C=58.91; H=7·00. CH10O3 requires C-58.89; H= 6.97 per cent.

The sodium carbonate washings from the ether extract of the product of hydrolysis of the cyano-ester gave nearly pure trans-aadimethylglutaric acid on acidifying and extracting with ether. The acid separated from hot water in needles and melted at 141°. On analysis:

0.2170 gave 0.4171 CO2 and 0·1503 H2O. C=52.49; H=7·63.
CH1204 requires C=52·50; H=7.50 per cent.

The acid gave no anhydride on treatment with acetyl chloride
THE OWENS COLLEGE,

MANCHESTER.

LXXXIV.-The Persulphuric Acids.*

By T. MARTIN LOWRY, D.Sc., and JOHN H. WEST. MUCH attention has been paid during the past two years in the Chemical Department of the Central Technical College to the study of the behaviour of solutions of sulphuric acid on electrolysis, as well as to the manner in which solutions of persulphuric acid undergo hydrolysis. This work has been undertaken in order to elucidate the part which persulphuric' acid plays in the accumulator (compare Armstrong and Robertson, Proc. R. S., 1891, 50, 105), and to ascertain to what extent the production of this acid must be taken into account in discussing the electric conductivity of solutions of sulphuric acid, as well as in determining the origin of the products such solutions yield when electrolysed under various conditions (compare Armstrong, Presidential Address, Trans., 1895, 6, 1156).

In the course of the inquiry, it became necessary to repeat and extend Berthelot's observations on the interaction of hydrogen peroxide and sulphuric acid (Ann. Chim. Phys., 1878, [v], 14, 345).

The experiments have led to results which appear to have an important bearing on the interpretation of the nature of the product formed by the interaction of sulphuric acid and potassium persulphate, which Caro (Zeit. angew. Chem., 1898, 845) has shown to be possessed of special properties, and which Bamberger and von Baeyer and Villiger have recently made use of as an oxidising agent under the name of 'Caro's reagent.'

Von Baeyer and Villiger have suggested that the acid characterised by Caro is a persulphuric acid composed of hydrogen peroxide and sulphuric anhydride in the ratio H2O2: SO. The experiments described in the following pages appear to favour the conclusion that Electrochemical Studies. No. I, from the Chemical Dept., C.T.C.

*

its composition may be H2O2: 4SO, whilst the acid, of which salts were first prepared by Marshall, is of intermediate composition, H2O: 280. The members of such a series of persulphuric acids may be provisionally distinguished as 'persulphuric,' 'perdisulphuric,' and 'pertetrasulphuric acids.

The substance analysed by Berthelot in 1878 was the anhydride, SO, which he obtained by combining oxygen with either of the lower oxides of sulphur by means of the electric discharge. He assumed that the corresponding acid was formed on dissolving this anhydride in water, and as he found that when moderately concentrated solutions of sulphuric acid were electrolysed they acquired properties similar to those possessed by solutions of the anhydride S2O7, he came to the conclusion that the same acid is formed during electrolysis. The correctness of this assumption was apparently placed beyond doubt by the discovery made by Marshall in 1891 of a series of well-defined salts derived from the anhydride S2O,; at all events, it has not been called in question up to the present time.

It was expressly pointed out by Berthelot in his original paper (loc. cit.) that persulphuric acid "se forme également, toujours à l'état dissous, lorsqu'on mélange avec précaution et en évitant toute élévation de température une solution d'eau oxygénée avec l'acide sulfurique soit concentré, soit étendu d'une quantité d'eau inférieure à 1 équivalent. Mais la combinaison n'a point lieu quand l'acide sulfurique est étendu à l'avance de 2 équivalents d'eau, ou davantage. Dans tous les cas, elle demeure partielle, c'est-à-dire qu'il subsiste une portion de l'eau oxygénée."

It will be noticed that in this passage he states that the peroxidation is confined to acids containing but a very small proportion of water. In a later paper, he discusses the formation of persulphuric acid on electrolysing solutions of sulphuric acid of various degrees of concentration, and also makes frequent reference to the production of hydrogen peroxide both during electrolysis and as a product of the decomposition of persulphuric acid; but he nowhere recognises that the proportions in which the two compounds are formed are in any way interdependent. In fact, in concluding his paper (Ann. Chim. Phys., 1880, [v], 21, 193), he remarks, "Cette formation d'eau oxygénée ne paraît pas due à une réaction lente de l'eau contenue dans les liqueurs et à un équilibre résultant entre les deux composés suroxygénés, comme on aurait pu le penser d'abord. En effet, la liqueur diluée avec 20 volumes d'eau, et qui s'est conservée presque sans variation pendant neuf jours (98 mgr. reduits à 90 mgr.), n'a donné lieu à aucune formation appréciable d'eau oxygénée. "La formation de l'eau oxygénée est donc composition lente de l'acide persulfurique.

simultanée avec la déElle en est probable

ment corrélative, comme nous l'avons déjà admis plus haut pour rendre compte de sa formation par electrolyse."

Our experiments, however, lead us to conclude that in a system comprising persulphuric acid' and hydrogen peroxide together with sulphuric acid and water, there is a definite state of equilibrium between the two peroxidised products, and that under constant conditions the ratio of hydrogen peroxide to 'persulphuric acid' is entirely determined by the ratio of water to sulphuric acid; the ratio of hydrogen peroxide to persulphuric acid' is therefore ultimately the same in a mixture prepared from hydrogen peroxide and sulphuric acid as in a solution of equal strength prepared by electrolysis.

Experimental Method.-The method used in determining the ratio of hydrogen peroxide to persulphuric acid was that used by Berthelot, and depends on the fact that whilst potassium permanganate is reduced by hydrogen peroxide, it is not affected by 'persulphuric acid,' and that the latter readily oxidises ferrous sulphate to ferric sulphate. In dilute solution at low temperatures, 'persulphuric' acid is very stable, and is not decomposed to any appreciable extent, either into sulphuric acid and oxygen, or into sulphuric acid and hydrogen peroxide; a mixture containing about 10 volumes of water to 1 volume of sulphuric acid was found to be stable at 0°, no change being perceptible in the amount of hydrogen peroxide or of 'persulphuric acid' after a week. In analysing the solutions, it is necessary that the acid should be rapidly diluted, and that no heating should take place during the dilution; if this condition is not fulfilled, the equilibrium is disturbed, and the proportion of 'persulphuric acid is lowered accordingly. For this reason, the acid mixture was slowly delivered from a pipette on to an excess of ice, so that the temperature fell below 0° during the dilution. Special care was necessary in the case of acids containing less than one equivalent of water; these were poured directly on to a large quantity of ice without any attempt at measuring the volume of the mixture, and portions of the diluted mixture were then titrated. After diluting, the hydrogen peroxide was first estimated by running in potassium permanganate from a burette until a permanent coloration was produced. When only a small amount of hydrogen peroxide was present, it was found to be advisable to add a few drops of a solution of manganese sulphate, as otherwise the first drop of potassium permanganate produces a pink coloration, which persists for some time in the ice-cold solution, and there is a danger of overlooking the presence of hydrogen peroxide. The interaction of hydrogen peroxide and potassium permanganate, in fact, appears to belong to the class of actions which only take place in presence of a catalytic agent. After destroying the hydrogen peroxide, the persulphuric' acid was esti

mated by adding an excess of ferrous sulphate solution, and titrating back in the usual way.

It was soon found that the ratio of persulphuric' acid to hydrogen peroxide was very largely influenced by minute differences in the proportions of sulphuric acid and water, and that any error in the estimation of the sulphuric acid would be reproduced five-fold in the persulphuric ratio. It was therefore necessary to take special precautions in standardising the sulphuric acid and in preparing the acid mixtures. It was, in fact, desirable that the value of the acid should be known, if possible, within 0.02 per cent., and this high degree of accuracy could only be obtained by determining the density (compare Pickering, Trans., 1890, 57, 64; Marshall, J. Soc. Chem. Ind., 1899, 18, 6). In all the later experiments, the acid used was drawn off by a siphon from a bottle containing about seven litres, and measured out from a repeating burette into flasks fitted with a rubber stopper and drying tube, the acid in the bottle and burette being also protected by drying apparatus. To determine the density, 10 c.c. of water were run into a dry flask and carefully mixed with about 80 grams of the acid from the repeating burette; the flask, water, and acid were weighed to 0.001 gram, and the density of the mixture was then determined at 18°, relatively to water at 18°, with the aid of a density tube about 5 c.c. in capacity. The concentration of the diluted acid could be deduced to within 0.01 per cent. from the table given by Marshall, and it was easy to calculate the concentration of the original acid from the value so found. As showing the accuracy of the method, the six values determined in the case of the acid used in the last series of experiments may be quoted, namely, 93.70, 93.81, 93 66, 93.78, 93.76, 93.76, mean value 93.75, mean error 0.04.

The mixtures to be titrated were prepared in a similar way, but 20 c.c. of hydrogen peroxide (20 vols.) were taken in place of the water, and the weighings were made to 0.01 gram only; the flasks containing the mixtures were closed by rubber stoppers each fitted with a tube drawn out to a minute capillary, in order to allow of the escape of oxygen, whilst preventing the entrance of moisture. In making the weaker mixtures, 40 c.c. of a dilute hydrogen peroxide solution was used, whilst in the case of those of higher concentration 5 c.c. of a "50 volume" hydrogen peroxide* was taken, in order to diminish the quantity of acid required, and so as to lessen the danger due to heating during the dilution, which has already been referred to. Commercial hydrogen peroxide always contains a considerable quantity of chlorine, which is given off on mixing it with sulphuric acid; the greater part of this may be removed by adding silver * We were indebted for this to Mr. Tyrer, who had kindly prepared it at Dr. Armstrong's request.