STUDIES ON THE ELECTROLYTIC OXIDATION OF PHEKOLS. 243 XXX.-Studies on the Electrolytic Oxidatioja of Phenols. Part I. By ARTHUR GEORGE PERKIN and FREDERICK MOLLWO PERKIN. OF the numerous substances which can be obtained by the oxidation of pyrogallol and its derivations, one of the most interesting is purpuro- gallin, a compound possessing well-marked tinctorial properties. Originally prepared by Gerard from pyrogallol (Bey., 1869, 2, 562) by the action of potassium permanganate and sulphuric acid, it has been244 PERKIN AND PERKIN: STUDIES ON THE subsequently obtained from the same phenol by means of a most varied series of oxidising agents. I n a recent communication, certain deriva- tives and decomposition products of this substance were described (Trans., 1903, 83, 192), from which i t appears t o have t h e formula C,,H,O,, and t o be a derivative of naphthalene, as was first indicated by Nietzki and Steinmann (Bey.., 1887, 20, 1277).The formation of a naphthalene compound from pyrogallol, although very remarkable, is not unique, for a somewhat similar reaction is described by Zincke and Branke (Annalen, 1896,293,l ZO), who prepared dibromo-/3-naphtha- quinonecarboxylic acid by the action of nitric acid on bromoproto- cntechuic acid. During the formation of these substances, it appears likely t h a t a certain quantity of the phenol is converted by oxidation into an open chain compound, which then condenses with the phenyl residues. Further work on the constitution of purpurogallin is in active progress, and it appeared possible that by a study of the electrolytic oxidation of pyrogallol some intermediate or secondary product might thus be prepared, and, moreover, it was anticipated also that if the colouring matter could be produced by this method, a considerably increased yield might be obtained.The following experiments indicate without doubt that by employing certain precautions the electrolytic method is the most suitable for the production of purpurogallin, for not only can the yield be thus augmented, but an almost chemically pure product is at once isolated. On the other hand,it has not as yet been found possible to prepare any intermediate compound which would be of service in elucidating the course of the reaction. EXPERIMENTAL. In selecting a suitable electrolyte, numerous substances were ex- amined, and although purpurogallin was obtained in all instances, it was frequently contaminated with a brownish-black impurity, which rendered the method of little service.The employment of a 15 per cent. solution of sodium sulphxte has been eventually adopted, and by this means a very pure bright orange-coloured product has been obtained. Solutions of sodium and ammonium acetates do not give satisfactory results, because when they are employed a somewhat poor yield of the impure alkali salt of purpurogallin is produced. In the earlier experiments, the anode and cathode compartments were divided by means of a porous cell, but i t subsequently appeared t h a t better results could be obtained if the anode and cathode were not separated.Satisfactory results are obtained only when the anode is made of platinum and caused t o rotate rapidly during the electro- lysis. When a stationary anode is employed, the colouring matter isELECTROLYTlC OXIDATION OF PHENOLS. PART I. 245 apt to settle on the electrode, and is thus subjected to further oxida- tion. A rotating anode of lead, even when previously peroxidised, was very unsatisfactory, for in this case from 10 to 16 per cent. only of very impure purpurogallin could be obtained." I n carrying out the oxidation, 28 grams of pprogallol dissolved in 500 C.C. of a 15 per cent, solution of sodium sulphate were placed in a rectangular glass jar, at two opposite corners of which were fixed two thin pieces of composition piping. On t o each of tbe pipes a copper wire was soldered, and these were connected with each other by means of a screw connection and formed the cathode.The iridio- platinum anode had a surface of 1.5 square decimetres and was rotated rapidly by means of a water turbine or electromotor. During the electrolysis, the current density was kept at 2 amperes and the E.H.F. was 4.3-4-5 volts. To avoid a rise in temperature, the cell was surrounded with cold water. As soon as the current commenced to flow, the solution became yellow and a yellow precipitate gradually separated. When the electrolysis had continued for 8 hours, the product was left overnight, then collected, washed with water, and dried on a porous plate, Satisfactory results mere also obtained when, instead of rotating the anode, a rapid stream of air was blown through tbe mixture.In this case, the cakhode consisted of a perforated 1ea.den tube placed between two anodes of platinum foil, having a total surface of 4 square decimetres. The air, which was blown through the cathode during the operation, appears to help in the oxidation of the pyrogallol. Thus 446 grams of pyrogallol dissolved in a solution of 1 kilogram of sodium sulphate (Na,SO,,lOH,O in six litres of water gave, after 35 hours' electrolysis, current density 9-10 amperes, X.M.F. 9.5-10.3 volts) 165 grams of purpurogallin or 37 per cent. The average yield was 10.2 grams or 36.4 per cent., although on two occasions 13 grams (or 46.4 per cent,) were obtained. Nitrous acid yields only 25 per cent. of the crtide product. That the substance thus obtained was practically pure is indicated by the analysis ( A ) ; the second estimation (B) was made on the product obtained with a lead anode, but this specimen was, however, first purified by crystal- lisation from acetic acid.In the preparation of periodic acid by the electrolytic oxidation of iodic acid, with smooth platinum electrodes as anode, the yield is extremely small. But when peroxidised lead anodes are employed a very much better yield is obtained. The explanation being that the potential at which the oxygen is yielded up at the anode is higher with a lead peroxide anode than with one of polished platinum (Ber., 1902, 35, 2662). In the case of pyrogallol, the probability is that when lead electrodes are used, the higher potential at which the oxygen is yielded causes the oxidation to be too vigorous.246 PERKIN AND PERRIN: STUDIES ON THE Found ( A ) C = 60*03 ; H = 3.60.C,,H,O, requires C = 60.00 ; H = 3.63 per cent. The colouring matter was further characterised by conversion into the acetyl derivative, which was obtained in the form of yellow needles melting a t 184-186O. Found C = 58.87 ; H = 4.04. After one crystallisation from acetic acid, this product was found to be chemically pure. Experiments were also tried with an alternating current, but although small quantities of purpurogallin were obtained, the method was not found satisfactory. (B) C = 60.19 ; H = 3.60. C,,H,O,(C,H,O), requires C = 58-77 ; H = 4.12 per cent. The Electrolytic Oxidation of Gallic Acid. It was previously shown that by the oxidation of gallic acid with potassium ferricyanide in the presence of sodium acetate a new substance, probab1y:a purpurogallincarboxylic acid, is formed.As the yield by this method was extremely poor (3 per cent.) and the compound difficult to isolate, the electrolytic oxidation of gallic acid was studied in the hope that as in the case of pgrogallol a larger yield of a purer product mould thus be obtained. Experiment showed that the purpurogallincarboxylic acid is rapidly obtained in this manner, and although the yield has up to the present been some- what disappointing, it is evident that sufficient can be obtained by this process for a complete study of its constitution. After numerous trials, the following method was adopted. It was found necessary to sepa- rate the anode and cathode com- partments. A rectangular glass vessel was employed as before, on two sides of which were fitted two porous cells.These cells contained a piece of sheet lead, to which was soldered a stout copper wire, In order to keep the cells in position, the wire was bent over against the side of the glass jar as shown in the figure. The anodes consisted of a platinum stirrer and a stout wire bent up at each side against the bottom of the porous cells. To this wire was also welded a piece of platinum foil nearly large enough to cover the bottom ofELECTROLYTIC OXIDATION OF PHENOLS. PART I. 247 the glass jar. The total active anode surface was about 1-75 square decimetres. Various electrolytes were tried, but the only one which gave satisfactory results was a 15 per cent.solution of sodium or ammonium acetate acidified with acetic acid. I n carrying out an electrolysis, 10-15 grams of gallic acid were dissolved in 600 C.C. of a 15 per cent. solution of ammonium or sodium acetate, 10 C.C. of glacial acetic acid were then added, and the mixture electrolysed. The cathode cells contained a similar solution, During the electrolysis, the cathode cells become too full and are apt to overflow ; as the solution in these cells becomes strongly alkaline, it is necessary to remove a portion every now and then by means of a pipette in order to prevent it flowing over into the anode compartment. As soon as the current passes, the solution commences to turn brown and finally become3 almost black. After about 6 hours, the current is stopped, the electrodes and porous cells removed, the mixture left overnight, and then filtered, when a dark bronze-coloured product is obtained. This consists chiefly of the alkali salt of the acid in a somewhat crude condition. The yield varies very greatly, from 10 grams of gallic acid usually about 2 grams (or 20 per cent.) of the crude product are isolated, but we have obtained as much as 5 grams, or 50 per cent. The product of the reaction was purified as follows : The substance ground into a thin cream with alcohol was diluted with some quantity of this solvent, a little hydrochloric acid added, and the mixture boiled for a minute or two ; the solution obtained was then poured into ether, the ethereal extract washed with water, boiled with a little animal charcoal, filtered, and then evaporated to dryness. The orange-yellow, crystal1ine:residue was collected, washed with water, then with ether, and dried. Found C = 54-64 ; H = 3.52. C,,H,07 requires C = 54.54 ; H = 3-03 per cent, It was found to be identical in all respects with the purpurogallin- carboxylic acid previously obtained from gallic acid by means of potassium ferricyanide.