95 X1V.-Action of Reducing Agents on Potassium Permanganate. By FRANCIS JONES, F.R.S.E. h a paper on Stibine (this JozcrnaZ, xxix, 642), read before this Society last year, I mentioned incidentally that this gas and its analogues decompose solution of potassium permanganate with sep: - ration of oxide of manganese. I also pointed out that the precipitate consisted only of oxide of manganese, and that the antimony, arsenic, &c., passed into solution." I have since examined these reactions more fully, as well as the action of hydrogen itself, which has long been known to act on permanganates (Gmelin's Han,dboolc, iv, 211). Tl-e results show that hydrogen, ammonia, phosphine, arsine, and stibine all decompose permanganates with separation of hydrated sesquioxide of manganese, and formation of salts of phosphorus, arsenic, &c.The experiments were made by passing the purified gases into a moderately strong solution of potassium permanganate until the dis- appearance of the purple colour. The precipitate was then allowed to settle, washed by decantation, and dried over sulphuric acid or on the water-bath. The liquid was poured off, filtered, and examined qualita- tively. The oxides were analysed in various ways, but generally by reduction in hydrogen to protoxide, or by ignition in air to manga- noso-manganic oxide. Occasionally the combined water was also estimated as a check. I. Action of Hydrogen. (a.) In Neutral Solutions.-When hydrogen is passed through a neutral solution of potassium permanganate, a light brown precipitate soon separates, but settles very slowly.After the prolonged action of the gas the precipitate becomes denser, and the superuatant liquid becomes perfectly clear and colourless. This clear liquid is alkaline to test-paper, and contains only potash ; the precipitate is hydrated sesqni- oxide of manganese. 2KMnOl + BH = Mn,O, + 2KH0 + 3H20. Tlie reaction may be thus represented :- Analysis of Oxide. A. 0.4465 gram 0.3658 gram = 63.44 6'2.50 (Mn203 + H,O) * E. S c h o h i g has since proposed this reaction as a means of purifying hydrogen from these gases, a method which I had likewise adopted with excellent results. J. pr. Chenz. [ 2 ] , xir, 289. Substance. MnO. Per cent. Mn. Calculated.96 JONES ON THE ACTION OF REDUCING AGENTS The absorption of hydrogen by solution of permanganate is tolerably rapid, and increases with the temperature and with the surface of liquid exposed t o the action of the gas.It may be shown experi- mentally, by filling a wide test-tube with hydrogen, and placing it mouth downwards in a solution of permanganate ; the solution will steadily rise in the tube until all the hydrogen has been oxidised. Or, if a tube filled with hydrogen and containing a few drops of a satu- rated solution of permanganate be sealed up and shaken from time to time, the whole of the hydrogen will be oxidised in two or three days, so that, if the end be broken off under water, the liquid will rise to the top of the tnbe. A similar result may be obtained in half-an-hour by placing the sealed-up tube for that length of time in a water- bath.(b.) In Acid Solutions.-The permanganate was acidified with a few drops of pure dilute sulphuric acid, and the hydrogen passed through. A precipitate soon separated, darker in colour than in the neutral soh- tion, and very dense. It consisted of manganese sesquioxide ; the fil- trate contained only potassium sulphate. Analyses of Oxide. Substance. MnO = per cent. manganese. Calculated. C. 0.4318 ,, 0.3408 ,, = 61.14 62.50 Mnz03 + H?O B. 1.8040 gram 1.455 gram = 62.50 62.50 D. 0.6220 ,, 0.5395X ,, = 62.49 62.50 (c.) In. Alkaline Solutions.-Pure caustic soda was added to the per- manganate ; the hydrogen acted much more slowly than in the pre- vious cases ; the liquid first became green, by reduction to manganate, and this colour disappeared only after the prolonged action of the gas.The precipitate was pale brown, and very similar in appearance to that produced in neutral solutions. 1 AnaZyyses of Oxide. Substance. MnO = per cent. manganese. Calculated. E. 0.5415 gram 0.4410 gram = 63.08 62.50 ‘1 Mnzos + H,o F. 0.5940 ,, 0.4850 ,, = 63.24 62-50 11. Action of Ammonia. The action of ammonia on potassium permanganate has been examined by several chemists. It is stated to precipitate hydrated peroxide of manganese and t o evolve nitrogen (Grnelin’s Handbook, iv, 237). Wohler (Jahresbericht, 1865, p. 150) found that nitrous acid was also produced in quantity. * Weighed as Mn304.ON POTASSIUM PERMANGANATE. 97 On adding a strong solution of ammonia to one of permanganate, the mixture becomes hot, nitrogen is slowly evolved, and a brown flocculent precipitate of manganese sesquioxide separates out.The filtrate contained potassium nitrite and nitrate. The reaction may be represented by the following equation :- 8RMnOa + 8NH3 = 4&tn203 + KNO, i KNOz + GKHO + 9HZO + 6N. Analysis of Oxide. Water. Substance. MnO Calculated. Found. Calculated. G. 1.0613 gram 0.1508 = 62.09 62.50 10.71 10.22 111. Action of Phosphine. Phosphine, prepared by warming phosphorus with solution of pot- ash, was passed through a series of flasks containing solution of potas- sium permanganate (the air had been previously expelled from the apparatus by a current of hydrogen). The escaping gas was not spontaneously inflammable so long as any pernianganate remained un- decomposed ; the completion of the readion was therefore easily noted by the gas regaining this property at the exit-tube of the apparatus. The excess of phosphine was expelled by hydrogen, and the precipi- tate collected and analysed.One portion of the oxide dried over sulphuric acid was found to contain a molecule more of water than the other portion, which was dried on the water-bath, but in each case the oxide obtained was the sesquioxide. The filtrate was quite clear and colourless, and had a slightly alka- line reaction. It gave the reactions both of phosphorous and phospho- ric acids. 6KMn04 + 4PH3 = 3Mn2O3 + 21C2HP03 + 2KH2P03 + 3H20 and 2KMnOa + PH3 = Mn,O, + KzHP04 + H,O. These changes may be thus represented :- Analyses of Oxide. H. 0.7804 gram 0.5674 gram = 56-31 56.70 (31n203 + 2H20) I.0.7995 ,, 0.6328 ,, = 61-31 62.50 (Mn,O, + HzO) Substance. MnO = per cent. of manganese. Calculated. IV. Action of Arsine. Arsine, prepared by allowing R solution of arsenic in hydrochloric acid to drop on granulated zinc, and purified by passing the gas through a dilute solution of soda, was passed into a neutral solution of potassium permanganate. The colour of this solution rapidly disap- peared, and a dark brown precipitate, which contained a trace of arsenic, separated out. The precipitate was found on analysis to be98 JONES ON THE ACTION OF REDUCING AGENTS manganese sesquioxide ; the filtrate, which was faintly alkaline, con- tained only potassium arsenate. The reaction may be thus repre- sented* :- 2KMn04 + ASH~ = Mn203 + K,HAsO4 + H20.Analyses of Oxide. J. 0.8250 gram 0,6135 grain = 57-60 56.70 (Mn203 + 2H,O) Substance. MnO = per cent. manganese. Calculated. K. 1.0200 ,, 0.8325 ,, = 63.19 62.50 (M11203 + H,O) V. Action of Stibine. This gas, prepared similarly to arsine, behaved like that gas, except that the precipitate was more flocculent ; the filtrate contained potas- sium antimonate and a small trace of manganese. The reaction may be thus represented- 2KRlnOi + SbH, = MU203 + KZHSbOa + HZO. Analyses. Substance. MnO. Per cent. Mn. Calc. Found. Calc. L. 0.9100 0.7465 = 63.53 62.50 9.81 10.22 M. 0.8895 0.7075 = 61.61 62.50 - -- So far then, the results show that ammonia and its analogues react on potassium permanganate in a closely similar manner, and that hydrogen agrees with them in so far as the nature of the precipitated oxide is concerned ; it ir, uniformly the hydrated sesyuioxide.On ex- tending the investigation to the action of other reducing agents on potassium permanganate, I obtained as before precipitation of the sesquioxide, but accompanied (in all the cases I have examined) by the Zibeyatio?a of oxygen. Water. I now give time results. VI. Actio$L of Oxulic Acid. The action of oxalic acid on permanganates has been made the sub- ject of investigation by many chemists. €3 e r t h e lo t (Juhresbem'cht, 1867, page 334) found that it is oxidised to carbon dioxide and water, not only in acid, but also (though slowly) in alkaline solutions. Ver- non H a r c o n r t (this Jouurttul [el, v, page 460) in a paper on the * S c h o b i g (Zoc.cit.) represents this reactiou by the following equation, but gives 10H385 + 16KMn04 = 5K34s04 + KRTnAsO4 + 2Mn3(As04)a + 9Mn(OH)2 The reaction has also been studied recently by P a r sons (ChmicaE News, 1877, page 236), who represents it by the following equation :- 3K2&h208 + ~ A S H ~ = 3Mn202(0H):: + 2As203 + 6KH0. no proof of the formation of the various products formulated- + 6Hz0.ON POTASSICW PERMANG ANATE. 9 9 action of oxalic acid on permanganate, gives the following equation for the reaction- 21(Mn04 + 3HzSOa + 5HzCz@4 = KzSOa + 2MnS04 + loco2 + 8HZO. I find, however, that this equation does not express all that occurs: since oxygen is evolved along with the carbon dioxide, both when sul- phuric acid is present and when it is absent. On adding solution of potassium permanganate to a warm solution of oxalic acid, the purple colour of the former rapidly disappears ; the liquid remaining colourless up to a cert.ain point, when oxalate of man- ganese separates out even in the warm liquid.Beyond this point, the further addition of permanganate produces a brownish coloration, and then precipitation of oxide of manganese, which continues until the whole of the oxalate has been decomposed. The reaction of the liquid, acid a t first, changes to neutral, and finally to alkaline. Carbon dioxide is evolved from the beginning to the end of the oxidation, and along with it oxygen. These two stages in the reaction may be represented by the follow- ing equations :- (a.) 2KAl'n04 + 8C2H204 = 2MnC204 + K2C2Oa + loco2 + 8H20, (b.) 2MnC204 + 2K1Sln04 = 2Mn203 + KZC03 + 3c02 + 0.The presence of oxygen in the evolved gases was ascertained by ex- periments made in one or other of the following ways :- (1.) Solutions of oxalic acid and potassium permanganate were separately poured into a long glass tube which was completely filled, then closed with the thumb, inverted and opened under water con- tained in a small beaker. The liquids soon mixed by diffusion and bubbles of gas rose to the top of the tube. When no more gas was evolved, the quantity obtained was decanted into another tube, shaken up with caustic soda to remove the carbon dioxide, and the residual gas tested for oxygen in the usual way. (2.) A flask was closed with an india-rubber stopper through which passed three glass tubes ; one of these reached to the bottom of the flask, and was connected with an apparatus evolving carbon dioxide, another conveyed the evo1-i-ed gases to a glass gas holder, and the third was wide enough to admit the point of a stoppered burette which was fitted air-tight into the tube.The flask contained the liquid to be examined (e.y., oxalic acid solution or oxalic mixed with dilute sul- phuric acid), the burette contained the permanganate. With this apparatus an experiment was made by first expelling the air with a current of carbon dioxide which was stopped as soon as the air was expelled. The permanganate was then allowed to flow in, and the gases produced by the reaction were received in the glass gas-holder100 JONES ON THE ACTION OF REDUCING AGENTS which contained solution of caustic soda to absorb the carbon dioxide.The unabsorbed gas was collected in a tube and tested for oxygen as before. I intend to make this reaction of oxalic acid on permanganate the sizbject of further investigation, and in the meantime need only say tha't in whatever way I varied the conditions of the experiment, oxygen was uniformly obtained as a product of the reaction; so that, whether there be excess of oxalic acid or of permanganate, whether the solutions be strong or dilute, whether another acid be present or not, in each case oxygen is evolved. It could scarcely be doubted that other reducing agents than oxalic acid might act similarly on permanganate, and liberate oxygen. I therefore examined its action on ferrous salts, and at once perceived that this was the case; oxygen is evolved in presence or absence of acids, in dilute or in strong solutions.Further, a few experiments soon showed me, that in the action of sulphuric acid upon manganese dioxide in presence of an oxalate, oxygen is evolved along with the carbon dioxide. Obviously, these observations are of importance in connection with the analysis of manganese dioxide, and with the use of permanganate in volumetric analysis. VII. Manganese. Chloride. When a strong solution of permanganate is added to one of man- ganese chloride, sesquioxide of manganese separates out and bubbles of gas are evolved. On examining this gas I found it to consist of a mixture of chlorine and oxygen. The reaction may be thus repre- sented- MnClz + RlfnOp = Mn,O, + KC1 + C1 + 0.On mixing dilute solutions of the same salts, oxygen was evolved, I also ascertained that oxygen is evolved when but not chlorine. manganese sulphate is added to solution of potassium permanganate. Analysis of Om&. N. 0.879 gram 0.505 gram = 62.13 62.50 (1\In203 + H,O) It will be notieed in the preceding experiments that the oxide of manganese uniformly precipitated is the sesquioxide, and not the dioxide, which is generally stated to result from the decomposition of permanganates. I f the dioxide be actually precipitated, I think it not improbable that it a t once decomposes into sesquioxide and oxygen (hence the oxygen in the above experiments), behaving in this respect like man- ganese tetrachloride, which is formed when cold hydrochloric acts Substance. MnO = per cent.of manganese. Calculated.ON POTASSIUM PERMANGANATE. 101 upon manganese dioxide, but which soon splits up into the dichloride and free chlorine. And if the dioxide be not obtained by the reduc- tion of permanganates, it seemed doubtful whether it can be obtained at all by precipitation. Some confirmation of this supposition was obtained by an examina- tion of two oxides prepared by methods which are said to yield the hydrated dioxide of manganese, viz., the oxide obtained by the addi- tion of bromine to a mixture of a manganese salt with ammonia, and the oxide obtained in Weldon's process for the recovery of manganese from the chloride. In the former case the oxide contained 62.36 per cent. of manga- nese, in the latter the oxide, after washing with dilute nitric acid to remove lime, &c., was found to contain 62.57 per cent.of manganese. Both these analyses agree closely with the formula, Mn,O, + H20, which requires 62.50 per cent. of manganese. Surnnaary. (u.) Hydrogen decomposes neutral, alkaline and acid solutions of potassium permanganate, with separation of sesquioxide of manganese in each case. (b.) Hydrogen left in contact with solution of permanganates is completely absorbed a t the ordinary temperature, more rapidly at 100" c. ( c . ) Ammonia decomposes potassium permanganate with liberation of nitrogen, separation of manganese sesquioxide, and formation of potassium nitrite and nitrate. (d.) Phosphine precipitates sesquioxide, and potassium phosphite and phosphate are formed. (e.) Arsine precipitates sesquioxide, and potassium arsenate is formed. (f.) Stibine precipitates sesquioxide, and potassium antimonate is formed. (g.) The first action of oxalic acid on potassium permanganate re- sults in the formation of manganese and potassium oxalates, water and carbon dioxide. After this stage has been reached, the further addi- tion of permanganate produces precipitation of manganese sesquioxide, formation of potassium carbonate, and evolution of carbon dioxide and oxygen. (h.) Oxygen is evolved along with carbon dioxide when sulphuric acid acts on manganese dioxide in presence of an oxalate. (i.) Oxygen is evolved when ferrous sulphate, manganese sulphate, or manganese chloride act on potassium permanganate. Oxygen is also evolved in presence of sulphuric acid.