ACTION OF BROMINE ON POTASSIUM FERRICYAKIDE. 7 6.i LX1.-The Action of Bromirze om Potassium Ferricyanide. By EDGAR J. REYNOLDS Student in the Laboratory of the Normal School of Science South Kensington. Im studying the action of the halogens on potassium ferricyanide with the view of obtaining the so-called superferricyanide of potas-sium I noticed that a black substance was produced in the later stages of the action. The formation of this was so general that the reaction seemed to merit further study; it was only observed, however when the bromine was in excess; when the ferricyanide was in excess a blue substance was formed very much resembling prussian blue but apparently it was not homogeneous and probably contained besides prussian blue Turnbull’s blue and the black substance already mentioned.It has been known for some time that when chlorine in excess is passed into a solution of potassium ferrocynnide or ferricyanide and the liquid is heated to boiling a cyanide of iron generally known by the name of Pelouze’s green is formed and that if the action is con 768 REYNOLDS THE ACTION OF BROMINE tinued this is converted into a blue compound (generally vaguely described as prussian blue). It is also stated that bromine acts in the same manner as chlorine. With regard t o the latter stage of the reaction namely the formation of prussian blue this is correct as it is also with regard to the first piqoduct of the transformation the so-called superferricyanide of potassium. But with regard to the intermediate stage preceding the formation of the blue compound, the statement is wholly inaccurate as instead of Pelouze's green we obtain a black brittle semicrystalline substance.I n my attempts t o obtain the superferricyanide I used iodine and bromine these being heated with potassium ferricyanide in sealed tubes at temperatures of 100" and 120" for several hours. With iodine the products were iodide of cyanogen potassium iodide, potassium superferricyanide and prussian (?) blue ; with bromine, bromides of cyanogen potassium bromide and a much larger quan-tity of the blue substance were formed but no trace of potassium superferricyanide could be discovered the action of the bromine being much more energetic than that of iodine. Blue compounds formed in reactions similar to this are generally vaguely described as prussian blue (as for example that formed by the continued action of chlorine On Pelouze's green).As no analyses of such compounds have been made I prepared a larger quantity by heating the mixture over steam in a flask connected with a reflux condenser as it was fonnd that the reaction went just as well as in sealed tubes. According to the proportion of bromine used and the time that the action was allowed to continue different results were obtained as follows :-I. When the potassium ferricyanide was in excess a blue com-pound was obtained not unlike Turnbull's blue in properties. 11. When bromine was in excess and the action was continued for a long time the blue substance obtained resembled prussian blue rather t.han Turnbull's blue.111. When bromine was in excess and the action was allowed t o proceed moderately and not continued too long a black substance was formed as already mentioned which proved to be a cyanide of iron. Analyses were made of these three products. The product obtained in reaction I was thrown on t o a filter then treated with dilute hydrochloric acid to free it from any hydrate or oxide of iron which it might contain and again thoroughly washed dried at loo" and analysed. The numbers obtained were as follows :-0.3137 gram gave 0.2318 gram CO, 0-2837 , 0.1560 , Fe20 = 38.49 , Fe. = 20.15 per cent. C. This gave the ratio of Fe to C as 1.91 1 a proportion somewha ON POTASSIUM FERRICYANIDE. 769 greater than it would have been had the substance been washed with cold water and dried at the ordinary temperature as during the drying at 100" there was a slight evolution of hydrocyanic acid.This would not however make any great difference and it was evident that the proportion of iron to carbon in this substance (which seemed homogeneous) was intermediate between that in prussian blue and Turnbull's blue respectively although nearer the latter. The product from I1 was treated in the same manner except that i t was dried at the ordinary temperature. Analysis gave the follow-ing results :-0.3225 gram gave 0.2401 gram C02 = 20.30 per cent. C. 0.2506 , 0.1853 , CO = 20.16 ,, 0.6028 , 0.3155 , Fe,O = 36.63 , Fe. This gives the ratio of Fe to C as 1.81 1 exactly that required for prussian blue Fe,CylB. It was in appearance very like pure prussian blue although slightly darker but probably contained also a little unchanged black cyanide of iron and a small quantity of iron oxide (due to washing with hot water).Before giving analyses of the product from 111 it will be advisable to give a more precise account of its production. It was not obtained in sealed tubes except when iodine was used the more energetic action of bromine speedily decomposing it. But when the potassium ferricyanide was heated with bromine in a flask connected with a reflux condenser the action could be moderated and a fair yield of this black cyanide obtained. A good proportion t'o use is 40 grams of bromine to 20 grams of potassium ferricyanide (a satu-rated solution may be taken) and to maintain the mixture at gentle ebullition for five or six hours.The flask containing the bromine and ferricyanide solution is connected with a reflux condenser the tube of which is of such a size that it fits accurately into the neck of the flask because as the action of the bromine has to be continued for a considerable time nothing in the form of a cork can be made use of. The junction of the tube of the condenser and the neck of the flask may be surrounded with caoutchouc tubing taking care that they fit sufficiently closely to prevent any of the caoutchouc which may be acted upon by the vapour of the bromine from falling info the flask (a better plan is to seal the inner tube of the condenser on to the neck of the flask). The substance so formed was treated with dilute hydrochloric acid, thoroughly washed and dried over sulphuric acid in a vacuum.It is advisable to wash with cold water as although the compound is com-paratively stable hot water seems to separate out a little oxide of iron and t o cause a transformation (though very slowly) into VOL. LIlI. 3 770 REYNOLDS THE ACTION OF BROMINE prussian blue. The separation of oxide of iron would appear from the fact that in testing the filtrate to see whether the washing was sufficiently complete no trace of i r m could be found in it yet after evaporating some of it to dryness over steam there was a separation of iron oxide. This could only arise from the water dissolving a small portion of the black substance which was thus decomposed on evaporation; hence the washing with hot water would probably decompose the substance although only to a very slight extent so slight that hydrocyanic acid is not always detected.For a similar reason that is a slight evolution of cyanogen it should not be dried at an elevated temperature but over sulphuric acid in a vacuum. The analyses were made as follows :-The carbon was estimated in the usual manner and the fixed residue after combustion was dis-solved in concentrated hydrochloric acid (with a little nitric acid), and then precipitated by ammonia. I n such a compound as this black cyanide it is highly important that the fixed residue should not be regarded as ferric oxide (as has often been erroneously done), because as I have pointed out in a previous paper the error may be considerable and in this particular case might very nearly amount to the difference between prussian blue and this black cyanide.In these analyses no estimation has been given of the nitrogen as it was evident from the method of formation of this substance and its decomposition by potash that it was a cyanide ; but as the analyses of M. J. A. Muller (Conzpt. rend. 104 934) had shown that the group CO occurred in some double cyanides I afterwards made several nitrogen determinations which showed that the nitrogen was combined very nearly in the form of cyanogen. It was however, remarkable that the proportion of nitrogen was always rather too low but the deficiency was too small to allow of its expression by means of formulae. The results of the analyses were as follows:-[O-3569 gram gave 0.2458 gram COz p-8278 , 0.3965 , Fe203 = 33.51 ,, 11.{ 0.8463 , 0.4323 , Fe203 = 35.76 ,, = 18.78 per cent. C. I. < 0.7352 , 0.3518 , Fe203 = 33.50 )) Fe. I 0.3783 , 0.2608 , COZ = 18.80 ,, This gives the ratio of Fe to C aa 1.78 1. 0.4455 gram gave 0.3379 gram COz = 20.69 per cent. C. This gives the ratio of Fe to C as 1.73 1. 0.3973 gram gave 0.3083 gi-ani COz 0.8217 , 0.4353 ,) FezOs = 37.08 , Fe. 1.0627 , 0.5638 ) Fe203 = 37-14! ,, This gives the ratio of Fe t o C as 1.75 1. = 31-16 per cent. C. I ?I ON POTASSIUM FERRICYAXIDE. 771 (0.3415 gram gave 0.2509 gram C02 I 0.6902 , 0.3485 , Fe203 = 35.34 , Pe. IV* 5 It was to I all appearance the purest specimen-appearing as a per-[ fectly black powder even when in the finest state of division, = 20.04 per cent.C. 0.3487 , 0.2570 , CO = 20.10 ,, This gives the ratio of Fe to C as 1.76 1. 0.4392 gram gave 0.3351 gram CO, 0.8804 , 0.4500 , Fe20 = 35-78 , Fe. This gives the ratio of Fe to C as 1-71 1. = 20.88 per cent. C. The mean of the above analyses gives the ratio of Fe to C a8 1-75 1 which is exactly the proportion required for the formula Fe3Cys. This formula does not include water as none of the speci-mens analysed were in comparable states. A specimen apparently dried as completely as possible over sulphuric acid gave numbers very nearly corresponding with the formula Fe3Cy,,4H,0 thus :-0.4218 gram gave 0.3338 gram CO = 21.58 per cent. C, 0.4545 gram gave 0.3592 , GOz = 21.55 , C, and0.0738 l HzO = 17.49 ,l H,O.and 0.0736 , H20 = 16.19 , H,O. The formula requires 21.43 per cent. C and 16.07 per cent. H,O. Properties and Reactions of the Compound. Its properties and behaviour with reagents generally resemble those of prussinn blue but it is more stable It is a semicrystalline powder, hard and brittle of a jet-black lustre and conchoidal fracture. The change in colour with cyanides of iron with the increase in cyanogen is isather curious thus we have-Ferrous cyanide (not isolated) FeCy, white. Turnbull’s blue Fe5CyI2 a fine blue but not dark. Prussian blue Fe7Cy18 a dark blue. Black cyanide of iron Fe,Cy, black. It is hygroscopic but not so much so as prussian blue. It is decomposed by potash into ferric hydrate and potassium ferro-cyanide and ferricyanide.We should expect that the reaction would be according to the equation-6Fe3Cye + 30KHO = 5Fe2(OH) + S-K8Fe2Cy, + K6Fe2CyI2, but the proportion of ferric hydrate separated is rather greater than this equation shows. Thus from a sample of the product from (11) when simply decomposed by potash without heating 19.76 per cent. Fe 3 ~ 7 72 REYNOLDS THE ACTION OF BROMINE separated ; when boiled with potash 20.70 per cent. Fe separated out of a total of 33.51 per cent. Fe. This greater proportion might arise from some having been already separated and partly from the fei-ri-cyanide formed being further decomposed which according t o Skraup takes place in accordance with the equation-3Fe2CylaK8 + 10KOH = 2FeaCy12K8 + Fe2(0H) + 1OCyK + 2CyOK + 2HzO. The decomposition by potash was precisely similar to that observed by Williamson on the green compound obtained by him by the con-tinued action of oxidising agents on the ferricyanide of iron and potassium namely first a separation of ferric hydrate and a brown liquid (containing potassium ferrocyanide and ferricyanide) which on boiling further deposited ferric hydrate.Hence the discrepancy in the two results given above (1 per cent.). After long-continued digestion concentrated hydrochloric acid dissolves it completely yielding a mixture of ferrous and ferric chlorides. Even the strongest nitric acid has very little effect but when digested with it for a long time it gradually decomposes it by oxidation. On the addition of cold concentrated sulphuric acid it is converted into a white pasty mass which would appear t o be the same product as that formed when prussian blue is treated in a similar way for when water is added to this a blue substance is formed.When boiled with excess of strong snlphuric acid it dis-solves entirely but a blue substauce is precipitated on adding water. Chlorine and bromine gradually decompose it yielding prnssian blue. By the prolonged action of tho air its colour changes to a dull blue, the substance being apparently transformed into prussian blue. It' map at first seem anomalous that both this black cyanide and Turn-bull's blue should be transformed on exposure to the air into prussian blue which is intermediate in composition between the two. But whereas in the latter case it is a process of oxidation in the former it appears to be due to the presence of moisture as if kept in a dry place it remains unchanged for st very long time.The reaction is probably as follows :-3Fe3Cye + 6H20 = Fe2(0H) + Fe,Cy, (prussian blue) + GHCy, whereas in the case of Turnbull's blue it is-6Fe5Cyl2 + 3 0 = 4Fe7Cy, + Fe203. When heated it gives off cyanogen and in presence of moistur ON POTASSIUM FERRICYANIDE. 773 hydrocyanic acid and ammonia forming ammonium cyanide. When heated more strongly in air it burns and leaves oxide of iron. Is this compound to be regarded as ferrosoferric ferricyanide, Fe1’3Fe’’’2(Fe2Cy12)2 or as a cyanide of iron similar to the magnetic oxide viz. Pe2CI,,FeCl2? The method of its formation does not afford us much assistance in answering this question as it consists essentially in the abstraction of potassium by the bromine.The whole molecule of the potassium ferricyanide is split up ferric bromide is formed which acting on the undecomposed ferricyanide might give ferric cyanide FeCy3. But the Fe atom would seem under ordinary conditions incapable of holding three cyanogen-atoms, and hence is immediately decomposed into a stable cyanide (ap-parently the most stable of the cyanides of iron) namely Fe3Cy8. Wyrouboff does indeed claim to have obtained the compound FeCyy, but has not published his analyses; and he gives the following equation to explain the reaction -2CysFeK4 + 8NH4C1 + Aq = 6CyNH4 + Cy6Fe2,3H,0 + 8KC1 + 2NHa (Ann. Chem. Phys. 47 16 284) in which NH4 stands by itself which is absurd.Endeavours to obtain this ferric cyanide FeCy, were not successful. Attempts to determine whether the view taken above of the method of formation of the black cyanide was correct (namely, the action of bromine on ferricyanide) did not meet with success. The plan adopted was by acting with bromine 011 the substance pro-duced by the action of ferric chloride in excess on potassium ferri-cyanide. Prussian blue was formed after some time but at no stage could the formation of the black cyanide be observed. The nature of the decomposition of the black cyanide by potash strongly favours the view that it is a ferrosoferric ferricyanide. This same view is also supported by the results of some analyses I made of the compound before I was aware that potassium was not one of its constituents. Thus one analysis showed 6.59 per cent. I(, 34.16 per cent. Fe and 22.83 per cent. C. Here by adding to the iron found the amount of ferrous iron equivalent to the potassium, we get the ratio of Fe C as already found that is 1.75 1 ; and it would thus seem to indicate that the displacement of potassium by ferrous iron (Fe“) takes place in the later stages of the reaction, which is precisely what we should expect