ACTION OF HEAT ON NITROSTL CHLORIDE. 73 X-Action of Heat on ATitrosyl Chloride. By J. J. SUDBOROUGH, B..Sc. (Lond.), A.I.C. (Associate of the Mason College), and J. H. MILLAR. IT is well known that nitric peroxide, Nz04, begins t o dissocide at temperatures just above the boiling point of the liquid. When the gas is heated, dissociation into nitrogen dioxide is far advanced, according to Playfair and Wanklyn at a temperature of 97*5", and is complete, according to Deville and Troost, a t 140" (Compt. rend., 64, 237). Richardson has studied the action of heat upon this compound n t higher temperatures, and has found that a t 620" nitrogen dioxide is cornpletely dissociated into nitric oxide and oxygen (Trans., 1887, 51, 397). We determined to examine the action of heat i n like manner upon nitrosyl chloride, the only known oxychloride of nitrogen, as i t seemed probable that it would dissociate into nitric oxide and chlorine more readily than nitrogon dioxide splits into Eitric oxide and oxygen.TOL. LIX. a7.3- SUDBOROUQH: ACTION OF HEAT To elucidate this point, we undertook several series of experinien ts, of which +his paper contains a brief account. Prepnration .f Nitrosy1 Chloride. As a means of preparing the chloride, we invai-iably iised nitroayl sulphate (chamber crystals) and sodium chloride. The sulphate we obtained by passing the red gas evolved from copper and nitric acid or white arsenic and nitric acid into sulphuric acid. The red fnmes were first passed through a Liebig conderiser with a Woulff's bottle attached, so as to condense any nitric acid which might pass over, and all traces of moisture were removed by interposing a calcium chloride tube.The purified gas was then passed into the sulphuric acid, That these precautions were necessary was proved by t,he fact that our first sdphate--prepared by merely paseing the nitrous fumes into sulphuric acid without previous drying-contained a small amount of nitric acid, and the chloride prepared from it always contained traces of nitrogen peroxide and hydrochloric acid, the former of which was readily detected by its absorption spectrum. When the nitrous fumes had been passed into the sulphuric acid for some eight or nine hours, small crystals of nitrosyl sulphate appeared ; the reaction was then stopped, as the sulphate in a semi- liquid form was much more convenient.The sodium chloride used was merely common salt thoroughly dried. The two salts were mixed together in a Wurtz's flask, and the nitrosyl chloride, which was evolved on gently warming, was passed through a calcium chloride tube. That the gas thus evolved was practically pure nitrosyl chloride, was proved by estimating t,he amount of chlorine in a given weight of the gas. Vapour Density at Ordinary Te?npemtu?.es. The vapour density of nitrosyl chloride at ordinary temperatures had been previously determined by Tilden (Joum. Chem. SOC., 1874, 27, 632). From his results, i t was clearly established that the mole- cule of the gas at ordinary temperatures is represented by tlie formula NOCl, and iiot by any multiple of this.Density found. 33.25 Calculated for NOCl. 32-67 Our investigations consisted of determinations of the vapour density at temperatures ranging from 15" t o 985". Where possible, the uensity was obtained by weighing the gas, and checked by esti-ON NITROSYL CHLORIDE. 75 mating the chlorine; but, in some cases, we had to rely mereIy on the amount of chlorine found in order to calculate the density. In order t o fill the bulbs with the gas, we used one of the two following methods : - (1.) Condensing the gas by placing the bulb in a freezing mixture, till several cubic centimetres of the liquid were formed, and then allowing it to boil off (b. p. -8O). In this case, we found it very requisite t o have the bulb perfectly dry, and to have the outlet tube as far removed from the freezing mixture as possible, as the moisture which condensed around the cold tube readily decomposed the chloride as it passed into t'he air, and the nitrous and hydrochloric mids thus formed would readily find their way back into the bulb. (2.) By passing the gas through a bulb open at both ends t'ill all the air was expelled.We usually let the gas pass through for from 30 to 45 minutes. ,Sulphuric Acid Bath. Our first set of experiments was made at temperatures between 15" and 165". The method used was that of condensing the gas in the tube. The kind of tube we found most convenient, both for fittirig in the bath and also for obtaining the liquid free from moisture, is that represented in the figure. The bulb had two openings, bhe larger one, which was wide enough .to allow an ordinary piece of glass tubing to pass down, was ground, and had a glass stopper to fit, whilst the smalier one consisted of a long narrow capillary.The gas was passed in through the wide 6 276 SUDBOROUGH : ACTION O F HEAT neck by means of a piece of glass tubing which reached t o the bottom of the bulb. When a quantity of liquid was condensed, the neck was stoppered, and a cap placed over the end of the capillary. The bulb was cleaned, and the cap removed. It was then put icto the sulphnric acid bath, which was large enough to cover the bulb up to the stopper. The temperature was slowly raised, and the bath kept at the required temperature for 10 minutes. The cap was then replaced, the bulb removed, cleaned, and inverted in distilled water ; the cap was removed, and the water allowed to enter.The solution thus obtained was mixed with excess of ammonia, and evaporated on the water-bath to dryness in order t o get rid of all nitrite; the residue was then dissolved, and the solution titrated with decinormal solution of silver nitrate. From the amount of chlorine found, the vapour density was readily determined. Three cloterminations made in this way at 15' gave a mean yesult of 32.5. Experiments conducted in the same way at 65", 115", and 165' gave results which showed that no dissociation had taken place. In all cases, the density came out above 32. The density according to theory is 32.67. Bath of Methyl Xnlicylate Vapour. We next proceeded to heat the gas in a bath of methyl salicylate vapour, the temperature of which was 222".In this set of experi- ments, the gas was allowed to pass through the bulb for about 45 minutes. The bulbs used were merely large boiling tubes of about 150 cubic centimetres capacity drawn out at both ends into fine long capillaries. When the gas had passed through for the allotted time, one end was sealed off; the bulb was then lowered into the bath, and the open end allowed to project through the cork; this open end was fitted with a smaller bulb containing nitrosyl chloride, in order to prevent diffusion during the experinien tr. The salicylate was then boiled, and after the bulb had been in the vapour for about 10 minutes, the open end was sealed off. The bulb was then removed, allowed t o cool, and weighed; and from this weight the density was calculated. The end of the capillary tuba was afterwards broken off under water, the chloride dissolved, and the chlorine estimated as before; this served as a check on the density determined by weighing.Three experiments were conducted at this temperature ; the resulting density was 32.3, hhus indicating that no dissociation had taken place. Bath of Sulphur Vapour. Our next experiments were conducted in a bath of sulphur vapour. The vessel in which the sulphur was boiled was not of very largeON NITROSYL CHLORIDE. i 7 dimensions, and consequently the vapour became superheated. It was, therefore, necessary to determine the temperature by means of an air thermometer. This was readily d-one by taking a dry bulb of approximately the same size as the one used for the nitrosyl chloride, and fusing off when it had been in the vapour f o r 10 minutes.When cool, it was opened under water, and the volume of residual air determined ; from this and the total capacity of the bulb the tempe- rature was easily calculated. Three temperatures determined in one day only varied by 7" ; the numbers found were 690*6", 692.3", and 697.6". Four determinations of the density were conducted in the sulphur vapour. I n the first two, we liquefied the gas in the bulb, and allowed all excess to boil away after the neck had been drawn out into a long capillary. At first, we experienced great difficulty in keeping the bulbs free from moisture whilst the gas was passing in ; we eventually got over this difficulty by fixing a long T-tube on t o the neck of the bulb, and leading the gas through a capillary which passed down the T-tube, and reached to the bottom of the bulb, where it was condensed.I n the last two experiments conducted at this temperature, we merely passed the gas through the bulbs for about 30 minutes. The four results agreed fairly well, ranging from 32.3 to 33.1, and thus they indicate that no dissociation had taken place. Ah-bat h. We next made two determinations of the density in an air-bath, which was heated by means of four large Buzlsen burners. The temperature was determined both before and after each experi- ment by means of air thermometers. The bulbs were filled with the chloride by passing the gas through, and the two following results were obtained : - Temperature...... 796" V. d. . . . . 31-36 Temperature.. .... 816" V. d. ..... 91.00 These numbers indicate a percentage decomposition of about 13. Combustion Furnace. Our last experiments were conducted in a combustion furnace. We first used a method similar to that adopted by V. Meyer and Ziiblin in their investigations on chlorine, oxygen, &c., at high temperatures (Berichte, 12, 1430). The tube used was an ordinary piece of combustion tubing, the same length as the furnace ; t o one end was fixed a tap, whilst the other was drawn 08 into a capillary.78 SUDBOROUGH : ACTION OF HEAT A determiuation of the temperature was made both before and after each experiment. The tube was first dried by aspirating dry air though it4 for half an hour whilst tho tube was red hot.The tap was then turned off and the tube allowed to remain at the temperature of the furnace f o r 10 minutes ; the residual air was then measured by driving i t into a Schiff’s nitrometer, filled with strong caustic potash solution. A stream of carbon dioxide, dried by passing through two sulphuric acid tubes, was nsed for driving the air over. The carbon dioxide, in its turn, was displaced by the nitrosyl chloride, which was allowed to pass through the tube for about half an hour. A flask was attached to the capillaryat the end of the tube in order t o prevent diffusion. That a certain amount of dissociation did take place was proved by the fact tthat the gas which passed out of the flask turned red on meeting the air, thus showing the presence of a certain amount of nitric oxide.When we considered thatall the carbon dioxide had been displaced, the tap was again turned off and the tube allowed t o remain for 10 minutes. The nitrosyl chloride was then driven out by means of a stream of carbon dioxide into a Varrentrap’s bulb, containing either distilled water or ammonia. When all had been absorbed, the solution was treated as in the former experiments, and the chlorine estimated. Another determination of the temperature mas then made by aspirating dry air through the tube, allowing it to expand, and measuring the residue in the nitro- meter. We found that the temperatures were generally very concor- dant, rarely differing by more than lo”, and often coming within 1 or 2 degrees of each other.Three determinations of the vapour density were made in this way with the furnace at its full heat. The results obtained were 29.73, 29.1, and 29.1. In all these cases we found that the nitrosyl chloride acted on the glass tube forming chlorides of the alkalis. These were volatilised by the great heat, and deposited in the capillary. As the gas was kept passing through the tube for some time, the amount of these chlorides was quite appreciable. We, therefore, made several determinations in the combustion furnace, the gas having previously been liquefied in the tube. This we accomplished by taking a piece of combnstion tubing nearly the same length as the furnace, rounding off one end, and drawing the other out slightly. Three or four cubic centimetres of the chloride were then condensed in the tube, and whilst it was still in the freezing mixture the open end of the tube was drawn out into n very fine capillary; the excess of nitrosyl chloride was allowed t o boil away, then the tube was warmed by placing i t on the bricks over the furnace.A small bulb was fixed on to the capillary to serve as a79 ON NITROSTL CHLORIDE. reservoir, and thus prevent diffusion. pushed straight into the hot furnace, the bed of which was lined with asbestos paper in order t o keep the tube from sticking, and also to facilitate the pulling in and out of the tube. The tube was left in the fiirnace for about a quarter of an hour; the capillary was the11 sealed off and the tube pulled out. TiYlien cool, the end was broken off under water, and from the solution thus obtained the density was calculated by estimating the chlorine in the usual way.The hem- perature was determined both before and after each experiment, by means of air thermometers. These consisted of tubes approximately of the same length as that used for the nitrosyl chloride. They were first thoroughly dried, then pushed into the hot furnace, left for 10 minutes, sealed off, pulled out of the furnace, and, when cool, opened under water. An experiment conducted in this way, with the burners only partly on, gave a mean temperature of 784", and a density of 31-77. Another experiment at 928" gave a result of 29.0. Two more experiments gave the following :- When warm, the tube Xean temperature.. .. 968" V. d. ..... 27.3 *Mean temperature.... 985" V. d. ..... 27.0 This indicates a dissociation of practically 50 per cent. of the nitrosyl chloride molecules at a temperatiire not much below 1000". As we were not able to obtain higher temperatures with the appa- ratus at our command, and using glass tubes, we coiicluded our ex- periments at this stage. The table (p. 80) gives a r6ssurne' of the results of our experiments. From these results, it is evident that nitrosyl chloride behaves in a, very different manner fyom nitrogen dioxide when subjected t o high temperatures. Thus, at 620", a t'emperature at which nitrogen dioxide is completely dissociated, nitrosyl chloride shows not the least trace of dissociation ; and near lOOO", only about 50 per cent. of the molecules are dissociated. This fact would seem to point to a, difference in the constitution of the molecules of the chloride and the oxide of nitrosyl, as nitrogen dioxide may be called.Nitrosyl chloride, in its reactions with water, behaves as the chloride of nitrous acid, and inasmuch as nitrous acid not only forms nitroso-compounds, but by acting on the group ,CH, produces oximes, TCX-OH, nitrous acid would appear to have the formula, - * The following are a few of the results with the air tubes; they show that the (1.) 925.2" (1.) 967" (2.) 930.0" (2.) 969" estimations of temperature were fairly concordant, and may be relied upon :-80 ACTION OF HEAT ON NITROSTL CHLORIDE. Vapour Density of Nifrosyl Chloride between 15" and 985". Met,hod adopted for obtaining bulb full of gas. Passing gas for 45 mins... Liquefying.. ........... ,, .............. ) ) ............. ,) ............. Passing gas.. ........... ,) ............. )) ............. Liquefying ............. ,) ............. , ) ............. Passing gas.. ........... Ziquef ying ............. Passing gas.. ........... ), ............. ,) ............. )) ............. Liquefying ............. )) ............. Bath. ---- Air.. ............ Sulphuric acid ..... >1 1, 7 1 ..... ..... ..... Nethyl salicylate ... Sulphur vapour.. .. >l 7 1 1 1 ..... ..... ..... Combustion furnace Air-bath .......... . . . . . . . . . . . . Combustion furnace Pempera. ture. -- 15" 15 65 11 5 165 222 693 693 693 693 7% 796 815 928 964 965 965 968 985 V. d. (H=l). -- 32.5'7 32 '50 33 *04 32 *50 33 *18 32 -30 32 -90 32.31 33 -24 33 *12 31 -77 31 *36 31 *00 29 -00 29-73 ? 29 *10 ? 29 -10 ? 27 -30 27-00 Per cent.amount of dissociation. 0 0 0 0 0 0 0 0 0 0 8.09 11 -86 15 -17 33 -51 49 '21 51 -97 V. d. calculated for NOCl .................. 32-65. 21-78. V. d. calculated for complete dissociation . . , . O:N*OH, and the chloride therefore O:N-Cl. Now, NOCl is incapable of directly combining wit'h oxygen, and the compound NOzCl seems to have no existence (Williams, Trans., 1886, 49). Hence, we must suppose that that part of the valency of nitrogen which in NO is con- cerned with linking on another atom of oxygen t o form NO, or N,O, (according to temperature) is already occupied by chlorine in NOC1. Since nitric oxide, NO, and nit,rosyl chloride, NOC1, show no ten- dency to polymerise, the union which is est,ablished between NO, and NO, at temperatures below 140" is probably owing to the oxygen. We have been accustomed to regard nitric peroxide as nitroso-nitric anhydride, representing it by the formula O:N.O*K<~, but since the molecules of NO,, concerned in the process of combination, are all alike, an unsymmetrical formula seems improbable. The formula O:N*O*O-N:O seems to satisfy the requirements of the case, as it would account for the format8ioii of nitric, as well as of nitrous, acid by the act'ion of water,CURVE SHOWING VAPOUR DENSITY OF NITROSYL CHLORIDE BETWEEN 600" AND 1000" C. DENSITIES. hARRl5ON L S l N S L l T H 5' MABTIHS. I A N F W rTHE FERMENTATION OF CALCIUM GLYCERATE, ETC. 81 and this leads t o the formula 0:N-O- for the dioxide. This may, perhaps, account for the instability of this oxide at high temperatures by representing it as due t o the unsaturated condition of the oxygen, while the more stable chloride may owe its greater permanence at high temperatures t o the fact that the chlorine is not in the same degree unsatuyated. In conclusion, me have t o thank Dr. Tildec for suggesting the work and supervising most of the experiments. 3Iuson College, Birrning ham.