72 THOSlAS ON CUPROUS CHLORIDE AND THE X.-Orrz Cuprozcs Chloride and the Absorption of Cnrboizic 0,:iJe and Hydrochloric Acid Gas. By J. W. THOMAS. SONE years ago I noticed considerable irregularity in the determination of carbonic oxide by absorption when using the reagent in the liquid form, simply passing the solution by the usual pipette into the labora- tory tube or absorption tube, and I came to the conclusion that the incorrectness of my determinations was due to the strength of the acid solution of cuprous chloride which I employed. On trying the cuprous chloride solution, supersaturated with ammonia, concordant results were obtained, and I have since used the re-agent in this form. Although the percentage of carbonic oxide (CO) in a gaseous mix- ture is very accurately determined by explosion with oxygen, the direct determination of the same by absorption is very important when it is required to know the constituents of a gas of unknown composition containing various carbon compounds.I n order there- fore to discover the reason of the irregularity in the absorption of CO by cupi-ous chloride, I undertook the experiments the results of which I now have the honour of submitting to the Society. The method recommended for making cuprous chloride ( Cu,CI2) for use in gas analysis (Sutton’s Volumetric Analysis, p. 28.5) is to strongly acidulate a saturated solution of cnpric chloride (CuCl,) with hydrochloric acid (HCl), and place the solution iu a bottle with copper filings or turnings, leaving it until the liquid is colourless. Thia end can be accomplished by adding almost any quantity of acid, provided the solution is allowed to stand for a sufficient time.In endeavouring t o find a more expeditious method by which a Cu,CI., solution could be made and the acid strength known comparatively, IABSORPTION OF CARBONIC OXIDE, ETC. 73 tried that of Wohler (Watt's Diet., 1st sup., p. 493), dissolving the cuprous chloride in hydrochloric acid of known strength. Although a good method, the process is tedious. By accident, when trying some experiments with hydrochloric acid of different but unknown strengths upon CuCI,, I succeeded in making a colourless solution of Cu2C12 jnstantly, hut although I have tried many times since I have not suc- ceeded in doing it again. The result of successive trials brought out the following method, which is very expeditious, although little more than an old method perfected.A long, narrow stoppered bottle, hold- ing about 4 ozs., is filled thvee parts full with closely-packed. copper tnrnings, previously boiled in caustic soda to remove grease. Intro- duce 6 grams of the ordinary hydrous cuprous chloride, and add 20 C.C. of hydrochloric acid (ordinary strong acid, mur. pur.), and shake for a minute until the cupric chloride is dissolved and forms the black cuproso-cupric chloride, then add 10 C.C. of water and let it run into the bottle slowly so as to float on the surface of the acid liquid; re- place the stopper of the bottle and give the latter a violent shake, when the solution becomes colourless instantly, and entirely converted into cuprous chloride, a large quantity of the white compound being deposited. This is a striking illustration of the action of water in break- ing up unstable chemical compounds.The colourless solution, which is very strongly acid, is saturated with cuprous chloride, and after the addition of a further 30 C.C. of water, is ready for use (sol. B). The fol- lowing is another good method for making cuprous chloride rapidly, the operation being conducted as before in a bottle three parts filled with clean and, in this instance, dry copper turuiigs closely packed. Six grams of anhydrous cupric chloride is placed in the bottle and 20 c.c, of the strongest hydrochloric acid added; on shaking for a minute or two the liquid will become colourless and the acid saturated with cuprous chloride (sol.A). For the purpose of finding the best strength of cuprons chloride solution for use in gas analysis, and the most accurate method of using it, 1 prepared some pure carbonic oxide from formic acid. Soh- tion A and R, and in fact any solution of' cuprous chloride in hydro- chloric acid, no matter how concentrated or reasonably dilute, absorbs carbonic oxide with facility, and the irregularity in the estimation of CO by absorption is due to the subsequent treatment of the solution with potash or soda to neutralise the free hydrochloric acid. When solution A or B, or one of intermediate or weaker strength was used, the absorption of carbonic oxide was rapid, especially if the absorption tube was agitated: when, however, a solution of potash was added, notably if passed from a quick-delivering pipette, a considerable evolu- tion of gas takes place.Much of the gas so liberated is carbonic anhy- dride, expelled from the potassium carbonate present in the potash-solu- .74 THOMAS ON CUPROUS CHLORIDE AND TEE tion by the free hydrochloric acid, but I invariably found that some carbonic oxide was set free, and that the quantity liberated depended upon the manner in which the KHo was added. If the potash (a saturated solution was employed) was passed up slowly and allowed t o form a stratum underneath the acid solution (separated by a film of cuprons hydrate) and, when it was added in excess, the absorption tube is briskly and suddenly agitated, as much as 63 per cent.of the carbonic oxide can be set free and will remain unabsorbed. The car- bonic acid liberated is of course absorbed when the canstic alkali is in excess. When the potash is slowly added and the mercury is allowed to drop through the liquid, or the absorption hbe is feebly agitated, the quantity of CO liberated is very much decreased, and it is possible to reduce it to 4 per cent.; the reaction is not, however, very con- trolable. I next tried whether the liberation of carbonic anhydride caused a similar evolution of carbonic oxide, and, to this end, removed as much carbonic anhydride as possible from the potash solution by quicklime, but did not obtain any better results. Potash-solntion of varying strengths was tried in order to see if the saturated solution was too concentrated, but the results obtained were very similar, car- bonic acid being liberated according to the manner in which the acid solution was neutralised, and the carbonic oxide libeyated in 6 experi- ments amounted to 42% per cent,, 37.4 per cent., 14.4 per cent., 5-2 per cent., 3.8 per cent., and 23-8 per cent.I endeavoured to liberate as much as possible in the two first experiments. Apart from the action of potash, the absorption of carbonic oxide by cuprous chloride, although very rapid at first until from 90 to 95 per cent. has been taken up, is by no means complete under half an hour if left at rest, or 15 minutes if well agitated, as the last traces are very slowly absorbed, more especially when mixed with a large volume of hydrogen, which gas appears to retard the absorption of carbonic oxide more than any which I have tried.Although the hydrochloric acid solution of cuprous chloride may be used to detect the presence of carbonic oxide, it cannot be relied upon for a quantitative determination when potash or soda is used subsequently to neutralise the free hydro- chloric acid. The above experiments relate only to the use of cnprons chloride in the liquid form in apparatus of the Frankland and Ward model, and have little bearing on the absorption of that gas by a papier machi6 ball saturated with the re-agent, and the acid vapour subsequently removed by a potash ball. I found, however, that by allowing the cuprous chloride solution to wet the sides of the absorption tube and the potash ball afterwards to come in contact with the liquid, that the accuracy of the absorption, even by this method of procedure, was perceptibly impaired.ABSORPTION OF CARBONIC OXIDE, ETC.75 The method which I long employed was to add ammonia to the cuprous chloride solution until it was strongly alkaline and the pre- cipitate at first formed waB redissolved, and after the absorption of carbonic oxide had taken place, to add very dilute sulphnric acid until the blue colour of the absorbent was nearly destroyed. There is, how- ever, a liability to introduce error by proceeding in this manner, especially if the hydrochloric acid solution of cuprous chloride is strongly supersaturated with ammonia, and if the latter contains much carbonate, on account of the carbonic anhydride which would be disengaged by the sulphuric acid and the risk of adding too much of the latter to set hydrochloric acid free.I have recently made some experiments with a view of using the ammonio-cuprous chloride solu- tion in a form which shall be su5ciently neutral to dispense with the necessity of employing any re-agent to remove the vapour of ammonia. The following method was found to give very concordant results when tried on the same sample of a gaseous mixture. A small pipette (straight) is required, having a long narrow point with an orifice amaller than a wash-bottle jet. Some ammonia (strength one NH3, -880, to one of HzO) is sucked into the pipette and a little of the cuprous chloride solution (B) is poured into a small porcelain crucible or a test-tube on foot, cut short, and stirred with the pipette containing ammonia.As soon as the slightest shade of blue remains permanent, on stirring, fill the usual bent pipette with the solution and introduce the latter into the absorption tube at once, the gas being previously t-eady for the absorption. The ammonia when added to the hydrochloric solu- tion of cuprous chloride forms ammonio-cuprous chloride (in addition to ammonium chloride) which soon separates in the form of very minute transparent crystals, and only a very little, of what appears to be a hydrate, is formed, and the slightest excess of ammonia over that required to neutralise the acid is indicated by the liquid assuming a blue colour. It is well known that when ammonia is added in excess to a colourless solution of cuprous chloride, in the absence of oxygen, the solution remains colourless, but if the cuprous chloride solution is exposed to the air for a few seconds, even some of the cuprous chloride is oxidised, and as quickly converted by the free acid into cupric or cuproso-cupric chloride, which gives a blue colour with ammonia as soon as the free acid is neutralised.By proceeding thus there will be no free ammonia to pervade the gas (the first blue colour being due to ammonio-cuproso-cupric chloride), and consequently no necessity to add acid subsequently, and although much of the copper compounds are precipitated, the re-agent absorbs carbonic oxide nearly as well as the hydrochloric acid solution of cuprous chloride. It is unnecessary to give details of the experiments made, but I may mention that they76 THOMAS ON CUPROUS CJXLORIDE AND THE were tried on mixtures of carbonic oxide and hydrogen, carbonic oxide and marsh-gas, carbonic oxide and ethane, and coal gas (free from carbonic anhydride and oxygen), and in all instances gave very good results, but the absorption is not complete under half an hour, unless the vessel is briskly agitated. A large volume of a mixture of hyditogen and marsh-gas was measured and then transferred over the ammonio-cuprous chloride solution prepared as above in order to see if the pressure of the gas varied, the difference in the two readings was *03 per cent.The hydrochloric acid solution of cuprous chloride (B) answers well, but it should be kept colourless, and to this end the copper turnings must always be above the liquid and the bottle well stoppered, else the solution soon becomes less acid, owing to the formation of fur- ther quantity of cuprous chloride, through oxidation and subsequent reduction.Since the above was written I have made further experiments, with a .view of using phosphoric acid in the place of sulphuric acid for neutralising an alkaline (NH,) solution of cuprous chloride, and I have also endeavoured to arrive a t some method of neutralising the acid solution of cuprous chloride by other reagents than potash and soda, Phosphoric acid does not appear to liberate hydrochloric acid from ammonium chloride in the cold, and only traces of hydrochloric acid are given off when a concentrated solution of ammonium chlo- ride is boiled with phosphoric acid.The free hydrochloric acid in an acid solution of cuprous chloride is neutralised by phosphate of ammonia, phosphate of soda, oxalate of ammonia, oxalate of soda, sulphate of soda, sulphate of ammonia, and alkali tartrates. It is stated (S utton’s Volumetric Analysis, p. 367) that sodic phos- phate and sodic snlphate owe their absorbent powers (for HC1 gas) to the water of crystallisation which they contain, and in Watts’s Diet. of Chem., i, 282, “the ball of sulphate of sodium is recom- mended to be large, else, if much HC1 gas be present, the sulphate of sodium is apt to become deliquescent.” I find that phosphate of soda and sulphate of soda do not absorb HCl gas by virtue of their water of crystallisation, but by giving up half their Na to form acid salts and sodium chloride, When HC1 is added to a saturated aqueous solution of sodic phosphate a t the boiling temperature, the whole of the acid combines with the sodium of the phosphate and forms sodium chloride, providing there be two equivalents of sodium to one of HCl in the solution.From experiments, it appears probable that in the presence of water at the ordinary temperature hydrochloric acid displaces all the sodium i n hydric-disodic phosphate. Either sodic sulphate or disodic phos-ABSORPTION OF CARBONIC OXIDE, ETC. 77 phate may be used to neutralise the €ree hydrochloric acid in the solu- tion of cuprous chloride after the absorption of carbonic oxide, but although a saturated solution of the former contains more sodium, a solution of the latter does not apparently contain free hydrochloric acid when the quantity of that acid present is nearly equivalent to the total sodium.A saturated solntion (aqueous) of ammonium sulphate is, however, the best adapted for neutralising the free acid in a cuprous chlo- ride solution, as this salt is very soluble in water, and also possesses the advantage of holding the cuprous cliloride in solution without destroy- ing its absorbent properties, so that the solution of ammonium sulphate might, if required, be added to the acid solution of cuprous chloride before introducing it into the laboratory tube. The hydrochloric acid solution of cuprous chloride (B) requires about 80 per cent. of its volume of a saturated solution of sodium sulphate to neutralise the hydrochloric acid, but it is best to add volume f o r volume.Per- fectly dry ammonium sulphate absorbs HC1 gas very readily. When R satmated solution of sodium sulphate or disodic phosphate is added to a HCI solution of cuprous chloride after the absorption of carbonic oxide, no carbonic oxide is liberated, but a white crystalline precipitate of cuprous chloride falls down and is afterwards partly dissolved in the sodium chloride formed. If, after a saturated solution of disodic phosphate is added, the tube containing the cnprous chlo- ride, &c., is allowed to stand over mercury for several days, the undis- solved cuprous chloride becomes converted into cupric phosphate by the reduction probably of orthophosphate to pyrophosphate, but no carbonic oxide is liberated. When a gaseous mixture containing hydrochloric acid is analysed in an apparatus on Dr. F r a n k l a n d or Prof. M'Cleod's model, a few drops of a saturated solution of sodium sulphate may be used to re- move hydrochloric acid gas, and potash and pyrogallic acid may be subsequently introduced to absorb carbonic anhydride and oxygen without washing out the laboratory tube, but owing to the moisture which invariably remains in the eudiometer the result will be much more accurate if the absorption of the hydrochloric acid gas is con- ducted in an ordinary absorption tube, and the remainder of the analysis done in the apparatus mentioned. The experiments relat- ing to the action of hydrochloric acid upon sodium sulphate, &c., will be described in another paper.