DR. ROSCOE OX K-On the Absorption of Chlorine in Wkter. By HENRYE. ROSCOE,B.A. PH.D. AT the beginning of this century Dalton and Henry set up the hypothesis that the amounts of gas dissolved by a liquid vary as the pressure under which the absorption takes place. As however this relation between the absorbed gas and the pressure could not be deduced from Dalton and Henry’s own experiments and still less from the later ones of Saussure it has been regarded by chemists as an ungrounded hypothesis until Professor Bunsen,* in his late research showed that it had a foundation in a true law. A series of very careful experiments which Dr. Carius and Dr. Schonfeld have carried out with the absorptiometcr described by Bunsen not only give fresh proofs of the exactitude of the law but show beyond doubt that it is applicable to gases of very great solu- bility.It thus appears of great interest to examine the absorptiometrical relations of gases at the limits of the temperatures at which the same are capable of entering into chemical combination with the solvent. Dr. Schonfeld ’has examined sulphurous acid in this respect and has found that the law is followed even at temperatures which ap- * Phil. Mag. Web. and Mbch 1856 ; Ann. Ch. Pharm. xciii. 1. TBE ABSORPTION OF CFILORINE IN WATER. proach the point where this acid forms a crystalline hydrate with the solvent. In the following research I shall describe the absorptioinetrical relation which exists between chlorine and water at temperatures approaching that at which hydrate of chlorine is formed.As the absorption-coefficient of this gas has already been accurately deter- mined at Schiinfeld I have been able to confine myself to the examination of mixtures of gas of known composition containing chlorine. The first mixture of gases examined was that evolved by the elcctrolysis of concentrated hydrochloric acid. The electrolysis was conducted in a small flask of about 100 cubic centimeters’ capacity filled with hydrochloric acid into which two poles of conducting carbon dipped. A glass tube with the upper end drawn out was fastened on to the neck of the flask by means of a caoutchouc ring and through the tube were melted two platinum wires which conirnuni- cated below with the carbon poles and above with the battery.The gas obtained by a current of four of Bunsen’s elements was washed by passing through a series of bulbs containing water blown on a glass tube and placed in an oblique position. The composition of the gas thus obtained by electroIysis must first be determined. For this purpose the gas was dried over fused chlo- ride of calcium and lcd into a tube of known capacity drawn out at both ends until there could be no doubt that the last traces of atmospheric air were driven out. After accurate observation of the temperature and pressure the tube filled with the mixture of gas was closed with the necessary precautionary measures and one end opened under a solution of iodide of potassium ; and in order to effect the rise of the liquid this was done at a lower temperature than that at which the gas was collected.The iodide of potassium was imw-diately absorbed and a quantity of iodine equivalent to the free chlorine present was separated out. From this free iodine the amount of chlorine present in the tube was determined by Bun sen’s volumetric method.* Two experiments with gas collected at separate occasions gave- 1. 11. t a = 0.0024869 a = 0.002443 t =92.0 t =67.0 t,=59.9 t =58.3 n = 5. n = 2. * Ann. Ch. Plinrm. lxxxvi. 265 ; Chem. SOC.Q.u. J. vi. 90. t The sigiiification of the various lettere will be seen by reference to the origiual research. DR. ROSCOE ON From these numbers the volume of chlorine V C. 0' reduced to and 0.76 pressure of mercury contained in the tubes used in the experiments is found in cubic centimeters by means of the formula- a (nt-tJ c1 0.0031823 I.= ' in which Og0O328.23is the weight in grms.of 1 cubic centimeter of chlorine at 0" C. and 0.76 pressure of mercury. The first experiment gave 16.24; the second 87.36 cub. cent chlorine at 0" C. and 0.76 pressure of mercury If the total capacity of the tube be called C,the barometric pressure at the time of closing P and the temperature during the same time I theincontained0.76andC 0' gas reduced to ofthe total volume tube is found by the following formula :-C. P (') (1+0.00366 T)0*76= For Experiments (1) and (2) the following values were found :-I. 11. C . . 38-81 c n 190.24 Y .Om*7415 . . . Om*7265 T . . 21O.6 C. . . . 11"-1C. This gives the total volume of the first tube 32.58; of the second 174.65. If now the respective volumes of chlorine found by gradu- ation be subtracted from the total volumes the volume of hydrogen gas present in the mixture will be obtained. The coinposition of the two mixtures of gas was therefore,-I. 11. Calculated. Chlorine . 49.85 50.02 50.00 Hydroges . 50.15 49-98 50.00 100*00 100*00 100*00 As the liquid subjected to electrolysis only contained hydrochloric acid and water the products of decomposition formed could only contain chlorine hydrogen oxygen or the oxides of chlorine or hydrogen. The absence of .free oxygen can be safely inferred from the experiments just cited for every volume of oxygen which is set free by the electrolysis of water is necessarily accompanied by two volumes of hydrogen whilst chlorine and hydrogen are set free in equal volumes by the electrolysis of hydrochloric acid.If therefore water were decomposed in the above manner the analysis would not THE ABSORPTION OF CHLORIXE IN WATER. have shown equal volumes of chlorine and hydrogen but an excess of the latter which as already stated was not the case. For the same reason peroxide of hydrogen cannot be formed in the decomposition as the presence of this body would cause a still greater proportional excess of hydrogen. It only remains therefow to be shown that in the mixed gas no oxygen-compounds of chlorine are present. Let us in the first place to take a particular case examine if the gas could contain hypo- chlorous acid.2 vols. of hypochlorous acid consist of 2 vols. of chlorine and 1vol. of oxygen ; 1 vol. of oxygen is equivalent to 2 vols. of chlorine and sets free in the volunietric process exactly as much iodine as 2 vols. of chlorine. This process leaves it therefore quite undecided whether 4 vols. of chlorine or 2 vols. of hypochlorous acid were present ;and further because in the electrolytic decomposition of 4 vols. of hydrochloric acid as in electrolytic formation of 2 vole. of hypochlorous acid exactly the same amount 4 vols. of hydrogen must be set free it is clear that the volumetric process will always show equal volumes of chlorine and hydrogen whether the gas be rendered impure by the presence of hypochlorous acid or not.The question as to the presence of this latter gas is however easily answered when a direct estimation of chlorine with solution of silver is made together with a volumetric determination. The silver detcr- mination shows only the amount of chlorine and not the oxygen of the hypochlorous acid and therefore may give only half as large an amount of chlorine as the volumetric process. The two following experiments show that the amount of chlorine found by the volu- metric method agrees so exactly with that found by the silver deter- mination that the absence of hypochlorous acid may be certainly deduced. By similar reasoning the absence of all other volatile oxides of chlorine can be proved.Three tubes were filled with the gas as formerly described. The first was opened under iodide of potassium and analysed by the volumetric process ; the two others were opened under tolerably con- centrated sulphurous acid by means of which the whole of the chlorine was reduced to hydrochloric acid and precipitated in presence of excess of nitric acid as chloride of silver. The elements for the first tube were- cc =0.0024869 n =2 1 =59.3 t =76.1 and T=14O.7 P=Orn*74G4 C=43.20 From these are obtained- VOL. VlII.-NO. XXIX. C DR. ROSCOE ON I. Volumeof chlorine at 0" C. and 0.76 found by the volumetric process . . 20.290 cc. 11. Ditto ditto calculated . . 20.131 , The elements for the second tube were- Weight of chloride of silver.. . 0-3980 ,) silver (with ash) . . 0*00:31 and T=14"*7 P=0*7464 C=66*70 From these are obtained- I. Reduced volume of chlorine found by silver determination . . 31-26 cc. 11. Ditto ditto calculated . . 31.018 , The elements for the third tube were- Weight of chloride of silver. . . 0.3834 , silver (in ash) . . 0-0026 and T=14°g07 P=0*7464 C=64.27 From these are obtained- I. Reduced volume of chloriue found by silver determination . . 30.034cc. II. Ditto ditto calculated . . 29.949 , After all these experiments and considerations it may be fairly concluded that the electrolytic gas really consists of a pure mixture of equal volumes of chlorine and hydrogen. As the absorption-coefficients of chlorine and hydrogen for water are known a simple volumetric determination of an aqueous solution saturated at a particular temperature with the gaseous mixture is all that is required to determine if chlorine obeys the law of absorption and if so up to what distance from the point at which hydrate of chlorine is formed.For if u represent the absorption-coefficient of chlorine P the barometer pressure v the volume of chlorine anbv the volume of hydrogen contained in the mixed gas which is passed through a volume h of water until it is saturated the amount of chlorine P dissolved in the water must have the following value if the law of absorption is applicable :-h.u.Pv =v* (3) 0.76 (v+ vl) THE ABSORPTION OF CHLORINE IN WATER. It will be as well to give here for reference the absorption-co- efficients for chlxine and water (t~) as found by Dr.Schonfeld :-C. 0' Coefficient. C. 0' Coefficient. C. 0' Coefficient. 10. . . . 2'5852 21. . . . 2.1 148 31. . .. 1.7104 11. . . . 2.5413 22. ... 2-0734 32. . . . 1.6712 12. . .. 2-4077 23. . .. 2.0322 33. .. . 1'6322 13. . . . 2.4543 2d. . .. 1.9912 34. . . . 1.5924 14. . . 2.4111 25. . . . 1.9504 35. . . . 1*5S50 15. . . . 2.3681 26. . . . 1.9099 36. . . . 1.5166 10. .I. 2-3253 27. . . 18895 37. . . . 14785 17. . . . 2'2828 28. . . . 1.8295 38. . . 1-4406 18. . . . 2.2405 29. . . . 1.7895 39. . . . 1-4029 19. . . . 2'1984 30. ... 1.7499 40. . . . 1.3655 20. ... 9.1565 Three experiments," made at different temperatures with the same volume (9.834 cc.) of saturated solntion gave the following results :-(1) (2) (3) n= 2-n = 1.n = 1-t,=48.9 t,=10.0 t,= 55 t =58.6 t =6'7.6 t =60.3 T-14'*4 I'= 21O.O i? =25O.O u =00024430 Calciilated from formula (3). Hence the vol. chlorine at 0"C. and 0-76at 14'04 is 14.70 . . 1 1.65 > 1 1) , , 21O.O ,,12.643.. 11.35 9 9 9 > , 25O.O ,,11.99 .. 9.36 These figures show that the amounts of chlorine found in the satu- rated solution differed considerably from the amount which should be contained therein according to the law of absorption. Let us now proceed to the consideration of the caizses which might possibly effect this increased absorption of chlorine. It has been already shown that the gas employed in the experiments did not contain any amount of oxides of chlorine which could" possibly pro- duce this greatly increased coefficient of absorption.It is however quite possible that chlorine should act towards water as it does towards so marly bases and a formation of hydrochloric acid and * It is almost unnecessary to state that all the experiments on mixtures of chlorine and hydrogen were conducted in a darkened room a candle being the only light present. DR. ROSCOE ON oxide of chlorine take place. It is possible that the compounds thus formed were not present in the gaseous mixture because they were retained in solution by the liquids with which they came in contact. Such a partial decomposition of the water by chlorine into hydro- chloric and hypochlorous acids would most satisfactorily account for the above irregularities.This question may be easily settled by an experiment founded upon the law of absorption. If we suppose that when chlorine is dissolved in water hydrochloric acid and any volatile oxide of chlorine is formed it is easily seen that not only the volu- metric process but also a direct silver determination must give exactly the same results as would be found if the liquid contained only free chlorine. A totally different result will however be obtained if any gas which obeys the law of absorption-as for instance carbonic acid- be passed into a saturated solution of chlorine in water. If merely chlorine be present it will be driven out by a stream of carbonic acid and replaced by this gas in the proportion of their relative absorption- coefficients.If on the contrary hydrochloric acid and a volatile oxide of chlorine are present together with free chlorine the chlorine and oxides of chlorine will be driven out in an amount different from that of the hydrochloric acid which when dissolved in alarge quantity of water is not volatile. Thus a relation between the components will be brought about by which the volumetric and silver determinations cannot give like results because the original relation by which the hydrochloric and hypochlorous acids are present in the proportion capable of forming chlorine and water does not now exist. The following experiment in which a stream of carbonic acid was passed in the dark through a solution of chlorine freshly prepared without access of light shows that after the current of gas had passed through for three hours the amounts of chloriiie obtained by volu- metric and silver determinations agreed exactly.The gas was first passed into a bottle containing the chlorine solution next into a second bottle containing distilled water and the resulting solution in both bottles was examined. ANALYSIS FROM THE FIRST BOTTLE. Volumetric method- a=O.0024869 (1) n=2 ti=50'5 t-72-0 (2) n=2 t,=53.4 t=71*0 THE ABSORPTION OF CHLORINE IN WATER. Silver determination- (1) Chloride of silver . . 0-2617 Silver . . . 0*0010 (2) Chloride of silver . . 0.2443 Silver .. 0.0050 These elements give- Reduced volume of chlorine found by the volumetric method . . (1) 20.421 (2) 19.962 Ditto ditto silver determination 20.517 19.599 ANALYSIS FROM THE SECOND BOTTLE. Volumetric method- OL =O.OOfZ4869 (1) n=2 t,=46*8 t=71*9 (2) n=2 t,=48*0 t=71*7 Silver determination- (1) Chloride of silver . . 0.2701 Silver . . 0.0012 (2) Chloride of silver . . 0-2594 Silver . . . 0-0005 These elements give- v (1) Reduced volume of chlorine found by the volumetric method . . 21.185 20.836 Ditto ditto silver determination 21.346 20.303 The supposition of a decomposition of water by chlorine to account €or the observed phenomena is therefore likewise unfounded. As an objection might be raised to this experiment that the oxides of chlorine are not volatile enough to be carried over from their solu- tion by a foreign gas I have examined the action of carbonic acid on a mixture of oxides of chlorine.The mixture of' all the various oxides of chlorine which is obtained by heating chlorate of potash with con- centrated sulphuric acid was dissolved in water and a known volume of the solution submitted to volumetric analysis ; this volume was found to be equivalent to 50.3 burette divisions of normal iodine solution. A rapid stream of carbonic acid was then passed through the solution which after fifteen minutes was again volumetrically analysed and the same volume of solution was found to be equal to DR. ROSCOE ON 24.0divisions; after the current of gas had passed for thirty minutes more the same volume corresponded to only 3.7 divisions.The rapidity with which the values of the volumetric determinations de- creased with the amount of gas passed through shows how easily the oxides of chlorine are expelled from their solutions by other gases and hence the former objection is entirely removed. A similar result is arrived at when the mixture of hydrogen and chlorine after being washed is allowed to saturate a volnme of water. It is here also easy to show that no oxides of chlorine have passed over for in the following experiment the same amount of chlorine was obtained by volumetric and by silver determination :-Chlorine and hydrogen absorbed in 9.843 cc. of water at 38’ and 0”-7339pressure Volumetric method- u =0*0024430 (1) n=l t,=31*8 t=72.4 (2) n=l t,=31.9 t=72.0 Silver deterrnination- ( 1) Chloride of silver .. 0.1117 Silver . . 0.0016 (2) Chloride of silver . . 0.1087 Silver . . 0.00‘48 Mean reduced volume of chlorine from volu- metric method . . 8.700 cc. Ditto ditto silver determination . 8.8966 , In order that no possible cause may be left undetermined I have examined the action of free hydrochloric acid upon the solution of chlorine. It was possible that the formation of hydrochloric acid from the hydrogen and chlorine might induce a larger absorption of chlorine and thus the phenomena be explained.. It was however found that the presence of hydrochloric acid lessened instead of increasing the absorption-coefficient of chlorine.Water containing T&$h of ils bulk of concentrated- hydrochloric acid was saturated with chlorine at 14O and Om.7366 pressure and the absorption-coefficient calculated according to the formda,- THE ABSORPTION OF CHLORINE IN WATER. The experiment gave when h=9.834-(1) n=2 t,=62*1 t=74*9 a=0*0024869 (2) n=B t1=60*0 t=73*9 Hence the Coefficient obtained is . . 1.9786 Coefficient for pure water . . 2.3911 One assumption alone remains after all these experiments namely that near the temperature at which the formation of hydrate of chlo-rine begins the atoms of chlorine exert an attraction on those of the other gas present and on the water similarly to the law of Mariotte at the point of condensation and that thereby the accuracy of the law of absorption is lessened.In order to form an idea of the amount of this molecular disturbance it is possible to calculate the volume of chlorine which for any given temperature does not obey the law of absorption. The equations for this calculation are obtained from the volumes of chlorine which are absorbed in water firstly for pure chlorine and secondly for a known mixture of this gas with hy-drogen. Let Y be the reduced volume of chlorine absorbed in A volumes of water when pure chlorine is used; Frl the volume of chlo-rine dissolved in h volumes of water when the mixed gas is used ; a the amount of chlorine in the mixed gas ;vl the volume of hydrogen in the latter ; P the observed barometric pressure; y the reduced volume of chlorine obeying the law of absorption which is contained in the unit of water; x the reduced volume of chlorine which by reason of the molecular action is supposed to be withdrawn from the law of absorption.The following equations give the values of x and y :-(5) V=hy+hx (6) ""0.76 h,yPv (v+v,) +h,x Y (7) x= x-y By meansof these formulz the values of x and y for various tem- peratures have been calculated from the following determinations :- DR. ROSCOE ON Chlorine and Hydrogen-a=0.0024869 h =9.834 I. T"=13'*5 P=0°7431 n=2 t =39*2 t=53.85 11. T0=140*3 P=0*7414 n=2 t,=48.9 t=58*6 111. T0=2l0.0 P=0.7402 n=l t,=10.0 tz67.6 IV. TO=25O-O P=0.7431 n=1 t,= 5.5 t=60*3 V. To=3O0-0 P=0.7320 n=l t,= 4.8 t=53*2 VI.To=38'.O P=0*'7339 n=l t,=31*85 t=72*2 Hence for- 13'05 . . y=1 7831 ~t?=0.6496 14O.3 . . y= '1.7641 x =0.6291 20°*1 . . y=1.6721 x:=0.4880 2PO . . y=1*6287 x =0.4861 25O.O . . y=1.5984 z=0.3589 3OO.O . . y=1.3633 x =0.3866 38O.O . . y=1*0625 x= 0.3771 In order to determine whether the amount of this molecular dis- turbance was dependent upon the nature of the gas with which the chlorine is in contact mixtures of known vohmes of chlorine and carbonic acid were examined and in a similar manner the values of x and y calculated from the experiments. The great difficulty of mixing a known volume of chlorine with a known volume of another gas was overcome by the following simple arrangement :-A large glass tube of about 130 to 150 cubic centimetres' capacity was drawn out the glass thickened at either end and pieces of glass rod ground to fit air-tight into the apertures.The capacity of the tube was then accurately determined and it was afterwards completely filled with carbonic acid and closed. The tube thus filled was opened under a saturated solution of chlorine freshly prepared in the dark and a part of the carbonic acid driven out by the saturated solution. The tube containing the mixture of chlorine water and carbonic acid gas was next well shaken in a water-bath of known temperature and one of the stoppers partly opened to allow the excess of gas to escape. By means of this agitation the statical equilibrium of absorption was established between the chlorine and carbonic acid dissolved in the water and the chlorine and carbonic acid present in the free gas.An effect of this process was an increased volume of free gas. This increased volume was allowed to escape and thereby the original pressure obtained and the agitation and other operations were re- peated until no more gas was evolved and the pressure remained THE ABSORPTION OF CHLORISE IN WATER. constant; or in other words until the equilibrium ensued. The free and absorbed gas must be present in a proportion which may be cal- culated from the law of absorption. This proposition is found by the formulz 1 and 2 used in the former case. The experiments were made in the following manner:-After the tube had been com-pletely agitated it was weighed in order to obtain the volume of water employed and the amount of dissolved chlorine was determined by the volumetric method.The chlorine contained in the gas was also estimated in the same way the tube being cooled with ether and opened under iodide of potassium. A deduction was also made for chlorine con-tained in the residual water the volume of which was found 'by a second weighing after the volumetric examination of the water. By these observations all the data for the calculation of x and y are given. To obtain from a series of experiments the values of x and y the same formulze were used viz. y Y h y= 1- h,. Pv ' Y x=--yh 0.76 (u+ u,) where is as before the volume of pure chlorine dissolved in the h unit of water ;5 the volume of chlorine dissolved in the unit of water hl from the mixed gases chlorine and carbonic acid v the volume of chlorine in the gas and ul the volumeof carbonic acid.Two experiments thus conducted with varying volumes of chlorine carbonic acid and water at the same temperature show very closely approximative results :-EXPERIMENT I. 2"=29'*5 P=0*7428 Capacity of tube=82-62 cc. ; Weight of tube empty= 16.745 ; Weight of tube water and gas =33.005 ; Volumetric analysis of the solution n = 1; t,= 3-5; t= 70.8 ; Weight of tube and residual water= 17.192; Volumetric analysis of the gas n= 2 ; t,= 44.8; t=64.8. From these elements we obtain- T,=15*098 U= 18.120 h,= 16.265 ~+~,=67.455 Hence y= 1.141; x=0.6287.26 DR. ROSCOE ON THE ABSORPTION OF CHLORINE IN WATER. EXPERIMENT 11. E29O.5 P=0*7514 Capacity of tube and weight of tube empty same as I. Weight of tube and solution =34.69 ; Volumetric analysis of the solution :%= 2 ; t,=58.0 ; t=68.0 ; Weight of tube and residual HO=17*160; Volumetric analysis of the gas n=2; t,=43*0; t=68-1. From these elements we obtain-V,=17455 w =19.939 h,3.17.945 W+ V,=64.775 Hence y=1*1456 x=0*6241. The following table shows the values of x and y for various tem- peratures as calculated from the experiments with chlorine and carbonic acid :-13O-5 . . . y=1.7940 . . . x=0.5955 14O.4 . . . y=1*7948 . . . x=0.5963 17O.5 . . . y=1*7990 . . . x=0.4599 20'05 . . . y=1*4024 . . . x=0*7638 22O.O .. . y=1*3129 . . . x=Oo7605 25O.O . . . y=lm2214 . . . x=0.7191 29O.O . . y=1*1022 . . . 3=0*6675 290-5 . . . y=i.i438 . . . x:=ome4 36O.O . . . . y=0.8230 . . . x=0.6283 Mean value of x for all the experiments 06399. From these experiments it is clear that for the same temperature the amount of chlorine not obeying th; law of absorption varies with the nature of the other absorbed gas and that the absorption-co- efficient of chlorine is also altered by this circumstance. Itl is seen from the first table of the coefficients for chlorine an4 hydrogen that the amount of chlorine present a8 not obeying the law diminishes as the temperature increases from the point at which hydrate of chlorine is formed. In the determinations with carbonic acid and chlorine on the other hand this diminution is not seen at the temperature at which the experiments were made.