摘要:

PRIDEAUX THB E2lTEU11 OF SEA-SALT ETC. 1223 CXK-The Efect of Sea-salt o n the Pressure of Carbon Dioxide and Alkalinity of Natural Waters. By EDMUND BRYDGES RUDHALL PRIDEAUX. THE total “uncompensated base” or alkali present as carbonate and hydrogen carbonate in sea-water has been determined by Schlesing and Dittmar and by many later investigators. The method used is nearly always titration with standard acid and an indicator of the methyl-orange class and is subject to the errors of such titrations. The results expressed in milli-equivalents per litre are 2-48 (Schloesing) and 2.41 (Dittmar). Moore Herdman, and the author found values ranging from 2.36 to 2.50 (extreme values) average 2-44 in the water of the Irish sea from November, 1912 to July 1913 (Herdman “Report on the Lancashire Sea-Fisheries Scientific Investigations for 1914,” Trans.Biol. SOC. Liverpod 1915 29). Fox (Tram. Farday SOC. 1909 5 SS) by boiling with excess of standard hydrochloric acid and titrating with standard sodium hydroxide and phenolphthalein found the equivalent of 40 milli-grams of hydroxylion on the average or 2.35 milli-equivalents. The author in June 1913 using the same method found 2-30, VOL. cxv. 3 1224 PRIDEAUX THE EFFECT 03 SEA-SALT ON THE The total carbonic acid may be obtained by boiling with an excess of sulphuric acid in a current of air and absorbing the carbon dioxide in standard barium hydroxide solution. In nearly all cases however the alkali of the carbonate and hydrogen carbonate and hence the total carbon dioxide has been obtained from the titrations with phenolphthalein and methyl-orange the difference between these (in equivalents) being numerically equal to the number of molecules of carbonic acid present.The amount of water taken in the work referred t o above is 100 c.c. to which five drops of 0.1 per cent. phenolphthalein are added and the acid (#/loo- or N/50-hydrochloric acid) is run in until the colour completely disappears. Then five drops of 0.1 per cent. dimethyl-aminoazobenzene or methyl-orange are added and the titration is continued to a decided reddish-orange. With regard t o the phenolphthalein titration it may be remarked that in fresh water AT f 10- down to N f 1000-hydrogen carbonate solutions are still slightly alkaline to phenolphthalein ( h = 8 - 4 t o S - a ) and the titration should therefore be continued until a weak colour remains.I n the case of sea-water however ths % by extrapola-tion of the values given later and by direct measurement (see below) is 7.6. This owing to the presence of salt will appear to have a pH of about 7.8. Consequently the titration should be continued until the solution is quite cdourless. The methyl-orange titrations in dilute solutions and especially in such as contain salt are likely to be high. This was verified in the standard carbonate solutions and neutral sea-water and it was found that from 1 to 2 C.C. of N/lOO-hydrochloric acid should be subtracted from the titrations of 100 C.C. Thus the total alkali is probably 2-24 to 2-34 milli-equivalents per litre. Since the aumber of C.C.of NllOO-acid required in the phenolphthalein titration is from 1 to 3 for 100 C.C. of water that is 0.1 t o 0.3 milli-equivalents per litre the value of R =equivalents of alkali i molecules of carbonic acid is from 2*3/2.2 = 1-04 to 2.3/2*0= 1.15. It may be noted that Scldcesing found carbonic acid equivalent to 98.3 milligrams of carbon dioxide and alkali equivalent to 99.3 milligrams of SO, from which R = 1.12. The alkalinities corresponding with all stages of neutralisation of carbonic acid a t different concentrations have already been expressed as general equations ( P ~ o c . Roy. sot. 1915 [ A ] 91, 535) and those which refer t o the dilute solutions contain only the dissociation constants of carbonic acid and the water constant, together with the expekental quantities c = total Concentration of carbonic acid and R the ratio of alkali to acid.From the curve (ZQC. &.> connecting pE and R which agrees well with th PRESSURE OF CARBON DIOXIDE ETC. 1225 cxpixneutal alkalinities of hard water etc. the following values have been taken in the row of calculated values: c. R= 1.06 1.08 1.10 1.12 0401 p ~ = 9.0 9-15 9-28 9.36 (calculaiad) 0.002 I)R= 8.7 8.9 8.96 8.9 (observed) Obviously the alkalinities are much greater than those found in the case of the same carbonate solutions present in sea-water. It seemed desirable to check experimentally this part of the curve in the case of fresh water. Standard carbonate sollutions were prepared by methods in-dependent of indicators and titrations.Some ill] 10-sodium carbonate was made in the usual way. The factor from the weight of sodium carbonate was 1.015 by conversion into sodium sulphate 1.016 and by titration using methyl-orange 1.018. Some N/lO-hydrochloric acid made by dilution from a known standard and again analysed by conversion into silver chloride had a factor of 1.000. Mixtures (I) (a) (3) (4) were made from these by adding to 50 C.C. of the iMllO-sodium carbonate 47-75 c.c., etc. of the hydrochloric acid in order to produce solutions of ratios 1.06 etc. as above. These were kept in tightly corked flasks. So many C.C. were taken as contained the equivalent of 2.2 C.C. of N/lO-alkali [for example of mixture (1) 4 c.c.] and made up to 100 C.C. with distilled water free from carbon dioxide.These solutions were all 0*0022N with respect to alkali and 0.0021 t o 0-00196M with respect to carbonic acid. Measured amounts of phenolphthalein were added and the samples matched against the Sorensen borate-hydrochloric acid standards. The results are given in the third row above. The curve for C=O*OO2 should fall slightly above that calculated for C = 0.001 ; actually i t is slightly below. The connexion was then determined between p and R of sea-water in which the value of R was accurately known. The same standard carbonate solutions were diluted with neutral sea-water instead of with distilled water. Many experiments carried out with the view of depriving sea-water or sea-salt of its “uncompensated base” and leaving it in a neutral condition were not sufficiently successful to warrant the use of such water in investigating the effect of adding such small quantities of carbonates and hydrogen carbonates.An artificial brine was therefore made from the crystallised salts which possessed the average composition given in Clarke’s “ Data of Geochemistry,” but omitting the potassium salts and others present in such small amounts that they were likely t o have a negligible effect on the 3 a 1226 PRIDEAUX THE EFFfCCT OF SEA-SALT ON THE alkalinities. This watar was tested for neutrality by means of rosolic acid which gave a yellow colour quite indistinguishable from that of a standard neutral solution having pH=7*07. The addition of 0.2 C.C. of 2v/ 100-hydrochloric acid and N / 100-sodium hydroxide to this water (10 c.c.) produced a perceptible change of colour showing that the brine contained no appreciable amount of hydrolysed salts or hydrion regulator.The solutions were then made up from the standard carbonates and this salt water as before. To 10 C.C. were added eight drops of 0-1 per cent. phenolphthalein or six drops of 0.04 per cent. a-naphtholphthalein. The salt error of the former is about 0.2 in pH and that of the latter is the same to a sufficiently close approximation (Sorensen). The hydrion standards used for cam-parison were (1) the borate-hydrochloric acid mixtures of Sorensen (2) the partly neutralised mixture of phosphoric acetic, and boric acids which were previously found by the author (Proc. Roy. SOC. 1916 [A] 92 463) to replace suitably the borate standards from %=8.3 t o 8.0.The results are as follows: R== 1.06 1.08 1.10 1-12 p~ found ............ . ..... 8.1 8.3 8.36 8.3 p~ corrected for salt ... 7-9 8-1 8.15 8.1 As in the case of the water without salt there is practically no change in the alkalinities between R = l . 0 8 and R=1*12. This was the case also both in the fresh and salt solutions having R=1*16. On the average then is 8.9 in fresh water containing hydrogen carbonate and carbonate of these ratios and carbonic acid a t a total concentration of about 0.002 mol. whilst it is 8.1 in salt water which is identical with respect to alkali and carbonic acid. This effect might of course be referred t o the less com-plete primary ionisation of the hydrogen carbonate and carbonate in the presence of so much salt leading to; a diminished hydrolysis.The experimental fact however in itself seems to involve bio-chemical and geochemileal consequences which will be considered after some results on the equilibrium with atmospheric carbon dioxide. The question whether sea-water is or is not saturated with atmospheric carbon dioxide has been much discussed. Fox (Zoc. cit.) has made a most useful series of observations of the actual pressures of carbon dioxide in equilibrium with waters of varying salinity. The values of hydrion were not then fully available as a t present. Obvioudy there is a relation between hydrion con-tent and pressure of carbon dioxide and indeed from the dat PRESSURE OF CARBON DIOXIDE ETC. 122’1 at present available it should be possible to calculate the latter from the values for hydrion R etc.There are still however, difficulties in the way of doing this except by empirical equations. Soino simple experiments carried out in the spring of 1919 a t Port Erin throw some light on this question. Pure air taken from outside the laboratory was bubbled through sea-water for periods varying from a few hours to twenty-four and twenty-seven. The value of p, found in this water was always 8.1. Numerous blank experiments were being made during this period on the sea-water freshly drawn and the pH values varied between 8-2 and 8.3. This alkalinity was preserved when the water was allowed t o remain motionless for a day or more in the laboratory. Thus the sea-water when exposed to a current of air gained carbon dioxide in every case and this although the alkalinity was distinctly low for the time of year since the surface water had been mixed with the lower layers and brought into better contact with the air by recent stdrms.The usual range in sea-water is from 7.95 to 8.35 but values below pEI=8*1 are exceptional in the case of surface waters although they have been found in the Skager Rack and off the coast of Norway. According to the above experiments all waters of higher alkalinity than pH=8.1 that is nearly all surface waters will gain carbon dioxide from the air. The experiment was repeated on the artificial sea-water which was made up with carbonates so that p was 8.2. By saturation with a current of air this was reduced t o 7.9 or 8.1 as measured by different standard solutions and indicators.From the measurements of the neutral sea-water containing known amounts of alkali and carbonic acid it is seen that ph=8*1 in salt water corresponds with a ratio of about 1.08, and therefore any sea-water containing less than 2.2 / 1.08 = 2.04 millimols. of carbonic acid to 2.2 milli-equivalents of alkali should gain carbon dioxide from the air. This result may be compared with a deduction from Fox’s measurements of the absorption of carbon dioxide in salt waters of varying salinity and alkalinity. The I f physically dissolved ” carbon dioxide is expressed as C.C. of the gas “a” dissolved a t different temperatures for each 0.01 per cent. of carbon dioxide in the air. Thus for his highest salinity=20 per cent.of chlorine, which is nearly equal to that of the artificial salt water of the present work pa=3 x 0.0875 =0.2625 C.C. per litre of water a t t = = 1 6 O and p m -0.0003 atmosphere. The carbon dioxide combined with alkali “ t ” is given for each milligram of alkali expressed as hydroxyl and for each preesure and temperature of carbon dioxide. A t t = 16O and pco =0.000 1228 PRIDIGAWX TEE EFFECT OF SEA-SALT ON THE atmosphere b ie 1.18. Taking FOX'S value of the total alkali= 40 milligrams of hydroxyl or 2-35 milli-equivalents the total carbon dioxide= 0.26 + (1.18 x 40) C.C. : R - ~ equivalents alkali - - 40 22.4 = 1.11. mols. of carbonic acid 17-2-47.46 If the sea-water has a total alkalinity of 2-2 milli-equivalents, or 37.4 milligrams of hydroxyl R becomes 1.05.The value of R in equilibrium with air is not greatly affected by changes of temperature; thus 1-11 a t 16O becomes 1.09 a t 1 2 O and 1.115 a t 18O. Thus the direct experimental result is confirmed that in a water of total alkalinity equal to 2.2 milli-equivalents the lowest ratio and the highest acidity that is normally encountered in sea-water is only just in equilibrium with atniaspheric carbon dioxide and that in all other cases the water will gain carbon dioxide. The surface waters of the sea gain carbon dioxide not only from the air but also by mixture with the subjacent layers in which low values of pE are almost invariably found and likewise by the decay of plants. It is apparently only the photosynthetic action of seaweeds plankton etc.which maintains alkalinities of more than 8.1 in the surface waters. There is no known in-organic chemical agency which is capable of doing this. A t the same time since the pressure of carbon dioxide in the sea becomes greater than that in the atmosphere a t alkalinities below pH = 8.1, the alkalinity of surface waters cannot fall much below this point. The alkalinity of bicarbonated fresh waters is also kept a t about p,=8 but by a different chemical equilibrium. I n these cases, the ratio R as defined above is 1.00. and this from the curve already quoted curresponds with p13 = 8.2. Solutions which have the higher ratios found in sea-water namely 1.06 t o 1.16 and contain quite an appreciable amount of carbonate could not long exist in fresh water; t'hey give a strong pink colour with phenol-phthalein as already determined.and would rapidly gain carbon dioxide from the air and probably also in many cases deposit calcium carbonate or basic magnesium carbonates. The converse effect is best seen by noting the probable course of event's when a fresh water of temporary hardnes 11 consisting of hydrogen carbonates having the same total alkali concentration as the sea, becomes mixed with i t . The static acidity would be greatly raised ; t!hus if the salt could be added t o the original hydrogen carbonate, fi would fall to 7-6. This was iouncl by a11 experiment in which the calculated quantities of the skandard carbonate and hydro PRESSURE OF CARBON DIOXIDE ETC. 1229 chloric acid were added to the neutral sea-water and the colour given with a-naphtholphthalein was matched against a mixed acid” alkali standard.By the mixture of equal volumes of the fresh water and sea-water the alkalinity would assume an inter-mediate value and on0 lower than that in normal surface waters, whilst the pressure of carbon dioxide would be correspondingly high and carbon dioxide must be given up t o the air or t o other large quantities of water until R rises again to 1-05. It is quite possible that the higher acidities and pressures of carbon dioxide which have been observed everywhere below the surface are due mainly to this cause-the continued addition of bicarbonate waters -without any adequate opportunity of yielding up the excess of carbon dioxide from such immense masses of water.It may also be noted that were it not for the effect of salt in raising to such a high degree the acidity and pressure of carbon dioxide which correspond with a given proportion of alkali and carbon dioxide, the sea would contain much more carbon dioxide than it actually does. Thus if it consisted of a dilute hydrogen carbonate solu-tion such as a hard water in equilibrium with the air it would contain about 5.5 per cent. more carbon dioxide combined with the same quantity of alkali. Since the total carbon dioxide is about 50 C.C. per litre every cubic metre of sea-water would con-tain 2.25 litres of carbon dioxide more than at present. The relative availability for plant life of the carbon dioxide in fresh and salt waters is a different question.As shown by Moore and his collaborators (lac. c i t . ) the flora of the sea can use the carbon dioxide of the hydrogen carbonate t o a certain limit which approximately corresponds with the production of carbonate. Now owing to the effect of the salt sea-water in equilibrium with the air already has a ratio of 1.04 to 1-06; there is some carbonate present. The alkali present in fresh water in equilibrium with the air and of almost the same alkalinity is practically all as hydrogen carbonate. Therefore in the fresh water of the same total alkalinity for example 2- 2 milli-equivalents per litre there is more carbon dioxide available before the carbonate point is reached. On the other hand far more carbon dioxide can be abstracted from sea-water without an excessive rise s f alkalinity. A sea-water of original ratio = 1-06 (= &G) can by abstraction of carbon dioxide have R changed to 1-16 ( =- o.i62) losing 8.7 per mnt. of the original carbon dioxide and still only have an alkalinity corresponding with p,= 8.35. A fresh water of original ratio=1-00 can have this changed to R=1*06 by losing 5 - 5 pe 1230 BURBOWS THE RATE OF HYDBOLYSIS OE’ METHYL c a t of the original carbon dioxide but this changm the alkalinity to the high value corresponding with p==8-7. The author desires to make grateful acknowledgment of the assistance given by the Percy Sladen Memorial Trust towards this research. MARIXE BIOLOUIOAL STATION UHXVER8ITY COLUEUE, POET ERIN ISLE OF MAN. NOTTINUHAM. [Received July 2274 1919.

ISSN:0368-1645    

DOI:10.1039/CT9191501223

出版商:RSC    

年代:1919

数据来源: RSC

摘要:

1230 BURBOWS THE RATE OF HYDEWLYSIS OE’ METHYL CXII1.-The Rate of Hydrolysis of Methyl Acetate by Hydrochloric Acid in Water-Acetone Mixtures. By GEORGE JOSEPH BURROIWS. IN a previous paper (T. 1914 105 1260) the author recorded some experiments on the inversion of sucrose by acids in water-ethyl alcohol mixtures. It was there found that the rate of inversion a t first decreased slightly as water was replaced by alcohol up to about 50 per cent. of alcohol and then increased. It was concluded from the results that the catalytic activity of the acid was really greater in alcohol than in water and that the addition of water had a depressing effect on the rate of hydrolysis. A t the same time the author was unable to explain the decrease in the rate of inversion which results from the replacement of water by alcohol up to 50 per cent.mixtures without assuming that the mixed solvents had some specific effect on the rate of reaction and it was there suggested that this was due to a change in the fluidity of such mixtures causing a variation in the rate of catalysis similar t o that observed in conductivity. In obtaining the results in that work the reaction was treated as a unimolecular one the concentration of the water being neglected. This assumption was made in view of the fact that even in a solution containing 75 per cent. of alcohol by volume the ratio of the number of molecules of water t o molecules of sugar present a t the commencement of the reaction was 45 so that the decrease in concentration of the water during the reaction was only 1/46 of its original value.In a 50 per cent alcohol mixture the ratio was 91. Under these wnditions it was considered that the concentration of the water could be omitted from the equatien and the reaction considered as unimolecular ACETATE BY HYDBOCHLORIC ACID ETC. 1231 If however the concentration of the water is considered the value of k, calculated from the bimolecuiar equation b(w - x) log - I t(w - b) w\b - x) 1 k =.-(where b and w are the number of gram-molecules of sugar and water respectively in the solution a t the beginning of the reaction), is of course much smaller than the corresponding value of k obtained from 1 b t b-x k = -log--In this case the total concentration of water is considered to represent its mass as one of the active substances irrespective of the fact that it is present in large excess and only a small fraction of the total concentration is used up during the reaction.The difference between the values of Ic and k is shown in the following table : Sucrose (10 per cent.) and N/2-Hydrochlon~c ,4cid at 25*0°. Alcohol (volume per cent.). 0.0 16.7 26.0 40.0 50.0 60.0 75.0 k. 0-00219 0.002 13 0*00204 0.001 92 0.00176 0.00186 0.00208 k,. 0.0000427 0.0000491 0-0000619 0*0000607 0*0000667 0.0000877 0.0001603 It will be seen that whereas the values of k pass through a minimum for a certain mixture the values of k steadily increase. It appeared of interest to see if similar results would be obtained in the case of ester hydrolysis and with this end in view experiments have since been performed on the rate of hydrolysis of methyl acetate by hydrochloric acid.Owing to its effect on the equilibrium alcohol could not be used as one of the solvents in this research so that the experiments were performed in mix-tures of acetone and water. Acetone undergoes change under the influence of the acid used in the hydrolysis but it has been assumed that this has no effect on the rate of hydrolysis of the methyl acetate. The acetone used in these experiments was dried over and dis-tilled from calcium chloride. The methyl acetate was freed from acetic acid by means of sodium carbonate and was distilled from calcium chloride; it was neutral to litmus. The volume of the 3 A 1232 BURROWS THE RATE OF HYDROLYSIS OF METHYL solution was adjusted a t the temperature of the experiment which was 25O in every case.The acetone percentages given are by volume; thus 70 per cent. acetone was prepared by mixing seven volumes of acetone with three of water. In preparing a 5 per cent. solution of methyl acetate for hydrolysis by N / 2-hydrochloric acid in 70 per cent. acetone the following method was adopted 10 C.C. of methyl acetate were diluted to 100 C.C. a t 2 5 O with 70 per cent. acetone ( A ) 20 C.C. of 5N-hydrochloric acid were mixed with 46.7 C.C. of anhydrous acetone and the mixture was diluted to 100 C.C. at 2 5 O with 70 per cent. acetone ( B ) . Equal volumes of ( A ) and ( B ) were then mixed in a dry flask and placed in a thermostat a t 25O.The weights of water and acetone used in preparing the solutions were also determined so that all concen-trations can be expressed in terms of gram-molecular weights. The reaction was followed in the usual way by titrating 5 C.C. of the solution under investigation with baryta solution after different intervals of time the increase in titre indicating the concentration of the acetic acid produced. At first the reaction was treated as unimolecular and the values of k were calculated from the equation ' * (1) 4bn + 2nx( J K - 1 - 1) 4bn - 2nx( J4bn + 1 +l> -~ log ____ 1 t JTbT+l k = obtained by integrating !@= k(b - 5) - k2x2. dt In this equation m= - '-' where ,$ represents the amount of ester which actually undergoes hydrolysis up to the equilibrium, b is the initial concentration of ester and x the amount hydrolysed in time t.It was found that the above equation gave values of k which were quite constant for any particular solution. How-ever in thO case of a series of experiments in different water-acetone mixtures it was found that as the water was replaced by acetone the value of k a t first decreased and then increased. Owing to thO relatively small concentration of water in solu-tions containing a high percentage of acetone it was then decided to calculate all results also from the equation for a bimolecular reaction. ThO equation used was that given by Griffith and Lewis P (I". 1916 109 869), obtained by integrating 2 = k,(b - x)(w - 2) - k2x2, d ACETATE BY HYDROCHLORIC ACID ETC.1233 where k =rate of hydrolysis, k =rate of esterification, b=initial number of gram-molecules of methyl acetate in 1000 C.C. of solution, = initial number of gram-molecules of water, z=number of molecules of ester decomposed in time t . I n equation (2), y = (20 + b) A = d(20 + b)2 + 4(K - 1)zob K == 5 = ______.___ Cwater at equilibrium. Cacetic acid C/deol~ol and The values of both k and k obtained in any solution were found to be constant. This is seen in tables I and 11 which contain the results for the hydrolysis of 5 per cent. methyl acetate by N/2-hydrochloric acid in 80 and 90 per cent. acetone (by volume) respectively. TABLE I. 80 Per c e n t . Acetone. b = 0.6267 ; w = t. 0 48.5 78.5 111 138 168 228 268 318 401 0: 10.20; X=5*14.2. -0.0606 0.0819 01096 0.1314 0-1675 0.1992 0.2200 0.2592 0.2983 0.4963 Mean ... k x lo6. 7.41 7.44 7.43 7.34 7-44 7.32 7.36 7.47 7-31 -7.40 3c x 104. -7.51 7.79 7-66 7-44 7-56 7-40 7.43 7-52 7.35 Mean 7-51 TABLE 11. 90 Per cent. Acetone. b =0.626’i ; w =4*752 ; K = 8-97. t. 0 50 85 10 140 206 263 324 400 484 a 2. -0.0636 0.099 1 0.1262 0.1523 0.2018 0.2436 0.2695 0-2967 0-3 192 0-3608 Menil ... k x lo5. -19.77 18-95 19-35 19.12 18.91 19.80 1.9-34 19.16 19.63 -18-56 m 104 -. 9.35 8-91 9-08 8.94 8.80 9.18 8-94 8.82 9.00 Mean 9-00 1234 BURROWS THE RATE OF HYDROLYSIS OF METHYL In table I11 is given a summary of results obtained for the hydrolysis of 5 per cent.methyl acetate by N / 2-hydrochloric acid in different water-acetone mixtures. The value of b is 0.6267 in each case. The value of K was determined by ascertaining at the completion of the reaction the amount of acetic acid produced and then calculating the equilibrium constant. It was found that the numbers so obtained varied with the different solutions. This method of determining the conditions a t equilibrium by analysing the solution actually used for the rate of hydrolysis is undoubtedly open to error as a small amount of the volatile substances must escape each time the flask is opened to determine the concentra-tion of acetic acid a t each particular time. The amount lost in this way during a complete experiment would probably be sufficient to introduce an error into the value of C (acetic acid) and this might lead t o quite a large error in the value of R.The varia-tions in K are possibly due to a small error in the values of the concentrations of acetic acid found for the various solutions. A t the same time it has been found that a comparatively large differ-ence in li produces only a small difference in the value of k,. Thus the mean value of k for a solution containing 90 per cent. of acetone is found to be 19.56 when R=8.97 and 18.7 when R is taken its 5 the approximate value found in the other solutions. In the latter case however the numbers obtained for b decrease regularly whereas if the experimental values of R are taken there is no such decrease.For this reason it has been decided to employ the values of K actually found although it is realised that the variation in its value for the different solutions may be due in part at least to the method of determination. In the following table the values under k are calculated from equation (l) those under k from equation (2). TABLE 111. Methyl Acetate (5 per celat.) and NI2-Hydrochloric Acid. Acetone (volume per cent.). 0 20 40 60 70 80 90 Grem-molecules of acetone per litre. 0 2-665 5-39 8-10 9.42 10.66 11-84 W. 52.32 42.46 32.08 21.23 15.72 10.20 4.762 K . k x 106. 5.40 2.76 5-10 3-03 4.38 3-38 4.47 3.86 4.61 4.66 5.14 7.40 8-97 19-56 k x 10'. 13.68 12.28 10.33 7.91 7.06 7-61 9.00 It will be seen that the value of k increases regularly as water The values in the sixth column is replaced by acetone as solvent ACETATE BY HYDROCHLORlC ACID ETC.1236 under k however a t first decrease as wa er is replaced by acetone and then increase. A similar result was obtained for the hydrolysis of 2.5 per cent. methyl acetate by N/10-hydrochloric acid. Owing to the small amount of methyl acetate actually present a t equilibrium the value taken for K for each of the solutions in this series was the value found in the corresponding solution with 5 per cent. methyl acetate and N / 2-hydrochloric acid. TABLE TV. Methyl Acetate (2.5 per cent,) and NIIO-Hyd7-ochloric A c i d . In all cases b = 0.3133 gram-molecules per litre.Acetone. Per cent. 0 20 40 60 70 80 90 Gram-molecules of acetone. 0 2.73 5.63 8.3 1 9-68 10.06 12.19 W. b x 10'. 54-00 4.78 43-97 6-40 33-44 5.76 22.26 6.40 16.62 7.42 10.95 10.23 5.351 23.60 k x 104. 2.60 2-39 1.90 1.42 1.22 1.13 1.26 Discussion of Results. I f the values of k given in tables I11 and IV are plotted against water concentrations they are found to lie on a rectangular hyper-bola. In the accompanying diagram a graph is also shown for the results obtained for sucrose inversion given under k in the table a t the beginning of this paper. It will be seen that the three curves are similar in shape and indicate the decrease in catalytic activity of the acid which accompanies an increase in the water concentration.Snethlage (Zeitsch. physikal. Chem. 1913 85 253) Acree (Amer. Chem. J . 1912 48 352) and Taylor (Zeitsch. Elektro-chem. 1914 20 201) have advanced the theory that the undis-sociated molecule of the acid also acts as a catalyst the activity of the undissociated molecule of hydrochloric acid varying for different reactions up to a value twice that of the hydrogen ion. In order to account f o r the present results according to this theory, it would be necessary to assume that kz is about 20 (km being the activity of the undissociated molecule and kEt that of the hydrogen ion). Figures are not available for the degree of dissociation of hydrochloric acid in water-acetone mixtures but in discussing the results obtained for the inversion of sucrose (loc.cit.) the author hx 1236 BURROWS THE RATE OF HYDROLYSIS OF METHYL gave values of Q (HC1) deduced for water-aloohol mixtures. In the following table the values observed for k for sucrose inversion 0 0 10 20 30 40 50 (calculated as a bimdecular reaction) are compared with the total catalytic activity of the acid expressed as t,he ratio k.L being taken as 20. k, + k H t = kH"a + 20(1 - 41, ACETATE BY HYDROCHLORIC ACID ETC. 1237 TABLE V. Alcohol (volume k x lo6 per cent.). a (HCI). [a+2O(l-a)]. k x lo5. [a+20(1-a)]. 0 0.86 3.66 4-27 1-17 26 0.83 4.23 5.19 1-23 40 0.80 4.80 6.07 1-26 50 0.72 6.32 6.67 1.06 60 0.65 7.65 8-77 1.15 75 0.46 11.45 16-03 1.11 The numbers in the fifth column approximate to a constant. Although results cannot be calculated in this way for the hydro-lysis of methyl acetate in water-acetone mixtures owing to lack of data it is considered from the results obtained for k in these solutions that if the value of a (HCl) were available a similar constancy for "1 would be obtained for a value of x in the neighbourhmd of 20.It follows from the above that the activity of the catalyst in these mixtures can best be expressed by k,=k,+k, when the ratio km is considered to be about 20 but this value is not supported by results obtained by Snethlage and others for the catalytic effect of hydrochloric acid in other reactions. The shape of the curves in the diagram suggests that in the mixtures investigated the variation of the activity of the acid with the water content is approximately expressed by k,(w + m) =n, where m and n are constants and k the observed rate of hydre lysis in a solution containing w gram-molecules of water.An equation of this type is capable of various interpretations but taken in conjunction with the marked decrease in k with increase in water concentration for low values of the latter it indicates that the water plays a very markedly anticatalytic r81e compared with the other constituents of the solution that is the value of the constant m must be very sniall compared with w. This may be interpreted (compare Lapworth and Fitzgerald T. 1908 93, 2168) as being due to the conversion of active free hydrogen ions to inactive ions by the solvent the effect of the water in this respect being far greater than that of the acetone.Or it may be interpreted as indicating that the reactive substance in ester hydrolysis is a complex between the ester and water which is readily dissociated by the excess of water in the solution (Grif-fiths and Lewis Zoc. cit.). It may also be interpreted as repre-senting the dissociating effect of the solvent on a complex between the ester and the catalyst which is considered to be the reactive a+x(l -a) 1238 THE RATE OF HYDROLYSIS OF MIETHPL AOE!PA!i% ETC. aubstance. Such a complex may either be dissociated by the solvent into the original substances or else hydrolysed by the water into the final products of hydrolysis. For the concentrations usually employed in hydrolysis the amount of this complex is very small owing to the high dissociating power of the water.As the latter is replaced by acetone or other liquid the dissociating power of the medium is decreased and the concentration of the complex is increased so that the observed rate of hydrolysis is also increased. In any of the above cases it is possible to assign a small value to the term m which really represents the relatively small power of the other constituents of the solution of converting an active sub-stance into one which is inactive in catalysis. The following discus-sion is independent of the nature of m ; it is concerned rather with its magnitude. W0 will assume that it is simply a function of the concentration of the other variable in the solution namely ace-tone. I f we substitute in the equation k,(w + m) = n the values of k and w in table I11 for solutions containing (a) 52-32 gram-mole-cules of water and no acetone and ( 6 ) 4.752 gram-molecules of water and 11.84 of acetone we obtain the value m=0*22 (in terms of 1 gram-molecule of acetone in the solution).Substituting this value for the other solutions the following results are obtained. TABLE VI. Acetone, 0.0 2.666 6-39 8-10 9.42 10.66 11.84 Acetone x 0.22 W. - 52.32 0.59 42-46 1.19 32.08 1.78 21-23 2-07 15-72 2-35 10.20 2.60 4.762 k x 109 2.76 3.03 3-38 3-86 4-86 7.40 19.66 n= k,(w+ 0.22 x metone), 144 130 112 89 81 93 144 The numbers in the last column pass through a minimum value, and vary in a similar manner to the values under k in table 111.Furthermore for any small value of m in the equation k,(w+ m) =n the values of n will be found to vary as in table VI. The similarity between the variations of k in table I11 and n in table VI appears to justify the use of the equation for a unimole cular reaction in mixtures such as those used in this work in which the total concentration of water varies considerably in the different solutions. It would appear that any specific effect of the solvent on the rate of hydrolysis other than that du0 to its disso-ciating power is more directly shown by disregarding the change in the concentration of the water in the different solutions. It is considered that the minimum value obtained for k (or n) has a definite meaning. It is interpreted as indicating a specific influ-ence of these mixtures on the rate of catalysis similar to tha VELOUITIES OF COMBINATION OF SODIUM DERIVATIVES ETC.1239 observed in the conductivity of certain electrolytiea dissolved in them or in the fluidity of the mixtures themselves. The solution of either alcohol or acetone in water is accompanied by contrac-tion and it is considered that the condensed state of such a mixture has a retarding influence on catalysis. Summary. The rates of hydrolysis of methyl acetate by N / 2 - and N/10-hydrochloric acid have been measured in various water-acetone mixtures. The velocity constants have been calculated according to both unimolecular and bimolecular equations the reverse reac-tions being considered in both cases. It has been found that if the total water concentration is taken as representing its active mass then the rate of hydrolysis increases as water is decreased in the mixtures throughout the series. The results obtained approximate t o k = k + kH if 5 is taken M k d 20 (approximately). The variation of k with water concentration indicates the anticatalytic function of the latter and if the values of El are corrected for this effect the numbers so obtained are found to vary in a manner similar to those obtained by using the ordinary unimolecular equation and pass through a minimum value for a certain mixture. This minimum is considered t o have a definite meaning representing a retarding influence of such mixtures on catalytic reactions. CHEaOcazl LABORATORY, UNIVEB,SITY OF SYDNEY. [Received Augucrt 23r4 1919.

ISSN:0368-1645    

DOI:10.1039/CT9191501230

出版商:RSC    

年代:1919

数据来源: RSC

摘要:

VELOUITIES OF COMBINATION OF SODIUM DERIVATIVES ETC. 1239 C X 1V.- The Velocities o j * Combination of Sodium Derivatives o f Phenols with OEeJine Oxides. Part 11. By DAVID RUNCIMAN BOYD and MISS DORIS FELTHAM THOMAS. A PREVIOUS communication (Boyd and Marle T. 1914 105, 2117) contained an account of experiments on the combination of ethylene and propylene oxides with the sodium derivatives of a variety of phenols. The nature of the reaction investigated is indicated by the equation (?=">(I + C,H,*ONa + zC,H,*OH = VH,*O *C,H CH,*OH CH* + C,H,*ONa + 2 - lC,H,*OH 1240 BOYD AND TIZIOMAS THE VELOUITIES OF COMBINATION OF From the results of these experiments the conolusion was drawn that combination probably in the first instance takes place between the phenoxy-ion and the olefine oxide molecule thus.the initial additive product afterwards reacting with the excess of phenol to give a glycol aryl ether and a new phenoxy-ion, CH,*OR $!H,*OR CH,-O’ CH,-OH + R*OH = I + ROO. A comparison of the values f o r the velocity constant of the reaction with different phenols indicated that the speed of com-bination diminished with increase in the acidity of the phenol, and the suggestion was made that a certain analogy exists between the two reactions C6H,*0’ + H’ =C,H,*OR and C,H,-O’ + yH2>0 = C,H,-OX?H,-CH,*O’. CH, I n other words where the tendency for a phenoxy-ion t o change into undissociated phenol is great the speed of combination of the phenoxy-ion with the ethylene oxide molecule should be relatively high and vice versa.The data available a t the time, however with regard to the relative acidities of different phenols were comparatively few. An investigation of the extent to which the sodium derivatives of alkyl substituted phenols are hydrolysed in aqueous solution was afterwards carried out by one of us (T. 1915 107 1538), and it then became possible to consider in more detail the rela-tionship existing between the acidity of a phenol and the speed with which its sodium derivative combines with ethylene oxide. As a result certain regularities with regard to the behaviour of ortho-substituted phenols came to light and it appeared desir-able to extend the observations so. as to confirm if possible the generalisations which had suggeshd themselves. The prment paper therefore includes experimental data for the velocity con-stants of three additional alkyl substituted phenols namely, o4-xyleno1 m-6-xyleno1 and mesitol.In all results for eighteen phenols are dealt with. The facts are summarised in the accom-panying diagram where the velocity constants for the ethylene oxide reaction are plotted against the values for the percentage of hydro-lysis of the sodium phenoxides in aqueous solution. A reference to this diagram will show that the velocity constants for phenol an SODIUM DERIVATIVES OF PHENOLS WITH OLEFINE OXIDES. 1241 its meta- and para-substitution products lie approximately along a straight line whilst the constants for the alkyl substituted phenols containiag one alkyl group in an ortho-position lie approximately on a second straight line which runs nearly parallel with the first and some distance below it.It appears therefore that whilst in general the speed of the ethylene oxide reaction increases approximately in proportion to the degree of hydrolysis of the sodium phenoxide a retarding influence makes itself felt in those cases where an ortho-placed alkyl group is present in the phenol molecule. This retarding influence it will be observed is superimposed on the factor which in the main determines the speed of the reaction. It becomes apparent only as the result of such an analysis of the phenomena as is here suggested. A comparison of the actual magnitude of the velocity constant for phenol with that of the constant for any of the alkyl ortho-substituted phenols reveals no such retardation.Thus thymol in the mcolecule of which an iwpropyl radicle is present in the ortho-pition to the hydroxyl group has a velocity constant one and a half times as great as that of phenol. Mesitol, with two ortho-placed methyl groups has a constant nearly twice as great as that of phenol. None the less the retarding influence of the ortho-placed alkyl qoups is operating in both instances. The case of mesitol is of particular interest in this connexion. Since in mesib1 two ortho-placed methyl Lgroups occur the retardation might be expected to be more pronounced than in cases where only one such group is present. A consideration of the diayram shows that this anticipation is fullv borne out by the exnerimental results. On the other hand the position of the cwhlorophenol constant, and more especially of that for 2 4 6-trichlorophenol indicates that negative substituents-or a t least chlorine atoms-in the ortho-position act in an exactly opposite way the velocity of com-bination being accelerated by their presence; and it is noteworthv that just as two ortheplaced alkyl groups cause a very pronounced retardation of the speed so two urtho-placed chlorine atoms have an accderating effect much more powerful than that due to a sincle chlorine atom.Whilst it is natural to attribiite t o steric hindrance the retard-ing influence of the ortho-placed alkyl groups. a final decision on the question is not yet possible in view of the quite different effect produced by ortho-situated chlorine atoms.Further experimental evidence is also required to decide whether the accelerating influence of the chlorine atom should be ascribed t o its residual affinity or to its polar quality. I n the meantime it may be pointed out that in the case of guaiacol which contains an ortho 1242 VELOUITIEIS OF COMBINATION OB SODIUM DERIVATIVES ETC. situated group of ill-defined. polar charaoter the velooity anstant liea some distance above ths line for phenols containing ortho-groups of well-marked positive type. EXPFRIMENTAL. The velocity constants for o4-xyleno1 m-6-xylenol and mesitol were determined a t 70'4O in 98 per cent. alcohol according to the method previously described (T. 1914 105 2117). It was found necessary to employ a mixture of light petroleum with ether (1 1) in separating glyco! m-6-xylenyl ether from unchanged m-6-xylen01, since this phenol is not completely extracted from an ethereal Percentage hydroJy& OJ sodium phew* in aquwzcs SOl&On at 25".solution by repeated shaking with aqueous potassium hydroxide. The same method had to be employed in the case of the mesityl ether. Summary of Results. Pqrcentage yield of glycol ether. 1 log 100 A /- -3 t 100-2. o -4-Xylenol. on-6-Xylenol. Mesitol. 2 hours. 76-3 76.1 2 hours. 74.5 74-8 2 hours. 82.3 81.8 82.6 1 hour. Maximum. 51.0 98.5 0.3060 - - -1% hours. Maximum. Maximum. 58.5 98.9 0.3005 - - -- 99.0 0-3746 - - w- AND JAMES MOLECULAR REFRACTIVITY ETC. 1243 GEycoZ o-zyEyZ ether C8H9=O-CH,*CH2*OH is a colourless oil boiling a t 1 5 9 y 18 mm. Its p-mitrobenzoate crystallises from alcohol in pale yellow plates melting at 225O: 0-1518 gave 6.5 C.C. N (moist) a t 23O and 762 mm. CI7Hl7O,N requires N = 4.62 per cent Glycol m-sylyl ether C,H,*O*CH,*CH,*OH was obtained as a whib solid which after recrystallisation from light petroleum, melted a t 57": 0.1621 gave 0.4289 CO and 0.1252 H,O. C=72*16; H=8*58. Glycd mesityl ether CgH,,*O*CH,*CH,*OH separates from 0-1556 gave 0.4196 CO and 0.1263 H,O. C=73-54; H=9.02. N=4*82. C,,H,,O requires C = 72.23 ; H =8*51 per cent. light petroleum in glistening white crystals which melt a t 60° : C,,H,,O requires C = 73.25 ; H = 8-97 per cent. THE UNIVERSITY COLLEGE OF SOUTEAlUPTON. [Received July 23rd 1919.

ISSN:0368-1645    

DOI:10.1039/CT9191501239

出版商:RSC    

年代:1919

数据来源: RSC

摘要:

w- AND JAMES MOLECULAR REFRACTIVITY ETC. 1243 CXV.-Molecdar Refractivity of Cin.ncxmic Acid Derivatives. By GRIC WALKER and THOMAS CAMPBELL JAMES. THE work of previous investigators has shown that the molecular refractivity of organic compounds is affected to a slight extent by constitutive influencee within the molecule. In this investigation we have determined the amount of variation produced by change of constitution in a series of closely allied derivatives of cinnamic acid and the results are tabulated below. a-Chlorooinnamic acid ......... a-Chlorodlocinnamic acid ...... Ethyl a-chlorocinnamste ...... Ethyl a-chloroaUocimamate.. . a-Bromocimamic acid ......... a-BromoaUocinnamic acid ...... Ethyl a-bromocinnamate ...... Ethyl a-bromoa2Zucinnamat.. .B-Bromocinnamio mid ......... B-BromoaEtocinnamio acid ... Dibromocinnsmic acid ......... DibromarUocinnamic acid ... Ally1 a-chlorocinnamate.. ....... AUyl a-chlbrarltocinnamate . . , [MI,. Difference. 50.24 -49.00 1-24 58.33 -57-61 0.72 62.90 -52.34 0.56 60.85 -60-64 0.3 1 51.98 -51.63 0.45 58.94 -62-13 -60.52 2.61 7 -[MI,. Difference. 50.96 -49-50 1-37 58.98 -58.19 0.79 53.62 -53.06 0.46 61.51 -01-14 0.37 52-81 -52-10 0.51 59-84 -82-82 -61-08 1-74 - 1244 WALKER AND JAMES MOLECULAR REFRACTIVITY OF An examination of the above figures confirms Bruhl's observa-tion that the more stable isomeride has the higher molecular refractivity and shows also that in the *substituted compounds, the difference decreases with increase of molecular weight.It is also seen from a consideration of the bromo-acids that the a-sub-stituted acid has a greater refractive power than the 13-substituted acid. EXPE,RIMEN T A L. (a) Refractivity Measurements. Materials.-The acids and esters used in this investigation were all specially prepared and the greatest care was taken to use only highly purified specimens for refractivity measurements. Details of new preparations are given at the end of the paper. The solvent used for the acids was ethyl alcohol rendered abso-lute by distilling the 99-8 per cent. commercial product over lime and afterwards over metallic calcium. Appratus and Method.-The solutions of the acids were made i ~ p by direct weighing to determine their percentage composition.In all cases approximately 5 per cent. solutions were employed. The esters were all liquids and were determined directly. Densities were determined by means of a Sprengel-Ostwald pyknometer of about 5 C.C. capacity. The pyknometer was suspended in a thermostat kept a t 2 5 O (*0*05O) for about fifteen minutes. The following formula was used: W D 0~0012(w' - W ) -7 d*'" = ~ - ~ ' O w W where d y =density of solution a t 25O referred to water a t 4*, JV" = weight of solution filling pyknometer, W =weight of water filling pyknometer, D=density of water a t 2 5 O (0.997073). Refractive indices were measured by means of a form of Pulfrich refractometer furnished with an ordinary thermostat The temperature of the liquid in the cell could be maintained a t 2 5 O (k0-05O) for a considerable time.The refractive index of each liquid was measured for the C and D lines. The zero error of the refractometer amounted to + I / and all readings were corrected for this. The temperature correction amounted to about one unit in the fifth place of decimals in the refractive index and could therefore be neglected. The refractive index of the alcohol used as solvent was deter CINNAMIC ACID DERIVATIVES. 1245 mined several times during the course of the work. values were obtained : The following nC= 1'35787 1.35784 1.35774, nD = 1.35967 1.35967 1.35954. -The specific refractivity r of the solute was calculated by the formula where n and d are refractive index and density o,f the solution a t 2 5 O mu and d are the corresponding quantities for alcohol a t 25", and p is the percentage composition by weight of solution.For the pure esters the formula n2-1 1 r = - . -na+2 d was employed. I n the tables which follow p and d have the same significance as above nc and nD denote the indices of refraction for the C and 1) lines respectively rc and rD refer to the specific refractions and [MI and [MID to the molecular refractions for these lines. I n the column headed '' solution," the number refers to the specimen of substance used whilst the letters denote the various solutions made up from the same specimen. TABLE I. a-Brornacinnamic A cid. Constants for alcohol d=0-7851; mc= 1.35787; nD= 1.35967. Solution. p . d. n . n,. rc . ~b- [MJc. [MI,. la .........5.041 0-8069 1.36562 1.36742 0.2332 0.2361 52-94 83-37 2a ......... 5.192 0-8076 1.36580 1.36780 0-2332 0.2364 52-04 53.66 2b ......... 4.677 0-8053 1.36493 1-36691 0.2327 0.2358 52-83 63-63 Means ... 52.90 63-52 TABLE 11. a-Bromoallocinnamic A cid. Constants f0.r alcohol as in table I. la ......... 5.330 0.8077 1-36556 1.36757 0.2310 0.2342 52.43 53-16 lb ......... 4.824 0.8066 1.36483 1.36681 0.2304 0.2338 52.30 63-07 IC ......... 6.248 0.8075 1.36647 1.36747 0.2304 0-2333 52-30 62.96 Means ... 62-34 53.0 1246 WALKER AND JAMES MOLECULAR REFaACTIVI!I'Y OF TABLE 111. a-Chlorocinnamic A cid. Constants for alcohol as in table I. Solution. p. d. mc. n D - r . 7,. [Ed],.. MD. la ......... 5.190 0-8034 1.36686 1.36893 0.2754 0.2795 50.26 51.01 l b .........4.984 0.8026 1.36645 1.36851 0.2751 0.2793 50.20 50.97 IC ......... 5.136 0,8031 1.36673 1.36879 0.2755 0.2796 50-28 51.03 2a ......... 5-469 0.8042 1.36724 1-36930 0.2751 0-2785 50-21 50.83 Means ... 50-24 50.96 TABLE IV. a-Chloroallocinnamic A cid. Constants for alcohol d=0*7851; no= 1.35774 ; n = 1.35954. la ......... 5.008 0.8021 1.36560 1.36758 0.2685 0.2716 49.00 49.57 l b ......... 5.051 0,8024 1.36570 1-36767 0.2681 0.2711 48.93 49.48 2a ........ ; 5-206 0.8024 1.36576 1.36776 0.2689 0.2725 49.07 49-73 Means ... 49.00 49.69 TABLE V. fLBrmocinnamic A cid. Constants for alcohol as in table IV. la ......... 4.877 0.8062 1.36489 1.36681 0-2296 0.2317 52.12 52.60 2a ......... 4-512 0.8045 1.36426 1.36623 0.2284 0.2318 51-86 62-02 Means ...51.98 62-61 TABLE VI. B-Bromoallocinnamic A cicl. Constants for alcohol as in table IT. la ......... 6.044 0.8061 1.36464 1.36649 0.2270 0.2295 61-63 62.10 TABLE VII. Dibromocinnamic A cid. Constants for alcohol as in table IT. la ......... 6.206 0.8107 1.36412 1.36603 0.1926 0-1949 58-94 59.64 TABLE VIII. Ethyl ocChlorocinnamate . d . nc . nD* rc. rD- [MI,. [MID. 1.1719 1.86292 1*67064 0.2771 0.2802 58.33 68.98 TABLE IX. Ethyl a-Ckloroallocinnapnate. 1.1869 1.54697 1.66246 0-2737 0.2764 57.61 58.1 QINNAMIC ACID DEBIVATIVf68. 1247 TABLE X. Ethyl a-Bromocimnmate. a. nc . %? rc . TD. [MJo. WID-1.3886; 1.57685 1.58460 0.2386 0.2412 00.86 01.61 TABLE XI. B t F y 1 a-Br maoallocinmmat e . 1.3713 1.56474 1.56973 0.2374 0.2398 60.64 61.14 TABLE XII.A llyl a-Chlorocinnamate. 1.1702 1.56718 1.57483 0.2792 0.2823 62.13 62.82 TABLE XIII. Allyl a-Chloroallocinlzamate . 1.1457 1.53563 1.54164 0.2720 0.2746 60.52 61.08 Preparations. B thy1 a- chloro cinmma t e C,H,*CH CC1 CO,* C,H,. B. p. 161-162O/8 mm. 0.3467 required 1.621 C.C. N-AgNO,. Ethyl a-chlwoallocinnamate. B. p. 157-158O/lO mm. 0.2985 required 1.392 C.C. N-AgNO,. AlZyZ a-chlorocinnamate C,R,*CH:CCl*CO,*CH,-CR:CH,. B. p. 162-163O/11 mm. 0.3340 required 1.470 C.C. N-AgN03. Allyl a-chloroallocimamate. B. p. l7lo/28 mm. D2'1*146. 0.2841 required 1.258 C.C. N-AgNO,. DY 1.172. CllHl,O,C1 requires C1= 16.84 per cent. DT 1,157. Cl1H,,O2C1 requires C1= 16.84 per cent. Di5 1.170. Cl2H1,O2Cl requires C1= 15.93 per cent. C1= 16.60. Cl= 16.54. C1= 15.61. C1= 15-70. Cl,HUO,Cl requires C1= 15-93 per cent. From the densities of form. TKE EDWARD above it will be observed t h a t the boiling points aiid the alZo-esters are always below those of the trans-DAVIE~ CJSEMICAL LABOBATOIUEB, ABERYSTWYTH. [Receiutld August 16th 1919.

ISSN:0368-1645    

DOI:10.1039/CT9191501243

出版商:RSC    

年代:1919

数据来源: RSC

摘要:

1248 WHITE AND PRICE THE DETERMINATION OB CXVI .-The Detewnination of Ignition-temperatures by the Soap-bubble Method. By ALBERT GREVILLE WHITE and TUDOR WILLIAMS PRICE. THE soap-bubble method for the determination of ignition-tem-peratures was first described by McDavid (T. 1917 111 1003). I n the method as finally adopted the experimental error is assumed to be less than *3O and the results obtained for the ignition-temperatures of the various gases tested are given as : Coal gas-air .............................. 878' Ethylene-air .............................. 1000 Hydrogen-air .............................. 747 Carbon monoxide-air .................. 931 Petrol (fraction 0-SO0)-air ............ 995 Benzeneair .............................. 1062 Ether-air .................................1033 The method is said t o eliminate so far as is practically possible, the time factor. The temperature of ignition is taken to be that temperature to which the gaseous mixture must be heated by the application of a hot body so as to cause instantaneous ignition. McDavid's results are very high in comparison with other recent determinations. Finding it necessary to determine the ignition-temperatures of certain ether-air mixtures the authors decided to use the appara-tus described by McDavid on p. 1005 (Zoc. cit.). The igniting coil was mounted between two binding screws fixed into a small wooden stand and was kept a t the requisite temperature by means of current drawn from a battery of accu-mulators. A variable resistance of nichrome ribbon formed part of the circuit which was completed by means of a sliding contact-maker.The current was measured by means of a delicate ammeter reading to 0.01 ampere. The soap solution was prepared by dissolving sodium oleate in water with the addition of a little glycerol. I n all the experiments in which ether was used the ether-air mixture was made by weighing the requisite. quantity of ether into an exhausted aspirator from which it was displaced later by means of mercury. It was found that the presence of mercury caused a slight change in the concentration of the solvent above it. The glass pipe from which the bubbles were blown was connected to the aspirator by means of a ground-glass joint so that there was no possibility of change in the concentration of the ether due t o absorption of the vapour by the leading tube which was of glass.I n the other experiments the vapour-air mixture wa IGNITION-TEMPERATURES BY THE SOAP-BUBBLE METHOD. 1249 made by weighing the requisite amount of liquid into an exhausted 15-litre aspirator and was in this case displaced by means of water. The hydrogen-air mixture was made up by volume in the same apparatus. Standardisatio n of A ppara t us. The coils were standardised as described *by McDavid ; the salts used were carefully purified and were kept in an exhausted desic-cator over phosphoric oxide. Before use the salt was crushed and only minute fragments were placed on the coil. When these precautions were taken the standardisation could almost invari-ably be repeated to 0-01 ampere.The number of amperes neces-sary to keep the coil a t such a temperature t’hat when in a steady state the salt was just melted was taken to be the amperage necessary to heat the coil to the melting point of that salt. The melting points are given in table I and are the same as given by McDavid except in the case of lead chloride which he did not use. TABLE I. Melting point given in Molting point “ Chemist’s Year Salt employed. used. Book ” for 1916. Potassium sulphate.. ....... 1072’ 1050” Sodium chloride ............ 800 820 Potassium iodide. ........... 687 705 Lead chloride ............... 485 485 For comparison the figures supplied in the “Chemist’s Year-Book” for 1916 are given. It will be seen that the difference between these and tjhe figures used is by no means negligible; the figure for lead chloride was taken from this source.Every effort was made to wind the coils as uniformly as possible but in almost every case it was found that patchesrather hotter than the average could be found. I n Fig. 1 are given the standardisations of some of the coils. It will be noticed that these curves are not straight lines as was that determined by McDavid i n the case of the coil used by him, but have a distinct curvature. Stamiardisation of the Method. Little work appears to have been done as regards standardising the soap-bubble method. McDavicl used a soap-bubble 3-7 cin. in diameter but 110 figure3 are given showing the effect of the siz 1250 WHITE AND PBICE THE DETERMINATION OF of the bubble or the place or the method of striking.All the standardisation undertaken in connexion with this method was done with coil No. 1 and most of it was carried out with ether-air mixtures although in some cases hydrogen-air mixtures were also used similar results being obtained. Throughout the work an attempt was made to keep the room temperature within 2 O of 18O. For some time it was found exceedingly difficult to obtain accurate and steady ammeter readings but the erection of a draught screen round the apparatus and the gradual elimination of poor connexions remedied these faults. It was also found that owing to the comparatively high currents used the nichrome resistance FIQ. 1. Showing the standardisatkn curves connecting the ammeter reading and temperature various coiEs.Temperature. required frequent cleaning And that the sliding contact-maker had to be silvered ur fitted with a platinum contact. So far as could be judged from the steadiness of the ammeter reading no appreciable change in the temperature of any coil took place when it was brought into contact with a soap bubble blown with air. When using platinum-wound coils a t the tem-perature necessary for the experiments a very slight draught caused a distinct rise in the ammeter reading. When determining an ignition-temperature the amperage was taken a t which a t least two ignitions were obtained out of three attempts whilst 0.01 ampere below a t least three siiccessive failures to ignite were registered IBNITION-TEMPERATURES BY THE SOAP-BUBBLE METHOD.1251 Size of Bi&hle and Place of Striking. It was found that the size of the bubble used and the place of striking affected the resulh very appreciably and various experi-ments were carried out in order to determine the magnitude of the variations due to these factors. The results of some of these tests are given in table 11. TDLE 11. Mixture containing 7-5 per cent. of ether in air. Number of amperes necessary to ignite. Position of striking 2.5 cm. 3.7 cm. 5 cm. surface on bubble. bubble. bubble. bubble. Struck at top ......... 5.29 5-05 5.02 Struck at middle ... 5.29 5.05 5.02 Struck at bottom ... 5.41 6.08 5.03 Mixture containing 12-6 per cent. of ether in air. Number of amperes necessary to ignite. 2.5 cm.3.7 cm. 6 cm. bubble. bubble. bubble. 5.31 6-20 6-18 5-32 5.20 6.18 6-43 6.22 6.20 .A / \ On plotting the above results graphically as in Fig. 2 it will be seen that the correct result appears to be about 0.01 ampere below the current necessary for the ignition of a 5 cm. bubble. Similar results were obtained using other concentrations of ether vapour in air. It scarcely appears to matter whether the bubble is struck at the top or middle but striking a t the bottom invari-ably gives a lower result. In practice bubbles were always struck as near the middle as possible and a 5 cm. bubble was always used. This size of bubble gave some little trouble owing to a tendency to give delayed ignitions and to burst before use both effects being particularly noticeable with mixtures containing much ether.There was also a great tendency for the ether to leak through these bubbles. The method of striking was also found t o have a great influence on the result but it was found that a slow, but not too slow approach of the coil to the bubble or even better, of the bubble to the coil gave very consistent results. When the approach of the bubble was too slow the ether pouring out through the soap film heated the coil so that the ammeter reading did not register its correct temperature. Time betfween Ignitions. It was also found that a fair time had to elapse after one igni-tion before another could be attempted with any certainty of obtaining a trustworthy result. Even if a bubble was not ignited, heating of the coil took place and a second bubble was very ofte 1252 WHITE AND PRICE THE DETERMINATION OF ignited a t what was apparently the game temperature wileas timo were given for the coil to cool.At least one minute had to be allowed between each attempted ignition and often a longer time was necessary the interval varying from coil to coil. It was also found that the time necessary for an adjustment of temperature FIG. 2. Showing how the current necessary to ignite a definite mixture vaTies with the. size of bubble and position of strikiw. Size of bubble in cm. due to change in the magnitude of the current flowing varied greatly from coil to coil. Changes in the Coil. Even when experiments were carried out precisely as described in the above method it was found that a coil gave different results for the same combustible mixture even on the same day.Careful observation showed that these variations were generally all in one direction and a connexion was soon traced between them and the standardisation of the coil used. In table I11 are given the ignition-temperatures of a 4.9 pe IQNITION-TEMPERATURES BY THE SOAP-BUBBLE METHOD. 1253 cent. ether-air mixture determined with coil No. 1 a t various times after standardisation. TABLE 111. Within After After 20 Minutes after 20 minutes. 3 days' use. 3 ~ e e k s ' use. re-standmdising. 968" 985' 1040' 970" It will be seen a t once that the addition of salts to coil No. 1 lowered the ignition-temperature of the ether-air mixture as given by this method by 70°. Some such change might have.been anti-cipated from the later work of Meunier (Compt.rend. 1909 149, 924; 1910 150 781). That this was not due to a change in the resistance of the coil can be seen from the fact that throughout the six months during which this coil was in use no point on the standardisation curve varied throughout a range of more than 0-02 ampere except on one occasion when a variation of 0-03 was found. Precisely similar results were obtained in a test experi-ment on standardisation even without washing off the salts used, as was invariably done in practice. A further investigation was then undertaken in order t o discover how the ignition-temperature of a particular mixture varied with the treat'ment to which the coil was subjected. The ammeter readings taken when using the untreated coil were converted into temperatures after the coil had been standardised by the use of salts.This was considered legitimate as every pre-caution was taken to ensure that the current-temperature ratio of the coil did not change with use. It was found that the newly-wound coil gave ignition a t ammeter readings which increased continuously with time up to a steady value. The first ignition-temperature determined for a 27 per cent. hydrogen-air mixture, using coil No. 2 before tyeatment was 760". In a week's time the ignition-temperature had increased to 860° but after this time no appreciable change in ammeter reading could be detected, After standardisation or using salts it was found that a decided rise in the ignition-temperature had always taken place.This was invariably followed by an initially rapid fall which decreased until a steady state was reached the rate of decrease of tempera-ture varying considerably with the nature of the coil and its previous treatment. Thus an untreated wire would take a very much longer time to reach a steady state than a recently treated or even a well washed wire which had Once reached that state. For instance the ignition-temperature of an ether-air mixture as determined by means of a platinum coil treated for the first time would probably take several days to become constant whereas i 1264 WHITE AND PRICE TEIE DETERXINATION OF the case of a recently treated coil twenty minutes would be ample time. Two hours after coil No. 2 was first treated with salts the ignition-temperature of a 5 per cent.ether-air mixture as regis-tered by this coil had fallen from 1068O to 1019* whilst that of a 12 per cent. mixture after its preliminary rise from 986O had fallen to 1006O (see table IV). I n both these cases the ignition-temperature of a 5 per cent. ether-air mixture appears to have fallen much more rapidly than that of a 12 per cent. mixture as would be expected from the curves shown in Fig. 5. In order to maintain it coil in the steady state it was necessary to treat with salts occasionally. This caused a temporary rise in ignition-temperature but it was only a matter of minutes before the steady state was re-established. The manner in which the ignition-temperature changed when no salts were added to a coil in the steady state can be seen from table 111.The temperature a t which a steady state was attained after treatment was greater than without treatment except in the case of coils 1 and 2 which were of platinum. The variation was com-plicated in the case of the nichrome coil No. 3 by the fact that after being in the steady state for three days with a fairly con-tinuous treatment with salts a rather rapid fall in ignition-tem-perature became evident. When the fall became slower the two points shown by asterisks in Fig. 5 were obtained the one value, for a 5 per cent. ether-air mixture being the mean of two read-ings 0.04 ampere apart obtained before and after the determina-tion for a 12 per cent. ether-air mixture. It will be notioed that in this case the ignition-temperature of a 5 per cent.ether-air mixture has fallen much more rapidly than that of a 12.5 per cent. mixture. This experiment was unfor-tunately interrupted by the breaking of the silica tube on which the mil was wound making it impossible for the eiperiment to be carried to its logical conclusion; but it is a t least possible that a flattened curve well below the original one (without salts) would have been obtained here as in the case of No. 1 . The upward change shown in table I11 could be made to take place 0ven faster if the salts were washed from the coil and the coil used without treatment after drying. It was found that after eighteen hours’ soaking of the coil in water the ignition-tempera-ture of a 27 per cent. hydrogen-air mixture as indicated by coil No.1 fell from 7 8 7 O to 771O and that of a benzene-air mixture fell from 1060O to 1056O. After sixty hours’ soaking the result ob-tained for the hydrogen-air mixture was 773O and that for the benzene-sir mixture 1056O. In all the above cases the ignition IGNITION-TEMPERATURES BY THE SOAP-BUBBLE METHOD. 1255 temperature was determined immediately after the drying of the coil was complete. During the course of tbree days after this preliminary fall ignition-temperatures determined by this coil continued to rise and a t the end of this time when the rise had become very slow the figure found for the hydrogen-air mixture was 793O and that for the benzene-air mixture 1075O. Further results of a similar kind will be found later.The effects of the changes considered above are shown in Fig. 3. It will be seen that consistent results ar0 only to be expected under conditions corresponding with the three portions of the curve marked A B and C. Except where otherwise stated results detailed later correspond with these steady states. FIG. 3. Showing diagrammatically the eflect of various changes during the treatment of the wire on the ignition-temperatures registered by a coil. Time. Influence of the Material of the Coil. McDavid (loc. cit.) states that the ignition-temperature of a 20 per cent. hydrogen-air mixture as determined by a Eureka wire coil is 9 O to 30° lower than that obtained with platinum wire, and the statement is made that “The figures found by using platinum are higher than those obtained by using Eureka wire, indicating a catalysing effect in the case of the latter.It is how-ever probable that both substances exert a catalytic influence.” The results given above for coils Nos. 1 and 2 make it certain that the temperatures obtained are influenced very greatly by surface action in every case. It was accordingly decided to deter-mine the ignition-temperatures of various ether-air mixtures using coils of different materials. VOX. cxv. 3 1256 WHITE AND PEICE THE DETERMINATION OF Coil N o . 1 was made by winding platinum wire 0.038 cm. in Coil N o . 2 was a similar coil made with platinum wire 0.025 cm. Coils Nos. 3 and 4 were made by winding nichrome wire 0.05 cm. Coil No. 5 consisted of platinum wire 0.025 cm. in diameter, diameter round a notched strip of mica.in diameter. in diameter round a narrow silica tube. FIG. 4. Showing the ignition-temperatures of various ether-air mixtures obtained with coils of platinum wire wound on mica. 0 5 10 15 Percentage of ether. 20 wound inside a silica tube and having the platinum leads shielded by tubes of silica. Coil N o . 6 consisted of platiiium wire of the same diameter as used in coil No. 5 enclosed in a uniform hard-glass tube with the leads safeguarded as for coil No. 5. Fig. 4 gives the ignition-temperature curves obtained by plotting the results for various ether-air mixtures determined by means o IUNITION-TEMPERATURES BY THE SOAP-BUBBLE METHOD. 1257 coils Nos. 1 and 2. It will be noticed that the curve obtained when coil No.2 had attained a steady state before being treated with salts (or standardised) is close to that obtained when coil No. 1 had been allowed to attain a steady state after all the salts had been removed by prolonged washing and use. Curve C shows the ignition-temperatures obtained when coil No. 1 is treated with salts until the ignition-temperature obtained is constant. It differs from the other two curves inasmuch as the main body is very much flatter and that the difference in ignition-temperature between a 5 per cent. mixture and a 12 per cent. mixture of ether and air is very much less. The ignition-temperature of a 5 per cent. ether mixture as obtained from different curves on this diagram varies by looo whilst the maximum difference in the case of a 12 per cent.mixture is little more than 30°. With both these coils a more or less steady state was attained a t a point above the ignition-temperature before treatment and very much above in the case of the silica tube No. 5. I n this case as before, the treated coils give flatter curves and the maximum difference between the ignition-temperatures of 12 per cent. mixtures appears to be about 1’75O whilst the maximum difference obtained for a 5 per cent. mixture is about 1 6 0 O . An examination of the results given in these two sets of curves shows that the ignition-temperature of any particular mixture as obtained by this method can vary very greatly. F o r example the ignition-temperature of a 12 per cent. mixture of ether and air can be considered to be anything from 859O to 1035O.The explanation of course lies in the fact that a catalytic action-surface action-must be exerted by the coil as would naturally be expected of any hot body inserted into a combustible gas; but it is fairly evident that the greater catalytic action is not exerted by the coil giving the lower ignition-temperature as suggested by McDavid but by that giving the higher. The modzis operan& is probably as follows When the coil is brought into the gas the combustible portion near it is at once removed by surface com-bustion and this weakens the bmcentration of combustible gas near the coil to such an extent that it is only when the radiation from the coil is of sufficient intensity to ignite the gas outside this limited sphere that ignition occurs.This reasoning which appears t o be quite in keeping with what is known of the combustion of gases on surfaces explains why in McDavid’s experiments platinum gave a higher result than Eureka wire and in the experiments described above why in the normal state it gave results higher than either silica or nichrome when unaffected by salts. This Fig. 5 gives the results obtained using coils 3 and 5. 3 ~ 1288 WHITB AND PRIOE THE DETERMINATION OF would also explain the reason why a coil had to be used for some time before atitaining its maximum activity as described in the standardisation of the apparatus. FIG. 5. Showing the ignition-tewperaturea of variom ether-air dxturee obtained wiul coik of nichmme wound on silica and platinum inside silica.0 5 10 15 20 Percentage of ether. J7due of Method f o r obtaining Comparative results. Although under any particular set of conditions the soap-bubble method is capable of giving results consistent to within +5O it will be se0n that giving as it does such a range for the ignition temperature of any one gas-air mixture it could scarcely be capable of giving the true ignition-temperature particularly as surface action is bound to occur even when using the most indif-ferent material possible. It thus becomes a question as to whether even comparative results could be obtained by this method. It is seen from the curves given in Figs. 4 and 5 that t h w obtained with different coils occasionally cut one another indicating that even comparative results could scarcely be obtained assuming the experimental methods could be guaranteed accurate.As how-ever the melting points of the salts used for standardisation pur-poses are not known t o any great degree of accuracy and the standardisation curves are not straight lines this evidence can scarcely be regarded as final. It was therefore decided to test several combustible mixtures using the coils already made in order to see if comparative results could be obtained by this method. Accordingly the following mixtures were used in addi-tion to those mentioned previously : 1. A mixture of 27 per cent. of hydrogen in air. 2. A mixture of 5.7 per cent. of benzene in air 3. A mixture of 0.23 gram of light petroleum (b. p. 90-100") per litre of air.4. A mixture of 0.23 gram of light petroleum (b. p. 60-80°) per litre of air. It was found to be impossible to obtain a definite ignition-temperature for a mixture of carbon disulphide and air or to ignite any mixture of alcohol and air or to obtain a reasonable size of bubble with mixtures of acetoae and air. The results obtained are given in table IV. I n the course of these experiments it was noticed that it was easier to obtain an accurate result with hydrogen than with any of the other gases and it was also very obvious that coils Nos. 5 and 6 gave results sooner and more accurately than the others. With any gases examined when using these coils it was easy to redetermine any figure to within 0.01 ampere whilst with the other coils twice this variation was often found.Coils Nos. 1 and 2 scarcely changed a t all during six months' use but the result was far different in the case of coils wound over silica tubing. For example three nichrome coils had to be rejected before a coil was found which would not change on heating and this had to be kept a t an exceedingly high temperature for some time before it reached this constant state. To test the comparative accuracy of the method a second coil of platinum wound inside a hard glass tube was made and the ignition-temperature of a hydrogen-air mixture before treatment with salts determined TABLE IV. Ignition I Light Light petroleum petroleum (b.p. 9'0-100') (b.p. 60-80') Hydrogen-air -air 0.23 gram -air 0-23 gram Coil used. 27 per cent. per litre. per litre.No. 1. With salts ..................... 787 1035 1056 No. 2. Wit>hout salts ............... 860 1038 1040 With salts before constant ...... 820 1040 1055 I No. 3. Without salts ............... .............................. I - With salts -No. 4. Without salts 764 With salts .............................. 773 1025 1016 No. 5. Without salts ............... - - -With salts .............................. - - -No. 6. Without salts 715 With salts 752 After washing and use ............ 793 1067 1075 - -1003 999 ............... - - ............... - - .............................. - - After washing ........................ 73 IGNITION-TEMPERATURES BY THE SOAP-BUBBLE METHOD. 1261 The result was quite satisfactory as on standardisation it was found that the ignition-temperature obtained in this case was 7 1 2 O the result obtained using coil No.6 being 7 1 5 O . No simi-lar tests were carried out with the other coils as it was consi-dered that in all cases except that in which a platinum coil was wound inside silica or glass such a test would probably have been useless as the ratio of the areas of the two constituent surfaces of the coil was bound ,to vary for every coil made. The results are interesting as they show that the ignition-temperatures of two gaseous mixtures obtained by this method may not even be comparative when determined by two different coils. This was carefully tested in the case of the mixtures of the two variet'ies of light petroleum and air by using coil 4 without salts and coil 1 with salts.When coil No. 4 without salts was kept at 1 0 0 1 O and bubbles of the above mixtures were brought into contact with it alternately in every case the mixture containing the light petroleum of lower boiling point ignited whilst in only one trial out of five did the other one ignite. A similar experiment with coil No 2 showed that the ignition-temperature of a mixture containing the light petroleum of lower boiling point was higher than that of the other as measured by this coil. A similar result was obtained when dealing with a 5 per cent. mixture of ether in air as in every case except when determined by coil No. 2 before treatment with salts the ignition-temperature of this mixture was found to be lower than that of the light petroleum and benzene mixtures.I n the case of the coil specified, however the ignition-temperature found for a 5 per cent. mixture of ether in air was definitely higher than those of the other mixtures mentioned above. I n table V are given the maximum and minimum values obtained for the ignition-temperatures of certain of the mixtures used. TABLE V. W =Washed. W.S. =Without salts. S =After using salts. Mixture used. Hydrogen (27 per cent.) in air Light petroleum (b.p. 90-100') 0.23 gram per litce in sir Maximum ignition-temperature observed. Coil 2 W.8. 1067 Coil 1 W. 860' Minimum ignition -temperature observed. Difference. 715" 145' 1003 64 Coil 6 W.S. Coil 4 W.8 1262 WHITE AND PRICE THE DETERMINATION OF TABLE V. (contiwed). Masimwn ignition-temperature Mixture used.observed. Coil 1 W. 1075 Coil 1 W. Light petroleum (b.p. 60-80') 1075" 0-23 gram per litre in air Benzene (5-7 per cent.) in air Minimum ignition-temperature observed. Difference. Coil 4 W.S. 999" 7 6" 1025 50 Cbil 4 W.S. Ether (5 per cent.) in air 1068 907 - 161 Coil 2 W.S. Coil 5 S. Coil 5 W.S. Coil 3 W.S. Ether (12 per cent.) in air 1035 869 176 When it is considered that the glass and silica coils were only us+ for the hydrogen and ether mixtures respectively it will be seen that the results have an appreciable regularity as regards what is generally considered to be the catalytic order of the various substances concerned in the ignition. I n this connexion it must not be forgotten that the substance on which a metal coil is wound must affect the ignition-temperature obtained.A comparison of the ignitio,n-temperatures obtained using coil No. 2 with and without salts is also interesting. I n the case of the hydrogen-air mixture and the 5 per cent. ether-air mixture, the treated coil gives appreciably lower results than the same coil without salts but the reverse of this is true of all the other mix-tures except light petroleum (b. p. 90-100°) and air when the ignition-temperatures are practically tbe same in the two cases. On summing up it will be seen that the only advantages which the method possesses appear to be those due to its convenience and rapidity in use. It was easy to find thel apparent ignition-temperature of any mixture within less than an hour and using coils 5 or 6 it could often be clone in half that time.The term '' instantaneous ignition " is obviously used by McDavid to mean ignition without perceptible delay but refined methods of time measurement would certainly show that diff went retardations occurred with diff erent gas mixtures. Its disadvantages appear to lie in the fact that rn far as can be ascertained from these experiments i t cannot give the true ignition-temperature of any gas mixture and that the results given by it can scarcely be utilised unreservedly even when only required f o r comparative purposes. Such results are fairly certain t o be higher than the true ignition-temperatures and not as described by McDavid lower. The accuracy of the method is also adversely affected by the fact that the exact melting points of the salts used are still a matter for controversy and that the standardisation curve for any coil appear IGNITION-TEM-PERATURES BY THE SOAP-BUBBLE METHOD.1263 from these experiments unlikely to be a straight line. The results are exceedingly dependent on draughts as the method of estim-ating temperature is an indirect one. If the temperature of the coil falls during an experiment due to any such disturbance in outside conditions the ammeter reading and hence the tempera-ture indicated becomes higher instead of lower owing to the change in resistance of the platinum or other metal forming the heating coil. Any result obtained also seems to depend on the state of the surface causing ignition as used and unused wire give different ignition-temperatures for the same gas mixture and an examination under a good lens of wire that has been in us0 for some time shows it to have altered appreciably in appearance.Amongst its other disadvantages is that of limited applicability. It can only be applied to moist gases and to those which are more or less insoluble in water and can be obtained a t a fair concen-tration in air a t the ordinary temperature. For instance it was found impossible to ignite a mixture of amyl acetate and air mads up a t 20° by this method. Another disadvantage lies in the fact that the least concentration of combustible gas a t which ignition can be obtained by the soap-bubble method is much greater than that found by ordinary methods which generally give results below 2 per cent.for ether-air mistures. Thus the ignition-temperature of any fixed concentration of mixture cannot be determined by this method as could also be judged from the obvious permeability of the soap film to some of the vapours used. The results obtained during this investigation appear to indicate that the soap-bubble method of determining ignition-temperatures as described by McDavid gives values for the ignition-temperature which appear to be erroneous. A radical modification of the method could probably be made however which would give more satisfactory results. For instance a soap-bubble of the mixture under examination might be introduced into a vertical tubular furnace the bubble being shielded from radiation until a t the desired spot.In’ this case the results would be far more likely to be correct as the temperature a t which ignition took place would be known and surface action would be partly eliminated. Summary. The soap-bubble method described by McDavid (Zoc. cit.) has been applied to determine the ignition-temperatures of certain mixtures of ether benzene light petroleum and hydrogen with air. The condit,icms under which consistent results can be obtained are given. 3 B 1264 CHAPMAN AND WHISTON THE INTEEACTION OF After careful standardisation the results given by this method were found to be affected by the physical state of the igniting surface and the nature of the material of which it was made; even the addition of small quantities of the salts used for standardisa-tion purposes altered the results obtained. This seemed to show that the method could scarcely give the true ignition-temperature of a gas mixture. The ignition-temperature of a particular gas mixture as determined by two different coils often varied by more than 150° and results appeared t o indicate that the method was not strictly trustworthy even for comparative purposes. I n conclusion the authors desire to express their thanks t o Messrs. Nobel’s Explosives Co. Ltd. for whom the work was carried out and t o Mr. Wm. Rintoul Manager of the Research Section f o r permission t o publish the results and to Mr. A. W. Sanderson for his kind assistance in carrying out some of the experimental work. THE RESEARCH LABORATORIES, ARDEER FACTORY, STEV ENSTON. [Received September lst 1919.

ISSN:0368-1645    

DOI:10.1039/CT9191501248

出版商:RSC    

年代:1919

数据来源: RSC

摘要:

1264 CHAPMAN AND WHISTON THE INTERACTION OF CXVII.-The Interaction of Cldor-ine and Hydrogen. The Irtjluence of Mass. By DAVID LEONARD CHAPMAN and JOIIN RESINALD HARVEY WHI STON. AFTER investigating the precautions which must be taken in order to obtain trustworthy actinometric measurements of the velocity of the photochemical action of chlorine and hydrogen Chapman and Underhill (T. 1913 103 496) examined the influence of the concentration of hydrogen on this reaction. The determinations were made with mixtures containing small measured quantities of oxygen and in the same series of deteriniiiations the conceiitra-tims of chlorine and oxygen in the mixtures were kept constant, whilst the concentration o f . hydrogen was varied. They found that (‘as the partial pressure of the hydrogen is increased from zero the rate of formation of hydrogen chloride per unit volume of the mixture is at first almost proportional t o the concentration of the hydrogen but the ratio of partial pressure of hydrogen t o velocity of interaction rises continuously in value as the propor CHLORINE AND HYDROGEN.1266 tion of hydrogen is increased and when the pressure of hydrogen has attained a definite value the rate of formation of hydrogen chloride becomes a maximum and then as the proportion of hydrogen is still further increased the rate of interaction of chlorine and hydrogen falls very slowly.’’ It had previously been shown by Chapman and MacMahon (T., 1909 95 959) that the sensitiveness of electrolytic gas a t atmo-spheric pressure is approximately inversely proportional tc the concentration of the oxygen it contains for proportions of oxygen varying between 0.08 and 1 per cent.by vo,lume of the electrolytic gas. These two results were interpreted by a theory advanced by Burgess and Chapman (T. 1906 89 1433). It was our intention to proceed with the investigation of the influence of the concentration of the chlorine on the rate of form-ation of hydrogen chloride. However before we were able to complete our investigation Bodenstein and Dux (Zeitsclz. ph ysikal. Chem. 1913 85 297) published the result that the reaction is of the second order in the case of a mixture containing equal volumes of chlorine and hydrogen and a small fixed proportion of oxygen. Our results mentioned above on the influence of the concentrations of hydrogen and oxygen were in the main confirmed by these authors.It will be remembered that Wildermann (Phil. Trans. 1902, 199 337) also found that the rate of formation of carbonyl chloride when a mixture containing equal volumes of carbon mon-oxide and chlorine was exposed to light of constant intensity was proportional t o the square of the pressure of the interacting gases. From Bodenstein and Dux’s result and the facts that the rate of interaction is inversely proportional to the concentration of the oxygen and nearly independent of the concentration of the hydrogen (provided that this concentration is not too small) it can obviously be deduced that the rate of formation of hydrogen chloride when a fixed volume of electrolytic gas is exposed to homogeneous light of constant intensity is given approximately by the formula in which k is a constant and Z the intensity of the radiation.Bodenstein and Dux confirmed the above formula with measure-ments made with mixtures containing unequal volumes of hydrogen and chlorine. It may be stated immediately that we are unable t o confirm the work of Bodenstein and Dux on the influence of the concen-3 P.* 1286 OHAPMAN AND WHISTON THH INTERACTION OF tration of the chlorine on the rate of the change. In fact we find that within wide limits of concentration of the interacting gases the rate of formation of hydrogen chloride is given with fairly close approximation by the expression k . I. C'l,] or in [%I other words since I .[Cl,] is proportional to the radiation absorbed per second the hydrogen chloride formed is nearly directly pro-portional to the radiation absorbed and inversely proportional to the concentration of the oxygen. The result embodied in the above expression for the rate of formation of hydrogen chloride in mixtures of compositions and pressures within the limits used in our experiments can be easily interpreted by the hypothesis of Chapman Burgess Gee and Underhill. Briefly the hypothesis in question postulates that the radiation is absorbed by the chlorine molecules. The absorbed energy in the molecules is gradually degraded. I n the earlier stages when the energy is still in a highly efficient form the degradation is accomplished very rapidly and completely by the agency of oxygen and the other inhibitors.The degradation accomplished in this way is so complete that the resulting degraded energy is no longer capable of assisting the union of the chlorine and hydrogen. The absorbed energy in the chlorine molecules which escapes being degraded in the earlier stages by the inhibitors is transformed to lower forms of energy which although capable of activating the chlorine are t o a much less degree and possibly not a t all degraded by the inhibitors. I n other words the life of an activated molecule is not shortened by the agency of the inhibitors. Consider unit volume of the mixture of gases exposed to light of intensity I . The rate of accumulation of efficient energy will be k,. I.[ClJ k being now and below taken t o represent a constant.The loss of efficient energy is due to two causes namely i t s degradation during the impacts of the chlorine molecules with oxygen molecules and the conversion of the chlorine molecules into the activated form. When however the proportion of oxygen to electrolytic gas is as high as 1 per cent. (as it was in all the experi-ments to be described below) the sensitiveness is almost one hundred times less than that of the purest electrolytic gas we could prepare and therefore with such a large proportion of oxygen present the loss of accumulated efficient energy must be almost entirely due to the first of the above-mentioned causes. Now the number of impacts per second between chlorine mole-cules and oxygen molecules is given by the expressio CHLORINE AND RYDROOEN.1267 k, [O,) [Cq. Further if it be assumed that in the mean a constant; proportion of the efficient energy is degraded during an impact of a chlorine molecule with an oxygen molecule the degradation of energy during a single impact will be in the mean, k ~ E being the mean efficient energy of a chlorine molecule after the steady state is reached. The rate of loss of efficient energy will be the product of the loss during a single impact and the number of impacts per second, which is 6 . k . E . [O,] [Cl,]. Equating this to the rate of accumulation of efficient energy, we obtain k . k,. E . [O,] [Cl,] = k1 . I . [a,] and therefore Now if we make the probable assumption that the rate of formation of active chlorine molecules is proportional to the pro-duct of the concentration of the chlorine and mean efficient energy of a single molecule,++ the number of active molecules formed per second should be but when the pressure of the hydrogen exceeds 15 cm.and possibly a t much lower pressures almost all the active molecules of chlorine combine with hydrogen (Chapman and Underhill Zoc. c i t . ) and the rate of formation of hydrogen chloride becomes equal to the active molecules of chlorine produced per second and almost in-dependent of the pressure of the hydrogen. Therefore the number of molecules of hydrogen chloride formed per second is also given by the expression That is the hydrogen chloride produced is proportional to the radiation absorbed and inveksely proportional to the concentration of the oxygen.I f as we have reason to believe is the case hydrogen is a weak inhibitor (Chapman and Underhill Zoc. cit.) the expression for the rate of formation of hydrogen chloride molecules would become k . k I. [Ul,] k2T9 a C(,23+k,[H,1’ * This assumption is equivalent to the amumption that the tendencyof each quantum of efficient energy to change its form is independent of the concentration of the ohlorine 1268 THE INTERACTION OF CHLORINE AND HYDROQEK. in which k is a constant of small magnitude. Since in our experiments however the pressures of the hydrogen and the oxygen have been varied in the same ratio our results are in equal agreement with both formulze and fail to distinguish between them. E x P E R I M E N T A I,. Experiments with the Moist Gases.-The apparatus used was a form of Bunsen and Roscoe’s actinometer in which the pressure of the contained gas could be varied very similar to that used by Chapman and Underhill.The precautions taken in making the measurements and to ensure purity of materials were the same as those observed by Chapman and Underhill: The mixture of gas experimented with was electrolytic gas which contained about 1-3 per cent. of oxygen by volume. The sensitiveness of the mixture a t a pressure of one atmosphere and half an atmosphere was measured and the results are tabulated below. Sensitiveness Sensitiveness Number of at one a t half an experiment. atmosphere. atmosphere. 1. 1000 1029 2. 1000 973 3. 1000 1060 Accordingly qith the moist gases the total hydrogen chloride formed in unit time is almost independent of the pressure, whereas if Bod’enstein and Dux’s result were correct the rate of formation of hydrogen chloride ought to be proportional to the pressure.Experiments wit JL t h e I)ry Gases.-For these experiments the apparatus was modified in one or two respects. The mixture of gases before use was confined in a gas-holder over sulphuric acid. The sulphuric acid which the holder contained was saturated with chlorine and boiled several times and the gas-holder was so con-structed that the sulphuric acid did not come into contact with the air of the laboratory. The actinometer was the same as that used in the experiments with the moist gases with the exception that the index contained sulphuric acid and the insolation vessel was horizontal and contained anhydrous copper sulphate spread uniformly over the bottom of the tube to serve as an absorbent of the hydrogen chloride.The copper sulphate was carefully dehydrated in the insolation vessel and heated in a current of chlorine a t about 300° for an hour. The internal diameter of the insolation vessel was 2-6 cm. The results are tabulated below AUTO-COMPLEXES IN SOLUTIONS OF CUPRTC CHLORIDE ETC. 1269 Sensitiveness Number of at one experiment. atmosphere. 1. 1000 2. 1000 3. 1000 4. 1000 5. 1 do0 Sensitiveness Sensitiveness at a at half an quarter of an atmosphere. atmosphere. 963 755* 93 6 - 776* 944 -- 758* - 752* * At this low pressure it is possible that many of the activated chlorine molecules lose their activity before making fruitful impacts with the hydrogen molecules (compare Chapman and Underhill Zoc. cit. ). Bodenstein and Dux admit that inhibitors were gradually pro-duced in the insolation vessel during an experiment and these inhibitors would tend to increase the apparent order of the reac-tion. They claim t.0 have eliminate,d the effect of these inhibitors by the method of conducting their experiments but it seems to us most likely that the discrepancy between their results and ours is due t o this cause. We take this opportunity of gratefully acknowledging a generous grant awarded to one of us by the Board of Scientific and Industrial Research. THE Sm LEOLINE JENKINS LABORATORIES, JESUS COLLEGE, OXFORD. [Received Ochbw 13th 1919.

ISSN:0368-1645    

DOI:10.1039/CT9191501264

出版商:RSC    

年代:1919

数据来源: RSC

摘要:

AUTO-COMPLEXES IN SOLUTIONS OF CUPRTC CHLORIDE ETC. 1269 C X VII1.-A uto-camplexes in Solutions of Cupric Chloride und Cupric Bromide. By STEWART BYRON WATKINS and HENRY GEORGE DENHAM. IN a paper by Donnan and Bassett (T. 1902 81 939) the view is expressed that the changes in colour brought about in solutions of copper and cobalt salts by changes in concentration and tempera-ture as well as by the addition of certain other haloid salts are t o be attributed to the formation of complex anions as exemplified in the following equations : CU" + 2C1' + zcuc1, 2K' + 2C1' + xCoC'l 1270 WATKINS AND DENHAM AUTO-COMPIJEXES IN the colour of the anions so formed differing essentially from the colour of the cupric and cobalt ions. Quantitative support to this theory was first afforded by Kohlschutter's migration experiments (Ber.1904 37 1193) whilst the investigations of one of the authors (Zeitsch. physikal. Chem. 1909 65 64) have shown that in concentrated aqueous and alcoholic splutions of copper and cobalt haloid salts the cations have a migration number which is distinctly negative. The present paper is an extension of the investigation and deals with the effect of temperature and con-centration on the migration number of the cupric ion. The experimental method adopted was practically the same as in the former paper and the results obtained are of the same order of accuracy. Experiments were nearly always carried out in duplicate and the middle layer was repeatedly analysed. Below are the experimental results : TABLE I.Copper Chloride in Aqueous 8olirtion. TenzperntirrP. 3 5 O . Before electrolysis : Water grams .................. Copper grams .................. per litre ........................ Cakhode liquid grams ......... liquid grams .................. Voltameter Ag .................. Equivalent Cu" ............... Copper on cathode ............ Cuprous copper in solution ... Cupric copper in solution ... Concentration in gram-mols After electrolysis : Copper found in cathode Cuprous chloride in solution Cupric chloride in solution ... Total salt! ........................ Water .............................. Migration number ............ Expt. 1. 13.6121 0-3455 0.25 36.0216 0.8176 0-2272 0-0669 0.0735 0.0604 0.0941 0.81 12 1.7 170 1-8111 34.2105 0.34 Temperature , Before electrolysis Expt.1. Copper grams .................. 0-3455 Concentration in gram-mols. per litre ........................ 0.25 Water grams .................. 13.6131 After electrolysis : Cathode liquid grams ......... 32.01 80 Copper found in cathode liquid grams .................. 0.6848 Voltameter Ag .................. 0.3659 Equivalent Cu" ............... 0.1052 2. 9.1827 0.4368 0.75 35.4874 1.4775 0-2718 0.0808 0.0832 0.0784 0-1221 1-3991 2.9605 3-0826 32.4048 0.24 4 5 O . 2. 9.1827 0.4368 0.75 41,1732 1.7113 0.3048 0.0898 3. 7.931 1.0695 2.1 41.201 6 4.2870 0.2750 0.0810 0-0252 0.1368 0.2126 4-1702 8-8256 9.0382 32.1634 - 0.06 3.6.8970 0.9 180 2.1 43.9082 4.5420 0-261 1 0.0770 4. 2-1160 0.5273 4.0 46.1836 7.5189 0.1994 0.0587 nil. 0.1173 0.1828 7.4016 15.6942 15.8770 30.3066 - 0.57 4. 8- 1700 1.6600 3.2 48.4132 6.8680 0.3105 0.09 1 SOLOTTONS OF CUPRIC CHLORIDE AND CUPRIC BROMIDE . 1271 TABLE I . (continued) . Copper Chloride in A queous Solution . Temperature. 4 5 O . After electrolysis : Copper on cathode ............ Cuprous copper in solution ... Cuprous chloride in solution Cupric copper in solution ...... Cupric chloride in solution ... Total salt ........................ Water .............................. Migration number ............ Expt . 1 . 0.1161 0.0942 0.1467 0.5906 1.2500 1.3967 30.6213 0.22 TABLE I1 .2 . 0.0978 0-0820 0-1277 1-6293 3.4460 3. 5737 37.5995 0.135 3 . 0.0322 0.1217 0.1896 4.4203 9.3520 9.5416 34.3666 - 0.01 4 . nil . 0.1829 0.2849 6.6851 14.1500 14.4349 33-9783 - 0.40 Copper Bromide in Aqueous Solution . Temperature. 3 5 O . Before electrolysis : Water. grams .................. Copper. grams .................. per litre ........................ Cathode liquid. grams ......... liquid. grams .................. Voltameter Ag .................. Equivalent Cu" ............... Copper on cathode ............ Cuprous copper in solution ... Cuprous bromide in solution Cupric copper in solution ...... Cupric bromide in solution ... Total salt ........................Water .............................. Migration number ............ Concentration in gram.mols . After electrolysis : Copper found in cathode Expt . 1 . 21.976 0.5036 0-36 42.5353 0.8739 0.0893 0.0272 0.0394 0-0194 0.0368 0.8590 3.0200 3.0568 39.4785 0.32 2 . 6.8659 0.3302 0.76 41-5343 1.6396 0.1891 0.0558 0.0908 0.0207 0.0467 1.6189 5.6901 5.7368 35.7975 0.16 Temperature. 4 5 O . Before electrolysis : Water. grams ........................ Copper. grams ........................ Concentration in gram.mols . per litre .................................... After electrolysis : Cathode liquid. grams ............... Voltameter Ag ........................ Equivalent Cu" ..................... Copper on cathode ..................Cuprous copper in solution ......... Cuprous bromide in solution ...... Cupric copper in solution ............ Cupric bromide in solution ......... Total salt .............................. Water .................................... Migration number .................. Copper found in cathode liquid. grams ................................. Expt . 1 . 10.7732 0.2518 0.35 40.841 1 0.7959 0.2205 0.0651 0.1068 0.0233 0.0526 0.7726 2.7169 2.7695 38.0716 0.19 3 . 2.2323 0.2246 1.6 48. 6760 3.6298 0.1857 0.0547 0-0005 0.1089 0.2461 3.5209 12.3752 12.62 14 36.0546 0.05 2 . 2.2323 0-2246 1.6 49.4290 3.6765 0.1673 0-0489 nil 0.0977 0.2206 3-6788 12.5821 12.8027 36.6263 4 .1.2527 0.2847 3.6 57.9704 7-3307 0.2000 0.0589 nil . 0.1177 0.2660 7.2130 25.3589 26.6249 32.3455 - 0.35 3 . 1.2627 0.2847 3.6 52.5873 6.6333 0-1937 0.0570 nil 0.1140 0-2576 6.5193 22.9200 23- 177 6 29.4061 -0.18 . 0.8 1272 WATKTNS AND DENRAM ATJTO-COMPLEXES IN TABLE I11 . Copper Bromide i n Bthyl Icohol . Ternperatwe. Before electrolylsis Expt . 1 . Alcohol. grams ........................ 10.3 7 13 Copper. grams ........................ 0.1441 Concentra€ion in gram.mols . per litre .................................... 0.21 After electrolysis : Cathode liquid. grams ............... Copper found in cathode liquid. grams ................................. Equivalent Cn ......................Copper on cathode .................. Cuprous copper in solution ......... Cuprous bromide in solution ...... Cupric copper in solution ............ Cupric bromide in solution ......... Total salt .............................. Alcohol ................................ Migration number .................. Voltameter Ag .......................... 13.3029 0.1464 0.1124 0.0332 0.0151 0.051 2 0.1156 0.0952 0.3346 0.4501 12.8528 - 0.51 Temperature. 4 5 O . Before electrolysis Expt . 1 . Alcohol. grams ........................ 10.3713 Copper. grams ........................ 0.1441 Concentration in gram.mols . per litre ................................. 0.21 After electrolysis : Cathode liquid. .grams ...............Voltameter Ag ........................ Equivalent Cu" ..................... Copper on cathode .................. Cuprous copper in solution ......... Cuprous bromide in solution ...... Cupric copper in solution ............ Cupric bromide in solution ......... Migration number .................. Copper found in cathode liquid. grams ................................. Total salt .............................. Alcohol ............................... 13.5570 0.1495 0.1033 0.0305 0.0098 0.0510 0.1152 0.0985 0.3463 0.4615 13.0955 - 0.77 2 . 9.0477 0.2517 0.43 13.7986 0.3244 0.0985 0.0300 0.0114 0.0466 0.1052 0.2788 0.9767 1.0519 12.7167 - 0.60 2 . 9.0477 0.2517 0.43 14.0426 0.3282 0.0822 0.0243 0.0103 0.0382 0.0862 0.2900 1.0038 1.0900 12.9526 -.0.90 1) is c I I .v s io I L of R es I I 1 t s . The boundary migration experiments described 3 5 O . 3 . 5.8932 0.5069 1.3 16.7181 1-0655 0.1065 0.0314 0.0176 0.0452 0.1021 1.0203 3.587 1 3.6892 13.0289 - 1.1 3 . 5.2739 0.3822 1.17 15.8708 0.8582 0.1229 0.0362 0.0270 0.0454 0.1025 0.8128 2.8576 2.9601 12.9107 - 1.35 by Steele (Zeitsch . physikal . Chenz. 1902. 40. 689) and by Donnan and Bassett (loc . cit.) supported by the quantitative migration experi-ments of Kohlschutter (7oc . c i t . ) and of Denham (loc . cit.) have afforded considerable support to the theory of complex-formation in concentrated solutions of copper and cobalt salts .I n Denham's experiments it was found that at 2 5 O the migration number o SOLUTIONS OF CUPRTO CHLORIDE AND CUPRIC BROMIDE. 1273 copper in a 5.29 solution of copper bromide reached the strikingly low value -0.39. This was interpreted by assuming that in such solutions complex anions of the type [CuBr,]// [Cu,Br,]// etc., carried copper out of the cathode compartment. Thus assuming that the equilibrium lies almost wholly to the right the migration number would be approximately zero whilst if appreciable [Cu,Br,]N ions were present as demanded by the equilibrium Cu" + 2CuBr2 + 2Br' Z Cu" + [Cu,Br,]//, Cu"+CuBrz+ 2Br' t Cu"+[CuBr4]// values less than zero would result. of the solution is shown in table IV. The parallelism between the migration number and the colour TABLE IV.Copper Bromide in Water. Temperature 2 5 O . Concen-tration. 0106 0.414 1.690 2.218 3.187 4-055 5-288 tfJ 11- Colour. 0.445 bluish-green. 0.440 green. 0.069 brownish-green. 0-052 9 9 - 0-086 brown. -0.159 deep brown. - 0.392 Y Y On the other hand many have sought an explanation of this colour-change in the hydration either of the dissolved salt or of the copper cation (notably Biltz Zeitsch. physikal. Chem. 1902, 40 185; Jones and his pupils C'arnegie Pzcbl. No. 60; for com-plete bibliography see Zeitsch. ~ ~ h y s i k a l . Chenz. 1909 65 641). The solution round the cathode may thus become weaker not only by the wandering away of complex anions containing copper but by solvent molecules being transported into the cathodic compart-ment attached t o the cation.That such a transport of solvent molecules attached to ions may occur has been proved by Washburn (Tech. Qzcart. 1908 21, No. a) a1t)hough in the cases studied the effect of hydration on migration ratios is not a marked one. Moreover in their critical review oE migration numbers Noyes and Falk (,7. Amw. Chem. ,'Joe. 1911 33 1436) have shown that the true migration number is connected with the apparent o r hydrated migration number by the equation TTr= T -t ANo. N / N 1274 WATKINS AND DENHAM AUTO-COMPLEXES IN where AN is the number of molecules of water transported to the cathode per faraday; N=number of equivalents of salt in solution associated with No molecules of solvent ; I” = ordinary migration number (referred to solvent which is assumed stationary) ; T T r = true migration number referred to a non-migrating substance.If one accepts the hydration value obtained by Jones (Carnegie Publ. No. 60 p. SS) this correction in concentrated solutions is of the order of 5-10 per cent. and therefore negligible in the light of the marked negative values obtained. The effect- of such a ‘‘ hydration ” effect will naturally be the more noticeable in concentrated solutions but Bein’s results (Zeitsch. physikal. Chem. 1898 27 50) for calcium chlorides salt which according- t o Abegg and Bodlander’s complex theory (Zeitsch. anorg. Chem. 1899 20 453) should not form complex anions and according to Jones (Zoc. clt.) has a strongly hydrated cation-show how small this probable hydration effect is.Chlorine Tempera-per cent. ture. ma. 0.039 2 2 O 0.447 0.42 24 0.406 0.99 21 0.390 Donnan and Bassett’s theory postulates that the formation of the complex anion in solutioi~s of copper and cobalt salts is attended by the absorption of heat that is it is favoured by a rise of temperature. Thus a solution of cobalt chloride in alcohol which is blu0 a t the ordinary temperature becomes pinkish-red on cool-ing to - 7 9 O . If this is the case an increase in temperature should bring about a decrease in the migration number due to the eqhilibrium being driven t o the right. On th’e other hand the. effect of increasing temperature on a hydrated ion will be in all probability t o cause a dissociation into a less hydrated or even an anhydrous ion (Lewis Zeitsch.physikal. Chem. 1905 52 222; 1906 56 223; Biltz Zoc. cit. ; Jones Zoc. cit.) . Jones and West have measured the temperature-coefficients of a large num‘ber of salts of varying degree of hydration and they have concluded that : (1) The temperature-coefficients of aqueous solutions of electro-lytes are greater the greater the hydrating po,wer of the electrolyte SOLUTIONS OF CUPRIO CHLORTDE AND CUPRIO BROMIDE. 1275 (2) This large increase in conductivity with rise of temperature, in the case of salts forming hydrated ions is due in part to the decreasing complexity of the hydrates formed around the ions. Consequently the equilibria Cu(H,O)," Cu(H,O)iL + yH,O =C u" + H20. will be driven t o the right and abnormally low migration numbers must consequently tend to approach the normal value (0.4 approx.) with rise of temperature should the abnormality arise from a hydration effect.0.2 0.4 0.6 0.8 1.0 1.2 Concentration. I n the diagram the experimental results for copper bromide in alcohol are plotted as a typical example of the definite influence of temperature on the results. These curves and the tables bring out clearly the rapid decrease in the migration number as the temperature of the solution rises in precise agreement with the demands of the complex theory in direct contradiction to the demands of Jones' hydration theory. Moreover the following results of Bein for salt solutions wherei 1276 WATKINS AND DENHAM AUTO-COMPLEXES IN complex formation is scarcely to be expected bear out the conten-tion that the temperature has very slight influence on the cathodic migration number even where the evidence of Jones and others indicates that such ions are more or less hydrated.Calcium Chloride. Chlorine Tempera-per cent. ture. 0.42 24" 0.42 97 0-77 20 0.85 94 UCa. 0.405 0.426 0.395 0.451 Lithium Chloride. Chlorine Tempera-per cent. ture. ULi. 0.036 20" 0.37 1 0.036 97 0.389 0.20 25 0.324 0.20 97 0.381 The most convincing evidence of the presence of auto-complexes in the solutions under discussion is however afforded by a con-sideration of the cathodic copper deposit. I n a solution of a copper salt tbe mechanism of the electrode process may be repre-sented by any one of the three equations: .. . . . . . . CU" -+ c u e + @ (1) CU' + c u +@ ( 3 ) CU" + c u +2@. (3) . . . . . . . . . . . . . . . or assuming hydration : . . . CU(K,O)," -f C'U(H,O)i + (X - y)H20 + @ ( 1 4 Cu( H,O),' + Cu + ~ ~ € 1 ~ 0 + @ ( 2 4 Cu(H,O)," + Cu + xH,O + 3 0 ( 3 4 . . . . . . . . . . . . . . (see Foerster and Seidel; Zeitsch. anoyy. Chent. 1897 14 106; Foerster and Coffetti Zeitsch. Elektrochent. 1904 10 736 ; Bose, ibid. 1898 5 163; Heiberg ibid. 1903 9 137; Abel ibid. 1903, 9 268; Bodlander and Storbeck Zeitsch. anorg. Chem. 1897 14, 106; Luther Zeitsch. physikal. Chem. 1900 34 488; 1901 36, 385 ; Wohlwill Borcher's I' Elektrometallurgie," 3rd Ed. p. 198). Should equation (3) or (3n) represent the cathode process the equations would lead to the value 0.590.Table V shows ho SOLUTIONS OF CUPRIC CHLORIDE AND CUFRIC BROMCDE. 1277 much the experimental values differ from these. In this table the last column gives Salt, CuBr, 9 7 9 9 7 9 9 9 9 9 9 9 CUCl, 9 9 9 9 7 7 7 9 9 9 9 9 9 9 CuBr, 9 3 9 9 9 7 7 3 9 7 Solvent. water 7 9 7 9 7 9 9 ) 7 9 9 9 9 7 9 9 9 9 7 9 3 9 9 9 9 9 7 7 alcohol 9 9 9 9 9 9 9 7 9 9 weight of copper deposited weight of silver in voltameter' the ratio TABLE V. Tempera- Concen- cu ture. tration. ?hi*,,. Ag. 35O 0.36 0.32 0.44 35 0.76 0.16 0.48 35 1.60 0.05 0.002 o*ooo 35 3.60 - 0.35 45 0.36 0.19 0.44 45 1.60 -0.18 0.00 45 3.60 - 0.89 0.000 35 0.26 0.34 0.324 35 0.76 0.24 0.308 35 2.00 0.181 0.090 35 4.00 - 0.57 0.000 45 0.26 0.225 0.317 45 0.75 0.135 0.177 46 - 0.01 0.1 23 45 - - 0.4 0.000 35 0.21 -0.51 0.134 35 0.43 - 0.60 0.115 35 1.30 - 1.1 0.1 65 45 0.21 - 0.77 0.095 45 0.43 - 0.90 0.125 45 1.17 - 1.35 0.219 These figures show that in dilute aqueous solutions much of the current is carried by the discharge of cupric t o cuprous ions and, indeed actual observation showed that the deposits consisted of a heterogeneous mixture of copper and cuprous haloid salt.I n such solutions the electrode processes included in the equations (1) and (3) or assuming hydration ( l a ) and (3a) prevail the cuprous ion being immediately precipitated as the insoluble chloride or bromide. The extent t o which either process occurs has been shown by Seidel and others (Zoc.cit.) to depend on the variables tempera-ture concentration current density and concentration of acid, and will in no way be influenced by the presence of hydrated ions in solution whether those solutions are concentrated or dilute. I n c u many of the more concentrated solutions however the ratio __ shows no quantitative agreement with Faraday's laws wh,ether the reduction to cuprous ion or t o copper itself occurs or both. I n Ag c u a number of experiments the ratio - - falls to zero that is no * 1278 AUTO-COMPLEXES IN SOLUTIONS OF CUPRIC CHLORIDE ETC. copper or cuprous salt whatsoever is deposited on the cathode. This abnormality incapable of explanation as it is by the hydra,-tion theory may be satisfactorily explained by the theory of auto-complexes.I n concentrated solutions such equilibria as (CuBr.,)% " Cu" + xCuBr,+ 2Br' Cu" + [ BrJ are postulated the undissociated salt of relatively weak electro-affinity being forced into a complex anion. As the cuprous ion, however possesses a more noble potential than does the cupric ion, that is have a weaker electroaffinity the cuprous salts according to Abegg and Bodlander's complex theory (Zoc. c i t . ) will be more readily forced into a complex than the corresponding cupric salts (Donnan Abegg's " Handbuch," Kupfer p. 517). I n concentrated solutions seeing that no metallic copper or copper salt is precipi-tated on the cathode the current must be wholly carried in the following way : (possibly Cu(H,O)Z + Cu(H,O)i+ (x - y)H,O + 0) ; the cuprous salt however instead of being thrown out of solution as happens in the more dilute solution and as demanded by the hydration theory forms a soluble cupri-cupro-salt with the cupric salt already in solution.It is precisely in those solutions where the absence of deposit on the cathode occurs that the migration number; is so strongly negative. The soluble nature of this cupri-cupresalt, coupled with the markedly negative migration number points very strongly to the assumption that such a reaction as the following occurs : CU" + CU' + 0 Cu" + 2Br' + sCuBr Cu" + [ c' B p ] " . I n further confirmation of the views here put forward a migra-tion experiment was carried out in a U-tube first with 4N-cupric chloride and secondly with 417-cupric chloride warmed with cuprous chloride these lower layers N/lO-cupric chloride. I n both cases the brown towards the anode. Summary. (1) The effect of temperature and concentration which had been being covered by boundary moved on the migration number of solutions (aqueous and alcohollic) of copper bromide and chloride has been investigated. (2) The effect of increasing temperature and concentration is t o cause a marked drop in the migration number of the copper ion, which in concentrated solutions approaches -1 in value COLLIODAT~ ELECTROLYTES SOAP SOLUTIONS AS A TYPE. I279 (3) The formation in concentrated solutions of soluble cupri-cupro-haloid salts at the electrode of the migration vessel has been proved. DEPARTMENT OF CHEMISTRY, UNIVERSITY OR' QUEENSLAND, BRISB ANE. [Received October I 1 th 19 19.

ISSN:0368-1645    

DOI:10.1039/CT9191501269

出版商:RSC    

年代:1919

数据来源: RSC

摘要:

COLLIODAT~ ELECTROLYTES SOAP SOLUTIONS AS A TYPE. I279 CXIX. -Colloidal Electrolytes Soap Solutions as a Type* By JAMES WILLIAM MCBAIN MARY EVELYN LAING and ALAN FRANCIS TITLEY. BASED on the extensive data obtained from the study of soap solutions in this laboratory since 1908 McBain and Salmon have defined a new class of compounds colloidal electrolytes to which a very large number of substances of great industrial importance may be expected to belong. This comprehensive group is defined as comprising salts in which one ion has been replaced by an ionic micelle of high valency mobility conductivity and degree of solvation. Regarded from another point of view this means that any colloid which carries electrical charges will in some measure approach the behaviour of a typical colloidal electrolyte.Light is also shed on the properties of colloidal solutions which contain acids bases or salts or to which these have been added. The consistent application of this point ,of view leads to an ionic micelle theory of all charged colloids ; the current assumption that the other charge is carried by the solvent is replaced by the hypo-thesis that free ions of charge equal and opposite to that of the charged colloid are present in the sol or gel. Soap was chosen as the subject for investigation not only because of its industrial importance but because of its known and definite chemical simplicity and constitution as compared with dyes or protein derivatives etc. Furthermore chemical literature con-tflined apparently irreconcilable data obtained by a number of well-known authorities such as Kraff t Smits and Kahlenberg .These either assumed that soap was an ordinary neutral colloid or else on the contrary an electrolyte which had suffered extreme hydrolysis with the formation of colloidal residue suspended in a strongly alkaline solution. Each investigator moreover con-sidered that the whole problem was solved. Now it is demonstrated through the measurements of con 1280 McBAIN LAING AND TITLEY COLLOIDAL ductivity osmotic activity and alkalinity of soap solutions com-municated from this laboratory that in concentrated solution the soaps are typical colloidal electrolytes. On dilution they gradu-ally break down into simple salts. I n extreme dilution acid soaps separate out through hydrolysis.Both catalytic and electro-motive force measurements have shown that except in extreme dilution hydrolysis is a very minor consideration the hydroxyl ion being present to the extent of only about N/1000. The argument for the existence of the ionic micelle was as follows I n all concentrations the conductivity is high and the osmotic activity which was measured by the unexceptionable method of dew-point lowering is only moderate. Hence in con-centrated soap solutions even if all the osmotic activity is ear-marked for the potassium sodium or ammonium ion nearly or quite half of the conductivity remains t o be accounted for. This conductivity must evidently be ascribed to some constituent that does not exert appreciable osmotic pressure and that therefore must be colloidal.One o f u s (J.W.M. Trms. Faradmy Soc. 1913, 9 99; Kolloid Zeitsch. 1913 12 256) has shown how this is possible and probable by applying the princide of Stokes's law t o the hypothesis of a heavily charged. heavily hydrated ionic micelle which would exhibit excellent conductivity and high viscosity a t the same time. Although our investigation of soap solutions has only reached its first stage it has been possible on the basis of the above reason-ing to set up a comprehensive theory which explains and reconciles all the mass of data of the most diverse sort which has already been accumulated. The varioixs details of this theory beyond the rough outline given above will be mentioned and discussed in turn as we come to the new and confirmatory experimental evidence described below.This communication presents measurements of the freezing point and conductivity of nearly all the soap solutions which can be studied at Oo. Such comprise solutions of the potassium salts of the saturated fatty acids np t o and including the laurate (Cl.) in all concentrations and of the sodium salts up to the octoate (CR). Further the values f o r potassium oleate can be measured up to 0.6N (0.8N is a solid jelly at the ordinary temperature) and sodium oleate up t o 0.4N a solution which is already quite viscous.* All other sodium potassium and ammonium soaps * The stable form of the sodium oleate solutions at 18" is a white curd. The conductivity at 18" of a very slightly alkaline 0.4N,0 sodium oleate solution which has solidified to a stiff white curd is still 95 per cent.as great aa when it is a clear oily liquid at the same tempemture.-M. E. L ELECTROLYTES SOAP SOLUTIONS AS A TYPE. 1281 gelatinise crystallise sediment or solidify a t temperatures between loo and 90° as the case may be. Even potassium laurate and decoate usually separate out a t above Oo. This is another reason why our chief experimental work hitherto has had to be carried out a t 90° in spite of the greatly increased difficulties entailed. The Conductivity of Soap Solutions at 18O.* I n all our experiments the precautions described in previous All instruments and vessels were communications were observed. ma. 1. L c/m, Cell for good con.ducwa. standardised. The soaps were likewise prepared from potassium drippings and fatty acids by methods previously described.Con-centrations are expressed in weight normality N (mols. per 1000 grams of water) and in volume normality N, the latter referring only to the temperature of the particular experiment. The conductivity measurements of solutions from N / 10 upwards were carried out in a simple and convenient cell of Jena glass of the form shown in Fig. 1 specially designed for excellent con-ductors. The large platinum electrodes were coated with grey * Expenmenfa by M. E. L 1282 McBAIN LAINO AND TITLEY COLLOIDAL platinum sponge and they were held firmly in position by glass rods fused to the cell wall and to each corner of the platinum. The mll constant wa8 13-20. The conductivity water posseased a specific conductivity of 1 - 2 x 10-6 for which no correction was made.Solutions from N'/10 downwards were studied in the boro-silicate cell described by one of us (M.E.L. T. 1918 113 245). The two cells gave identical values for the N / 10-solutions. The first column gives the weight normality N, of the soap; the second the volume normality a t 18O N ; the third the values obtained for the specific conductivity K the values given being the final results from wholly independent solutions ; * the fourth,+ the density D i8; the fifth the equivalent conductivity A ; the last column the apparent degree of dissociation a where a=pLv/pa poo being taken as equal to 85.4 for potassium salts except the n-octoate (88.2) and sodium oleate (64.3). TABLE I. Condmtivity of Potassium Laurate at 18*OOo.ThO results are recorded in tables I-IV. Nw; 2.000 1.500 1.000 0.750 0.500 0.400 0.200 0- 100 0.050 N,. 1.378 1.118 0.820 0.641 0.449 0.367 0.191 0.098 0.049 K: 0.05947 0.05131 { EE;) 0.03032 0.02042 0-01 622 : :EE } (8:XX;;f;) 0.0027 13 A. 43.14 45.80 47.09 47.21 45.44 44.22 41.77 44.03 54.89 a. 0.605 0.636 0.651 0.553 0.632 0.518 0.489 0.516 0.64-3 0.0247 0.024 0.001702 68.74 0.805 0.010 0.010 0.0007539 75.44 0.883 0.600 0-400 0-200 0.100 0.050 N/32 0-010 TABLE 11. Condcuctivity of Potassium 0.540 0.363 0.188 0.097 0.049 0.031 0*008 K . { x:8;:;;} 0.01304 0.00 6 24 8 0.002877 0.001 45 1 0.001035 0.00041 98 Oleate 1*0070 1.0030 0.9999 0.9996 0.9991 0.9989 0.9987 at 18*00°.A. 37-25 36.72 33.30 29-74 29.57 33.32 51.95 a. 0.436 0.430 0.390 0.348 0.346 0.390 0.608 * One preparation of oleic acid gave quite colourless clear soap solutions, t Omitted in Tables I and IV as new determinations were not required. the other pale yellow solutions ELECTBOLYTES SOAP BOLUTIONS AS A TME. 1283 TABLE 111. Conductivity of Sodium Oleate a t 18.00°. Nw. N v . K. Di8. A. a. 0.600 0.510 0.01 104 1.0050 21.67 0.337 20.80 0.323 0.007362 0-400 0’357 (0.007476) 1’oo20 0.200 0.197 {~:::~~~~) 0.9997 19.77 0.307 0-100 0.099 0.001922 0.9993 20.46 0.318 0.060 0.049 0.0009966 0.9987 20.59 0.320 0.010 0.010 0.0002966 0.9986 30.09 0.468 TABLE IV.Conductivity of Potassium n-Octoate at 18.00°. Nu?. N,. K. A. a. 2.000 1.601 0.0675 7 42-24 0.479 1.000 0.895 0-04208 48-60 0.661 0.760 0.690 0.03345 49.76 0.664 0.500 0.472 0-02465 63.00 0-601 0.200 0.198 0.01185 63.05 0.716 0.100 0-099 0.006456 69.60 0.788 The density in every case is very slightly greater than that of water as is always observed for potassium and sodium salh of the lower f a;tty acids. Some of these solutions have been measured a t neighbouring temperatures by other investigators with fewer experimental pre-cautions (Kahlenberg and Schreiner Zeitsch. physikd. Chcm., 1898 27 552; Dennhardt Diss. 1898; Ann. P&ys. Chem. 1899, [iii] 67 325; Kurtzmann Diss. 1914; liToEl. Chem. Beihefte, 1914 5 465 ; Reychler Eighth International Congress of Applied Chemistry).Their results differ more or less irregularly from ours by amounh varying from 0 to 30 per cent. in both directions. The smaller differences are chiefly attributable to alkali dissolved from the glass an important source of error in the study of soap solutions (see McBain and Taylor loc. cit. on whose exact study the validity of these experiments is based). On the whole the agreement is substantial and our results are probably accurate to about 1 per cent. with the possible exception of sodium oleate. The effect of time is of the order of magnitude of 1 per cent., but all our solutions were kept for several days a t the ordinary temperature to allow them to become quite constant before being measured. The data refer to clear solutions in every case except N/ZO-potassium laurate which is always cloudy.Thus many of these solutions are supersaturated or metastable with respect t 1284 McBAIN LAING AND TITLEY COLLOIDAL the separation of indefinite acid soaps. The latter dissolve on warming and the metastable solutions can then be preserved for long periods. The ExpJanatioiz of the Good Conductivity of these Colloidal Electrolytes tugether with their Anomalous ilfaxima and Nininza by Means of t h e Hypothesis of the Ionic Micelle. It will be recollected that hydrolysis has been shown to occur to only a slight extent in conoentrated solutions of soap so that in any case it cannot be adduced t o explain an appreciable part of the good conductivity exhibited by soaps (and by hexadecane-sulphonic acid and its salts).This must find another explanation. The conductivity results are best discussed by showing them in the form of a graph f o r comparison. This is done in Fig. 2 where the equivalent conductivities are plotted as ordinates against the concentration N, as abscissa The figure also includes curves for the results of the higher potassium soaps a t 90°. The most striking point about these curves is that they represent moderately good conductivities even in concentrated solutions. Indeed in many cases the conductivity curve after passing through a minimum rises on further concentration t o a maximum a phenomenon otherwise unknown in any aqueous solution,9 except for Reychler's hexadecanesulphonic acid which is also a soap (Bull.SOC. chim. Belg. 1913 27 113). This effect is more pronounced a t 1 8 O than a t 90°. Again for any one saturated fatty acid the potassium salt exhibits this behaviour to a distinctly greater degree than does the corresponding sodium soap. The oleates differ entirely from the soaps of the saturated fatty acids in that sodium oleate is a very much more typical soap in this respect. The existence of the minimum as an experimental fact has shown that on further increase of concentration the changes effected are such as t o increase the conductivity. This cannot possibly be due to dissociation suddenly increasing again for a limited interval, nor can it be due to free hydroxyl ions since the latter factor is negligibly small. The increase of conductivity with increase of concentration must therefore be due to the replacement of the simple laurate anion by an ionic mioelle of higher mobility.Thus the effect of decreas-ing dissociation is more than counterbalanced by this replacement. Finally however a maximum is reached where the steadily * Gloan ( J . Arner. Chem. Sm. 1910,32 946) finds that at 0" the equivalent conductivity of aqueous potassium iodide is constant over a considerable range of concentration ELECTROLYTES SOAP SOLUTIONS AS A TYPE. 1285 diminishing dissociation of the colloidal electrolyte itself over-balances all other factors. That the conductivity of the ionic micelle is greatest in concentrated solutions where its hydration is least will be discussed below. FIG 2. 8' 0.00 0.10 0*20 0.30 0.40 0 5 0 0.60 0.70 0.80 0.90 Equivalent conductivity of soap solutions at 18" and at 90".The position of the minimum a t the ordinary temperature is still only a t N / 5 but the greater formation of ionic micelle is evinced by the quite slow rise of the curve on dilution a t the solu-tion N/10 as contrasted with its rapid rise a t 90°. This behaviour was predicted by McBain and Salmon as was also the fact %ha 1286 McBAfN LATNG AND TITLEEP COLLOIDAL the minimum for potassium oleate occurs between N/10 and N / 2 0 owing t o the greater tendency of this higher soap to form micelle. The Temperature-coefficient of Conductivity of Solutions. The behaviour just described is in accordance with the abnorm-ally high temperature-coefficient of conductivity in soap solutions, a property which was formerly ascribed to rapid increase of hydro-lysis on heating but is now seen to be the result of diminishing hydration of the ionic micelle with rise of temperature.The influence of concentration on t,he temperature-coefficient of the conductivity may be analysed by the use of table V which gives the ratios between the values a t 90° and a t 1 8 O together with those of a few electrolytes for comparison. Kurzmann (Eoc. cit.) has already pointed out that for a O*GN-potassium oleate solution the conductivity rises threefold between 20° and 90°, whilst the viscosity falls four hundred-f old. TABLE V. Ratios bet,ween Conductivities a't 90° aPzd 1 8 O . Salt. O.OlN,. 0*05Nw. Q-lN,. 0.2Nw. Q.5-0*6Nw. l.ONw. 2.0h7,. Sodium acetate ...3-25 3.22 3.19 3.13 3-12 3.14 3.33 Potassium acetate 2-88 2.85 2.82 2.79 2.75 2.79 2.98 Sodium hydroxide 2.53 2.46 2-46 2.44 2.45 2-50 -Potassium 2.22 - hydroxide -2.95 3.03 3.18 3.06 - Potassium octoate . - - - - - -, laurate 3.09 3.56 3-62 3-45 3.22 3.04 2.86 It is evident from table V that for electrolytes the ratio between the conductivities a t 90° and 20° do not vary much with change of concentration although they make a flat curve with a minimum a t about half normal. Further the values for sodium salts are in every case 0.3 unit or about 10 per cent. higher than for the potassium salts. Soaps on the other hand possess higher values than the corresponding potassium electrolytes and they exhibit an opposite behaviour in that the ratio for the temperature effect is a maximum in medium concentration.Kurzmann's ratios as far as they go appear to parallel ours but are 0.3 to 0.4 unit lower. The customary temperature-coefficients of conductivity involve three separate factors for any electrolyte and five for a colloidal electrolyte. The first two are the specific mobilities of the two ions of which those with lowest mobility have the highest tempera-turewfficient the third is the change of dissociation with temperature whilst in the case of a colloidal electrolyte ther'e i ELEOTROLYTES SOAP SOLUTIONS AS A TYPE. 1287 the further factor of the chauge in equilibrium between ions and ionic micelle in addition t o thO temperature-coefficient of the latter itself. The more exact analysis is simple for electrolytes but must remain in abeyance for the soaps until we have obtained the further experimental data contemplated.The higher temperature-coefficient for sodium ion is in accordance with the recognised probability of its hydration being greater than that of potassium, and that the same factor of hydration would explain the data for soap solutions. The increase of colloid with lowering of temperature is also a factor which might conceivably operate in either direction according as to whether this is increase of highly mobile ionic micelle or increase of neutral colloid a t the expense of electrolyte and micelle. The other case of a colloidal electrolyte exhibiting an anomalous conductivity curve is hexadecylsulphonic acid (Reydhler Zoc. c i t .), which exhibits a minimum conductivity of 135 mhos.(to 149 mhos.) in N/30-solution a t 56O (or 55O) but it has been measured only up to N/15-solution. Indeed it is tantalising in the extreme to note how many promising cases of colloidal electrolytes have been studied only in dilute solution probably the reason why this type of behaviour has not been long familiar. The Osmotic and Preezinppoint Methods. These data are not nearly so accessible as one might expect for the case of colloidal electrolytes in which we are interested as a brief examination will show. Indeed the data of McBain and Salmon constitute the only satisfactory determinations hitherto recorded. The freezing-point method which is the subject of our present study is the osmotic method par excellence where it can be appliedj but it is here surprisingly limited in its range of application.Its trustworthiness is undisputed and further it may be made accurate. The boiling-point method is wholly untrustworthy in its application to solutions that froth and contain colloids on account of the effects of enclosed air which as McBain and Taylor proved experimentally entirely vitiate the indications of the method and may lead to large apparent lowering instead of rise of boiling point. The vapour-pressure method even in the hands of Smits led to equally erroneous results on account of the same unsuspected error. McBain and Taylor found that weeks of effort were required to obtain a single rather inexact measurement. Osmometer data again depend so much on the mode of inter-pretation that it is inadvisable to build upon such a foundation.VOL. cxv. 3 1288 MCBAIN LAING AND TITLEY OOLLOIDAL The lowering of dew-point method has been developed by McBain and Salmon and measuremente have been made of nearly a hundred soap solutiosns. It possessed the double advantage that it could be used at various temperatures and that the presence of air had no influence on the results. It is however an unfamiliar method which has not been previously applied t o the study of solutions. The existing literature consists of two measurements by Kahlen-berg and Schreiner of the lowering of freezing point of N / 8 - and N/lG-sodium deate and the dew-point data of McBain and Salmon mostly referring to a temperature of 90°. The freezing-point measurements here presented were carried out in order to study the effect of temperature and obtain further corroborative evidence by this independent method.The great difficulty is to measure the soaps while in the form of homogeneous solution. On cooling in the freezing apparatus, they usually become turbid with the formation of fine white or colourless crystalline flakes presumably of somewhat acid soap. This may often be avoided and the soaps can be measured in the metastable condition. It should be noted that they are only a few degrees below the temperature a t which the homogeneous form of the system constitutes stable reversible equilibrium. Indeed, perhaps the most important fact for the whole study of soap solu-tions is that the solutions constitute perfect reversible equilibria in which colloid micelle and crystalloid alike participate (see McBain and Taylor Zoc.cit.). Incidentally this would lead us t o infer that colloidal systems in general represent the equilibria much more often than is usually recognised owing to the conditions not being kept constant. Two methods were used namely the ordinary one of Beckmann and that of Richards (Zeitsch. physikal. Chem. 1903 44 563; J . Amer. Chem. SOC. 1903 25 291). The latter is the quick&, and probLibly also the most accurate of the precision methods and it had been shown to be capable of yielding results accurate to about 0'0003°. It consists of the use of a Dewar vacuum vessel surrounded by a bath of the freezing temperature and containing a solution which is full of finely powdered pure ice.When after stirring equilibria is attained a portion of the solution is with-drawn for analysis. F o r the latter we used either a Zeiss interfero-meter in conjunction with a graph prepared from standard solu-tions or else we evaporated the solutions to dryness and weighed thO residue. The former method for example is suitable for potassium acetate the latter for potassium laurate. Althoagh our object was not great accuracy but chiefly trustworthiness an ELECTROLYTES SOAP SOLUTLONE AS A TYPE. 1289 results free from distortion we used a standard thermometer graduated in five-hundredths of a degree the scale and readings of which were corrected to the readings on the international hydrogen thermometer a t the Reichsanstalt in 1913 and this thermometer was kept permanently a t Oo.Results by the ordinary Beckmann method are usually much distorted by the effect of the low convergence temperature which makes the lowering too great. I n concentrated solutions more over the alteration of concentration due to undercooling is very appreciable and tends in the same direction. This error cannot readily be allowed for by the ordinary method of calculation in the case of soap solutions although it is very marked probably on account of the protective action' of the colloid on incipient ice crystals. Thus unless an inoculating rod was used the under-cooling easily exceeded 7 O for an hour a t a time in spite of vigorous stirring. Great care was taken to minimise undercooling and to raise the convergence temperature.The important point to remember in what followed is that the lowering observed is never less than the true value but that on the contrary the osmotic effect is exaggerated. The Freezing-point Data. Table VI presents the results of the Beckmann freezing method. The results marked with an asterisk denote turbid solutions. The potassium myristate ( C14) separated out completely on cooling with marked evolution of heat; the liquid froze a few thousandths of a degree below Oo. TABLE VI.* Lowering of Freezing Point b y Beckmann Method. 0*2N,. 0-5N,. 1-ON,. 2*ON,. Potassium acetate. ..... 0.749' 1.948' 3-82' 8.56' Sodium acetate . . . . . . 0.704 1.774 3.739 8.10 Potassium n-octoa te... 0.742 1.860 2.519 3.146 , leurate ...{ !:% { !:!; t0.737 { :% - 9 9 decoate.. . 0,649 0.752 1-014 j 0.623 Also 0.75 sodium acetate 2.740°; 1.5 sodium acetate 5-83O; 0.05 potassium laurate 0 . 1 7 7 O ; 0.1 potassium laurate 0-212O ; 1.0 sodium n-octoate 2.445O ; 0.6 potassium oleate 0'348O ; 0.4 potassium oleate 0.215O; 0.4 sodium oleate 0.146O ; 0.2 sodium oleate 0.095"; 3.0 potassium n-octoate 4*71°. * Experiments by M. E. L. or 2. W. M. 3 c 1290 McBAIN LAINQ AND TITLEY COLLOIDAL Table VII contains the reeults of the Richards method in which it is a matter of great difficulty to avoid separation of soap with resulting turbidity. Here again such solutions are designated by an asterisk. Many unsuccessful attempts wer0 made to measure clear laurate solutions until finally in one case a solution which had been carefully f rmen without becoming turbid was introduced in this form and stirred very slowly until the temperature was constant; the liquid then withdrawn was clear.Lowering of Potassium acetrtte. +- 0*824X 2.034' 0.718 2.622 0-638 1.890 0.629 1430 0.488 1.712 0.469 1.636 0.426 1.490 0.378 1.324 0.362 1.230 - -TABLE VII.* Freezing Point (corr.) b y Rich0 Potassium acetate. 0-323Nw 0.316 0.292 0.268 0.234 0-206 0.169 0.146 0.113 -1*120° 1.102 1.007 0.882 0-800 0.690 0.588 0.507 0.404 -Potassium oleate. - 1*196N 0.117' 0,096 0.088 0.066 0.064 Potassium n-octoate 0-811NW 2.367 0-609 1.912 0.406 1-634 0.269 1.004 0.169 0.610 vds' Method. Potassium laurate.- 0- 684Nw 0- 38 lo 0.569 0.337 0.49 1 0,304 0.176 0.182 0.124 0.168 0.0829 0.164 0.0361 0.120 0.466 0.328 0.262 0.242 0.121 0.203 0.093 0.164 The values obtained by the Beckmann method are in many cases higher than the more accurate data of table VII. Further, the data for solutions which have become turbid through separation of solid are distinctly less than those for clear solutions. The lowering of freezing point indicates the total concentration of crystalloid. This is shown in table VIII for round concentra-tions in most cases making use of the data of table VII in prefer-ence to those of table VI and of course taking only the data for clear solutions. The numbers are obtained by dividing the lower-ings by 1.858 the cryoscopic constant for water.To tccl Concentration 0.1 N,. Potaesium acetate...... 0.19 Sodium acetate . . . . . . -Potassium octoate . . . -, decoate ... -, laurate ... 0.093 , oleate ...... 0.046 sodium oleate ......... -TABLE VIII. of Crystalloid in these Solutions. 0*2N,. 0.37 0.38 0.37 0.36 0.136 0.064 0.061 0.4Nw. 0*6N,. 0*6N,. l-ON,. %ON, 0.76 0.94 1.13 2-03 4.60 0.76 0.95 1.16 2.01 4.36 0.81 1.00 1.13 1.36 1.693 0.39 0.40 0.41 0.55 -0.164 0.20 0-24 0.40 0.79 0.116 0.16 0.19 - -0.079 - - - -* Experimenbs by A. F. T ELE6ROLYTES SOAP SOLUTIONS AS A TYPE. 1291 Also sodium octoate l.ON, 1.32 ; potassium octoate 3*ON,, 2.53. The lowering of freezing point is so large as to be indisputable, although 3 is in many of the solutions less than for an electrolyte of the same concentration.FIa. 3. K Oleate Oo Na Oleate 0” Nw + The EfJect of Temperature on the Osmoltic Activity of Soaps. The significance of these results is more apparent from a study of the results plotted graphically as in Fig. 3 in the form of the value of the van’t Hoff factor “i.” This is merely the ratio between the actual lowering and that predicted theoretically for a perfect non-electrolyte of the same concentration namely 1’858O per equivalent of normality. In Fig. 3 the values of “ i ” are plotted against total weight normality of the solution. Some of the values a t 90° are included for comparison. 3 u* 1292 McBAIN LAINO AND TITLEY COLLOIDAL First with regard to the absolute magnitude of the osmotic effect in concentration of the higher soaps above N / 3 it lies between 1/5 and 2/5 of that for a crystalloid such as sucrose or between 1/10 and 2/10 of that for an acetate.Further the osmotic effect tends to be constant for concentrated solutions. A t 90° on the other hand the osmotic effect is several times greater, and it decreases rapidly with concentration up to 1.5N or beyond. Both facts indicate a much more complete formation of colloid a t the lower temperature. Secondly the familiar general influence of hydration in magnify-ing osmotic effects is clearly apparent in the data for the acetates, which in turn stirnulab complete dissociation in concentrated solu-tion. The effect of hydration is much more in evidence a t Oo than a t 90°. The octoate is particularly interesting in its intermediate posi-tion in the homologous series as in its behaviour.I n solutions up to N / 2 it is quite like the acetate although with even greater apparent hydration. From N / 2 upwards however it rapidly and steadily falls like the decoate a t 90° until the osmotic activity is only 0.85 that of a theoretical (non-hydrated) non-electrolyte. The Coizcentmtion of Potassio;n or Sodim and the Mobility of the Ionic Micelle at 1 8 O . I n tables I to IV values were given for the degree of dissocia-tion deduced from conductivity based on the mobilities K’ = 64.7, Na’ = 43.6 C,’ = 34.7 C,’ = 23.5 C12’ = 20.7 C l i = 20.7 where the fatty acid ions are indicated by the number of carbon atoms they contain. Our data supplemented by those of Kohlrausch for the acetates when calculated lead to the provisional concentration of alkali ions given in table IXa.This tacitly assumes that even the ionic micelle exhibits the same conductivity as the ions from which i t originates. An alternative basis of calculation is con-tained in table IXb to be explained below. TABLE IX. Concentration of Potassion or Sodion at 180. (a) Assuming Ordinary Ionic Mobilities. Potassium acetate.. . . . . 0.084 0.159 0.298 0-360 0.422 0.634 1.002 Sodium acetate ...... 0.069 0.129 0.235 0-278 0.324 0.467 0.680 Potmium octoate ... 0.079 0-143 0.261 0.300 0.350 0.551 0.958 Substance. O-lN. 0.2N. 0.4N. 0-6N. 0.6N. 1.ON. 2.0N. , laurate ... 0.052 0.098 0.207 0.265 0.324 0.661 1-01 oleate ...... 0.036 0.078 0.172 0.217 0:262 - -80diL 01-b .......#. 0.032 0.061 0.129 0.166 0-192 - E T J ~ R O L Y T X ~ SOAP SQLUTTONS AS A TYPE. 1283 TABLE IX (continued). (b) If Ionic Micelle has t h e same Mobility as Potassion. Substance. 0.1N. 0 0 2 ~ . 0.437. 0 . W . 0-6N. 1.ON %ON. Potassiumoctoate ... 0.079 0.143 0-261 0.300 0.342 0.504 0-822 , laurate ... 0.034 0.066 0.137 0.178 0.21'4 0-36 0.66 , oleate ...... 0.023 0.051 0.114 0.143 0-173 - -Sodium oleate ......... 0.019 0.037 0.077 0.098 0.120 - I Taking first the tentative results of table IXn for the oleates and laurate and comparing them with the data of table VIII it is at once apparent that the supposed concentration of potassion or sodion greatly exceeds the total concentration of crystalloidal matter present except for the most dilute solutions.Thus for 0*4N-solutions the discrepancies amount to about 0-05N. We are forced to revise the assumption made with regard to the conductivity of the ionic micelle and to ascribe to it a conductivity equal to that of the potassion if conductivity and osmotic measure-ments are to be harmonised a t all. I n table IXb then the con-ductivity data are calculated on the assumption that the ionic micelle has an equivalent conductivity of 64.7 which is more than three times as great as that of the true oleate or oleate ion. This is however in agreement with the theoretical considerations advanced by one of us (J.W.M. Trans. Farday Soc. 1913 9, 99; E d l d Zeitsch.. 1913 13 56) and already applied by one of us in a previous communication with Salmon (Zoc.cit.). It was found necessary to make a similar assumption in the case of concentrated solutions a t 90°. Once again it is necessary to recall that 'the known sources of distortion of the experimental data operate in such direction as to emphasise the argument on which our conclusions are based, Thus owing presumably to hydration osmotic data in general are obviously magnified (see for example Landolt-Biirnstein '( Tabellen," where apparent dissociation frequently exceeds 100 per cent.). Again high viscosity is conceded to have the effect of diminishing apparent conductivity. Yet in spite of this the outstanding experimental result is that in soap solutions the osmotic effect is only sufficient t o explain about half of the con-ductivity exhibited.The effects just discussed for the case of ordinary electrolytes are exemplified in the usual unmistakable fashion by the results for the acetates and the more dilute solutions of the octoate. In theae cases the osmotic activity as measured considerably exceeds the total predicted from conductivity. I n the more concentrated solutions of octoate however the soap character predominate 1294 McBAIN LAINQ AND TZTLEY :. COLLOIDAL sufficiently to mask this and they show more than 50 per cent. of colloid (see below table X). In this intermediate case of potassium octoate in table IXb use was made of Pig. 3 in computing the average mobilities of the varying mixtures of ions and ionic micelle here present. For solutions between 0.5 and 2*ON it was simply assumed that A, varied linearly with the i value from 88-2 for i=2.00 to 108.3 for i=0.40.This is a first approximation pending the results of actual measurements of migration now being carried out with soap solutions. The Amounts of Crystaltoid and Colloid in Soap Solutions. ThO total amounts of crystalloidal matter other than potassion or ssdion are obtained by subtracting the numbers in table IXb for sodium or potassium from those of table VIII for total crystal-loids; the results are given in table X. Further these amounts subtracted from the total concentration leave the amounts which it is necessary to regard as colloid. Further since the total amount of say oleate present must be the sum of crystalloidal and colloidal oleate the amount of colloid is simply t h O total con-centration less the value for crystalloid given in table X.These values are collected in table XI and the latter includes for com-parison a few results found by the dew-point method a t 90°. TABLE X. Cryatalloidal Matter other than Potassion or Sodion. Substance. O-lN. 0-2N. 0.4N. O*5N. 0.6N. 1*ON. !&ON. Pottmssium acetate.. .. . . 0.106 0.21 1 0.462 0.580 0.708 1.396 3.698 Sodium acetate ...... - 0.251 0.526 0.672 0.836 1.543 3.680 Potagsiwn octoate ... - 0.248 0.659 0.700 0.790 0.866 0.868 , laurate ... 0.059 0.071 0.027 0.02 0.03 0.04 0.13 , oleate ...... 0.023 0.013 0.002 0.007 0.014 - - - - - Sodium oleate ......... - 0.014 0.002 -TABLE XI. Total Con'centration of Colloid. Substance. 0.137. 0-2N. 0427. 0.5N. 0.6N. 1.ON. %ON. Potassium octoate - -0.05 -0.06 -0.2 -0.19 +0.144 $10132 Potassium lamate 0.041 0.129 0.373 0.48 0.57 0.96 1.87 Potweium olea te...0.077 0.187 0.398 0.493 0.586 -Do. at 90° ...... - -0.02 - +0*07 - +0*28 -Do. at 90" ...... - -0.01 - +0-22 - +Om85 +la89 - - Sodium oleate . . . . . . - 0.186 0.398 ELECTROLYTES SOAP SOLUTIONS AS A TYPE. 1295 The results given in tables X and XI may be summarised in the statement that in all but the most dilute solutions of the laurate and still more so with the oleates the soap exists almost entirely as colloid. The 0.1N- and OW2N-laurate and the 2*0N-octoate contain comparable amounts each of crystalloid and colloid. It is instructive to note that the effects of hydration are much Iess evident a t 90° which agrees well with the extensive investi-gations of H.C. Jones and others on the solvate theory. The results with soap solutions in general afford strong support for a solvate .form of the dissociation theory. For instance the change in hydration with temperature explains the very high temperature-coefficient of the conductivity of soap solution. The second effect of temperature is that much more colloid is formed a t lower temperatures. This is very evident in the case of the higher soaps. A t 90° as a t the ordinary temperature very dilute solutions of soap contain but little colloid and the soap is essentially in the form of a simple electrolyte. A t the lower temperature however the formation of colloid sets in a t much lower concentration. The combined influences just referred t o are most apparent in the intermediate case of the octoate.Whereas a t the ordinary temperature apparently negative concentrations are deduced for colloid owing to the influence of hydration even up to O-GN-soh-tion a t 900 the negative value a t 0.2N is almost within the experimental error and the formation of colloid shown already a t O~5N-solution to the extent of 14 per cent. The Formulation and Concentration of the Tonic Micelle. As was shown in the previous communication the conception of the ionic micelle that appears most probable pending the resultr: of further experiments is that it consists essentially of an aggre-gation of ordinary ions retaining their original charges. Around this nucleus would be condensed a large number of molecules of water and probably also most of the neutral colloidal soap avail-able.(NaP),(P);(') . (H,O),. With increasing concentration or when the dissociation is diminished this must tend to alter towards neutral colloid This is represented in the formula (Nap) * (H,O)m. There was evidence for the conclusion that hydration diminishes We have shown how the approximate values of the concentra-and mobility increases with increasing concentration 1296 McBAIN IiAfNG AND TITLEY COLLOIDAL tions of the total crystalloid and total colloid can be obtained. We have now to allocate the total crystalloid between undissociated soap sodion or potassion and simple fatty ion and the total colloid between neutral colloid (Nap), and aggregated ions in the micelle (P)$). This is a t present possible only between certain limits set out in table XI1 below.The first figure in each pair allocates the total colloid so far as possible to micelle; this cannot of course, exceed the concentration of potassion or sodion and any excess must necessarily be ascribed to neutral colloid. The second figure in each case makes the opposite assumption allotting the total colloid to neutral colloid as far as possible. The maximum here is the total amount of undissociated soap obtained by subtracting the concentration of potassion or sodion from the total concentra-tion of the solutions; any excess of colloid must then be recognised as ionic micelle. Table XI1 includes previous data for the case of potassium laurate a t 90° for comparison. Results are expressed in mols.per 1000 grams of water. TABLE XII. Measured Limits of Concentration of Constitzients of Soap N w . 1.0 2.0 0.1 0.2 0.4 0.5 0-6 1.0 2.0 0.2 0-6 1.0 2.0 So 2.r~ tio ns. Neutral Simple Simple colloid Micelle ion undissociated Cation. (KWZ P;(’) P. KP. K. Potassium Octoate at 0-18O. 0.00-0-14 0.14-0.00 0.35-0.50 0*50-0*35 0-50 0.31-1.13 0.82-0.00 0.04-0.82 0.82-0*04 0.82 Potassium Laurate at 0-18O. 0.01-0.04 0*03-0.00 0.00-0.03 0*06-0*03 0.03 0*06-0.13 0.07-0-00 0.00-0.07 0.07-0.01 0.07 0.24-0.26 0*14-0*11 0*00-0.03 0*03-0*00 0.14 0 . 3 0 4 - 3 2 0.18-0.16 0*00-0*02 0*02-0*00 0.18 0.36-0.39 0-21-0*18 0.00-0.03 0.03-0*00 0.21 0.60-0.64 0.3G0.32 0.00-0.04 0*04-0*00 0.36 1.21-1.34 0.66-0.53 0*00-0.06 0.13-0*07 0.66 Potassium Laurate at 90°.0.00 0.00 0.10 0.10 0.10 0.00-0.22 0 . 2 2 4 . 0 0 0-04-0.26 0.2A0-02 0.26 0.33-0.48 0.52-0.37 0*00-0*15 0.16-0.00 0.52 1.00-1.11 0.89-0.78 O*O&O*ll 0*11-0*00 0.8 ELECTROLYTES SOAP SOTAUTIONS AS A !FYPE. 1297 TABLE XII. (continued). itfeasured Limits of Concentration of Constituents of Soap N,, . 0.1 0.2 0.4 0.5 0.0 0.2 0.4 Solutions. Neutral Simple Simple colloid Micelle ion undissociated Cation. (KP)z. P",'). P'. KP. K. Potassium Olente at 0-18O. 0.060.08 0 . 0 2 4 . 0 0 0.00-0.02 0*02-0*00 0.02 0*1&0-15 0.05-0.04 O*OO-O*Ol 0 . 0 1 4 . 0 0 0.05 0 . 2 8 4 . 2 9 0.1 1 0.00 0.00 0.1 1 0*35-0*36 0.14 0.00 0.00 0.14 0.41-0.43 0.17-0.16 0*00-0*01 0*01-0*00 0.17 Sodium Olewte at 0-18O.0*1&0*16 0*0&0*02 0*00-0*01 0.01-0.00 0.04 0.32 0.08 0.00 0.00 0.08 An inspection of thO data in table XI1 reveals that in the cases of the oleates and the more concentrated solutions of the laurates, the limits of concentration of each constituent are narrowly defined. These solutions consist almost entirely of colloid together with potassion or sodion. There is more than twice as much neutral colloid (KP), as of agglomerated ions P$); in these potassium soaps and in sodium oleate there is four times as much as of Pz(') whereas a t 90° the proportion varied between 3 / 2 and nearly equal amounts. Once more the difference between potassium and sodium soaps asserts itself a difference that is evidence for the inclusion of the neutral colloid in the micelle since otherwise the micelle must be of identical composition in the two cases.Although there is the same total amount of colloid in both cases the sodium soap con-tains only about half as much micelle. Comparisoii with solutions a t higher temperatures shows that there is more neutral colloid present a t low temperatures pre-sumably in the micelle and possibly in the same way that there is also greater hydration a t the lower temperature. This accords also with the abnormally high temperaturecoefficient of conductivity. Cornpurism with Restilts of Dew-point Menswements at 203. We have carried out a few dew-point measurements a t 20° for The data comparison employing the method previously described 1298 McBAIN LAINB AND TITLEY COLLOIDAL are given in table XIII.It will be noticed that th'e data agree exactly with the most accurate of the freezing-point data; the method is applicable a t all temperatures and therefore to all soap soh tions. TABLE XIII. Measzcrements of Lowering of Dew Point of Soap Solution.* soap. N,u. perature. Lowering. loid. " i." Potassium chloride ...... 1.ON 20" 0.58" 1.99 1.99 20 0.04 0.14 0.70 , laurate 0.2 , octoate ...... 3.0 20 0-70 2.41 0.99 oleate 0.6 20 0.07 0.24 0.40 20 0.12 0.41 0.51 0.2 Potassium laurate -0.6 , oleate ] Tem- Crystal -......... .. ......... Ammonium laurate ...... 1.ON 20" 0.17" 0.583 0-583 Y ? I? ... 0.6 20 0.08 0.276 0.65 , palmitate ... 1.0 20 0.06 0.206 0.21 ? 7 1.0 90 0-13 0.277 0.28 The complicated effect possible when soaps are mixed in solution has been discussed elsewhere.I n the case of the mixture in table XIII 0*6N in respect of oleate and 0-2N in respect of laurate the total concentration of crystalloid is 0*41N as com-pared with 0*38N the sum of that of the constituents separately. The conductivity of the mixture is also 5 per cent. above that of the constituents so that most of this increase may be attributed to formation of mixed ionic micelle. Ammonium Soaps. The study of ammonium soap prepared from palm-kernel oil (which is largely laurate with some higher constituents) by Goldsmidt and Weismann (Xolloid Zeitsch. 1913 12 18) has given very interesting results. Like the potassium soaps they exhibit fairly high conductivity which in this case cannot possibly be due to products of hydrolysis.The conductivity curve differs greatly from those of sodium and potassium soaps in that it rises steadily with concentration from 0.2N- up to 1'ON-solution by about 38 per cent. The rise is rapid a t first then more gradual. Dilute solutions were not measured. The rise is accompanied by an increase in viscosity of several hundred-fold. All these solu-tions are appreciably hydrolysed as is shown by the increase in conductivity eBected by addition of excess of ammonia which in itself is a poor conductor. These reeults can be interpreted as showing that the simple fatty * Measurementa by M. E. L BLECJTROLYTES SOAP SOLUTIONS AS A TYPE. 1299 ion exirrting in more dilute solutions ie being replaced in more con-centrated solutions by the ionic micelle which is a better conductor.This is quite in accord with our conclusions but the micelle in cdncentrated solutions must conduct as well as a good conducting ion such as potassion. Some dew-point measurements of ammonium soaps are given in the lower half of table XIII. The solutions were prepared from pure fatty acids. It was not considered necessary t o correct them for the effect of the partial pressure of the ammonia as it was evidently too small. In warm weather the N- and N/2-laurates were quite clear solutions showing none of the usual appearances of soap solutions except that they gave a very good lather. On cooling the N / 2 -laurate solution deposited feathery crystal-like flakes ; the N-soh-tions did not do this but in very cold weather they set to a semi-transparent jelly similar t o a potassium soap.The N / 2-palmitate was somewhat similar but the N-solution was practically solid a t all temperatures and resembled solid white foam. A glance a t the results shows that independent of the tempera-ture there is a very large difference between ammonium laurate and palmitate-very much greater than that observed in potassium soap solutions. The total crystalloid in. ammonium laurate is nearly 50 per cent. greater than in potassium laurate whereas in ammonium laurate it is much less than in ammonium palmitate. An appreciable fraction of the total crystalloid will be undissociated ammonia. Further investigation would evidently be well repaid particu-larly if pure fatty acids wer0 used for making the solutions and hydrolysis were avoided as suggested above.Enough has been presented to show that the relationships agree with our general conception of colloidal electrolytes. General Co~nclusim.~ and Summaq. The theory of colloidal electrolytes defined by one of us receives further confirmation from the measurements of conductivity freez-ing point and vapour pressure of soap solutions at the ordinary temperatures here communicated. The general theory is based on the conception justified on mechanical grounds that ions may be aggregated to form the nucleus of a colloidal particle termed the ionic micelle whilst retaining their equivalent electrical charges. Thus the ionic micelle exhibits conductivity as well as mobility even greater than the ions contained in it 1300 McRAIN AND TAYLOR THE DEGREE OF HYDRATION ETO. In the case of soap solutions such as those of potassium oleate, the ionic mioelle in concentrated solutions exhibits an equivalent conductivity three times greater than the oleate ion and equal to that of the potassion. Its general formula may be taken as (KOl) . (01):"). (H20b; the amount of water of hydration involved being least in concentrated solutions and a t higher temperatures. The change in hydration explains the abnormal temperature-coefficient of conductivity. The formation of colloid in the case of higher soaps is so com-plek a t the ordinary temperature that in all ordinary solutions the only other constituent is the potassion sodion or ammonion as the case may be. I n dilute solutions the soaps become simple crystalloidal salts and hydrolysis becomes appreciable. No other representative of this very numerous and important group of substances has yet been completely investigated but the data available for such instances as the silicates tellurates dyes, proteins salts of alkaloids gelatin or casein etc. agree with the requirements of the theory which affords a fresh interpretation and rec~nciliation of the results. In conclusion we desire to express our thanks to the Colston Society of the University of Bristol for substantial grants towards the purchase of materials and apparatus. CHEMICAL DEPARTMENT; THE UNIVEBEIITP BRISTOL. [Receitwd September 17th 1919.

ISSN:0368-1645    

DOI:10.1039/CT9191501279

出版商:RSC    

年代:1919

数据来源: RSC

摘要:

1300 McRAIN AND TAYLOR THE DEGREE OF HYDRATION ETO. CXX.-The Degree which Form the Determined in Salting Out. qf Hydmtion of the Particles Strwctural Basis oj' Soap Cwd, Experiments ori Sorption and By JAMES WILLIAM MCBAIN and MILLICENT TAYLOR. IN 1911 w0 published experiments on the salting out of sodium palmitate by sodium hydroxide thO results of which appeased t o us highly remarkable (Zeitsch. physikal. Chem. 1911 76 199 table IX). When the sodium palmitate was salted out the clear soh-tion that remained as mother liquor or lye contained a higher concentration of sodium hydroxide than before. The explanations advanced were either that the curd which was salted out contained some acid sodium palmitate or that ver MCBAIN AND TAYLOR THE DEGREE OF HYDRATION ETC.1301 pronounced negative sorption had occurred; and further experi-ments to decide this were outlined. It is the object of this cominunication t o present the results of the first of a series of investigations which show that salted out sodium palmitate is hydrated and that the curd as a whole con-sists of this hydrated solid together with entangled or enmeshed mother liquor. It is impossible to elucidate this problem merely by the direct analysis of the curd since the curd contains a quantity of concen-trated lye mechanically entangled and yet cannot be washed. If acid soap were present it would nevertheless be entirely masked by the large amount of free alkali in the solution clinging to it. Methods had to be devised for distinguishing the water chemicaily or physically combined (sorbed) from that of the enmeshed solution thus affording a knowledge of the composition of the curd itself.I n carrying out the present experiments the concentration of the lye before and after salting out was determined and this was sup-plemented by a complete analysis of the wet curds. The calculation is based upon the assumption that the sodium hydroxide is not appreciably sorbed by the hydrated curd. As will be shown this leads to values which are only slightly less than the true degree of hydration. Methods of preparation employed and precautions observed were those previously described. The Method of Calcddion of Hydratiom from Analysis of Lyes. I f a system is made u p by taking 1 mol. of sodium palmitate (Nap) and 1000 grams of water and 2 mols.of sodium hydroxide (thus 2.000 weight-normal N,) and the curd separates out quanti-tatively with say a composition corresponding with NaP,5&0, 5 mols. of water will have been abstracted from the lye. The removal of 5 x 18-02=90*1 grams of water from the 1000 grams originally present will have resulted in an increase of the weight-normality of the solution of sodium hydroxide from the value N=2-000 to a normality N' where 1000 N* N' = -1000- 90.1 Hence in general if 14' is the weight of water abstracted, W = 1000 1 - - grams. ( 3 If the weighbnormality of the original sodium palmitate had bee 1302 MCBAIN AND TAYLOR THE DEGREE OB HYDRATION ETO. N, the hydrate or sorbed water abstracted by N mols. of ~odiwn palmitate would have bwn .mols. of water, 18.02 x N8 (1 - 5) mols. of water to one of Nap. 1000 18-02 x Ns The rmiduad sodium palmitate i n solution in the lye amounted to only a few ten-thousandths normal which does not affect the results. Actually in most of the experiments pure palmitic acid was weighed into a large silver tube and a measured volume of a concen-trated solution of sodium hydroxide was added. After sealing with pure silver the tube was placed in a thermostat a t about 90° and shaken during periods ranging from ten days to three months. The contents were then filtered through perforated silver foil a t 90°. I n carrying out the calculation outlined above the water formed by the chemical reaction between the palmitic acid and the sodium hydroxide wag taken into account and volume was converted into weight-normality .Hydration Results from Andysis of Lyes. Table I presents the data of ten experiments in which sodium palmitate was salted out by sodium hydroxide. The first column contains the number of the experiment for further reference; the second the total amount in mols. of palmitate in 1000 grams of water calculated as sodium palmitate; the third the amount of sodium hydroxide in excess; the fourth the time allowed f o r inter-action; the fifth the weight-normality of the sodium hydroxide in the lye after the experiment; the sixth the amount of sorbed water to 1 mol. of sodium palmitate. In experiment 3 (a) previously prepared pure sodium palmitate was added t o the hydroxide; in all other cases palmitic acid was added direct.TABLE I. Hydration of Sodium Palmitate Curds at 90°. Original charge. - Hydration : No.of Nap. NaOH. Time Lye,NaOH mols.KO expt. N,. N . shaken. N'. t o 1 NaP. la 0.9823 2.9470 10 days 3.119 3.1 10 0.9930 2.9816 4 weeko 3.2189 3.7 Mem hydration= 8-4 H,O MCBAIN AND TAYLOR THE DE~QREE OF HYDRATION ETC. 1303 No. of expt. 1 3 6 7 3a 2a 11 12 TABLE I. (continued). Hydration of Sodium Palmitate Curds at 90°. Original charge. N&P. NaOH. N,. N. 0.9608 1.9216 0.9608 1.9216 0.9337 1.86713 0.4949 1.7919 P 045021 1.6066 0.496 1.485 0.6002 1.6006 0.6146 1.689 Time shaken. 4 months 3 10 we&s 10 8s Hydration : E:"N%0 Lye NaOH N . 2.084 4.8; 2.1004 4.9 2.042 6.0 1.898 6.3 Mean hydration = 6.2 H,O.1.611 7.2 1.696 8.0 1.566 4.7 1-686 6.2 Mean hydration = 6-6 H,O. The resulits recorded in table I show clearly that there is a definite amount of combined water in each case and that this depende on the concentration of sodium hydroxide present in equilibrium with the curd. Thus for 3.0 1-9 and 1'5N-sodium hydroxide solutions the hydration is 3.4 5.2 and 6.5 mols. of water to 1 mol. of sodium palmitate. In other words halving the concentration of the lye has doubled the degree of hydration. Of course i E some sodium hydroxide also is combined in the curd, the above numbers have to be slightly increased t o give the true hydration valuea. The result foand for curd in the presence of 1.9N-sodium hydr-oxide is confirmed by an analysis of the curd itself.One hundred grams of curd contained 57.66 grams of sodium palmitate 1.681 grams of sodium hydroxide and 40.66 grams of water; or to 1 mol. of sodium palmitab 0.2027 mol. of sodium hyd'roxide and 10.89 mols. of water. Since the lye wit8 shown to be 1.898N-sodium hydr-oxide 5-93 mols. of water are to be regarded as solvent (enmwhed lye). This leavea 5.0 mols. of water of hydration to 1 mol. of d l u m palmitate as compared with the mean result 6.2 mols., given in table I. Again it is of interest to note that this curd contained about equal amounts of enmeshed lye and combined water. Here the solution employed for salting out contained 2N-sodium chloride together with N / %sodium hydroxide and previously prepared sodium palmitate.The headings of the columns are as in table I exwpt that a Two further results are given in table 11 1304 McBAIN AND TAYLOR THE DEGREE 0.F HYDRATION NTC. column giving the original normality of sodium chloride has been inserted after the third. TABLE 11. Hydration of Sodium Palmitate at 90°. No. of Original Charge Time Lye Hydration expt. NEtP. NaOH NaCI. shaken. NaOH. molsH,O. 15 1.100 0.5069 2.000 2 weeks 0.6608 4.4 26 1.000 0.5069 2.000 14 days 0.6608 4.4 Mean hydration = 4.4 H,O. This result 4*4H20 to lNaP where the lye is 2.5N altogether, agrees excellently with the values in table I and would make it appear that the hydration of the curd may be governed more by the concentration of the lye than by the nature of the salt employed in salting oat.If this proves to be the case it will only be necessary to ascertain the molar concentration of soap lye in order to measure the hydration of commercial soap curds once a standard value for each type has been obtained. Results with sodium stearate are given in table 111 showing t.hat with 1.4N-sodium hydroxide the mean hydration is about 4.3H20, as compared with 6*5H,O for the palmitate. TABLE 111. Hydration of Sodium Steara,te Curd a t 90°. No. of Original Charge Time Lye Hydration expt. Na stearate. NaOH. shaken. NaOH. moh. H,O. 14 0.4679 1.404 48 ?? 1.469 6.3 13 0.4679 1.404 46 days 1.444 3.3 Mean hydration = 4.3 H,O. The Effect of great Pressure on the Hydration of Soap. I n order to leave no loophole for the alternative explanation of the phenomena here considered experiments were carried out in which the curd was subjected to a pressure of hundreds or thou-sands of atmospheres during filtration.The lye filtering through was of course concentrated sodium hydroxide and if the solid part of the curd really contained acid sodium palmitate and if the separation from the curd was carried far enough analysis of the residual curd must show a direct deficiency of alkali. However, we found that the curd was always alkaline no matter how extreme the pressure or how much lye had been forced out. The only ob-served effect of applying pressure to the curd whilst allowing the lye to escape was to diminish the amount of hydration. The press employed together with the thermostat in which it wa McBAIN AND TAYLOR THI DEGREE OF HYDRATION ETC.1305 fixd is sketched in cross-section in the figure. 7310 filter bed con-sieted essentially of sugar-carbon previously tr0aW with lye but arranged BO as to avoid contact with the soap curd as far as possible. The filtration was carried out in the stesl tube a fitted with a perforated steel filter disk b . On t.he top of this disk was a sheet of nickel gauze and above Preas e q l o y e d for diminbhing the hydration and imbibition of 8 0 q curd. that again a layer of perforated silver foil. The carbon bed C, some 1-1.5 cm. in depth rested on the foil and was covered with two more layers of the perforated silver foil. Before use the whole bed was moistened with lye and made thoroughly compact by pressure from the piston d.The soap curd from which the lye had been rapidly filtered, through a silver cone a t 90° was transferred to the filter and covered with two layers of perforated silver foil. Between this an 1306 McBAIN AND TAYLOR THE DEGREE OF HYDRATION ETC. the steel piaton was inserted a hollowed out thick-walled rubber washer of the form used in a Bramah press. Pressure was applied through the screw by means of a double set of levers about 80 cm. in length. The sample of curd for analysis was taken from the centre of the block of hard curd left in the press. I n satisfactory experiments it contained no carbon. Various methods were employed in the analysis but only the one foand most convenient is here described. The curd was dissolved in neutral boiled-out 80 per cent.alcohol containing phenolphthalein. The solution was titrated with N / 10-sulphuric acid the alcohol evaporated and excess of N-sul-phuric acid added. After decomposition of the soap had been com-pleted by alternate heating and cooling the solid palmitic acid was collected. The filtrate was titrated to ascertain the amount of sodium palmitate which ha’d been present. I n many cases the palmitic acid was determined directly by weighing or by titration in alcoholic solution. Water was always obtained by difference. The method of calculation was usually that. illustrated by the curd analysis accompanying table I above. This is termed “mixed” calcGlation in the table above. I n a few cases however the concentration of lye after salting out was not determined.Here the amount of lye and its concentration were calculated from the analysis of the curd by successive approxi-TABLE IV. Hydration o/ Curds after Great Pressure at 90°. Original charge. Curd composition : 100 grams of curd contain Method Hydration : No. of Nap. NaOH. \ of mols. H-0: expt. N,. N . Nap. NaOH. l o 0.994 4 0.997 1 0.961 2 0.961 3 0.961 6 0.934 7 0.495 Sodium stearate 13 0.468 14 0.468 2-98 83.644 1.99 77.580 1.92 (73.850 (74.424 1.92 (69.99 (67.23 (77.51 1.92 66.146 1-87 (68.404 (67-61 6 (81.913 1.79 73.804 1.404 70.682 1.404 72.916 0.899 1.154 1.286 1,383 1.021 1.306 1.153 1.235 1.329 1.314 2.084 0.785 1.227 0.905 H,O. calculition. 15.457 ‘‘ Mixed7’,’ 21.266 “ Curd 24.864) “ Mixed ” 24.193) 28.994) “ Curd ” 31-479) 21-342) 32.619 “ Mixed ” 30.207) “ Mixed ” 31.070) 16.003) 25.411 “ Mixed ” to 1N&-1.547 1-42 2.0 1-6 3.6 3.5 (1.3 pressed on tile) 4.3 3.3 3.4 - 1.8 (PIWSS~ on tile) 3.2 28.091 “ Mixed ” 1.6 26.179 “ Mixed ” 2.MCBAIN AND TAYLOR THE DEGREE OF HYDRATION ma. 1307 mations based on the hypothesis that all excess of hydroxide waa contained in the lye and not in the curd proper. As may be shown theoretically this leads to a perfectly defined result for the value of the hydration. The results for sodium palmitate salted out by sodium hydroxide are given in table IV. The pressure was different in each case, since it depended on the condition of the filter bed pressure being applied until the curd began to be forced through.(In one or two cases some carbon was mixed with the curd.) Several samples a t different levels were taken f o r analysis in some of the experiments. Table V gives similar data for the pressed curds of Experiments 16 and 16 of table 11. This we term the pure “ curd ” calculation. TABLE V. Curd composition : No. 100 grams contain expt. NaP. NaOH. NaCl. cdculetion. H,O to 1 NaP. h of ,- I Method of Hydration mols. 16 80.01 0.144 0-761 “Mixed” 2.2 16 80.43 0.160 0.922 “ Mixed ” 2.2 On examining the results of tables IV and V it appears that the very great pressure to which the curds were subjected while in con-tact with residual lye lowers the degree of hydration to about half of that of the unpressed curd.This is in accordance with the recognised property of colloids that the swelling pressure increases enormously as the solvent is removed although the pressures here employed fax transcend those hitherto investigated. The most effective way of dehydrating soap curd is to press it OL a porous tile thus making use of the swelling pressure of kaolin to abstract the water. Curds from Experiments 2 and 6 so treated and the composition calculated as before gave values for apparent hydration of 1‘3 and -1.8 mole. of water respectively; this appease to show that the residual lye had also been greatly concentrated in the process owing to hydroxide being left when the water wa13 absorbed. Summary . Marked negative sorption of sodium hydroxide occurs when soap is salted out by sodium hydroxide in concentrated solution. From the extent of this effect the amount of water contained in the solid part of the curd as distinguished from the entangled lye clinging to it may be deduced. Soap curd is thus shown to be a mechanical mixture of hydrate (or eorption compound) and enmeshed lye. VOL. cxv. 31 130s RAY QUHA AND DAS : The degree of hydration varies with the concentration of lye as followa : Percentage of Lye. Sorption oompound. Fatty acid. 3*ON NaP,3*2H20 76.28 2.8 NaP,4*4H20 71.67 1.0 NaP,6.2H,O 68.89 1.6 NRP 6-6HL0 64.81 Extreme pressure lowars the degree of hydration considerably. I n conclusion we have pleasure in thanking the Colston Society of the University of Bristol for a generous grant towards the purchase of materials and apparatus. T m CHEMIOAL DEPARTMENT, BBISTOL UNIVERSITY. [Received September 171h 19 19.

ISSN:0368-1645    

DOI:10.1039/CT9191501300

出版商:RSC    

年代:1919

数据来源: RSC

摘要:

130s RAY QUHA AND DAS : CXX-Reaction of the Potassium Salts of 2-Thiol 8-thio-4-phenyl-4 5-dihydro-I 3 4-thiodiaxole and 2 5-Dithiol-1 3 4-thiodiaxole with Halogenated Organic Compounds. By PRAFULLA CHANDRA RAY PRAFULLA CHANDRA GUHA and RADHA KISHEN DAS. GABRIEL (Ber. 1877 10 185) and Holmberg (Ber. 1907 40, 1740) treated sodium ethylmercaptide with chloroform and obhinwl ethyl orthotrithiof ormate CH(SEt),. In the present investigation it will be shown that the potassium atom of the complex cyclic mercaptides named in the title is singularly reactive tawards the halogen atoms of organic compounds of divers types. Thus chloroform bromoform and iodoform yield compounds wit% the potassium monomercaptide which may be represented by the general equation 3RK + CHX,= 3KX + CHR, where X represents a halogen atom and R the radicle of the complex mercaptide.Chloropicrin acts exactly like chloroform but is far more reactive thm the latter and the reaction may be expressed by the equation 3B3 + N02*CC1 = N02*CR3 + 3KC1. The behaviour of tribrornoresorcinol beiizylidene chloride mono-&loroawtic acid and ethylene bromide has also been found to be of an identical nature. The potassium salt of 2 5-dihhiol-1 3 4 REAOTION 0% THZ POTASSR?M SALTS ETC. 1309 tbiodiazole on the other hand acts almost as an inert substance towards chloro- brome and iodo-form. It is evident that the presence of the two -SK groups of the dimercaptide exercises a sort of inhibitory influence on the halogen atoms The reactivity of these halogenated compounds can how-ever be materially enhanced by substituting the remaining hydrogen atom by a nitro-Goup.For instance chloropicrin acts very readily on the dimercaptide even in the cold. The introduc-tion of an additional negative group has thus a marked effect. The reaction may be represented as follows: 3 2-4 +2*CCI,*NO = KSC CSK I - 1 I I NO ON0 Nitrous fumes escape during the reaction and an atom of oxygen, as shown above forms thO connecting link between the two carbon atoms of the residuee of two molecules of chloropicrin. Ethyleae bromide acts on the potassium monomercaptide; only one atom of bromide combinee with the potassium atom resulting in the formation of the compound, M onochl or oa =ti c acid b enzy lid en e chloride and ethylene bromide no doubt act on the dimercaptide but the products of the reaction are insoluble in the ordinary solvents and thus cannot be purified.E X P E R I Y E NTAL. thb&aaole and Idofornt Bromof orm and Chloroform. Three molecular proportions of the mercaptide were treated with one mdecular proportion of the halogenated compound in alcoholic solution the mixture being boiled under reflux on the water-bath €or several hours. An insoluble mass was obtained consisting of the potassium haloid and the organic derivative. The solution pas Potmsium Salt of 2-Thiol-5-thi(l.-4-ph~n~l-~ 5-dihydro-l 3 4-3 D 1310 d Y QUHA AND DA8: allowed to cool and then triturated with water. The aquems filtrate on evaporation gave crystals of the potassium haloid. The insoluble portion was dried and dissolved in benzene; on evapora-tion of the solvent an oily liquid was obtained which was redis-solved in benzene and precipitated by alcohol as an oil.On keep ing the oil solidified t o a yellow powder. As it was difficult to get rid of the last tram of iodoform the powder was repeatedly washed with alcohol and dried in the stSam-oven until the odour of iodoform was no longer perceptible. The substance melted at With bromoform exactly the same method was followed but as it is highly volatile the excess was easily removed from the yellow compound which meIted a t 66-68O. When the reaction mixture in alcoholic solution was heated at !210-220° in a sealed tube for several hours a tarry resinous IJILUS was obt'ained which was collected and dissolved in benzene.Addition of alcohol to the benzene solution gave the same tarry precipitate but not the yellow powder. The alcoholic mother liquor on concentration gave shining needleshaped crystals which melted sharply a t 62O. Analysis proved this product to be the alcoholate of the compound described above having the formula CHR,,2EtOH. The yield was very poor most of the product having evidently become resinified owing to the high temperature employed. No reaction took place when chloroform was heated under reflux with the mercaptide in alcoholic solution. The mixture was therefore heated in a sealed tube as in the case of bromoform. The filtrate separated from the resinifid mass, gave on evaporation crystals of the alcoholate melting a t 62O: 0.1520 gave 0.2532 CO,.C=45.23. 0.1101 , 11.4 C.C. N at 32O and 760 mm. N=11*26. 0.1260 , 0.3532 BaSO,. S=38.48.* C,H,,NBS,,2EtOH requires C = 44.6 ; N = 10.85 ; S = 37-50 per cent. It is thus evident that iodoform and bromoform are more reactive towards the mercaptide than is chloroform. Compound m. p. 66-68O from iodoform : 66-68'. * It may be necessary to point out that in these compounds the sulphur atom is linked both to an aliphatic and to an aromatic and cyclic residue. Carius's method gave an unusually low result owing to the formation of sulphonic acid. The excess of nitric acid ww therefore neutralised with sodium carbonate and evaporated to dryness and fused in a silver dish. The product waa evaporated several times with hydrochloric acid before the addition of barium chloride.Owing t40 the presence of a large amount of sodium chloride and unchanged sodium nitrate the result is generally a little too,high BEACTION OF THE POTASSIUM SALTS ETC 1311 0.0953 gave 0.1502 CO and 0.0202 H,O. 0.0876 , 0.2087 BaSO,. S=41.19. 0.075 , 8.3 C.C. N2 a t 24O and 760 mm. N=12%0. C2,H,,N,8 requires C=43.60; H=2*76; 8=41.80; N=12*21 per cent. The compound from bromoform gave C=43*11; H=2-03; C=42*98; H=2-33* N = 12.59 per cent. Potassium Salt of the Mercaptart and Chloropicrin. The alcoholic solution of the parent substances was heated a t 50° under reflux as before. A bulky yellow precipitate was obtained; after decanting off the mother liquor it was washed with alcohol and triturated with water t o remove potassium chloride.It crystallised from hot benzene in shining yellow crystals melting at 128-129O. The reaction was almost quantitative as was proved by weighing the potassium chloride formed : 0-1140 gave 0.1727 CO and 0-0307 H,O. 0.0948 , 11.4 C.C. N a t 3 2 O and 760 mm. N=13-18. C25H1502N7S9 requires C = 40.97 ; H = 2.05 ; N = 13-37 per cent. C=41-31; H=2*99. Potassium Salt of the Mercaptun and Tribromoresorcinol. The substancw were heated in alcoholic solution as before. The amorphous powder obtained was freed from potassium bromide by water dried and dissolved in a mixture of alcohol and carbon disulphide ; on evaporation shining crystals melting a t 166O were obtained : 0.0881 gave 0.1458 CO and 0.0236 H,O. 0.0724 , 0.1897 BaSO,. S=36.00.C30H,602N8S9 requires C = 46.03 ; H = 2.30 ; S = 36.55 per cent. C=45*13; H=2.97. Potassium Salt of the Mercaptan and Benzylidene Chloride. The components in alcoholic solution were heated under reflux 0.0708 gave 0.1321 CO and 0.0193 H20. 0.0849 , 8.6 C.C. N a t 22O and 760 mm. N=11*02. 0.0974 , 0.2550 BaSO,. S=35*83. C23H,,N4S requires C=51.11; H=2*96; N=10.40; S=35.55 per cent. for about an hour. The crystalline product melted a t 59-62O: C=50.87; H=3.03 1312 REACTION OF THE POTASSIUM SALTS ETC. Potassium Salt of the Mercaptan and Ethylene Bromide. After heating in alcoholic solution a8 usual the insoluble pro-duct was freed from potassium bromide by means of water dried, and dissolved in ether. On evaporation shining crystals melting a t 9 4 O were obtained: 0.1644 gave 0.2183 CO,.C=36-21. 0.0812 , 6.6 C.C. N a t 30° and 760 nun. N=8*97. 0.1303 , 0.0715 AgBr. Br=23.35. C,,H,N2BrS3 requires C = 36.03 ; N = 8-40 ; Br = 24.02 per cent. Potassium Salt of the Mercaptan and Monochloroacetic A cid. On mixing the components in aqueous solution the reaction takes The product place even in the cold and is completed on heating. crystallises from boiling water in silky needles melting at 145O : 0-1279 gave 0.1935 COz and 0.0414 H,O. 0-1697 , 15-00 C.C. N a t 2 4 O and 760 mm. N=9.99. 0.1580 , 0.3495 BaSO,. S=30*38. C,,H,02N,S3 requires C =42*06 ; H = 2.81 ; N = 9.86 ; S = 33.81 per cent. It is of interest to note that whilst monochloroacetic acid behaves in the above manner dichloro- and trichloro-acetic acids on the other hand behave like strong acids such as hydrochloric acid that is they simply regenerate the original mercaptan.C=41*26; H=3.60. Potassium Salt of 2 5-Dithiol-l 3 ; 4-thiodiazole and Chlwopicrin. The components were heated under reflux in alcoholic solution for several hours. Nitrous fumes were evolved and a yellow pre-cipitate was obtained. On cooling this was collected and washed with alcohol to remove adhering chloropicrin arid then triturated with water to move potassium chloride. The yellow powder was insoluble in nearly a dozen ordinary solvents and in mixtures of some of them. Two different preparations had however the same melting point (166-168O) and the same percentage composition : 0.1734 gave 0.1274 CO,. C=20*04. 0-0853 , 14.00 C.C. N a t 31° and 760 mm. N=18.03. 0.0899 , 0.3907 BaSO,. 5-59-68. C80N,S9 requires C=19.83; N=17.40; S=59*50 per cent. C H ~ c AL LAB OBATORY, COLLEQE OF SCIENCE, UNIVEMITY OF CALCTJTTA. [Recehved October 3184 1918.

ISSN:0368-1645    

DOI:10.1039/CT9191501308

出版商:RSC    

年代:1919

数据来源: RSC