摘要:

600 BOUSFIELD : IONISATLON AND THELX.-Ionisation and the Law of Mass Action.Part II. The Osmotic Data in Relation toCombined Water.By WILLIAM ROBERT BOUSFIELD.IN the first paper (T., 1913, 103, 307) it was found that the trueionisatdon law a t very high dilutions for electrolytes of the classt o which potassium chloride belongs was accurately(1 - a) = h-4 x ConstantLAW OF MASS ACTION. PART 11. 601as contrasted with the accurate law for weak electrolytes at veryhigh dilution, which is(1 - a) = h-1 x Constant.h is the total number of molecules of water per molecule ofsolute; and if n be the number of water molecules combined withthe solute, then at decinormal dilution, where the value of h isabout 555, the difference between h, which expresses the total water,and (h-n), which expresses the free water, may be of the orderof 1 to 3 per cent.In the same region of dilution the possibleerror in the ionisation-coefficients is of about the same order. Toget ionisationcoefficients correct t o a second approximation, saywithin 4 per cent., it is necessary to take this combined water intoaccount. I f ionisation is calculated from conductivity, it has tobe taken into account by considering the variation of mobilitywith the changing size of the hydrated ions, as has been done informer papers.” We now propose to inquire how the combinedwater comes into the reckoning when we approach the matter fromthe point of view of the osmotic data, and attempt to use these forchecking the ionisation values. The best data for this purpose areMorse and Fraser’s latest values for the osmotic pressure of sucrosesolutions, which are fully set out in Findlay’s excellent mono-It so happens that Morse and Fraser’s original set of values forthe osmotic pressure of sucrose solutions (Amer.Chem. J., 1905,34, 1) in the neighbourhood of 18O, very fairly satisfied the gasequation when V was taken as the total volume of water containing1 gram-molecule, which led to the view that the combined waterneed not be taken into account. The later values published byMorse and his ceworkers (Morse, Holland, Zies, Myers, Clark, andGill, Amer. Chem. J., 1911, 45, 602), however, which are believedby them to be more accurate, do not give the same result. Findlayhas shown (Zoc. cit., p.41) that by assuming a constant hydrationof sugar with 5 molecules of water, the new values approximatelysatisfy the gas equation in a modified form which is based onthermodynamical considerations. It is of interest to note that ina still more recent paper (Morse, Holland, Myers, Cash, and Zinn,Smer. Chem. J., 1912, 48, 29), Morse and his co-workers concludethat as a temperature of 80° is approached the osmotic pressuresconform to the gas law without taking into consideration anyquestion of water combination. Possibly a t this temperature the* See papers by the author with reference to the sizes of ions, namely : Zeilsch.physikal. Chem., 1905, 53, 257; Phil. Trans., 1906, A, 206, 101 ; Proc. Roy.SOC., 1913, A . , 88, 147.graph.i-f “ Osmotic Pressure,” by A. Findlny : Longmans Green & Co602 BOUSFIELD : IONISATION AND THEcombined water is reduced to so small an amount that it may beneglected without much loss of accuracy. For our purpose, how-ever, it is important to consider osmotic pressures in the neigh-bourhood of 18O, where plenty of conductivity data are available,a t which temperature it is known that sucrose is combined withfrom 5 to 7 molecules of water (Philip, T., 1907, 91, 711).The modified form of the gas equation on which Findlay baseshis calculations, which is deduced from thermodynamical considera-tions, presents no advantages as regards accurate expression of theresults over the simpler form that will be adopted. The idealisa-tion of the solution for thermodynamical consideration involvesa departure from actuality which has t o be corrected.From thepoint of view of the hydrate theory, the '( ideal " process of solutionis one in which a portion of the solvent seizes upon and combineswith the solute with it consequent evolution of heat and contractionof volume which are often large. We are driven from thermo-dynamical t o empirical considerations when we desire to get beyonda first approximation. Especially is this true when we are in-quiring whether the hydration of the solute proceeds progressively,as the change of mobility derived from conductivity measurementswould lead us to believe. It will be seen that a very simplemodification of the gas equation gives expression to the osmoticresults with an accuracy which is well within the limits of experi-mental error, and.gives support to the view that the hydration isprogressive.The gas equation in the ordinary form isIf N (as usual) expresses the number of mole. of solute per1000 grams of water, the weight of water (in kilograms) per mol.of solute is .1jA7. It is more convenient to base our computationon h, the total hydration of the solute. We have h=- -- -- , wheree=18'016, the weight of a gram-molecule of water. If for V , thehe volume of the solution in litres, we substitute -, which is the 1 oouweight of water in kilograms per mol. of solute, the gas equationas applied to osmotic pressure becomes1000Ne(2). - -- A a - R . . . , . . .1' . 1 oouIf we take the weight of free water instead of the total water,the equation becomeLAW OF MASS ACTION.PART 11. 603Or, if we put1000 - 0.0821 x 1000 ~ 4.657, If= n x --e 18.016the equation becomesP(?h-n)=R'T . . . . . . , (4).We shall find that this equation gives values of n which are inaccord with f reezing-point determinations within the limits ofexperimental error.Let us first see what values of R we get from equations (l), (2),and (3), for the last taking n a constant and equal to 6. I ntable I are set out the osm?tic pressures of sucrose for 20°, to-gether with the values of N and h, and the densities of sucrosesolutions a t 20°, which are required to work out V , the volume ofthe solution in iitres containing 1 molecular proportion of solute.There are set out in the table for comparison three sets of valuesof R, obtained from equations (l), (2), and (3) above.Now if, asis sometimes loosely stated, the simple gas equation applies toosmotic pressures, the values of P V / T should all be equal to theTABLE I.0.10.20.30.40.50.60.70.80.91.02.5905.0647.60510.13712.74815.38818.12820.90823.71726,638555.06277.53185.02138.7611 1.0192.5179-2969.3861-6755.511.011121.023471.035271.046601.057461,067891.077921.087561.096861.1057610-2285.2183-5502.71572.21471.88151-64321.46401-32491.21380.09040.09020.09210.09400.09640.09880-10170.10450.10720.1 1040-08840-08640-08650-08650.08700.08750.08840.08920.08990.09090.08740.08460.08370.08280-08230.08190-08170-08150-08120.0811gas constant R=0.0821.As a matter of fact, we get a series ofgradually increasing values, the last being about 20 per cent.greater than the first and 37 per cent. greater than the gas con-stant. If, instead of the volume of the solution, we take theweight in kilograms of total water containing 1 gram-molecule ofI' lcasolute, we get the improved set of figures set out under II - 1 l000'The values now only increase by about 6 per cent. from N = 0.3 toN=1*0, and the mean value of the constant is only about 6 percent. above the true value of the gas constant.If, however, we take the figures in the last column underP (h - 6)e , obtained on the assumption that n is constant and =6, ?loo0we see that there is now only a difference, of 3 per cent. betweenthe values for N = 0 - 2 and N=1*0.Moreover, the mean of thes604 BOTJSFIELD : IONISATION AND THEeight values is 0.820, which is very close to the gas constant. Weshall find that in the range from N=0.1 t o N=0*5, the value ofn is probably a little more than 6, whilst in the range from N=0*6to AT=l.O, it is a little less than 6.We will now test the relation by taking the accurate value ofthe gas constant R =0*0821 or R1=4.577, and seeing what valuesof 7~ are yielded by the equationP(h - n> = R'TTABLE 11.N .0-10.20.30.40.50.60-70.80.91-0h.555.06277-53185.02138.76111.0192.5179-2969-3861-6755-51P.2.5905.0647.60510.13712.74815.38818.12820-90523-71726.6381335P- -515.44263.63175.54131.70104.7286.7673.6463.8656.2950.12~- n,.39-6213-909-487-066.295.755.655-825-385.39n2'6-116.366.306-186-035-885.735.595-455.311335 *,* p x P'LO4.441-505.616-084.865.755-645.515-626.05521.17269-82181.73137.8411007492.6579.3869-4561-7455-42A t 20°, RlT=4.557 x 293 =1335, and the values of .n are givenbyThe values so obtained are set out in table I1 in the columnunder nl.It will be seen that in the more concentrated solutionsthey approximate closely to the accepted figure for the hydrationof the sucrose molecule in solution.I n the columns headed 9%and ng are given the values for n calculated in two different waysfrom freezing-point data. The methods will be explained furtheron. The accord is very close for the more concentrated solutions.*The agreement between the two sets of figures indicates thatin the more concentrated solutions, the osmotic pressures of sucrosesolutions obey, with considerable accuracy, the relationP(h - n> = R f l'.A t the more dilute end, however, the values of n yielded by theosmotic data are too high. Here a comparatively small error inthe value of P causes a large error in n, for example, if the valuegiven for P is 6 per cent. too great, this will make the value of nfor the highest dilution five or six times too great.I n the earlierpaper of Morse and Fraser (Zoc. c i t . ) , the mean value of eleven* I t i v intercsting to note that the values of 7~ found by Philip (loc. c d . ) were :h --. 95 44 38YL = 6.7 5 '9 5'LAW OF mss ACTION. PART Ir. 605determinations of the osmotic pressure at 18O for N=0*1 was 2.410.The present value is 7.5 per cent. larger. The old value wouldgive n= 1.2 for this concentration instead of 39.6. Furthermore,there is a small amount of ionisation in dilute sucrose solutionswhich would also tend to reduce the value of n.I n the last column are set out the somewhat remarkable set ofvalues for the expression1335y P20.Comparing them with the values of h we see that they are almostidentical, within the limih of experimental error, and they giveus another empirical form in which the osmotic law for sucrosemay be written, namely,P , = I??',Pto which reference will be made later.The Freezinppoint Data.It was shown in a previous paper (Phil. Trans., 1906, A , 206,149)by reference to potassium chloride solutions that, if we reckon themolecular depression of the freezing point on the free water in-stead of the total water, we get a constant value for the expression(5)9W~ -- - P .. . . . . AN( 1 + a) 1000where w is the uncombined water in 1000 prams of water. Since "w - k - 36 , putting F'=F E, we may write this ,Y=---- al;d ___- 1000hs 1000 /Iequation in the formwhich exhibits its analogy to the osmotic equation.Let us now show that this relation gives the same values for nas our osmotic equation, when applied to sugar, where i=l.Intable I11 are set out the values of A given by Morse and Fraser(Zoc. cit.). For comparative purposes it is best t o take the valueof F yielded by these figures.A(h-r&)=F'i,The best value which they give isF= 1.855.To reckon n, equation (5) may be writtenA h - n - - - 1.855,N h. . . . . . A (6). o606 BOUSFIELD : IONISATION AND THEN.0.10.20.30.40.50.60.70.80.91-0A.0.1870.3730.5740-7760.9701.1871.3981.6121.8372.082TABLE 111.h.555.06277.53185.02138.7611 1.0192-5179.2969.3861-6755.511 - 1*855N/0.00800.00540.03030.04380.04380.06220.07 110.07940.091 10.1090A.rind.4.441.505-616-084- 865.755- 645.515- 626- 05The values are given in the column n3. They are somewhatrough a t the dilute end, where accurate fqeezing-point values arealways difficult to obtain. A smoother series may be obtainedfrom an empirical relation which was discovered by Morse andFraser for sucrose, and has considerable interest for the presentinquiry. They found that their freezing-point values could beexpressed bywhere p is the density of the sugar solutions at their freezing points.A = 1- 85Np,The equation may be writtenA=& . . . . . . . . (7).Nand by comparing i t with th9 relat,:on expressed in equation (5)above, which becomes for sucroseA , h - n -‘V / IF, - --we see that we have as an empirical relation for sugar solutionsh-n-1 - --h p).. . . . . . or (8)-Let us apply this relation to calculate the values of n, whichis equivalent to using a series of Usingthe densities given in table I for 20°, we get the series of valuesfor 12 given under n2 in table 11.The close coincidence of the values of n reckoned from theosmotic equatioii and the freezing-point equation shows definitelythat the osmotic pressures satisfy the equationP(h - n> = R’Twith great accuracy for the more concentrated solutions, andwithin the limits of experimental error for the most dilute solutions.smoothed ” values for ALAW OF MASS ACTION. PART IT. 607W e now eee the reitson why they also satisfy the relationY.B=R'T.PIt is simply because.there exists in the case of sugar the relationgiven in equation (8)) which must for the present be regarded asan empirical relationship not capable of general application.The Vapour-pressure Relation.For the purpose of considering the relation of vapour-pressuredata to the foregoing results, we may take the well-known ap-proximate expressionP= p . s p . . . . . , , . (9).Uwhenceand equation (9) may beHence, if the osmoticHere 6p is the difference between the vapour pressure of thesolution and that of water, p is the mean density of the solution,and u the mean density of t.he vapour. Since saturated steamobeys the gas law in the neighbourhood of Oo to 20° we have, ifv be the volume of a molecule of steam in litres,1000v(r=e and pv=RT,writtenY=p- 8P .R'.Z'.Ppressure relation bethe vapour pressure relation takes the form:;(h - It) = i.Thus the density p comes into the general vapour-pressureI n the case of sugar, sincerelation.A - r). 1 -h 6we may write this expression asA- SP = 1,Pand if this were applicable to an electrolyte it would take the for608 IONISATION AND THE LAW OF MASS ACTION. PART 11.which is identical with thO relation originally propounded by van%Hoff, namely,;=2 8 .M -,P mwhere there are M mols. of solvent to m of solute.the equations to electrolytes that the density relationcannot be applied to eledrolybes. F o r a binary electrolyte, thelimiting value of i in dilute solution is 2, and in solutions ofmoderate concentration, where aatolytic conductivity is small, thevalue of i is much less than 2.The matter may be illustrated byfigures deduced from Tammann’s data for the vapour pressures oflithium chloride solutions,* from which, for the values of h givenbelow, we can deduce, in the neighbourhood of 1 8 O , the values ofGpjp and htspjp set out thereunder:It is clear, however, from a consideration of the application of(h - n) j h = 1 j ph= 7.278 10.40 16.25 29-196pJp=0-326 0.229 0.14 0.07hEp/p= 2-373 2.382 2.274 2.043Here we see that the values of i given by van’t Hoff’s equationare above 2 instead of below. The number of combined watermolecules in lithium chloride solutions increases rapidly with thedilution. Merely t o show the general result of bringing theseinto the account, and without any attempt a t getting the realvalues of n, let us take values for n which are approximately ofthe right order of magnitude:h=7-278 10.40 16-25 29.19n=4 5 7 9We then get:These figures, when multiplied by p, would give values for i ofapproximately the right order of magnitude.This short hypo-thetical reference is only introduced for the purpose of showingthat, in the application t o electrolytes of the osmotic equations asderived from sucrose, we must take them in the form which isindependent of the peculiar density relation which holds forsucrose. The consideration of the results of applying these equa-tions to electrolytes must be deferred to another occasion in con-junction with data from vapour-pressure experiments which havebeen in progress for some time, and are still incomplete.The result, then, a t which we arrive, is that the osmotic relationsQHQ Tmidolt aud Bomstein’s tables.(h-n)EpJp= 1-069 1.237 1.295 1.413THE DECOMPOSITION OF CARBAMIDE. 609as derived from a consideration of the sucrose data. may be putinto the form :whereI000 I000 R'=R--- and P'=F---,e eF and R being the ordinary freezing-pointconstants.TakingF= 1.86 and R =0*0821,and gas equationwe haveF'= 103.2 ; R'= 4.557.For sucrose, since we have the special relation ( h - n ) / h = l / p ,they may be put into a form which eliminates p in the vapour-pressure equation, but this is not the form in which we can applythem to concentrated solutions of electrolytes.ST. SWITHINS,HENDON, N.W

ISSN:0368-1645    

DOI:10.1039/CT9140500600

出版商:RSC    

年代:1914

数据来源: RSC

摘要:

THE DECOMPOSITION OF CARBAMIDE.LXL-The Decomposition of Carbamide.By GEORGE JOSEPH BURROWS and CHARLES EDWARD FAWSITT.A PAPER on this subject has already been published by one of US(Zeitsch. physikal. Chem., 1902, 41, 603), in which the decomposi-tion of carbamide by water alone, and by aqueous solutions ofacids and alkalis, was studied.It was shown then that whilst carbamide is decomposed in thesecircumstances to ammonia and carbon dioxide, the mechanism ofthe decomposition is not that of a simple hydrolysis. It wasshown, further, that the experimental results could be explainedsatisfactorily if the decomposition of carbamide is regarded as atransformation into ammonium cyanate, which is then convertedinto ammonia and carbon dioxide.In the period that has elapsed since the publication of this paper,other papers have been published, which have made us considermore attentively the mechanism of the decomposition of carbamide.We thought that some experimenh on the decomposition ofcarbamide in water-alcohol solutions might throw further light onthe reactions involved, and the results of some experiments on thedecomposition of carbamide in the presence of alcohol are recorde610 BURROWS AND FAWSITT :in the present paper.Whilst the results are not exactly what weanticipated, they are, we believe, a further proof of the theorythat the decomposition of carbamide by water and acids is not ahydrolysis, but a decomposition through ammonium cyanate.I n the tables given below, the results of experiments are givenin which the decomposition of carbamide was carried out in water-alcohol solutions.Experiments were carried out with and without addition of acid(hydrochloric).The method of following the rate of action wasthe same as that formerly given (Zoc. cit.), namely, titration of theammonia produced.Whenno acid is added, the method only gives approximately correctresults; a small part of the decomposed carbamide in this case ispresent as ammonium cyanate, and titration of ammonium cyanatewith acids results in the escape of varying amounts of undecomposedcyanic acid. The numbers here given for solutions to which noacid has been added are therefore probably correct only to about5 per cent. The numbers given for acid solutions are probablycorrect to about 1 per cent.I n preparing solutions which contain both alcohol and water, itmay be noticed that there are several ways in which the percentageof alcohol can be expressed.The following method has here been adopted to make a solutioncontaining x per cent.of alcohol: x volumes of absolute alcoholare mixed with 100 - x volumes of water. This mixture is then usedto dissolve the carbamide or other substance required.Aqueous alcohol containing as much as 90 per cent. of alcoholwas used in these experiments, but no investigations were madewith alcohol containing no water (100 per cent.).Ordinary absolute alcohol (99-33 per cent.) was used in makingup the solutions.In the details of experiments given here:t refers to the time in minutes measured from the beginning ofA is a measure of the concentration of carbamide a t thex is a measure of the concentration of the decomposed carbamideThis method is entirely satisfactory for acid solutions.the experiment.beginning.a t time t.1 At A - x ' k=- log*, -Most experiments were made in duplicate, and the mean value( A - x ) ~ was taken in calculating the constant k.THE DECOMPOSITION OF CARBAMTDE.611TABLE I.N / 10-Carbamide in water. Temperature, 71'25O.t. ( A - 4 1 * ( A - 4 2 . (A-4,.0 12.05 12.05 12.052760 11-27 11.27 11.275760 10.72 10.82 10.779960 10.27 10.27 10-2717280 9-60 9.60 9.6022920 9.00 9.00 9.00TABLE 11.N / 10-Carbamide in 40 per cent. alcohol. Temperature, 71.25O.t. ( A - 4 1 .(A-X),. (A -@m.0 12.05 12-05 12-056700 11.15 11-16 11.1511460 10.90 10.96 10.9220100 10.35 10.35 10.3523040 10.15 10.16 10.15The addition of alcohol diminishes the velocity of decompositionof the carbamide. Comparing the time occupied in the decomposi-tion of oneseventh of th'e carbamide in tables I and 11: for waterthe value is 9430 minutes; for 40 per cent. alcohol the number is20,300 minutks.The effect of alcohol in decreasing the velocity of the reaction isnot exactly the result that we expected. Addition of alcohol some-times decreases and sometimes increases the velocity of reactionsas measured for aqueous solutions, but it more usually increases thevelocity. In particular, it increases the velocity of transformationof ammonium cyanate into carbamide (Walker and Kay, T., 1897,71, 5041, and the decomposition of carbamide is closely related toits formation.The action of acids on carbamide is a somewhat simpler reactionto study practically than the decomposition of carbamids startingwith a neutral solution, as the ammonium cyanate formed isimmediately decomposed by the acid.The only detectable pro-ducts of decomposition in this case are ammonia and carbondioxide.The results for the action of acids on carbamide are given below.As one of us has already shown that the free carbamide in asolution of carbamide hydrochloride is the active carbamide in thedecomposition, experiments were made to discover whether additionof alcohol to a solution of carbamide and hydrochloric acid hadany marked effect on the hydrolysis of the carbamide hydrochloride.The amount of free acid was determined by the effect of thesolution on the hydrolytic decomposition of sucrose a t 40°.s 612 BURROW8 AND FAWSITTTABLE 111.lV/lO-Hydrochloric acid and 10 per cent.of sucrose in water.As before,t.010203650641 A k = jloglo - . A - xA-X. b.58.0 -54.8 0.0024661.5 0.0025846.5 0.0026642.6 0.0026839.0 0-00269Mean.. . . . .0.00265TABLE IIIa.N / 10-Hydrochloric acid + 10 peramide in water.t.0457486136170210A-X.73.756.647.441-132.626.621.1cent. of sucrose + M / 20-carb-k.0-002550.002590.002590.002610.002600.00269-Mean., . . . .0-00269From these figures the concentration of hydrogen ion in thesolution containing M/ZO-carbamide is 0.98 of that in the solutioncontaining no carbamide.TABLE IV.N/10-Hydrochloric acid and 10 per cent.of sucrose in 25 percent. alcohol.i0306080101120166A-X.68.767-943-037.933.926.148.1Mean.....k.0.002480.002680.002640.002660.002560-00265.0*00256THE DECOMPOSITION OF CAH3AMIbB. 613TABLE IVa.N / 10-Hydrochloric acid + MI20-carbamide with 10 per cent, ofsucrose in 25 per cent. alcohol.t.020457590150192A-X.68.861-353.344-741.029.122.8Mean..k. -0.002510.002460.002500.002500.002490.00250, . . .0.00250The concentration of hydrogen ion in the solution containingM/2O-carbamide is 0.98 of that in the solution without carbamide.TABLE V.N / 10-Hydrochloric acid with 10cent.alcohol.t.03650637896109134157182A-X.66.054.450.446.543.039.036.031-127-424.0per cent. of sucrose in 50 perk.0.002150.002210.002310.002300.002310.002360.002390.002390.00238Mean.. . . . .0.00230-TABLE Va.N/lO-Hydrochloric acid + X/ZO-carbamide with 10 per cent. ofsucrose in 50 per cent. alcohol.t.06276110134150175191227A-x k.64-6 -47.3 0.002 1844.9 0.0021 137.0 0-0022032.7 0.0022130.0 0.0022226.5 0.0022124.3 0.0022220.3 0.00222Mean ...... 0.00220The ratio of hydrogen ion in these two experiments is 0.956.s s 614 BURROWS AND FAWSITT :TABLE VI.N / 1 O-Hydrochloriccent.alcohol.t.041627392102110134150164acid with 10 per cent. of sucrose in 75 perA-X. k.62.9 -48.8 0.0026942.4 0.0027639.5 0.0027737-0 0.0025134.7 0.0025330.5 0.0028626-5 0.0028023.2 0.0028921.5 0.00284Mean.. ... .O-00274TABLE VIa.N l 10-Hydrochloric acid + M / 20-carbamide with 10 per cent. ofsucrose in 75 per cent. alcohol.t.02237666584101125148163177190A-X.62.255.350.945.143.037.834.429.825.622.621.119-8E.0.002330.00236;0.002490.002470.0025'70.002550.002560*002600.002700.002650.00262-Mean.. ... .0-00268The ratio of free hydrogen ion in these two experiments is 0-941.These experiments show that alcohol, in the concentrations used,has very little effect on the hydrolysis of the carbamide hydro-chloride.The next experimental work was on the velocity of decompositionof carbamide in acid solution containing alcohol.These experiments were carried out a t three temperatures,namely, 98.2O, 71*25O, and 61'05O.The results are given in the following tablesTHE DECOMPOSITION OF CARBAMIDE.615TABLE VII.N / 10-Carbamide + N/10-hydrochloric acid in water. Tempera-ture, 98-2O.t. (A-41- ( A - 4 2 - ( A - 4 m . Kale.0 19.70 19.80 19.75 -60 17.58 17-48 17.53 84.3141 15-10 15-10 15.10 82.7170 14.28 14-20 1 4.24 83.6268 11.80 11-90 11.85 84.6479 7.80 7.90 7.85 83.7601 6.30 6-30 6.30 82.6Mean ...... 83.6TABLE VIII.N / 10-Carbamide + N / 10-hydrochloric acid in 10 per cent.alcoholTemperature, 95-2O.t. ( A - 4 , . (A-x),. ( A - 4 m . K. 105.0 20.00 20.00 20.00 -71 17.55 17.60 17.58 78.9171 14-60 14-60 14.60 79.9261 12.44 12.45 12.45 78.9374 10.30 10-30 10.30 77-0576 7.25 7.20 7.23 76.7Mean ...... 78.3TABLE IX.N / 10-Carbamide + N / 10-hydrochloric acid in 20 per cent. alcohol.fernperattire, 98.2O.t.07514123036353 1636( A -4,. ( A - x ) ~ .20-28 20.2517-85 17-8015.92 15.85- 13.6010-98 11-008.27 8.336-95 7.05(A-4,n. K.105.20.27 -17.83 74.315-89 75.113-60 75.410.99 73.28-30 73.07.00 72.6Mean ...... 73.9TABLE X.N / 10-Carbamide + N / 10-hydrochloric acid in 40 per cent. alcohol.Temperature, 98.2O.t.070182307424544666(A-41.20.3018-3815-6012-9010.959-458.22( A -4*-20.4018.4015.6012.9411-059.458.20( A - 4 m *20.3518.3815-6012-9211.009.458.21Mean..R.105.63.163.564.363.161-269.1 .....62-4616 BURROWS AND FAWSITT :TABLE XI.N / 10-Carbamide + N / 10-hydrochloric acid in water. Tempern-ture, 71'25O.t. (A--a),. ( A - x)?. ( A - x ) m . I<-105.0 13.35 13.40 13.38 -2818 11.17 11.20 11.19 2-784718 9.80 9.90 9.85 2.828596 7.70 7.70 7.70 2.799223 7.50 7-50 7.50 2.73Meen......2-77TABLE XII.N / 10-Carbamide + N / 10-hydrochloric acid in 10 per cent. alcohol.Temperature, 7 1 -2 5 O.t. ( A - 4 , . ( A - 4 2 . ( A - x ) , n . K*lV.0 13-70 13.65 13-68 -2818 11.65 11.60 11.63 2.495696 9-72 9.78 9.75 2-588594 8-30 8.30 8.30 2-5310582 7- 66 7.55 7.55 2-44Mean.. ....2.51TABLE XIII.N / 10-Carbamide + N / 10-hydrochloric acid in 20 per cent. alcohol.Temperature, 71-25O.t. ( A - 4 , . (A--)!d* (A-x)m- K. 105.0 16.50 16.54 16.52 -2818 14-10 14.10 14.10 2.445700 11.85 11-88 11-87 2-528598 10.10 10.10 10.10 2.4910584 9.90 8-95 8.98 2.5012917 8-16 7.95 8.05 2.42Mean.. .... 2.47TABLE XIV.N / 10-Carbamide + N / 10-hydrochloric acid in 30 per cent. alcohol.Temperature, 7 1 2 5O.t . ( A - 4 , .0 20.302799 17.405697 14.707683 13.2810016 11.8012907 10.2216266 -T( A -42.20.3517.45147513.3011.7210.268.70( A --&a. K . 1 0 5 .20.33 -17-43 2-3914-73 2-4813-29 2-4011-76 2.3710.24 2.318-70 2-27Mean ......2.3THE DECOMPOSITION OF CARBAMIDE. 61'7TABLE XV.N / 10-Carbamide + N/10-hydrochloric acid in 40 per cent. alcoholTemperature, 71.25O.1. ( A -4. (A--2),. ( A - x ) ~ . R.105.- 0 20.25 20.25 20.252895 17-40 17-50 17-45 2.244881 15.50 15.50 15.50 2.387214 13.72 13-78 13.75 2-3313452 10-44 10-42 10.43 2.1417334 8.66 8.66 8.66 2.1310106 12.12 12.18 12-15 2.20Mean ...... 2-24TABLE XVI.N / 10-Carbamide + N / 10-hydrochloric acid in 49.67 per cent.alcohol. Temperature, 71.25O.t. ( A - 4 , . ( A - 4 2 * ( A -4** K. 105. - 0 20.40 20.40 20.401982 18-55 18.55 18.55 2.084315 16-60 16-50 16-55 2.107208 14.62 14.58 14.60 2.0210554 12-60 12.64 12.62 1.9814438 10.60 10.66 10.63 1-9617264 9-50 9.56 9-53 1.91Mean. .....2.01TABLE XVII.N / 10-Carbamide + N / 10-hydrochloric acid in 60 psr cent. alcohol.Temperature, 71'25O.t. ( A - 4 , . ( A -x)2. ( A -@m. K-105.0 20.65 20-65 20.65 -2331 18.60 18.60 18-60 1.955224 16-40 16.38 16-39 1-92857 1 14.30 14-28 14.29 1-8712455 12.30 12.21 12.26 1.8215281 11.00 11 00 11-00 1-7918166 9-84 9.84 9.84 1-77Mectn......145TABLE XVIII.N / 10-Carbamide + N / 10-hydrochloric acid in 70 per cent. alcohol.t. (A-4,- ( A - X ) , . ( A -4m. K. 105.Temperature, 7 1 *25O.0 20.45 20.45 20.45 -2890 18.33 18.33 18.33 1-646237 16.16 16.12 16-14 1.6510122 14.02 13.98 14-00 1.6312947 12-60 12.56 12.58 1.6315832 11-42 11-46 11.44 1-6920144 10.00 10.06 10.03 1.64Mean.. . , . .la6618 BURROWS AND FAWSITT :TABLE XIX.N / 10-Carbamide + N / 10-hydrochloric acid in 80 per cent.alcohol.Temperature, 71*25O.t. (A--s),. ( A - 4 P ( A - 4*- K-10"0 20.33 20.33 20.33 -3343 18.20 18-20 18-20 1.447228 16.20 16-26 16.23 1-3510053 14.80 14-74 14-77 1-3812938 13-60 13-50 13.55 1.3617261 12.00 12-00 12.00 1.3319139 11.36 11.42 11-39 1.31Mean.. .... 1.36TABLE XX.N / 10-Carbamide + N / 10-hydrochloric acid in 90 per cent. alcohol.t. ( A - 4 1 . ( A -&- ( A -x)rn* K-105.Temperature, 71.25O.0 20.27 20-27 20.27 -6762 17-00 17.00 17-00 1-149585 15.78 15.74 15.76 1.1413898 14.31 14.31 14.31 1-0915785 - 13-45 13.45 1.13Mean. ... .. 1.12TABLE XXI.N / 10-Carbamide + N / 10-hydrochloric acid in water. Tempera-ture, 61'05O.t.( A -XI,* ( A - x ) ~ (A-x)m- K. 10'.0 12-65 12.65 12-65 -9660 10.80 10.80 10.80 0.71118340 9.36 9.40 9.38 0.70828400 7.90 7.90 7-90 0,720Mean.. . . . . 0.7 13TABLE XXII.N/10-Carbamide + N / 10-hydrochloric acid in 30 per cent. alcohol.t. ( A - 4 , . ( A - 4 2 ' ( A -x)rn- K.10".0 13.05 13.05 13-05 -9660 11.47 11.47 11.47 0.58018340 10.20 10.18 10.19 0.58628400 8.80 8.84 8.82 0.599Mean. . . . . . 0.588Temperature, 61 -05O.TABLE XXIII.N/10-Carbamide + N/10-hydrochloric acid in 60 per cent. alcohol.Temperature, 61'05O.t. ( A -41. (A-X),. ( A - x)m- K . 105.0 13.88 13-90 23-89 -9660 12-38 12.40 12-39 0.51418330 11.28 11.30 11-29 0.49128400 10.02 10.06 10.04 0.496Mean.. . . . .0*50THE DECOMPOSITION OF CARBAMIDE.619TABLE XXIV.N / 10-Carbamide + N / 10-hydrochloric acid in 90 per cent. alcohol.Temperature, 61.05O.t. (A-41, ( A -42' ( A - x ) ~ . KO 105. - 0 14-16 14.09 14-129660 13.30 13.28 13-29 0.27118330 12.40 12.42 12-41 0.30628400 11-70 11.72 11.71 0.286Mean.. . . . .0.288Discwrsion of the Experimental Results and Conclusions Drawn.from these.It, is remarkable that throughout tables VII-XXIV the velocity-constants calculated by the '' first order " (logarithmic) formulaare so satisfactory. Apparently, alteration of temperature oralteration of the solvent by the addition of alcohol does not changethe mechanism of the reaction. With regard to the effect oftemperature on aqueous solutions, it has already been found byone of us that the temperaturecoefficient for loo for the decom-position of carbamide by water alone; and also by hydrochloricacid solutions, is 3.5 between 90° and 99O.Calculating now fromthe constants given in tables VII, XI, and XXI, the temperature-coefficient for loo between 71° and 98O is 3-53, or between 61° and98O 3-61.The temperaturecoefficient for the aqueous alcohol solutions isaiso almost identical with that for the aqueous solutions. Theseresults indicate that change of temperature and addition of alcoholdo not affect the mechanism of the reaction, although the velocityis altered.The detailed explanation of the decompoeition of carbamide maybe here stated as follows:When carbamide is decomposed by water, the initial decomposi-tion of the carbamide is one into ammonium cyanate, and the rateof this reaction is the same as the velocity of the decomposition ofcarbamide with acids.We therefore assume that the function ofthe hydrochloric acid is simply to destroy the ammonium cyanateas soon as it is formed.The conversion of ammonium cyanate into ammonia and carbondioxide is rapid, but we do not believe this decomposition to bean infinitely fast reaction. When ammonium cyanate is decom-posed by water alone, the reaction is much slower than when acidsare present, and the decomposition of carbamide in this case isregulated by the rate of decomposition of the cyanate into ammoniaand carbon dioxide620 BURROWS AND FAWSITT :We believe i t is possible that a small amount of decompositionof carbamide into ammonia and carbon dioxide, either direct orvia ammonium carbamate, may g o on, but the amount of this isquite insignificant in comparison with the amount decomposedafter passing through the intermediate stage of ammonium cyanate.I f there were any appreciable amount of decomposition ofcarbamide to ammonia and carbon dioxide via ammoniumcarbamate, it seems to us that addition of salts of calcium toaqueous solutions of carbamide should increase the velocity of thereaction, but we find that this is not the case.We feel compelledalso to believe that the amount of any direct decomposition ofcarbamide to ammonium carbonate is very small, and may beneglected. The same conclusion appears to have been formed byJ. Walker and others who have worked with carbamide andc yanates.Let i% be the velocity-constant of the reaction carbamide-pmmonium cyanate, and let k' be the velocity-constant of theopposing reaction, ammonium cyanate -+ carbamide. Then thetrue value of '' II: " is in our opinion approximated to very closelyby the velocity-constant which can be calculated from the decom-position experiments with acids.The values in tables XI and XXI are :2'77x10-6 a t 71*2O and 0.713~10-5 a t 61O.If the whole of the carbamide were free in these acid solutions,the numbers would be, probably, 2.83 x 10-5 and 0.73 x 10-5respectively. Corrected further by turning the logarithms used intheir calculation into natural logarithms, the values are 6.5 x 10-5a t 71.25O and 1.7 x 10-5 a t 61.05O.Calculating the velocity-coefficient for the reverse action,ammonium cyanate + carbamide, from the results of Walker andHambly (T., 1895, 67, 746), we have the values 0.445 a t 71.25Oand 0.157 a t 61'05O.The values (z) for the equilibrium constantsare therefore :From determinations of the quantity of carbamide in equilibriumwith ammonium cyanate by Walker andthat the equilibrium constants are :- _ 0.0065 x 0.00650.09350.0032 x 0.00320.0968Ic;oo. =Zsa =his colleagues, we calculate= 0.000457.= 0*000106THE DECOMPOSITION OF CARBAMIDE. 621We believe these results show very conclusively that the decom-position of carbamide by acids is none other than the decompositionof carbamide into ammonium cyanate.With regard to the effect of alcohol on the velocity of decom-position of carbamide, there is a gradual reduction in velocitycorresponding with the proportion of alcohol added.This is bestseen in examining tables XI-XX.It is not likely that the decreased velocity due t o the additionof alcohol is on account of a retarding effect on the reactioncarbamide -+ ammonium cyanate, for other work indicates that adecrease is hardly to be looked for there.Taking the numbers of Walker and Kay (T., 1897, 71, 489) forthe velocity (k’) of the reaction ammonium cyanate -+ carbamide,and also the proportions of the compounds a t the point ofequilibrium, it is possible to calculate the value of k a t 3 2 O .Percentageof alcohol.01030497290TABLE XXV.k’ .K. k (calculated).0.00596 0*0001058 0-63 x ~ O - ~0.00774 0*000092 7 0.72 x ~ O - ~0.01290 0.000033 0.42 x lo-‘0.02950 0.000026 0-767 X lo-‘0.09300 0*00002 1-86 ~ 1 0 ~ ~0.57500 0~00001 5.75 x 10-6The tendency is rather for the velocity-constant 13 to increasewith addition of alcohol.Now, although these calculated values of k are for 3 2 O , and theiesults ive have obtained are for temperatures not lower than 61°,we think that, even if the experimental results were for the sametemperature as the calculated values, a similar disparity would benot ice able.The effect of the alcohol in our experiments appears to us t o berather in the way of decreasing the velocity of decomposition ofammonium cyanate into ammonia and carbon dioxide.It is not possible t o prove this by a direct experiment onammonium cyanate, as this decomposes so rapidly into carbamide.We have tried, however, to discover whether alcohol has anyeffect in decreasing the velocity of decomposition of potassiumcyanate. Our experiments indicate that potassium cyanate doesnot decompose so completely into carbonate and ammonia in agiven time when alcohol is present as when it is absent.We believe that the mechanism of the decomposition ofcarbamide, when starting out with acid or neutral solutions, isillustrated by the following equations :CO(NH,), NH,’ + OCN’.. . . . . . (1)NH,’ + OCN’ + 2H,O + 2NH,’ + CO//, . . . (2622 THE DECOMPOSITION OF CARBAMIDE,We have not proved, in the second equation, that the ions arethe active molecules, but formulate the reaction as if such werethe case.We have not found any evidence to show that the reaction repre-sented by equation (2) is reversible.On heating normal solutions of ammonium carbonate for severaldays, no evidence of the formation of carbamide could be detected.Lewis and G.H. Burrows have recently shown (J. Amer. Chenz.SOC., 1912, 34, 1515) that on heating solutions of ammoniumcarbamate, about 1 per cent. of the carbamate is converted intocarbamide ; it seemed, theref ore, a possibility that ammoniumcarbonate might be formed from carbamide, according to theequations :HzO+CO(NH,)zX NH4O*OC*NH,. . . . . (3)NH,0*OC*NH,+H,0-+(NH4),C0, . . . . (4)To test this, we tried the effect of adding calcium nitrate to a(neutral) solution of carbamide, in order to find out whether anyacceleration in its rake of decomposition took place.It was foundthat this was not so, and we have concluded that no measurableamount of carbamide decomposes into ammonium carbonate byway of ammonium carbamate.The results of our experiments here and in a previously pub-lished research exclude the possibility of a measurable amount ofdirect hydrolysis of carbamide into ammonium carbonate. For ifthis were the case, the rate of decomposition would not be so in-dependent of the concentration of acid as i t is found to be.There is therefore a great mass of undoubted evidence that thewhole, or very nearly the whole, of the carbamide decomposesin neutral or acid solutions via ammonium cyanate.Chattaway has recently (P., 1911, 27, 280) made the suggestionthat an isocarbamide [ HN:C(OH)*NH,] is an intermediate sub-stance formed when aminoiiium cyanate decomposes into carbamide.E.E. Walker (Proc. Roy. Sac., 1912, A , 87, 539) has suggestedthat there is possibly an intermediate product of the formulaI f either of these is an intermediate substance in the formationof carbamide, we should expect it also to be an intermediate sub-stance in the decomposition of carbamide to ammonium cyanate.We do not think there is any real evidence for the existence ofsuch a substance.One of us (Fawsitt, Zeitsch. physilcal. Chem., 1904, 48, 585) hasshown that if an isocarbamide, containing a hydroxyl group, bepresent in aqueous solutions of carbamide, it can only be presentin very small proportions. We have, however, tried to contemplateC(OIrf),(NH,)*HULME : THE MECHANISM OF DENITRIFICATION. 623the possibility of a very small amount of such an isocarbamidebeing present in equilibrum with carbamide and with ammoniumcyanate, but find iihat it is not possible to use this supposition forany better interpretation of our results than we have given already.Summary of Results.(1) The apparent hydrolysis of carbamide by water alone or byaqueous solutions of acids is confirmed to be merely a trans-formation of carbamide into ammonium cyanate, which is thenchanged into ammonia and carbon dioxide.(2) Addition of alcohol to aqueous solutions decreases the velocityof decomposition of carbamide, whether starting from neutral oracid solutions.(3) Addition of alcohol up to 75 per cent. does not greatly affectthe amount of free carbamide in carbamide hydrochloride solutions.We have pleasure in expressing our thanks to Mr. E. A. Briggs,B.Sc., for conducting the experiments given in tables I and 11.THE UNIVERSITY,SYDNEY

ISSN:0368-1645    

DOI:10.1039/CT9140500609

出版商:RSC    

年代:1914

数据来源: RSC

摘要:

HULME : THE MECHANISM OF DENITRIFICATION. 623LXII. -The Mechanism of Denitrification.By WILLIAM HULME.THE first observation of a denitrifying process seems to have beenmade in the year 1868 by Angus Smith, who observed that nitrogengas was evolved from a solution containing nitrates and organicmatter. In the same year Schloesing also noticed that the lacticfermentation of sugar, in the presence of nitrates, evolved nitrogen.Neither, however, attributed these changes to bacterial action, butthought them to be purely chemical. E. Meusel, however, in 1875suggested that these and similar reactions were due to bacterialactivity ; and Gayon and Dupetit first isolated specific denitrifyingorganisms, which they designated Bacillus denitrificans a and B(Compt. rend., 1882, 95, 644, 1365).Following these, many other organisms showing denitrifyingproperties were isolated by different investigators, these organismsbeing differentiated according as the nitrate molecule was reducedto nitrite, nitrous or nitric oxide, nitrogen or ammonia.Prior t o Beyerinck and Minkman’s observation (Cent.Bakt.Par., 1910, 25, 30) that nitrous oxide could result from fsrmenta-tion, the conception of denitrification was represented by thefollowing equations (1) 2N03K + C.. . = 2N0,K + CO,,(2) 4N02E + 3C.. . = 2Nz + 2COSK + CO,,where C represents the carbon of the nutrient substance.Beyerinck revised the equation as follows :(1) 2N03K + 2C.. . = N,O + CO,K + CO,.(2) 2N02K + C . . . = N20 + C03K.(3) 2N20 + C... = 2N2 + COPMaze ( A i m Znst. Pasteur, 1911, 25, 289), however,conclusion that denitrification was due to a production ofdrew thehydrogenby the organism, which VGW had been previously put forward-byStoklasa and Vitek (Cent. Bakt. Par., 1905, ii, 14, 102, 183), whostudied the effects of different carbohydrates in their relationshipto denitrification.The following research, originally undertaken to investigate themechanism of the bacterial reduction of nitrates, pointed t o thesame conclusion, with the addition, however, of a certain amountof enzymatic reaction.The paper may therefore be broadly divided into two divisions,namely, (1) the bacterial reduction, and (2) the enzymaticreduction.The Bacterial Reduction.Isolation of the Denitrifying Organism.Previous experience having shown that “ Dried Slurry ” (sewagefilter “humus” pressed and dried with lime) from DavyhulmeSewage Works wits rich in denitrifying organisms, this substancewas thought to be a suitable source for their isolation.Accord-ingly, nitrate broth was placed in small stoppered bottles, inocu-lated with a small quantity of “dried slurry,” and after two daysa plate culture was taken. After two days a large number ofcolonies had grown on the plate, 90 per cent. of which consisted ofa non-liquefying, fluorescent organism. A replate of this organismproduced a pure culture, which on testing for denitrification bynitrate broth gave an excellent positive result.Grown in peptone broth f o r twenty-four hours a t 37O, theorganism is 3 .0 ~ long and about 1 . 0 ~ broad. Forty-eight-hourcultures showed the formation of endospores, readily stained bycarbolf uchsin.Gelatin.-After twenty-four hours a streak culture formed abeautiful, leaf-like growth, which showed a green fluorescence intransmitted light, and appeared bluish-green by reflected light.This growth gradually spread over the whole surface of the gelatinwithout showing any tendency to thicken. The gelatin was notliquefiedHULME : THE MECHANISM OP DENITRIFICATTON. 625Stab cultures w0re beaded, and only developed very slowly; non-Agur.-White, f ern-like growth, spreading gradually over thePotato.-Yellowish-white, slimy growth, slightly raised, andMiEk-Turned slightly brown, without showing any tendency toNeutral-red milk tube turned yellow on second day, that is,fluorescent.whole surface of the agar.showing a tendency to spread in a fern-like manner.clot.the organism produces an alkaline change in milk.Anaerobic Action of the Organism.In originally investigating the action of the organism, mediawere used containing dextrose, potassium nitrate, and potassiumphosphate, but no fermentation took place even after remainingin the incubator for three months. As denitrification, however,took place readily in nitrate broth, it was thought that perhapsthe presence of peptone was necessary for the development of theorganism.Accordingly, a medium was made up containing dex-trose, peptone, potassium nitrate, and potassium phosphate, whichwas found to ferment readily.A suitable medium having been found for the cultivation ofthe organism, its comparative effects, with and without nitrate,were investigated, with the view of finding out the mechanism bywhich the nitrate was reduced.The media used in these experiments were as follows:Flask 1.Flask 2. Flask 3. Flask 4.Peptone.. . 5 grams Peptone.. . 5 gram Peptone.. . 5 grams Peptone. 5 gramMeat Meatextract 5 ,, extract 5 ,, Dextrose. 5 ,, Dextrose. 5 ,,Potassium Potassiumeach made up t o 1 litre with tap water, being made weakly alkalinewith potassium carbonate when necessary.The apparatus consisted of a 600 C.C. Erlenmeyer flask (a) closedwith a two-holed rubber stopper containing the two delivery tubes(d) and (e).One (d) dipped down into the flask, whilst theother (e) was bent over, and terminated just level with the lowersurface of the rubber stopper of the flask (b). Near the top bendof the tube ( e ) a side-tube cf) was fused on, which could be closedwith a pinch-cock and piece of pressure tubing. The flask ( b ) wasgraduated in C.C. by means of a paper scale affixed to the flask,and was connected to a vessel (c) by the syphon tube (g), whichwould be closed by a pinch-cock and pressure-tubing joint.The mode of investigation was t o almost fill the flask (a) withthe medium 1, 2, 3, or 4 (about 550 C.C. sufficed for this purpose),nitrate. 5 ,, nitrate. 5 ,626 HULME : THE MECHANISM OF DENITRIFICATION.insert the stopper, plug the tubes (d), ( e ) , and oc> with cotton wooland filter-paper caps (the flask ( b ) being disconnected from thetube ( e ) ) , and then sterilise the whole.After sterilisation thefilter-paper caps were removed, and the tube ( e ) connected withthe flask ( b ) . The flask (a) was then inoculated with the organismby momentarily removing the stopper. The vessel ( c ) was nowfilled with a saturated solution of calcium chloride, which wasthen syphoned over into the flask ( b ) by applying suction a t ( f )until it had completely filled the flask ( b ) and had risen up thetube (e) as far as the tube cf). By closing the screw clip on (9)this solution was prevented from syphoning back. As the gasevolved from the flasks 1 and 3, that is, those containing nitrates,was expected to evolve nitrogen, and that evolved from flasks 2FIG.1.and 4 to contain hydrogen, and, moreover, as the fermentationswere to take place anaerobically, it was necessary to displace theair in the flasks by some gas which would not be evolved by thefermentation itself. Accordingly, methane was used to replacethe air in flasks 1 and 3, and nitrogen for this purpose in flasks2 and 4. This was effected by connecting the tube (d) with a gas-holder, and passing the requisite gas slowly through the liquid forabout a quarter of an hour, the displaced air escaping through thetube ( f ) . When all the air had been replaced, the tube was closedby a piece of pressure tubing and a pinch-cock, after which thetube (d) was sealed off in the blowpipe.The pinch-cock on (9)was now removed, and if the level of the calcium chloride solutionin the tube did not alter, this showed the absence of leakage intho apparatusHULME : THE MECHANISM OF DENITRIFICATION. 627The fermentations were then allowed t o take place a t theordinary temperature, the daily gas-evolution being measured bythe replacement of the calcium chloride solution in the flask ( b ) .The following results were obtained :Days.2468101214161820Flask 1.4101517242731343739Flask 2.-710151922242626Flask 3.4101416222730333536Flask 4...s163038464951525353As the fermentation had now apparently ceased, samples of thegases were removed for analysis by connecting a gas-burette to thepressure tubing on the tube (f>.The percentage composition of the gases was as follows:Flask.Hydrogen. Methane. Nitrogen. Carbon dioxide.1 - 24.43 73.17 2.392 50.85 - 37.50 14.653 - 24.51 74.41 1-084 60.22 - 18.80 20.98Recalculated in terms of the gases of fermentation only, that is,without the methane in flasks 1 and 3, and without the nitrogenin flasks 2 and 4, we get:Flask. Hydrogen. Nitrogen. Carbon dioxide.1 - 96.83 3.173 - 98.59 1-412 70.14 - 29.864 74-17 - 25-83Flasks 1 and 3 were now examined for nitrite.Flask 1 containedFlask 3 ,, 120 ,, ?, ?, ,,50 milligrams of nitrogen as nitrite per litre.The increased nitrite production in the latter case was probablydue to a chemical reduction of the nitrate by the dextrose, as itwas found by a test experiment that when solutions of potassiumnitrate and dextrose were sterilised together, a considerablequantity of nitrite was produced.The above results show: (1) that the gas evolution due to agiven organism depends on the composition of the medium;(2) that the chemical agent by which the nitrate is reduced isnascent hydrogen; and (3j that in the presence of nitrate the gasof fermentation consists almost solely of nitrogen.Further evidence for these conclusions was furnished by anVOL.cv. T 628 HULME : TPE MECHANISM OP bRNITRIPICATION,experiment in which a medium containing only a small percentage(0.1) of nitrate was fermented.Daily analysis showed that as long as nitrite was present in thesolution, the gases of fermentation were nitrogen and carbondioxide, but that as soon as this was reduced the gases evolved werenow hydrogen and carbon dioxide.Thus the mechanism ofdenitrification consists in a liberation of nascent hydrogen, eitherfrsm the organic foodstuff or from the water in which it is dis-solved (the oxygen being taken up by the organism), which reducesthe nitrate.The extremely small production of carbon dioxide when nitratesare prwent is probably due to its reacting with the Fra. 2.nitrite to form hydrogen carbonate,2E;NOz + 5H9 + 2CO2=2KHCO, + 4H2O + N2.The Enzymic Reduction. PART I .It was now thought that perhaps the organismsecreted some enzyme during the procese of denitrifica-tion, which acted along with the nascent hydrogen inthe reduction of the nitrates, and the four flasks, 1, 2,3, 4, were accordingly examined.Preparation of the Enzyme Solution.One hundred C.C.of each of the contents of the flasks1, 2, 3, 4 were precipitated with 500 C.C. of absolutealcohol and a few grams of common salt. After re-maining for half an hour the precipitate was collectedand dried in the incubator a t 37O until quite free fromalcohol, and finally redissolved in 30 C.C. of distilledwater. The solutions were rendered sterile by Cham-berland filtration.The appara.tus consisted of a 25 C.C. burette (a) fittedwith a rubber stopper containing two delivery tubes ( b )and ( c ) , ( 6 ) being closed by a piece of pressure tubingcontaining a short length of glass rod, whilst ( c ) wasclosed with a wad of cotton wool and covered with a filter-papercap.This portion was then sterilised, after which a sterile Cham-berland filter-candle was connected to ( b ) (by removal of the glassrod). The candle was in turn surrounded by a glass jacket (d),held in position by a stout piece of rubber tubing.The solution t o be filtered was introduced into the jacket (d),and by applying suction a t the tube ( c ) was rapidly transferred tHULME : THE; MECHANISM OF DENLTRIPICATIOS. 629the burette, from which measured quantities of the sterile enzymesolution could readily be obtained.The solution was tested for enzyme action by adding a few C.C.t o small tubes containing 5 C.C.of a sterile solution of potassiumnitrate (1 per cent.). The tubes were then incubated a t 37O fortwenty-four hours, after which the nitrite present was quantita-tively estimated by means of a-naphthylamine and sulphaniIic acid.Check tubes containing the equivalent quantity of enzyme solutionalone, and of potassium nitrate solution alone, were always treatedsimultaneously in the same manner, so that the nitrite resultingfrom the enzymatic action alone could be accurately determined.I n each series of experiments a tube containing the boiledenzymg solution was also investigated to determine the action, ifany, of heat on the enzyme.The results given with the “enzyme” solutions from the flasks1, 2, 3, and 4 were as follows:Milligram of nitrogenmask 1 (contained nitrate).&s nitrate.1 per cent. Nitrate tube alone .................... 0.005“ Enzyme ” solution alone ......................... 0.011 per cent. Nitrate + enzyme solution ............ 0.031 y y Nitrate+enzyme solution (boiled) 0.03Flask 2.1 per cent.,Ftrate tube alone ..................... 0.005“ Enzyme solution alone ......................... nil.1 per cent. Nitratefenzyme solution ............ 0.011 ,) Nitrate+ enzyme solution (boiled) 0.01“ Enzyme ” solution alone .........................1 ,, Nitrate + enzyme solution (boiled) 0.021 per cent.,Yitrate tube alone .....................1 per cent. Nitrate+enzyme solution ............Flrcsk 3 (contained nitrate).1 per cent. Nitrate tube alone .....................1 per cent. Nitrate+enzyme solution ............0.005nil.0-02Flask 4.0.0050.01“ Enzyme solution alone .........................nil.1 ,, Nitrate + enzyme solution (boiled) 0.01These results seem to show that the denitrification of a mediumcontaining nitrates and peptone under anaerobic conditions yieldsan enzyme, which has the power of reducing a 1 per cent. solutionof potassium nitrate. When dextrose was absent (flask 1) theenzyme produced was stronger in its action than when present(flask 3). The slight reduction obtained with the precipitates fromflasks 2 and 4 was probably due to a purely chemical reductionof the nitrate by the organic matter present in the so1utio;l.T T 630 HULME : THE MECHANISM OF DENITBIFLCATION.PART 11.The first series of experiments having shown evidence of the pro-duction of enzymes in a solution undergoing denitrification, accord-ingly further media were made up more fully t o investigate thisproperty.As denitrification was found to take place most rapidlyin nitrate broth under slightly aerobic conditions, this was alsothought to be the most efficient way of obtaining an enzyme.250 C.C. of a medium consisting ofPeptone .................. 10 gramsPotassium nitrate. ..... 5 .. Meat extract .......... 10 .. Common salt .......... 5 y ymade up to 1 litre with tap water and neutralised with potassiumcarbonate, were placed in a 500 C.C. Erlenmeyer flask, which wasthen sterilised and placed in the incubator a t 37O.After four days’ incubation 100 C.C.of the medium were re-moved with a sterile pipette, and were found to contain largequantities of nitrite. The medium was then precipitated byalcohol and salt, redissolved, and filtered as described above.The enzyme action was as follows:Milligram of nitrogenas nitrite.1 per cent. Nitrate tube alone ..................... 0.0051 per cent. Nitrate+enzyme solution ............ 0.031 ,, Nitrate+ enzyme solution (boiled) 0.03Enzyme solution alone .............................. nil.that is, a denitrifying agent appears to be present in the ‘‘ enzyme ”solution.The remaining 150 C.C. of the medium were allowed t o fermentuntil all the nitrate and nitrite had been decomposed. Onehundred C.C. were precipitated with alcohol and tested for enzymeaction as before.The following results were obtained :Milligram of nitrogeiias nitrite.0-0050.0051 ,, Nitrate+enzyme solution (boiled) .0.0051 per cent. Nitrate tube alone .....................Enzyme solution alone .............................. nil.1 per cent. Nitrate + enzyme solution ............that is, no enzyme action has taken place.These results show that the denitrification of nitrate brothunder semi-aerobic conditions involves the production of an enzymeas long as nitrate or nibrite are present in the solution, but it dis-appears when the latter have been decomposed.The emyme is not affected by boilingHULME : THE MECHANISM OF DENITRIFICATION. 631Comparative Tests with and without Znoculation.These tests were carried out to see if the reducing productisolated from denitrifying solutions was really due t o bacterialinfluence, or whether it would also be produced in a flask contain-ing. similar ingredients and treated in exactly the sam0 way, butwhich all the while remained sterile.For this purpose the following media were used:Medium 1.Medium 2.Potassium nitrate . 10 grams Potassium nitrate.. . 10 gramsTap-water. ............ 1OOO’k.o.Peptone ............... 10 ,, Peptone ............... 10 ,,Tap-water ............ 1000 c.c Dextrose ............. 10Two flasks of medium 1 and two of medium 2, all containing250 c.c., were sterilised, after which one each of media 1 and 2 wereinoculated. Then all four flasks were incubated a t 37O.After sixteen days’ incubation the flasks containing medium 1were investigated.The uninoculated flask was still clear andbright, but the corresponding inoculated one had become quitecloudy, and a considerable precipitate had formed.One hundred C.C. of each were precipitated with alcohol and salt,dried, redissolved, etc., as before.The solutions resulting from the precipitates from the inoculatedand uninoculated flasks are in the following table referred t o asEI and EU.The results were as follows:Milligram of nitrogenas nitrite.1 per cent. Nitrate tube done ..................... 0-0055 C.C. EU alone ....................................... nil.5 C.C. EI alone ........................................ 0.011 per cent. Nitrate+ 5 C.C.EU ................... 0-0061 ,, Nitrate+5 C.C. EU (boiled) ....... 0.0051 ,, NitrateS-5 C.C. EL.. ................... 0.0251 ,, Nitrate+5 C.C. EI (boiled) ............ 0.025Thus, whilst the precipitates from the inoculated medium con-firm the previous results, that. obtained from the uninoculatedflask is devoid of denitrifying power.After one month’s incubation the flasks containing the medium 3were examined in exactly the same way.The results were as follows:Milligram of nitrogenas nitrite.1 per cent. Nitrate tubes alone (Fresh samples) nil.2 C.C. EU done ....................................... nil.2 C.C. EI alone ........................................ 0.0071 per cent. Nitrate+2 C.C. EU ................... 0.0011 ,, Nitrate+ 2 C.C.EI. ..................... 0.011 ,, Nitratef2 C.C. EI (boiled) ......... 0.011 ,, Nitrate+2 C.C. EU (boiled) ......... nil632 HULME : THE MECHANISM OF DENITRIFICATION.The effect of using 10 per cent. nitrate tubes was also investi-gated t o determine whether the nitrite produced was dependenton the concentration of the nitrate.The results obtained were:10 per cent. Nitrate tube alone ......................... nil.10 .. Nitratef2 C.C. EU ......................... 0.00110 .. Nitrate+2 C.C. EI ........................ 0.01that is, the nitrite produced is independent of the concentrationof the nitrate.Thosewhich result from the inoculated flasks (enzymes) possess the powerof reducing a solution of potassium nitrate to nitrite, whilst thoseobtained from the uninoculated ones are devoid of this property.Hence the precipitate which reduced nitrates is due entirely t obacterial activity, and not to a purely chemical reduction.Thus both media 1 and 2 yield precipitates with alcohol.Con c Zusion .(1) I n the process of denitrification, the denitrifying organismdecomposes the organic foodstuff, with liberation of nascenthydrogen and carbon dioxide, o r perhaps acts in such a way as todecompose the water contained in the medium, liberating freehydrogen, and eliminating the oxygen along with some of thecarbon of the foodstuff as carbon dioxide.(2) When nitrates are introduced into the solution, hydrogen isno longer evolved, but now nitrogen, along with a very small per-centage of carbon dioxide, the nitrates in the solution beingsimultaneously reduced.(3) The production of nitrites in the medium seems to be dueto the organism excreting some enzyme which possesses the powerof reducing nitrates to nitrites, probably by combining directly withan atom of oxygen from the nitrate group, after which this oxidisedenzyme is again reduced to its original state by the nascent hydrogenevolved by the organism, when it in turn reduces more nitrate.(4) The nitrite formed then reacts with the hydrogen and carbondioxide evolved by the organism to form free nitrogen,2KN0, + 5H2 + 2C02 = ZKHCO, + 4H20 + N,,which accounts for the high nitrogen percentage in the gases evolvedfrom nitratecontaining tnedia.I n conclusion, the author wishes to thank Dr.G . J. Fowler forhis kind advice and friendly criticism throughout the course of thisresearch.THE UNIVERSITY,bl ANCHEPFE R

ISSN:0368-1645    

DOI:10.1039/CT9140500623

出版商:RSC    

年代:1914

数据来源: RSC

摘要:

THE SYSTEM M-XYLENE-ETHYL ALCOHOL-WATER. 633LXT 1 J.-The System m-Xylene-Ethyl Alcohol- Watey.By ALFRED HOLT and NORMAN MURRAY BELL.THE present communication had its origin in some observationsmade by one of us while attempting to separate a substance dis-solved in wet alcohol by means of m-xylene. It appears that inthe presence of a large volume of m-xylene a very small quantityof water enables extraction to be carried out satisfactorily, andsome rough experiments seemed to warrant the idea that thesystem m-xylene-ethyl alcohol-water was worthy of a further andmore detailed examination.The system resembles that of benzene-acetic acid-water studiedby Waddell (J. Physical Chem., 1898,2, 233) and benzene-alcohol-water. The latter system has been examined by Taylor (ibid.,1897, 1, 461), but the present investigation covers a somewhatmore extended field of the physical properties.Absolute ethyl alcohol and nearly pure m-xylene were employed,these substances, like ethyl alcohol and water, being miscible inall proportions.The question whether m-xylene and water aremiscible to any appreciable extent has been tested by the deter-mination of the freezing point of distilled water both before andafter shaking with xylene, and also by treating xylene previouslyshaken with water with anhydrous copper sulphate. I n both casesnegative results were obtained, and very recently Campbell (Phil.Mag., 1913, [vi], 26, 1044) has shown that moisture does not affectthe electrical resistance of m-xylene. For all practical purposes,these two liquids may, therefore, be regarded as non-miscible.The first series of experiments were directed to a determination ofthe amount of water which could be added to mixtures of xyleneand alcohol of varying composition and a t different temperatureswithout causing separation into two layers.The method of pro-cedure was as follows: a known volume of xylene was taken, andsuccessive equal small portions of alcohol were added. After eachaddition of alcohol, water was run in until a distinct milkiness wasproduced through separation of one phase in minute drops. Themixture was kept a t constant temperature and was shaken re-,peatedly. Experiments a t 9 8 O were not easy, and had to be carriedout in closed vessels, and it was also difficult t o determine the endpoint on account of the boiling of the liquid.Table I contains a summary of the results thus obtained, thevalues representing composition being expressed in mols.per cent. ;whilst in Fig. 1 these values are represented graphically, mols. oM-XYLENE-ETHYL ALCOHOL-WATER. 635water being plotted against mols. of alcohol in 100 mols. of alcoholand xylene. The curved lines represent the isotherms.Since, howeyer, mols. per cent. is an inconvenient way of re-presenting the system graphically, the results plotted in syntheticvolumes (thab is, vols. of water added to 100 vols. of a xylene-alcohol mixture) are given in Fig. 2, thus forming a direct com-parison with Taylor's figures for the system benzeneL-alcohol-water.OO.TABLE I.19O.41'.A- l c -A5kylene. Alcohol. Water. Xylene. Alcohol. Water. Xylene. Alcohol. Water. \ /49.18 41.12 9.70 48.36 40-43 11.21 46.08 38.53 16.3931.89 53.35 14.76 30.31 50.69 19.00 28.22 47.19 24-5923.23 68.27 18.60 21-31 53.44 25.25 19.79 49.63 30.5817.83 59.61 22.56 16-36 54-71 28.93 15.17 50.75 34.0814.46 59.07 26-47 13.18 56.37 30.45 12-14 50.75 37.1111.96 60.02 28.02 10-87 54.52 34-61 9.98 50.07 39.958.73 58.43 32.84 9.22 53.98 36.80 5.57 46.61 47.826.00 55.22 38.78 6.91 52-00 41.09 4.42 44.37 51.214.50 52.71 42.79 4.04 47.31 48.65 3.36 41.22 55-423.62 51.43 44-95 2.23 42.89 54.88 2-68 40.35 56.9763O, looo.Xylene. Alcohol. Water. Xylene. Alcohol. Water.44.35 37.08 18.57 18-81 39.56 41.6333.22 41-67 25.11 7-59 38.24 54-1728.14 47-06 24.80 3.75 40.97 55.2821-63 45-22 33.1518.36 46.06 35.5815.87 46-44 37-6912.48 45.64 41.889.93 45-66 44.417.66 44-95 47-394.19 40.35 55.46hr--h\ t 5The isotherms for mols.per cent. mixtures containing up toabout 70 per cent. of alcohol can be fairly represented by theexpression :Percentage of alcohol':/Percentage of water= constant ( K ) ,and for mixtures with a greater alcohol content by the expression:Percentage of alcohol x 'Ypercentage of xylem= constant (P),the isotherms consisting of two curves of the same general shape.The accuracy of the values is shown in Fig. 3, where plotting thelogarithms yields nearly straight lines.Further, these lines are about equidistant, and as the temper-ature intervals were almost the same, it follows that the constantvaries directly with the temperature.This relation is given by theexpressions :Kts = Kt - 0.085t636 HOLT AND BELL: THE SYSTEMandK't, = K't, + 4.40t,as will be seen from the following table:K.Temperature. Obs. Calc.0" 14.28 14.2819 12.41 12.6741 10.66 10.8063 9.20 8.9398 - -~ ~~~120 100 80 60 40M o k of water.FIG. 3.K'.Obs. Calc.314 314390 398480 494605 59 1765 74520 30- 4 0 50 60 i 0 80Mols. If alcoholM-XY LENE-ETHY 1, ALCOHOLWATER. 637The synthetic volume isotherms for mixtures with an alcoholcontent from 0 to 75 per cent. can also be closely represented bythe same expression:Percentage of alcohol,/Percentage of water=constant,the temperature relation being in this case:where, as in the expressions above, Rt, is the constant a t a tem-perature t,, and Kto the value at Oo.I n view of the fact that, under favourable conditions, it is notpcxssible to determine the end-point of the phMe of completemiscibility of the three liquids nearer than 0.1 t o 0.2 C.C.waterin a total mass of liquid of about 60 c.c., i t is seen that if the xylenecontent is large, the error i n the volume of water added may bevery considerable, whilst, as has already been pointed out, withlittle xylene the conditions are unfavourable for a sharp end-point.The above values are, therefore, in quite fair agreement under theconditions of the experiments. Four tie-lines were determined,and are represented by the lines TT', T,T,', T3T3/, and T,T,' iinFig.2. The composition of the conjugate solutions was arrived a tby different methods of analysis, and since the tie-lines have prac-tically a common vanishing point V , it follows that they mustpossess considerable accuracy. I n the first method the xylene wasseparated from the mixture by the addition of a large excw ofwater. Its volume was measured, and thence the percentagevolume in the phase under investigation, the amounts of alcoholand water being calculat6d by the formula already given.This procedure assumes that by the addition of a large volumeof water all the xylene is separated, a point to which reference willbe made later, but for mixtures rich in xylene i t probably givesfairly true values.When, however, little of the hydrocarbon ispresent, the great volume of water necessary for its separationtends to prevent accurate estimation through breaking the xylenephase into small globules which cannot readily be brought together.In a variation of this method the density of the phase was firstdetermined, and then the xylene was separated and its percentagevolume measured. From the data thus obtained, the density andcomposition of the mixture of alcohol and water, after removal ofthe xylene, could be calculated. Thus it was possible to arrivea t a direct measure of the volumes of each constituent liquid. Onaccount of the contraction resulting from the admixture of waterand alcohol, all calculations had to be carried out in terms ofweight per cent., and finally transformed into volumes.JVt, = K - 0*125t,638 THE SYSTEM M-XYLENE-ETHYL ALCOHOT--WATER.This method is probably most suitable for mixtures poor inxylene, for the amount of hydrocarbon left in the solution afteraddition of water would be relatively very small compared with thealcohol content, and hence analysis of the alcohol-water mixtureby density should give good rwults.The tie-line TT’ was deter-mined in a totally different manner. It was observed that whenwater was gradually added to mixtures of xylene and alcoholvarying from 75 to 25 per cent. by volume of xylene, the phase richin the hydrocarbon separated initially as the lower layer, but onincreasing the amount of water, a critical point could be obtainedwhen the conjugate solutions had the same specific gravity.Further additions of water placed the xylene layer on top.De-terminations with these different mixtures of xylene and alcoholshowed that this critical density composition lay in each case inthe same straight line, hence there is only one phase rich in xylene,and one rich in alcohol, which have identical specific gravities;consequently the line passing through the points is a tie-line. I nFig. 2 it is seen that it vanishes in the same point as the others.The letters R, Kl, lye, K,, and K4 in the figure represent thepoints when the tie-lines become non-existent, and the dotted lineshows that, at any rate, over a temperature ranging from Oo to 9 8 Othey lie on an almost straight line.From an inspection of theform of the isotherms it is, however, clear that a t temperaturesabove looo, the line joining them will have an upward trend.The following table gives the analyses of the conjugate solutionsfor each tie-line, and is expressed both in mols. per cent. andsynthetic volumes.TABLE 111.Xylem phase. - Mols. Val.Tempera. per cent. per cent.ture. of xylene. of xylene.O0 80.18 91-220 69-95 85-1019 82-12 91.2919 58-40 78.30Alcohol phase. - Mols. Val.per cent. per cent.of xylene. of xylene.0.93 3-402.29 7.201.05 3.805.27 15.26The connexion between the volume of xylene phase separatedand the amount of water added is given in the following tableKNIGHT: THE AGEING OF ALLOYS OF SILVER AND TIN.639TABLE IV.Composition of 100 C.C. of Composition of 100 C.C. of Composition of .lo0 C.C. of7511 C.C. of xylene 50'15 C.C. of xylene 2501 C.C. of xylene2489 C.C. of alcohol. 4985 C.C. of alcohol. 7499 C.C. of alcohol.Vol. of water Vol..of phase Vol. of water Vol..of phaseadded to rich in added to rick in added to rich inabove mixture. separates. above mixture. separates. above mixture. separates.xylene-alcohol mixture taken, xylene-alcohol mixture taken, xylene-alcohol mixture taken.A h r I / \ f 7 Vol. of water Vol. of phase100 C.C. of xylene which 100 C.C. of xylene which 100 C.C. of xylene which2-11 C.C. - 6.30 C.C. - 15.92 C.C. -2-65 ,, 85-00 C.C. 6.80 ,, 27.00 C.C. 18.45 ,, 6-25 C.C3.15 ,, 82.55 ,, 7.80 ,, 34.75 ,, 19.70 ,, 8.50 ,,3.79 ,, 80.75 ,, 8.80 ,, 38-50 ,, 22.20 ,, 12-25 ,,6.40 ,, 78.75 ,, 9.80 ,, 42.00 ,, 27.70 ,, 16.26 ,,8-79 ,, 78.25 ,, 11.30 ,, 44.50 ,,13.90 ,, 77.50 ,, 14.05 ,, 46.25 ,, Xylene phase changes22-80 ,, 49.00 ,, from bottom to topXylene phase changes with 17.2 C.C. of water.from bottom to top Xylene phase c h a n g e Ewith 4-39 0.0. water. from bottom to topwith 10.8 C.C. water.Reference to Fig. 2 shows that proceeding along the isothe5mfor 1 9 O the phase rich in xylene separates out as the top layerbetween the points ,4 and T', and as the bottom layer between Tand K, and that above about 640 two phases with the same specificgravity cease to exist.MUSPRATT LABORATORY OF PHTSIOAL AND ELECTRO-CHEMISTRY,THE UNIVERSITY, LIVERPOOL

ISSN:0368-1645    

DOI:10.1039/CT9140500633

出版商:RSC    

年代:1914

数据来源: RSC

摘要:

KNIGHT: THE AGEING OF ALLOYS OF SILVER AND TIN. 639LXlV.-The Ageing of Alloys of Silver and Tin.By WILLIAM ARTHUR KNIGHT.THIS communication deals with the ageing of filings * of alloys ofsilver and tin, and with the densities of unaged and aged filingsof the alloy Ag,Sn.One of the first suggestions with regard to ageing was that itwas due to superficial oxidation, and it must be admitted that thishypothesis is simple and attractive. Further than this, it is aID I n the experiments t o be described all filings were obtained by means of files oftho same degree of coarseness, and were freed from iron by means of a magnet. Thebars of alloy were always filed slowly and were cooled iu running water every tenseconds. A l l amalgams were prepared, and the excess of mercury was squeezed outexactly as doscribcd in previous comiiiunication (T., 1911, 99, 198), and the ratioHg : alloy should be takeu to indicateWeight of mcrcury retaiued by the flingsWeight of' liliugs of alloy takeu 640 KNIGHT! THE AGEING 01 ALLOYS OF SILVER AND TlN.matter of extreme difficulty-if indeed it is not impossible-toprove directly that superficial oxidation does not take place duringthe ageilig of these alloys, and the hypothesis had not beenabandoned here until investigations in this laboratory seemed torender i t untenable (R.A. Joyner, T., 1911, 99, 198; Knight andJoyner, T., 1913, 103, 2247).A. Further Experiments on the Ageing of Filings of Alloys ofSilver and Tin.(1) Is the ageing of filings of an alloy accompanied by a changein weight ?The alloy chosen was Ag3Sn, as with this the effect of ageing ismost pronounced, and the ageing was carried out by heating thefilings to looo in a current of coal-gas.I n the two experimentscarried out it was found that no change of weight greater than0.0001 gram fook place even after heating for one hour, and itwas shown that replacing the current of coal-gas by a current of airwas also without effect. The possible change of weight in theseexperiments was not greater than 0.002 per cent. of the weight ofalloy taken.*(2) Attempts to oxidise filings of the alloy Ag,Sn at roomtemperature.Three experiments were carried out for the purpose of deter-mining whether ozone, dry or moist, had any effect on these filings.It was found thak the value of the ratio Hg : alloy was not affectedby this treatment, the actual values of this ratio in the three casesbeing :(1) Hg : alloy= 7.178 : 3.525 = 2-04 : 1.(3) Hg : alloy= 5.263 : 2.525 = 2-08 : 1.(3) Hg :alloy= 6.157 : 3.062 =2.01: 1.Hence the filings were unaged by ozone, and, in addition, nochange in weight was detected.Further, no change in the colour of the filings could be noticed,although it waa found that moist ozone, under these same con-ditions, blackened silver almost instantaneously.A concentrated solution of hydrogen peroxide to which a few* From the point of view of the kinetic theory, 0.0001 gram of oxygen wouldactually have covered all the filings about 1 molecule deep, since their diameter wasfrom 0.02 to 0'03 mm.In the form of oxide, on the other hand, only aboutone-third of the surface would have been covered, and it has been shown thatcharacteristic physical properties are not exhibited until such a layer is manymolecules thick (see, for instancc, Oberbeck, A m . Phys. Chcm., 1887, [iii], 31,336)KNIGHT: THE AGEING OF ALLOYS OF SILVER AND TIN. 641drops of sulphuric acid had been added had no effect on the ageingof these filings, as it was afterwards found that:Hg : alloy = 4.788 : 2.289 = 2-09 : 1.If filings of the alloy Ag,Sn a t room temperature are so resistantto the action of energetic oxidising agents like ozone and hydrogenperoxide, it is difficult to believe that they should become oxidisedin coal-gas, in hydrogen, or in an exhausted vessel a t looo.Further, if the ageing of these alloys is due to superficial oxida-tion, it is reasonable to suppose that the effect would be mostpronounced in the case of the metal which oxidises most readily.Yet pure tin does not age (T., 1911, 99, 201).To sum up the evidence unfavourable to the explanation ofageing as being due to superficial oxidation: ageing is not accom-panied by a change in weight, although (see below) it is accompaniedby a change in density; it takes place alike on heating the filingsin hydrogen, in coa,l-gas, or in a vacuum; it is not induced byozone, moist ozone, hydrogen peroxide, or hydrogen sulphide (seebelow) at room temperature; it is not prevented by the presenceof magnesium ip the alloy; grinding in an agate mortar neitherages unaged filings nor renders aged filings unaged.It may beimpossible directly to disprove such a hypothesis, but it is evidentlyuntenable.Since bars of these alloys definitely do not age, the hypothesis (d)(T., 1911, 99, 203) will receive further experimental examination.(3) What is the effect of hydrogen sulphide on filings of thealloy Ag,Sn?Dry hydrogen sulphide appeared to have no action on the filings,and the undried gas also seemed to be without effect, so a quantityof the filings of known weight was moistened with water and ex-posed to the action of the gas for thirty minutes. After drying thefilings over phosphoric oxide at room temperature, the weight wasre-determined. The retmlts were as follows :The weight before treatment was 2.7454 grams, and after treat-ment was 2.7455 grams.It seems possible that the slight increasein weight (if any) may be due to a small amount of sulphur intro-duced by decomposition of the hydrogen sulphide by the waterprewnt.Further, it was found thatHg : alloy = 5.331 : 2.730 = 1-95 : 1,and, therefore, the filings had remained unaged.A t looo it was found that a pronounced change of colouroccurred on exposing filings of the alloy Ag,Sn to the action ofhydrogen sulphide. Filings aged in air or in coal-gas are grey, thos642 KNIGHT: THE AGEING OF ALLOYS OF SILVER AND TIN.aged in hydrogen sulphide are brownish-black. This shows thatthe ordinary phenomenon of ageing is not due to any small amountof hydrogen sulphide present in the gas (air, hydrogen, coal-gas,or vacuum) in which the filings are aged.(4) What is the effect on the ageing of dental alloys of thepresence of solid solutions of silver in the alloy Ag3Sn?Alloys containing less than 75 atoms per cent.of silver, if wellannealed, contain only the compound Ag3Sn and free tin (Petrenko,Zeitsch. anorg. Chem., 1907, 53, 200; and also Joyner, Zoc. cit.),and Joyner has shown that for these alloys the phenomenon ofageing is roughly proportional t o the amount of Ag,Sn present,and vanish= entirely for pure tin.It seemed desirable, therefore, to determine the effect on theageing of an alloy of the presence of solid solutions of silver in thecompound Ag3Sn.For this purpose the value of the ratio Hg : alloy was determinedfor filings (unaged and aged) of an alloy containing 85 atoms percent.of silver.The following values were found:Unaged: Hg : alloy= 1.99 : 1aged: (1) Hg:alIoy=1*57:1(2) Hg : alloy= 1.68 : 1Hence the effect of the presence of solid solutions of silver in thecompound Ag3Sn is to make the difference between the mercuryretained by aged and unaged filings less pronounced, correspondingwith the influence of the free tin in those alloys which containless than 75 atoms per cent. of silver.It was noticed that the amalgams resulting from the alloy con-taining 85 atoms per cent. of silver were quite soft and wouldnearly flow, although the excess of mercury had been squeezedout exactly as in previous experiments.After three and a-halfhours the mass had become quite hard, although the edges couldstill be rubbed off with the fingers.For pure silver it was found that Hg : Ag= 1.26 : 1, but it wasobvious that the filings retained, mechanically, a considerable quan-tity of mercury, which could not be squeezed out through chamoisleather.The resulting amalgams were so fluid that they would readilyflow on the hand, yet when squeezed they seemed to become harder,and appeared to take up more of the adhering mercury, Theirbehaviour in this respect was exactly similar to that of moist sand.When left overnight these amalgams became quite hard and dry.(5) Can ageing be due to catalytic action of the iron or productKNIGHT: THE AGEING O F ALLOYS OF SILVER AND TIN.643of iron introduced during the filing (compare Knight and Joyner,loc. cit., p. 2250) 1A bar of the alloy AgsSn was heated in an atmosphere of coal-gas to 418O for six hours, and then to 300° for a further thirty-sixhours, in order to obtain a loose crystalline structure. The barwas then wrapped in chamois leather, and broken into severalpieces. One of these pieces was placed between two silver plates,wrapped up in chamois leather, and wi~s beaten up quite fine. Thesmall particles thus obtained were ground up in an agate mortar,and the value of the ratio Hg:alloy was determined for unagedand aged particles.The values found were :Unaged .- Hg : alloy = 1.90 : 1Aged: (1) Hg:alloy=0*63:1(2) Hg : alloy= 0.61 : 1These values show that particles obtained in this way, which hadnot been in contact with iron, age as do filings.Hence, the hypo-thesis that ageing is due to catalytic action of the iron introducedduring the filing is untenable.(6) Time required for the ageing of the alloy Ag,Sn.Separate lots of filings of the alloy were heated a t looo in coal-gas for five, ten, twenty . . . . . . minutes, and the correspondingvalues of the ratio Hg: alloy were determined. The values aregiven below:Timeof heating ....... 0 5 10 20 40 80 320min.Ratio Hg: alloy ....... 2-15 1.33 1-21 1.14 1-08 1-12 1-12Hence, the alloy is completely aged in twenty minutes at looo,the last three values above differing from the value after twentyminutw by less than the experimental error.B.Density at 2 5 O of Filings of the Alloy Ag,Sn, add Contractionin Vohme o n Ageing.The densities of unaged and of aged filings of this alloy weredetermined in the usual way by means of the specific gravity bottle.All the water used in the determinations was air-free distilledwater, and efforts were made to free the filings from the airadhering to them before filling the specific gravity bottle withwater for the final weighing. The values obtained were as follows:Unizged Filings.Weight of flings taken.Grams. q 5 .1.8029 9-7810.3473 9.7916.921 6 9- 85(1)(2)(3)VOL. cv. u 644 KNIGHT: THE AGEING OF ALLOYS OF SILVER AND TIN.The values (1) and (2) were obtained for filings of the samesample of alloy, whilst (3) was obtained from a different (muchcoarser) sample.Fenchel (Dental Cosmos, January, 1910) foundthat the specific volume of the alloy Ag,Sn was Om102 (temperaturenot stated), giving a density of 9-80.Aged Filings.Weight of filings taken.Grams. u 2 5 .7.8029 9.8911.6688 9.8827.5410 9.8616.5507 9-934(1)(2)(3)(4)I n the first of these determinations for aged filings the ageingwas carried out under water in the specific gravity bottle. I n theother cases the ageing was carried out in coal-gas. The last value,9.93, was obtained from filings of the same sample of alloy as wasused in obtaining the last value, 9.85, for the density of unagedfilings.The above results show that the ageing of filings of the alloyAg,Sn is accompanied by an increase in density, and, as i t hasalready been proved that there is no detectable increase in weight,it follows that there must be a contraction in volume.Dilatometer measurements showed that this is the case, althoughthe results did not agree well amongst themselves or with thoseindicated by the density determinations.The indicating liquid inthe first of the dilatometer nieasurements was paraffin oil, in theother cases, xylene was used for this purpose. The ageing of thefilings was carried out in the dilatometer, and in each case, after thealloy had been aged, the dilatometer was left for twenty-four hoursa t 2 5 O in order to recover its original volume, a, blank experimenthaving shown that eighteen hours was sufficient.. The contractionin volume of the alloy w a found by weighing the amount ofmercury which could be contained between two marks on thecapillary of the dilatometer-the upper mark showing the level of theindicating liquid a t 250 before the alloy was aged, and the lowermark showing the corresponding level after the alloy had been aged.The results were as follows:Weight of filings in Contraction (as athe dilatometer. percentage of theGrams. original volume).6-70 0-884.742 0.4525-620 0.3026.952 0.3THE PRODUCTION OF HIGH VACUA, ETC. 645Sum ma my.1. The ageing of filings of alloys of silver and tin, whether ificoal-gas or in air, is not accompanied by a change of weight, butthere is a contraction in volume amounting t o about 0.4 per cent.2. The, evidence now accumulated appears to render untenableaiiy explanation of ageing as being due to superficial oxidation ofthe filings.I desire to express my indebtedness to. Dr. James W. McBainfor constant advice throughout this investigation.My thanks are also due to the Research Fund Committee of theChemical Society for a grant towards the purcliaee of materialsand apparatus.THE CHEMICAL LABOI~ATOI'LY,~TNIVEtlSI'rT OF BRIS1'OL

ISSN:0368-1645    

DOI:10.1039/CT9140500639

出版商:RSC    

年代:1914

数据来源: RSC

摘要:

THE PRODUCTION OF HIGH VACUA, ETC. 645LXV.-The Production of High Vacucc by n i e u ~of Finely Divided Copper.By THOMAS RALPH MERTON.THE production of high vacua by the absorption of gases was firstintroduced by Dewar (Proc. Roy. SOC., 1904, 74, 122), who dis-covered the reinarkable absorption of gases by charcoal a t thetemperature of liquid air, a method now widely used in themanipulation of gases.Soddy (I-'roc. Boy. SOC., 1907, -1, 78, 429) has found that calciumvapour, obttained by 'heating calcium in a porcelain tube, absorbsgases with great rapidity, and has used the method for preparinggases of the helium group, highly purified from diatomic impurities,since these gases are not absorbed.Gehlhoff (Ber. Ueut. ph?/siknl. Ges., 1911, 6, 271) has found that,with the exception of the helium group, all gases can be completelyabsorbed by passing them through a vessel containing potassiumvapour, through which the discharge from an induction coil ispassed.Pfund (PhLysikaZ. Zeitsch., 1912, 13, 870) has described a methodof producing high vacua, based on the absorption cf gases bycharcoal under the influence of an electric discharge, without theuse of liquid air.I have recently found that finely divided copper, which can beobtained by reducing a solution of a copper salt, and is sold corn-mercially as '' precipitated copper," absorbs gases with great readi-u u 646' THE PRODUCTION O F HIGH VACUA, ETC.ness, and that the vapour pressure of the gases thus absorbed is sosmall that under suitable conditions i t may be used for the pro-duction of high vacua.The precipitated copper is very dark brown,almost black in colour, a,nd the commercial article always containsa large quantity of water, in addition to the occluded gases, whichare given up on heating. It may be used for the production ofhigh vacua by sealing a bulb containing a few grams of the coppert o the vessel to be exhausted. The vessel is then partly exhaustedwith an air-pump, the copper being heated t o about 250O. Oncutting of the air-pump and allowing the copper t o cool, the residualgases are rapidly absorbed.It has been found possible to exhaust a small vacuum tube to anon-conducting vacuum in a very short time with t w o or threegrams of copper, using a Fleuss pump, with a small calcium chloridedrying tube, for the preliminary exhaustion.The absorption of gasesis not due t o a cheiiiical combination, thegases being evolved again on heating the copper.In exhaustinga tube in this way, it has been found that the carbon bands dis-appear first, then nitrogen, and lastly hydrogen.Experiments have also been made with helium, but this gas doesnot appear to be appreciably absorbed.It is necessary t o observe a certain number of precautions inusing this method. I f the copper is too strongly heated i t under-goes a change in colour and loses its absorbing power. The coppermay be contained in a soft glass bulb, and if the bulb is heated ina Bunsen flame it should never be allowed t o become sufficientlyhot t o impart a yellow colour t o the flame. After it has been usedfor a number of times, the absorbing power of the copper appearsto diminish, and exposure for a long time to the mercury vapourfrom a mercury pump appears to have a similar effect.Great care should be taken when copper which has not beenused before for the purpose is heated for the first time. Theoccluded gases are sometimes evolved with such violence that thecopper is blown through into the pump.For this reason, the bulb containing the copper should never befilled to more than, a t most, one-third of its capacity, and the tubeleading from it should be tiglitly packed for a short distance withglass wool.If these precautions are observed, there appears to be no reasonwhy this method should not prove useful in a variety of ways inthe manipulation of gases, more especially where liquid air is notavailable.25, GILBERT STKEP,~',LONDON, W

ISSN:0368-1645    

DOI:10.1039/CT9140500645

出版商:RSC    

年代:1914

数据来源: RSC

摘要:

A STUDY OF THE VAPOUR PRESSURE OF NITROGEN PEROXIDE, 64'7LXV1.-A Study of the Vapour Pressure ofNity-ogen Peroxide.By ALFRED CHARLES GLYN EGERTON.THE measurement of the vapour pressure of ammonia a t lowtemperatures, by passing over i t a known volume of hydrogen a ta known pressure, was made by Brill in 1906; Dr. Brill and thepresent author (in 1908) used a similar method in some unpublishedwork to determine the vapour pressure of bromine. The followingwork is an application of the method to the determination of thevapour pressures of nitrogen peroxide. A considerable number ofinvestigations has been published on the vapour pressure of thisgas, notably by Guye ( J . Chim. Phys., 1910, 8, 473), and Schefferand Treub (Proc. R. Akad. Wetemch. Amsterdam, 1911, 14, 536).The present work begins where these finished, and the determina-tion of the smaller vapour pressures a t low temperatures has beenattempted.The Method.-Hydrogen from an electrolytic generator is passedover solid nitrogen peroxide immersed in a bath a t the requiredtemperature.I f P is the pressure of hydrogen and T; its volume,?J the volume of nitrogen peroxide collected in the absorptionpipette during the passage of the volume Tr of hydrogen, and p thev . P vapour pressure of the nitrogen peroxide, then p = ~- (V+ v).The first point to be considered in such a method is whether thehydrogen is saturated with the vapour of the substance a t the par-ticular temperature. I n the work on bromine and on ammonia (byDr. Brill) it was found that constant results were obtained if thecurrent of gas was sufficiently slow, and the surface of the vapor-king subst'ance sufficiently great. I n the case of the vaporisationof solids, two actions go on: first, the vaporisatiop process, and,secondly, the dissociation of the more complicated molecules ; bothtake time before equilibrium is finally established.As a rule,therefore, lower pressures are obtained by the dynamical methodthan by the statical, when the velocity of the second process isslow. I n the method used in these experiments, the volume ofnitrogen peroxide a t Oo is calculated from the weight found bytitration or by colorimetric estimation, so that if the evaporationprocess is rapid enough to saturate the hydrogen, the dissociationprocess of slower rate does not come in.The vapour pressureobtained is that which would be given if the evaporated moleculeswere all N,O,; if some of the molecules were more complex tha648 EGERTON: A S'I'UDT OF THEthis, the actual saturation pressure would be less than the numbersdetermined in this way, but those numbers would agree with thestatically determined pressure. At the temperatures dealt with,practically none of the molecules will be NO,; some dissociationof N204 molecules may occur during t h passage to the absorbingvessel, but that does not interfere with the results, for the totalamount of nitrogen peroxide is determined by weight, the volumea t normal temperature and pressure being afterwards calculated.The preparation of the hydrogen by the electrolytic method makesit possible t o vary the speed of its passage easily; by making experi-ments a t different speeds, it is possible to obtain evidence that thehydrogen, as it passes over, is being satisfactorily saturated withthe vapour.The volume of hydrogen is determined electrolyticallyby means of a copper voltameter placed in series with the generator.The pressure is the barometric pressure, together with the over-pressure in the apparatus; this is measured by adjusting the heightof the end of a tube which dips in the water contained in a tallcylinder, so that the levels of ths sodium hydroxide solution in theU-shaped generator remain in line; the oxygen passes out throughthis tall tube and bubbles through the water contained in thecylinder.The method employed in these experiments for maintaining lowtemperatures consisted in using a Dewar vessel containing petro-leum or pentane as the thermostat, and regulating the temperatureby the replacing of warm o r cooled petroleum, either by hand orautomatically; this can be done a t any temperature by means of aspecially designed apparatus. Carbon dioxide and ether were usedfor experiments a t -7S0, and carbon dioxide and alcohol attemperatures above - 78O.The necessary alcohol, o r cooled alcoholand carbon dioxide, was added to keep the temperature constantto within a degree, a portion of the liquid in a vacuum vessel beingremoved previously by a syphon. This was found better than theaddition of small quantities of carbon dioxide, which was apt t ocause too great local cooling.Stirring was accomplished by meansair cooled in a tube immersed in carbon dioxide and alcohol, o racetone. The temperature was measured by means of a pentanethermometer, which was checked by a thermocouple, as well as byimmersion in melting ether, nitrous oxide, carbon dioxide andether, mercury and boiling ammonia.The Appratus.-A large U-tube contains a pure 10 per cent.solution of sodium or barium hydroxide. The U-tube contains twolarge nickel electrodes connected with nickel rods, which passthrough gas-tight stoppers. The U-tube is attached by a mercury-sealed joint to the gas-drying tubes filled with sulphuric aciVAPOUR PRESSURE OF NITROGEN PEROXIDE.649and phosphotic oxide. The ends of the nickel rods are connectedthrough an ammeter, an adjustable lamp resistance, and a Leblancvoltameter, which can be switched in or out of circuit, as desired, toa 66-volt main. The phosphoric oxide tube connects with the glassapparatus which is to contain the nitrogen peroxide. Thisconsists of a small bulb a t the end of two long tubes, one of whichis sealed internally, so that gas passing down it would bubblethrough the liquid in the bulb.The absorption vessel consists of a slanting tube containingthe reagent by which the nitrogen peroxide is absorbed andcoloured. The absorption apparatus consists also of a flask, intoFIG 1.which the long exit tube dips, and the liquid in this absorbs anyresidual gas.The silvered Dewar vessel holds the bulb containing nitrogenperoxide, also the pentane thermometer and the thermo-couple, thearrangement for pouring in cooled petroleum and syphoning outthat which is replaced by it, and the stirring tube.Method of Carrying Out an Experiment.-The current is firstadjusted so as to give a slow stream of hydrogen (about 0.5 ampere).After the hydrogen has been passing for a short time, the capis removed from the vessel containing the nitrogen peroxide, andis connected by the ground joint to the phosphoric oxide tubethrough which the oxygen is passing; the other cap is then re-moved, and the small phosphoric oxide tube replaced in its stead;the current is then increased so that the hydrogen agitates thenitrogen peroxide, which is then suddenly cooled and frozen ; in thi650 EGERTON: A STUDY OF THEway the beads become well covered with solid nitrogen peroxide.The current is readjusted t o its former value, the temperature ofthe bath is suitably adjusted, and the hydrogen is passed throughfor a t least three-quarters of an hour, the tube leading to the absorp-tion vessel being warmed so as to rid it; of any trace of condensednitrogen peroxide.I n the meantime, the solution iYliich is t o absorbthe nitrogen peroxide is placed in the absorption vessel, and thevoltameter plate weighed and inserted. The small phosphoric oxidetube is then removed, and the absorption vessel replaced in itsstead, the voltameter switched into circuit, and the over-pressureadjusted.The ground joints are waxed on with non-reactive waxso as t o prevent any leak, and the current is kept as constant aspossible during the experiment. When the reagent has attainedthe right tint, or when the experiment has run sufficiently long tomake it possible t o titrate the solution, the voltameter is switchedout, the phosphoric oxide tube replaced, and the absorption vesseltipped up and washed into the flask; it is now ready for measure-ment, either by titration or colorimetric estimation. The vesselcontaining the nitrogen peroxide can be removed and quickly sealedup by the aid of ground caps, which are held on by wire embeddedin non-reactive wax.Reagent, etc.-The nitrogen peroxide was prepared by Ramsayand Cundall’s method, distilled through pure phosphoric oxide andover dry anhydrous strontium nitrate twice, and then again into aflask with a long, narrow tube, through which i t could be distilleddirectly into the bulb of the apparatus containing tlie glass beads.Solid white nitrogen peroxide was obtained, which melted a t- 1 0 ‘ 5 O ( - 9 .6 O according to Guye, - 1 0 . 8 O according to Sclrefferand Treub). Several preparations were made ; if tlie nitrogenperoxide solidifies with any yellow or greenish tinge it is not pure,and this fact acts as a fairly delicate test of its purity. The boilingpoint was found t o be 22.5O. Scheffer and Treub have found thatthe melting point is a better gauge of the purity of nitrogenperoxide than the boiling point, which depends on the rapidity ofheating.A solution of pure potassium iodide was used for theexperiments a t higher temperatures in order t o absorb the nitrogenperoxide.2KI + 2N,04 = 2KN0, + I, + 2N0was titrated in a stream of hydrogen by means of N/200-thio-sulphate solution. Griess’s reagent (a-naphthylamine and sulph-anilic acid) was used a t the lower temperatures, the nitrogenperoxide being estimated colorimetrically.The iodide liberated according to the equatioVAPOUR PRESSURE OF NITROGEN PER.OXIDE. 651TABLE Irr.Tempera- Am- min- Pressure Vapour pressureTimeinture. pere. utea. in mm. Volume and constant. in mm.- 78.0' +_0-5' 0.30 67 773.0 0.70 C.C. a-Naphthvlamine 0.037- 78-5 f0.5 0.29 28fr 771.0 1-00 ,, (0-00236) 0.042(0.00589) 1 - 79.0 f0.5 0.29 308 773.0 1-00 ,, (0.00236) 0.0410.0120- 78.0 k0.5 0.23 45 774-0 1.00 ,, (0.00236) 0.035- 78.0 20.5 0.29 26i 762.0 1.00 ,, (0*00236) 0.043- 78.0 k0.5 0.28 25 756.5 1.00 ,, (0.00236) 0.042- 90.0 k1.0 0.90 81 771.0 1.00 ,, (0.00236) 0.00431 - 90.0 k1.0 0.90 65 762.0 1-10 ,, (0.00236) 0.0059 j 0.0°5'-105.5 t l .0 0.92 67 763.0 0.50 ,, (0.00236) 0.0030)-105.5 k1.0 0.92 82 762.0 0.55 ,, (0'00236) 0.0023 0.0024-1120.0 k1.0 0.96 191 760-0 0.077 ,, (000589) 0.00033-105.6 +_1-0 0.96 86 759-0 0.50 ,, (0.00236) 0.0019fTABLE l b .Tempera-ture.- 34*0°+_ 1.0"-36.0 +_ 1.0-37.5 1.5- 38.0 +, 1.0-44.0 +_ 1.0-45.0 +_ 1.0-45.5 f l . 0- 50.5 +_ 1.0-50.5 +Om5-52.0 k 2 . 0-58.0 k2.0-57.5 * 1.0-57.5 * 1.0- 60.5 f0.5-69.7 f0.4Am.pere.0.680.620-610.900.620.620.610.650.600.970.480-930.930.860.96Timeinmin-utes.25271241767735647967118557297546349VapourPressure pressurein mm.Volume and constant in mm.760.5 9.94 C.C. Thiosulphfite (0.2227) 22.47743.5 11.33 9 7 (0.2227) 16.21757-5 29.16 9 , (0.2227) 13.15749.5 7-51 9 , (0'2227) 12-93766.0 9-23 9, (0.2205) 6.62755.0 9-61 9 , (0'2205) 6.46754.5 6.36 9 9 (0.2205) 5.31767.5 5-76 9 9 (0.1124) 1-76768.0 5.48 9 9 (0.1124) 1.86773.0 11.00 Y9 (0.1120) 2.36760.0 1-64 7, (0.2205) 0.975774.0 1.60 0.883761.0 4.14 (0.1124) 0.891763.0 6.1 C.C. Thidklphate and Per-763.0 1.1 C.C. a-Naphthyl-manganate (0.0309) 0.453amine) (0*00236) 0-150The values obtained a t about - 50° are not very satisfactory; theresults with the a-naphthylamine reagent are more trustworthybhan those with the potassium iodide.Some experiments done a t- 7 8 O with potassium iodide instead of naphthylamine gave erraticresults. The iodide is oxidised by the nitrogen peroxide, and toohigh values for the vapour pressure are obtained when the resultsare calculated according t o the formula given, depending on theformation of nitrous acid. A t temperatures below -105O thenaphthylamine colours only very slowly, and does not colour t o agood tint, and so the results a t - 120° cannot claim much accuracy,but the value is not far from that calculated from the simpleempirical formula. I n one experiment, a t -60°, a potassium per652 EGERTON: A STUDY OF THEmanganate solution was decolorised, and, after the addition ofpotassium iodide, was titrated with thiosulphate. The valueobtained was in fair agreement with those given by the potassiumiodide reagent.FIG.2.Yapour PIWSZL~L: of nitrogen yiwoxide.Curve I. : - 60" to - 10". Curve 11. : - 80" to - 30".Curve 111. : -80" t o -55".x = Schrfer and Trcub's r)aeusurcments.o = Anthor's measarements.15090-80" -70" -60" -50" -40" -30" -20" - 10"l'cmpernt itre.The curve from the mean of the above results gives a vapour-pressure curve closely following the equation I, = abe, whereTABLE 11.10gp=G.9166 + S(0.0604).Temperature.- 30"- 35- 40- 45- 50- 55- 60- 65Pressure in mm.39.2419-689-774-872 441-210.6050-302Temperature.Pressure in mm.- 70" 0.151- 76 0.075- 80 0.036- 85 0.019- 90 0.0093- 95 0-0047- 100 0.0023The above table, calculated from the foregoing equation, givesthe vapour pressure of solid nitsrogen peroxide for every 5 O below-30° down t o -loooVAPOUR PRESSURE O F NITROGEN PEROXIDE. 653These investigations were carried out some years ago, the primaryobject of the work being to standardise some drops, which were usedby the author for the measurement of the nitrogen peroxide evolvedfrom explosives (J. SOC. Chem. Ind., 1913, 32, 331). An oppor-tunity t o continue the work has not arisen, and in the meantimethe interesting paper of Scheffer and Treub has been published.Several points appear when the above work is compared with theirresults, and it seems worth while to publish these preliminaryinvestigations.The following table gives the vapour pressures obtained fornitrogen peroxide (solid) obtained by Scheffer and Treub, and byGuye and Drouginine :TABLE 111.Tempera-ture.- 36.0"- 28.0- 26.0- 25.4- 23.0- 21.0- 20.2S .& T. G. & D. R. & Y, - - 183442 66 68.67053 - - 84.691- -- -- -- -Tempera-ture. S . & T . G . & D . R . & Y .f 96.0lo2 '\91-8 -18-0" -- 15.5-14.0 114 - 114- 10.8 146153 147 - 10.2 -115 - -- -The results of Scheffer and Treub are in good agreement withthose of Ramsay and Young for the undercooled liquid. Those ofGuye and Drouginine also agree well for the undercooled liquid atlow temperatures, but are somewhat higher than those of Schefferand Treub and Ramsay and Young for temperatures above - 2 O O .This fact is ascribed by Scheffer and Treub to the method employedby Guye and Drouginine for the measurement of statical vapourpressures.However, the nitrogen peroxide used by Guye andDrouginine must have been pure; the melting point ( - 9 * 6 O ) thatthey obtained is the highest recorded. The experimental resultshere given nearly agree with those of Scheffer and Treub fortemperatures above -30°, but the rate of the decrease of thevapour pressure is much greater. Scheffer and Treub's resultswould indicate that solid nitrogen peroxide would have a vapourpressure a t - 4 5 O of 8 mm., and a t - 5 O O of about 5 mm.Guyeand Drouginine's results decrease still less rapidly; a t - 30' thevapour pressure would be about 32 mm.The curve from the author's measurements cuts the solid nitrogenperoxide curve of Scheffer and Treub about - 34O, and the curveof undercooled liquid about -26O, also the curve has a greaterslope than would be expected from the measurements at the lowertemperatures a t which the author experimented.A t the time of doing the experiments, in order to check theresults obtained by the dynamic method a t -50°, a few statica654 EGER't'ON: A STUDY OF THEmeasurements were made, using paraffin (D 0.8889) instead ofmercury in the narrow-gauge tube of a tensimeter. The bulb ofthe tensimeter contained charcoal, whilst into the opposite bulbpure nitrogen peroxide was distilled.The nitrogen peroxide wascooled in liquid air, and the whole apparatus pumped out, all theair being finally removed by immersing the charcoal bulb in liquidair and sealing off the junctions to the side of the tensimeter con-nected to the nitrogen peroxide. The nitrogen peroxide was thensurrounded by a bath of alcohol cooled by carbon dioxide, and thetemperature gradually allowed to rise, stirring before takingmeasurements. The temperatures were read by means of a pentanethermometer from - 60° t o - 40°.TABLE IV.PressureTemperature. in mm.- 60.4" 0.93- 58.7 1.06- 56.5 1.42- 53.5 1.62-61.5 2-34- 49.3 2-93PressureTemperature. in mm.- 46.2 4.10- 45.8 4-43- 42.8 6.20- 40.5 8.06- 38.0 9.29- 36.0 12.70The above table gives the mean of the result of two experiments;the table is open to error, due to the solubility of the nitrogenperoxide in the oil, and also to the small trace of gas occludedin the nitrogen peroxide when it was frozen in the bulb of &hetensimeter.The values obtained between - 4 5 O and -55O agreeclosely with those obtained by the former method, and serve tocheck those results. Whether the errors mentioned can account forthe discrepancies above and below these temperatures is a matterthat can only be settled by further experiments. It seems improb-able, however, that, even taking into consideration these errors, thevapour pressure can be as high as 5 mm.a t -50°, which is thevalue t o be expected if the curve continued according t o thatobtained by Scheffer and Treub.The above results are substantiated by comparing the vapourpressure of nitrogen peroxide with that of another substance accord-ing t o Ramsay and Young's law.TABLE V.P. 8'. e.Undercooled liquid. NO,. H,O.85 263 321128 260 330200 268 340e p ' .1.2691.2691.26VAPOUR PRESSURE OF NITROGEN PEROXIDE. 655P.Solid S. and T146705334189.804.581-960.2880.030TABLE V. (continued).8.262-2253.0250.0245.0237.0NO,.e.H,O. e p .332.5 1.26s317.0 1.255312.3 1-247304.5 1.243293.5 1.240233 284 1.219227 273 1.203221 263 1.191207 243 1.174192 223 1.161The above table gives the ratio of the absolute temperatures a twhich water, or ice and nitrogen peroxide, have the same vapourpressures.The ratio of the undercooled liquid, as the rule requires,is quite constant. The ratio of the measurements f o r the solid ofScheffer and Treub decreases, and a straight line is not obtained.However, if the point a t - 3 4 O is neglected, a straight line can bedrawn through the other points. The points obtained by plottingthe values of the ratio as given by the author's results lie near theprolongation of this straight line. The vapour pressures calculatedapproximately by Ramsay and Young's equation compare well withthe experimentally determined values from - 3 5 O to -80°, as isshown in table VI. It does not appear that there is any dis-continuity in the curve about -35O, but rather that the determinedvalues between -30° and -36' are slightly too high.A t the sametime, i t is possible that there may be some change in the physicalcondition of the solid about this temperature, a change which mightalso account for the difference in the melting points and vapourpressures obtained by various observers.TABLE VI.Temperature. + 30" - 34.0 - 40.6 - 45.4 - 50.0 - 55.0 - 60.0 - 65.0 - 70.0- 75.0- 80.0Calculated fromR. & Y. equation.26.718.09.85.292.951.540.750.350.140.060.03From o w e .-9.7744302-441-210.600.300.1610-0750-036If the vapour pressures of the nitrogen peroxide are calculate656 EGERTON: A STUDY OF THEaccording t o van der Waals' equation, logfollowing values :TABLE VII.T.43-20"39.0027.4021-4515.007.70- 0.60- 6.90- 10.00 - 23.00-223.00 solid-36.00 ,,-40.00 ,,P mm.1982-01668.01007-0770.0565.0293.0256-0180.0146.070.053.018.09.8a. 7 .r.4.32 171.2" 1474.30 (utmos.)4.274-044.224-324.224.204-144-124-514.594.48All the values, except the last, are taken from Scheffer andTreub's measurements. The last value lies on the author's curve.The formula does not hold for small values of 1) t o the sameaccuracy as for larger values. The mean of the values for liquidnitrogen peroxide gives a=4'2. Now Nernst has shown that the'' chemical constant " of the vaporisation is empirically 1.1~; con-sequently, C for nitrogen peroxide should have the value 4.6.Measurements of the latent heat of vaporisation of nitrogenperoxide have only apparently been attempted once, namely, byBerthelot and Ogier, and they obtained a t 2 9 1 O abs.the value93.2 cals. (therefore the molecular heat of vaporisation A, takinginto accountthe dissociation of N,04, would be about 8000 cals.).NO value of a has been experimentally determined. Calculationfrom the vapour-pressure curve, according t o the forniula(€Agives as value kl=7821 cals.equationIf this value is substituted in thethe molecular heat of the liquid a t 18O=26'9 cals., whence- @-14.4, therefore A, will be 9458 cals. Substituting inNernst's equation :dT-9458 0435T+ 4.6, logp = -- +1*7510g7'- -4-57 1 1' 4.57 VAPOUR PRESSURE OF NITROGEN PEROXIDE.657A similar equation for the solid gives the equation for its vapourpressurewhere h, is calculated from the value (9586 cals.) between -23Oand - 26O, as calculated from the vapour-pressure curve of the solid,giving valuesThe equations do not fit the experimentally determined vapour-pressure curves with satisfaction. Owing to the fact that neitherthe molecular heat of the solid or liquid, nor the change in thelatent heat of vaporisation, has been measured, it is only possibleto obtain an approximation of the value A, and e, according to theabove calculation. It is hardly to be expected, then, that goodagreement should be obtained for the calculated results with thoseobserved. It will be necessary to determine the molecular heat ofthe liquid and the solid, and the latent heat of fusion of the solid,in order to check the observed vapour pressures a t low temperatures.Summary.(1) An account has been given of a method of determining thevapour pressure of solid nitrogen peroxide from -35O t o -looo.(2) The vapour pressure decreases somewhat more rapidly thanwould be expected, and the measurements lie on the curve(3) The results are compared with the results of other investi-gators above - 3 5 O . The slope of the curves do not agree, althoughthe individual measurements about - 30° are in agreement.(4) The vapour pressures between -60° and -40° have beenmeasured statically, and agree with those measured dynamically.(5) The results agree with those calculated from measurementsfrom the vapour pressure of water according to Ramsay and Young'slaw.(6) The values obtained agree with those of Scheffer and Treubfor the solid, if their two lowest values, which do not obey theRamsay and Young equation, are considered slightly too high.(7) An attempt t o calculate the chemical constant, the latent heatof vaporisation, the molecular heat, etc., according to Nernst'stheory, has been made, but further data are required.log p = 14.9166 + e(o-0604).ROYAL MILITARY ACADEMY,WOQLWICR

ISSN:0368-1645    

DOI:10.1039/CT9140500647

出版商:RSC    

年代:1914

数据来源: RSC

摘要:

658 CRYMBLE : THE ABSORPTION SPECTRA OFLXTrII.-I’lze A bsorption S’xctra of #onw MercuryCompourds.By CECIL REGINALD CRYMBLE (1851 Exhibition Research Scholar).THE results of an examiiiatioii of the absorption spectra of somesimple salts of mercury have been published recently by Ley andFischer (Zeitsch. a n o ~ g . Chem., 1913, 82, 329). The present work,which was carried out previous t o their publication, confirms someof their observations, and includes various compounds not yetexamined.Attention has been drawn by Ley and Fischer to the greatdiversity in the absorptive power of the mercuric haloids. Theeffect on the spectrum of change of solvent was noted in the caseof mercuric chloride, and the greater absorption of alcoholic solu-tions of this salt compared with that of aqueous solutions wasattributed t o the weakening of the spectrum in this region causedby the absorption of the solvent itself (alcohol).This explanationcan scarcely be accepted in view of the fact that a similar differ-ence has been observed in solutions of mercuric bromide, althoughthe absorption boundary of these solutions is much nearer thevisible region. The curve for an aqueous solution of the bromideis given in Fig. 1; comparative readings for 10 cm. ofM / 100-mercuric bromide solution are :Aqueous solution transmits t o 1/A 338Alcoholic ,, 9 , 9 , l / A 332The absorption curves for various aqueous solutions are given inFig. 1, and the curves for the acetate, bromide, and nitrate havebeen added to those previously described.The absorption spectrumof the propionate is similar to that of the acetate, but the curveis dispersed slightly towards the visible.The absorptive power of the highly ioriised mercuric chlorateaiid perchlorate varies slightly with the amount of acid added.The curve drawn in Fig. 1 (full curve) represents the absorption ofsolutions containing 8 molecular proportions of acid to 1 of salt,at which concentrations the two salts give the same spectra.Decrease in the concentration of the acid is accompanied by aslight increase in absorptive power, which is probably connectedwith the formation of basic salts.The difference between the absorption of the mercuric andmercurous ions has already been commented on (Zoc. cit.).The absorption spectra (Fig.2) of the two nitrates are of interesSOME MERCURY COMPOUNDS. 659owing to the presence of the selective absorption due to the nitrategroup, which gives rise to a characteristic band in N / 10-solution.I n the mercuric salt the selective absorption of the nitrate groupis apparent, because the absorption due to the mercuric ion a t thatdilution does not extend far enough towards the visible t o maskFIGS. 1 AXD 2.Hg' (ion) in H,O. -- Hg," (@?I,) i n H20.-..- HgUr, i i ~ H,O.- - - - - Hg( NO,), in H,O.- . _ HgCI, in H,O....... ... Mg(C2H302j2 i n H20.Oscillation friqucncies.28 3000 32 34 36 38 4000 42 44_ _ - _ _ Hg(N03), in H,O. - Hg,(NO,), in H,O.-- Hg," (i09~) in H,O... .... ... NaNO, in H,O.the band.I n the mercurous salt the nitrate band has almost dis-appeared, as the general absorption of the mercurous ion extendsup to, and partly overshadows, the region occupied by the band.The absorption of mercuric nitrite' stands in strong contrast to thatof the nitrates. The low conductivity of its aqueous solution showsthat the degree of dissociation is very small. This is in agreementVOL. cv. x 660 CRYMBLE : THE ABSORPTION SPECTRA OFwith the absorption curve (Fig. Z), in which there is no trace ofthe selective absorption peculiar t o the nitrite ion as seen in solu-tions of sodium nitrite, which possess a band in N/lO-solution witha head of frequency 1 / A 2900 (Fig. 2).The intense absorption of the iodine and bromine molecules com-pared with the weak absorption of single atoms of these elementsas ions or in alkyl compounds, suggests that the strong absorptionof mercuric haloids might be due solely to the close proximity inthe molecule of two halogen atoms.It has been shown (Crymble,Stewart and Wright, Ber., 1910, 43, 1183) that the introductionof a second iodine atom into a carbon chain causes no essentialchange in the spectrum unless the two halogen atoms are bothattached to the same carbon atom, when a marked alteration inthe absorption appears. F o r example, the series CH,I*CH,I,CH,*CHI,, and I,, represents both increase in absorptive powerand increase in proximity of the iodine atoms.It might be thought that the strong absorption of ethylidene andmercuric iodides is simply due to the adjacent position of thehalogen atoms, and that the central carbon or mercury atom playsno active part in the production of colour.To test this sugges-tion, some other metallic di-haloids were examined, but the resultsobtained indicate that the absorptive power of a halogen atom isnot necessarily increased by the introduction of a second similaratom into the molecule, even if both halogen atoms are attachedto the same central atom. For example, both 3N-calcium chlorideand 3N-potassium chloride are quite transparent. I n normal con-centration bromides absorb the extreme ultra-violet region, but noincrease in the absorption corresponding with one bromine atom isproduced by the presence in the molecule of a second bromine atom.This is seen by comparing solutions of sodium and magnesiumbromides in alcohol, in which solvent the dissociation is small :1 em.MgBr,/lO transmits to 1 / ~ 4402 em. NaBr/lO )) ), l/h 426Twice the length of sodium bromide solution is taken so as t ocompare lengths of solution equivalent with regard t o the bromineatom.Further confirmation is afforded by the comparison of alcoholicsolutions of potassium and barium iodides :5 cm. BaI,/1000 transmit to l / A 40810 em. KI/1000 ), ), l / h 406Dihalogen metallic derivatives may therefore be divided into twogroups, one in which the central metallic atom has no effect on theabsorption of the haloid atoms attached, for example, barium anSOME MERCURY COMPOUNDS. 661.magnesium, and a second in which the central and attached atomsform a system showing marked increase in absorptive power beyondthat due to a purely additive effect of the individual atoms.It issuggested tha't a characteristic property of the central atom in thelatter caseis power to vary in valency.These conclusions are applicable in similar manner t o otherfeebly ionised salts, such as the acetate and nitrite.Coinp/ex Mercuric Salts.Complex illercriric Chlorides.-In the experimental methodemployed solutions were photographed which contained constantstrengths of mercuric haloid and varying amounts of haloid acidor alkali haloid. The absorption curves of the complex chloridesin alcohol are given in Fig. 3. It will be seen that the absorptionof mercuric chloride is greatly increased by the addition of hydro-chloric acid.If a solution is taken in which the concentration ofthe mercury is Hg/lO, it is clear that the increase depends on theconcentration of the hydrochloric acid, and a maximum absorptionis attained when the mercuric chloride and the acid are presentin equimolecdar amounts. The solution then has a compositioncorresponding with HgCl,,HCl. Further increase in the concentra-tion of the acid-causes no change in the absorption curve, and i tmust be assumed, therefore, that the increased absorption is dueto the presence of the complex HHgCI,. Additional proof of thismay be obtained from a study of the curves, whence it appears thatin solutions of the concentration Hg/lO the formation of thiscomplex is quantitative.This becomes apparent on comparison ofthe absorption of solutions in which the concentration of the acidis varied with that shown by a solution of the compositionHgCl,,HCI, which is the solution giving the maximum absorptiona t this dilution. F o r example, 10 cm. of a solution (concentrationHg/ 10) of the composition HgC1,,1/ 1OHC1 gives the same absorp-tion as 1 cm. of a solution (concentration Hg/lO) of the compositionHgCl,,HCl. Consequently, it may be assumed that the samenumber of absorbing particles is present in both cases. Thesimplest explanation of this lies in the supposition that each mole-cule of acid added combines with a molecule of mercuric chloride,forming the complex HHgCI,. This continues until all ths chlorideis converted into complex, when further addition of acid causes nochange.As it was desirable to confirm this result by other experimentalmethods, the molecular weight of mercuric chloride in alcoholbefore and after the addition of ammonium chloride was deter-mined by the elevation produced in the boiling point of the solu-X X 662 CRYMBLE : THE ABSORPTION SPECTRA OFtio,n.Walker apparatus :The following experiments were carried out in the Lumsden-Ammo-No.of B. p. of chloride. chloride. B. p. of solution, M.W.Mercuric nium Volume.ofexpt. solvent. Grams. Grams. solution. C.C. HgCI,= 271I. 78.2' 2.409 79.40" 12 26179.05 16 2760.4743 78-70 27 278I1 78-15 2-7124 0-9134 79-125 20-5 211111. 78.35 2-166 79.40 12 2680.8584 79.175 22 18679.00 27 192FIG.3. PIG. 4.Alcoholic solidion. Apiicozis Yolutiov.Oscillntiou frcqiLemies.- HpCIp HgCI, -t MC I. ...... HgCl,+ 1/1OMC'1. ~ HgCl, -F 10MC1SOME MERCURY C9MPOTTNDS. 663The figures in the .last column are derived by calculating themolecular weight of the mercuric chloride from the different risesin the boiling point, irrespective of the ammonium chlorideadded.I n experiment I the values obtained for the molecular weightare 261 and 276. One molecular equivalent of ammonium chloridewas now added, and the molecular weight redetermined. Theresultant value, 278, showed that the number of active units insolution remained the same, indicating the formation of the com-plex NH,Cl,EgCl,.I n experiment I1 the amount of ammonium chloride added wasmore than equivalent to 1 molecule of mercuric salt.A markedfall in the molecular weight occurred, showing that the numberof active units had greatly increased, and that a portion ofammonium salt remained uncombined.A concluding experiment was made (111) in which two molecularequivalents of ammonium salt were added. As in experiment 11,a marked increase in the number of active units was apparent, andthe values obtained were found to agree very closely with thatcalculated for a solution of the composition NH,Cl,HgCl, +NH,Cl (188).The conclusion from both methods is the same, namely, that thecomplex HHgC1, or MHgC1, is formed quantitatively a t theseconcentrations in alcohol, but no higher complex is formed.The complex, although stable a t decimolecular concentrations,becomes dissociated a t greater dilutions, the degree of dissociationfor a given concentration of mercury depending on the concentra-tion of the added chloride.The curve for the solution of the com-position HgCl,,lOHCl is continuous until a dilution of Hg/ 10,000is reached. On further dilution a break in the curve is observed(Fig. 3), showing that dissociation of the complex has commenced.As the concentration of the added chloride decreases relative tothe mercury, the complex begins t o dissociate much sooner. Froman examination of the curves i t is possible t o determine approxi-mately the concentration of the complex at a given dilution.The absorption curves of the complex formed in aqueous solutions(Fig.4) are more complicated owing to the high dissociating powerof the solvent. I n decimolecular solutions of mercury salts, withincreasing concentration of the added chloride, the limit of generalabsorption is gradually displaced towards the red, and a maximumabsorption curve is not attained until a ratio of 30 molecules ofchloride to 1 of mercuric salt is reached. No further increase inabsorption can be attained, even with a solution of mercuricchloride in concentrated hydrochloric acid. This maximum curv6G4 CliYRlRLE : THE ABSORPTION SPECTRA OFis almost identical with that for alcoholic solutions, lying just onthe ultra-violet side of the latter.Differences between the alcoholic and aqueous solutions appeara t higher dilutions, where the greater dissociating power of thewater causes disruption of the complex.I n an alcoholic solutionof the composition HgC1, + lOMCl the absorption curve shows nobreak until a dilution Hg/10,000 is reached, whereas in an aqueoussolution of similar composition a marked break in the curve occursFIG. 5. FIG. 6.Oscillation f rrpiicncil s.28 3000 32 34 36 38 4000 26 25 3000 33 34 36 38 4000FIG. 6 . FIG. 6.HgRr, in EtOH. -- HgBr,+ KBr in EtOK- . - HgBr, in H,O.-..- HgHr, + 200KBr i n H,O.HgI, in EtOH._.- HgI, + KI in 1ttO H.Hgl, + 100 li I i i ~ H,O.on passing from Hg/lO to Hg/100, indicating that dissociation ofthe complex has commenced in the latter dilution.Itifluence of t h e Ccrtiorr oil t h e Spctw.--The absorption spectrumof the complex is independent of the nature of the cation added,provided that the base itself does not absorb light.Identicalresults have been obtained in alcohol with hydrogen, ammonium,and sodium chlorides, and in water with hydrogen, potassium,sodium, and ammonium chloridesSOME MERCURY COMPOUNDS. 665Complex Mercuric Bromides and Iodides.-As will be seen frolnFigs. 5 and 6, the behaviour of the complex bromides and iodidesis quite analogous t o that of the chlorides, only the absorptioncurves are more displaced towards the red end of the spectrum.The quantitative formation of the complex MHgX, can be followedwith both the bromide and the iodide, and the maximum absorp-tion curve is reached (in alcohol) when the two salts are present inequimolecular proportions, further addition of alkali salt producingno change.The maximum curves for the aqueous solutions lie on the ultra-violet side of those for the alcoholic solutions, and the differencebetween the two solutions, which is very slight in the case of thecomplex chlorides, is quite marked in the complex bromides a,ndiodides.I n accordance with the Abegg and Bodlander rule, the stabilityof the complex (that is, its power to resist dissociation) depends onthe electro-affinity of the acidic ion, and increases in the seriesCl-Br-I.This stability can be compared by noting the propor-tion of haloid salt t o mercuric salt necessary to maintain the maxi-mum absorption a t a given concentration. For example, themaximum curve for aqueous solutions a t Hg/ 100 concentration isattained with 2 molecular proportions of potassium iodide to 1 ofmercuric iodide, 10 molecular proportions of potassium bromideto 1 of mercuric bromide, or 100 moleculw proportions of potassiamchloride to 1 of mercuric chloride.The nature of these complex mercuric haloids in aqueous solutionhas been studied extensively by various methods.Evidence in favour of the existence in solution of the complexHgX2,2MX has been adduced from solubility data by bittie (A1211.Chim.Phys., 1881, Lv], 22, 551) and Richards and Archibald(Zeitsch. physikal. Chem., 1902, 40, 385) ; from observations on thedepression of the freezing point by Le Blanc and Noyes (;bid.:1890, 6, 389) ; and from E’.Jl.F.determinations by Slierrill (ibid.,1903, 43, 705). I n addition to this complex many other types arecapable of existence, and Jander (Zeitsch. EZe?ctrochem ., 1902, 8,688) concluded from solubility measurements that in dilute solu-tion a t least the complex HgCbC1’ preponderated. Sherrill (Zoc.cit.) states, however, that the assumptions on which Jander’s con-clusions are based are not justifiable. I n alcoholic solutions thespectroscopic evidence leads to the conclusion that the complexHgX,,MX alone is formed, and this has been confirmed in the caseof the chlorides by molecular-weight determinations. In aqueoussolutions the evidence is not so clear, as objections can be raisedapinst the comparison OC the absorption curves of aqueous an666 CRYMBLE : THE A13SORPTION SPECTRA OFalcoholic solutions.It should, however, be noted that, a t the con-centrations examined, the maximum absorption curve of theaqueous solution coincides with, or is on the ultra-violet side of, thesimilar curve for alcoholic solutions. The alternative conclusionsto be drawn are that the addition of an extra chlorine atom form-ing the complex HgX”, has 110 effect on the absorptive power ofthe molecule, o r that the complex HgX’, forms the preponderatingconstituent of aqueous solutions a t the concentrations considered.It must be admitted that the latter deduction is a t variance withthe bulk of the evidence mentioned above.Some Orgaikic Derivatives of Mercury.-The dialkyl mercuryderivatives are colourless, volatile liquids, devoid of any metallicproperties. The alkyl salts (RHgX, X being an acidic radicle)are white, crystalline solids, the lower members of which dissociatein solution t o a certain extent, yielding the ions RHg’ and X.I n these compounds the group RHg- behaves like a univalentmetallic radicle.Quite recently Kraus (1. Amer. Chem. SOC.,1913, 35, 1732) has succeeded in isolating the groups CH,Hg,C,H,Hg, and higher members of the series, and he describes themas black deposits of high conductivity, which cohere on pressureand exhibit metallic reflection.It is of importance to observe how the absorptive power of themercury atom is influenced by these changes in its chemical nature.The absorption curves of alcoholic solutions of mercury methylchloride, bromide and iodide, mercury ethyl chloride and iodide,and mercury dimethyl and diethyl, are reproduced in Figs.7 and 8.Ethereal solutions of mercury dimethyl and diethyl possess thesame absorptive power as t!ie corresponding alcoliolic solutions.Mercury methyl and ethzl chlorides have already been examinedby Ley and Fischer, and they have pointed out that the replace-ment of one chlorine atom of mercuric chloride by an ethyl ormethyl group causes a decrease in the absorptive power of themolecule. A similar change has been noted in the iodides andbromides, and is apparent on comparing the ciirves for these com-pounds with those previously given for mercuric bromide andiodide. An aqueous solution of the base CH,Hg*OH was pre-pared according t o the method described by Dunhsupt (An tinleu,1854, 92, 381), and was found to be diactinic in decimolecularconcentration.The compounds of the base CH,Hg may now be arranged inorder of increasing absorptive power as follows : CH,Hg-OH,-C1, -Br, and -I, the absorption of the alkyl haloid beingalways less than that of the corresponding dihaloid compound.Ley and Fisclier attribute this change t o an hypsochromic actioSOME MERCURY COMPOUNDS.667of the carbon-mercury linking, and they account for the diminu-tion in absorption on passing from the mercuric ion to mercuriccyanid,e by the presence of two carbon-metal linkings in the lattercompound. Applying this interpretation t o the dialkyl derivatives,one would expect that in the series C1-Hg-C1, C1-Hg-R,R-Hg-R the dialkyl compound would show even less absorptionthan the intermediate compound C1-Hg-R.Reference t o Figs.FIG. 7. F.G. 8.Oscillation frequencies.32 31 36 38 4000 42 34 36 38 4000 420-FIG. 7. FIG. 8. - - - CH3'HgC1 i / ~ EtOH.CH3'Hg*CH, zn e i ? ~ . ~ .- - C,H;llgCI in &OH.- . - C2H,*HgI it( EtOH.C2H,'Hg'C,H, i r j EtCH. - . - Cti3*I1gBr in ElOli. -- - . - CH,*HgI it& EtOH.7 and 8 will show that this is not the case, and, in fact, whilstmercury dimethyl possesses much the same absorptive as mercuricchloride, mercury diethyl is noticeably more absorptive. It isevident that the introduction of the second alkyl group producesthe reverse effect to that caused by the entry of the first.Since the methyl and ethyl groups themselves do not possessabsorptive power, the absorption of the dialkyl derivatives isprobably allied t o the electrochemical change undergone by th668 THE ABSORPTION SPECTRA OF SOME MERCURY COMPOUNDS.mercury atom which takes place on formation of these compounds,and is associated with the consequent disappearance of its metallicproperties.EXPERIMENTAL.The metallic salts employed were carefully recrystallised beforeuse.It was found that alcoholic solutions of barium iodide decomposeon keeping, and absorption bands begin to appear with heads a tfrequencies similar to those described for various carbon-iodinecompounds. This difficulty was avoided by photographing thesolution as rapidly as possible after preparation.The solution of mercuric nitrite was obtained by treating a silvernitrite solution with an equivalent of mercuric chloride, filteringoff the precipitate, and diluting t o the required volume.Solutions of the oxy-salts were prepared by dissolving the purifiedoxide in the desired excess of acid.Mercury dimethyl (redistilled) boiled at 91--92O, and mercurydiethyl (redistilled) a t 1 5 7 O .The alkyl haloid salts were recrystallised from alcohol severaltimes.The bromides and iodides were prepared by boiling solu-tions of t.he alkyl chlorides with silver oxide, removing the silverchloride, and precipitating the salts from the solution of the freebase by the addition of bromides or iodides.Mercury methyl chloride melted a t 170°, mercury methylbromide a t 160°, mercury methyl iodide a t 145O, mercury ethylchloride a t 192*5O, and mercury ethyl iodide a t 1 8 2 O .Gonclusions.(1) The absorption spectra of mercury compounds varies greatlywith the nature of the attached groups.The absorptive power ofthe molecule is increased by the different groups in the followingorder : (CN),, Hg" (ion), Cl,, (C2H302)2, (C3H50&, Hg2** (ion),(2) It has been shown that the absorptive power of the com-pound does not reside solely in the attached groups, but dependson an interaction between them and the mercury atom.(3) The entrance of the mercury atom into complex haloid saltsis accompanied by an increase in the absorptive power of the com-pound, which is independent of the second positive atom (forexample, H, Na, K, NH,).The composition of the complex formed can be determined, andits progressive dissociation on dilution followed.(4) The haloid complex HgX,,MX is formed quantitatively a tdecimolecular concentrations in alcohol. There is no spectroscopicBrz, (NO,), 12RELATIOS BETWEEN THE ABSORPTION SPECTRA, ETC. 669evidence of the formation of any higher complexes such asHgX,,ZMX in either alcohol or water.(5) Replacement of one chlorine atom in mercuric chloride by amethyl or an ethyl group causes a decrease in absorptive power, butsimilar replacement of the second chlorine atom causes an increase.The author desires t.0 express thanks to the Research Fund Com-mittee of the Chemical Society for a grant in aid of the expenseof this investigation.UNIV EI:SI'L'Y COLLEGE,GOWET1 ST., W.C

ISSN:0368-1645    

DOI:10.1039/CT9140500658

出版商:RSC    

年代:1914

数据来源: RSC

摘要:

RELATIOS BETWEEN THE ABSORPTION SPECTRA, ETC. 669LXVII1.-The Relation Between the Absorption t i j x c t r aof Acids and their Sults. I’aPSt If.By ROBERT WRIGHT (1851 Exhibition Scholar).IN a former communication (T., 1913, 103, 528) the absorptionspectra of several acids and their sodium salts were compared, andit was shown that in many cases change in absorptive power takesplace on neutralisation, even when alteration of molecular structureis hardly possible. Hantzsch and Scharf (Ber., 1913, 46, 454, 3570)have further compared various acids with their salts and esters inseveral different solvents, the results-in as far as they are incommon-being in agreement with those published in Part I. ofthis paper; whilst Henri (Ber., 1913, 46, 464, 3650), by measure-ment of extinction-coefficient, has shown that the sodium salt offormic acid is about three times, and that of acetic acid two anda-half times, less absorptive than the corresponding acid.I n the present paper the work has been extended to metals otherthan sodium, care being taken that the metal used does not of itselfshow any absorption in the range of spectrum investigated. I n allcases the salts were prepared by neutralising a definite quantity ofthe acid, either with the metal itself, as in the case of magnesium,or more usually by means of the oxide or hydroxide, so that thedifferent salts of a given acid all contain equal quantities of thesame sample of acid, and are therefore of equal purity as far asthe acidic radicle is concerned. The only exception to this methodof preparation was in the case of zinc acetate, when a recrystallisedspecimen of the salt, was used670 WRIGHT : THE RELATION BETWEEN THE ARSORPTIONTABLE I .N-Formic Acid .Acid ............................Li salt .......................Na .............................K ...........................NH..............................Ca .............................Sr .............................Ba .............................M g .............................10 cm .Spectrum trans- Spectrum trans-mitted to h1 cm .mitted to A2520 24002420 23202430 23302420 23202430 23252390 23202400 23252400 23202390 2380TABLE I1 .N-Acetic Acid .10 cm .mitted to A1 cm .mitted to ASpectrum trans- Spectrum trans-Acid ......Li saltNa y yK ..NH.11Ca .. Sr .. Ba I ?Mg .................................... ................................. .................................................................. ................................. ................................. ............................................................... .............................. ..............................2460240024102410242023902430240024102400TABLE I11 .N-PTopionic Acid ............................... .............................. .............................. .............................. ... C .......................... .............................. .............................. ............................................................10 cm .Spectrum trans-mitted to A24802460246024602465247024602465246023502320232023202320232523302325232523201 cm .Spectrum trans .mitted to A23602340233523402336233023402340234SPECTRA o F ACIDS AND THEIR SALTS. PART 11. 671TABLE IV.N-Bu tyric A cid.10 cm.Spectrum trans-mitted to hAcid .................................... 2490Li salt .............................. 2465Nrt ), .............................. 2465K : ............................... 2460NH, ,) .............................. 2460Sr ,) .............................. 2465Ba ,) .............................. 2470Mg ), .............................. 2460Ca , , ..............................24601 cm.Speotrum transmitted to A239023302330233023302330233023402340It will be seen that in all cases the neutralisation of the acid, bywhatever base, produces the same effect, although in slightly vary-ing degree-that is, the salt is always more diactinic than the aciditself. Thus the quality of the change produced is independent ofthe metal used, and so must be due to some common property ofsalts which is non-existent in the corresponding acids. It seemspossible that the change in the degree of ionisation which takesplace on neutralisation may be the explanation, as in all the abovecases the salts are much more highly dissociated than the acids fromwhich they are formed.In order to determine _whether there is any definite relationbetween the change in absorptive power and the change in thedegree of ionisation which takes place on neutralisation, the absorp-tion spectra of a number of inorganic and organic acids of differentstrengths have been compared with those of their sodium salts.The comparison in all cases was made in aqueous solution, and acidsof a complicated nature, which are more or less liable to change ofstructure on neutralisation, were avoided.For the sake of convenience, the acids examined are roughlygrouped according to their strengths.Stroug A cids.-Most of the strong inorganic acids and their saltsare too diactinic to admit of investigation.These diactinic acidsinclude sulphuric, hydrochloric, chloric, perchloric, hydrobromic,and the various phosphoric acids and their salts; again, others suchas hydriodic decompose so readily that their examination isimpossible.Nitric acid has been shown by Hartley (T., 1902, 81,556) to have a spectrum identical with that of its potassium salt,and in Part I. of this paper the same has been shown to be trueof the moderately strong picric and chloroacetic acids. Benzene-sulphonic acid, the absorption-curve of which is shown in compari-son with sulphurous acid, has the same absorption as its sodiu672 WRIGHT : THE RELATIOX BE'L'WZEN THE ABSORPTIOI?salt., and the same is seen to be the case with bromic and iodic acid,the absorptions of which are here given:TABLE V.10 c1n. 1 cm.mitted to mitted toSpectrum trans- Spectrum trans-Bromic acid (N/10) ............... 2700 2490Sodium bromate ...................2700 2490Iodic acid (N/10). .................. 2870 27 10Sodium iodate ..................... 2870 2710The conclusion may be drawn that strong acids which do notchange their structure on neutralisation show the same absorption astheir sodium salts.,4 cids of Moderate Strerzgth.-In this group are considered acidsof a strength comparable with acetic or formic acid, or, in otherwords, such acids as obey Ostwald's dilution law and give anionisation-constant. The fatky acids have already been dealt with,and i t is seen that, no matter what metal is used, provided its ionis non-absorbent, the salt is always more diactinic than the acid,the greatest difference between the transparency of acid and saltbeing found in the case of the strongest acid, falling off as weakeracids are dealt with.The same result was found in the seriesbenzoic, phenylacetic, and phenylpropionic acid, the spectrum ofthe last being almost identical with that of its salt (Part I). Theoxalic acid series was also examined in Part I of this paper, but asthe two hydrogen ions of the dibasic acids have different ionisation-constants, i t was considered advisable to examine the effect ofneutralising each separately.TABLE VI.10 cm.Spectrum trans-mitted toOxalic acid (N/10) ................ 2900Na, salt ( N / 10) ......................Malonic acid ( N ) ................... 2520HNa salt ( N ) ........................2500Na, salt ( N ) ......................... 2485Succinic acid ( N ) .................. 2470HNa salt ( N ) ......................... 2470Na2 Salk ( N ) ......................... 2470Tartaric acid ( N ) .................. 2530Na, salt ( N ) ......................... 2470HNa salt (N/10). .................... 29002740HNa salt ( N ) ........................ 25101 cmSpectrum trans-mitted to25002500244024302410237523602350234024702430m nA'.10.00.1630.00660.09SPECTRA OF ACIDS AXD THEIR SALTS. PART 11. 673It will be seen that oxalic acid behaves in the same way as astrong acid in the neutralisation of the first hydrogen ion; this,indeed, has been shown by Hantzsch and Scliarf in the paperalready mentioned, and it is only on replacement of the second ionthat an increase in transparency takes place.This result is inagreement with the view that the first ion is much more highly dis-sociated than the second, and explains why a strong acid like oxalicFIG. 1.Frepzicncics.26 28 30 39 34 31 38 40 42Upper continnons curve = HNO,.Upper dotted curve L- NaNO,.Lowcr jigrrre.I. =NaHSO,. 11 I. = C6H,*HS0,.Strdight liiie = NR,SO, 11. = H,S03.can behave like its weaker homologues, but only with regard to itssecond or weaker hydrogen ion. I n malonic acid both of thecarboxyl groups are feebly ionised, so we find a gradual change onthe neutralisation of each. The spectra of succinic acid and itssalts are practically identical, but on strengthening the acid by theintroduction of hydroxyl groups we again find an increase in trans-parency on neutralisation674 WRIGHT : THE RELATION BETWEEN THE ABSORPTIONThe absorption curves for nitrous and sulphurous acid and theirsalts are shown (Fig.1). In both cases the results are similar tothose obtained for moderately strong organic acids, that is, the saltsare more diactinic than the acids. It seems probable, however, inthe case of sulphurous acid we have a change of structure takingplace, for the absorption band, which is well marked in the acid, isabsent in the neutral salt. I n view of the doubtful constitution ofsulphurous acid, the result is of interest, and would probably repayfurther investigation. The curve for nitrous acid is of somewhatdoubtful value, as this substance decomposes very readily; the acidwas prepared by adding excess of hydrochloric acid to the requisitequantity of sodium nitrite, the solution being kept a t zero by meansof ice.The conclusion may be drawn that in acids of moderate strengththe sodium salt is more diactinic than the acid itself, the differencebeing greatest in the case of the stronger acids, and falling off tonothing with the weaker homologues. It should be noted that thisgradation only holds good for the acids of a given series; forexample, we cannot compare succinic or phenylpropionic acid withacetic.Very Feebly Acidic Substances.-This group comprises the phenolsand their derivatives, together with such acids as arsenious acid andhydrogen sulphide.Owing to their want of absorptive power, boricacid and hydrocyanic acid cannot be examined. As in the case ofthe oxalic acid series, the plan has been followed of neutralising thedibasic acids in stages, so that the effect of each replacement ofhydrogen can be noted separately.TABLE VII.10 cm. 1 cm.Spectrum trans- Spectrum trans-mitted to mitted to K .Phenol (N/100). .....................Na salt.. ...............................p-Cresol (N/100) .....................Na salt.. ...............................o-Cresol (N/100) ...................Na ...................................Thymol (N/lOO) ...................Na salt .................................Guaiacol (N/lOO) ..................Na salt .................................o-Hydroxybenzoic aoid (N/100)C02H neutralised ..................CO,H and OH neutralised ......2890314029803230291031602930319029403160.3415340035002840 O*O,l32930 O-O,lI306031802850 0-0542309028703090288030903310 0- 10232603390-SPECTHA OF AtrIDS AND ?HEIR SALTS.PART 11.TABLE VII. (corztinued).10 cm.Spectrum trans-mitted top-Hydroxybenzoic acid (N/100) 2980CO,H neutralised .................. 2940CO,H and OH neutralised ...... 3170Arsenious acid (N/10) ............ 2390H2Na salt ............................ 2530HNa, salt ............................ 2560Na3 salt .............................. 2560Hydrogen sulphide (N/10) ...... 2480HNa salt ............................2790N+ salt .............................. 27901 cm.Spectrum trans-mitted t o2930287030908250238024102410233026802680(i75K .0.0029I0 ~ 0 ~ 5 7The conductivity constants are those given by Kohlrausch, withthe exception of the constants for phenol and hydrogen sulphide,which are taken from a paper by Walker (T., 1900, 77, 5). It willbe seen that the neutralisation of a very feeble acid causes anincrease in the absorptive power. Further, the more strongly acidiccarboxylic groups of the hydroxybenzoic acids, more especially inthe case of the para-compound, behave like the moderately strongacetic or benzoic acids, in that their neutralisation causes a decreasein absorption, and it is only on neutralisation of the feebly acidicphenolic group that an increase in absorptive power takes place.This effect is much more marked in the para- than in the ortho-compound, possibly because the two groups are more separated inthe former case, being thus less liable to affect each other.Itshould also be noted that the neutralisation of the final hydrogenatom in arsenious acid and hydrogen sulphide produces no changein their absorption spectra; this is probably due to the high degreeof hydrolysis of these fully neutralised salts, and points t o the exist-ence in solution of the acid salts alone. I n order to show the effectof gradually increasing the acidic properties of a phenol, the absorp-tion curves are given for o- and p-nitrophenols, 2 : 4-dinitrophenol,and picric acid (Fig.2). It is clear that as the strength of the acidincreases, the relation between the absorption of acid and salt under-goes marked change; thus, the large increase in absorption producedon neutralising o-nitrophenol ( K = 0*000043) and pnitrophenol(K=0-000012) has almost vanished in the case of 2 : 4-dinitro-phenol ( K =O-OOS), whilst the strong picric acid shows the samespectrum as its salt. Again, it should be noted that we cannotcompare acids of widely different structure, for although2 : 4-dinitrophenol is stronger than acetic, yet i t behaves in thereverse way on neutralisation, becoming more absorbent instead ofmore diactinic.VOL. cv. Y 676 WRIGHT : THE RE1,ATION BETWEEN THE ABSORPTIONIt is very generally held that the change produced on the neutral-isation of phenolic substances, especially those containing a nitro-group, is one of structural rearrangement (Baly, Edwards, andStewartj T., 1906, 89, 514; also Buttle and Hewitt, T., 1909, 95,1756), and it is quite probable that in some cases this view is correct.VIG.2,1+cq7se i~cies.20 22 24 26 28 30 32 34 33 35 40 42 4.1 46Continuous curves =acids.Dotted cnrves =sn7ts.I st curve : o-ATitrophcnol.2nd ,, p-Nitrophenol. 4th ,, Picric acid.3rd curve : 2 : 4-Dikh-ophcnol.On the other hand, it should be borne in mind that very often nofundamental change is produced in the form of the absorptioncurve, neutralisatdon neither creating nor destroying a band, butonly altering its position and intensity.It is, indeed, a matterof some difficulty to decide when a change in structure does osPECI'KA OF ACIDS AND THEIR. SALTS. PART 11. 677does not take place; for example, 2 : 4-dinitrophenol is shown to besimilar to, but slightly more diactinic than, its salt. Buttle andHewitt have, however, shown (loc. c'1.t.) that on adding hydrochloricacid (100 mols. HC1: 1 inol. 2 : 4-dinitrophenol), not only is thisdiactinity greatly increased, but the form of the absorption curveis also changed,. I n fact, the question would seem t o resolve itselfinto one of degree: How great a change in absorptive power is per-missible w'ithout assigning change of structure ? If the phenoxidesdiffer in structure from their parent substaiices, is the same true ofacetates and f ormates ?The conclusion may be drawn that very feebly acidic substancesare less absorptive than their sodium salts.Although i t is certainly - desirable that more extended observa-tions, preferably of a quantitative nature, should be made beforeany definite conclusions are drawn, still it must be borne in mindthat comparisons can only justly be made between acids of a similarnature, and that the absolute certainty of absence of change ofstructure can never be depended on, more especially in the case ofcomplicated substaiices.The following explanation has at leastthe claim of simplicity, and i t is in agreement with the so farobserved facts.A simple electrolyte in solution consists of free ions and non-ionised molecules; the latter may be considered t o be in a state ofstress, as they 'break down on further dilution.The assumptionwhich will bf made is that these stressed molecules have an in-creased power of absorption, being not only more absorbent thanfree ions, but also than molecules incapable of ionisation, and there-fore free from what may be called ionising stress.Now in strong acids the degree of ionisation approximates moreor less with that of the salt, that is, the percentage of stressedmolecules is nearly the same in the two cases, and therefore, sinceneither the acid hydrogen nor the metal which replaces it showany absorption in the range of spectrum investigated, the absorp-tion spectra are identical. With weaker acids, such as formic,acetic, etc., the proportion of non-ionised, and therefore stressed,molecules is much greater than in the corresponding highly ionisedsalt; we therefore find that the acid is the more absorbent, oxalicacid presenting the interesting case of an acid with one strong andone weak hydrogen atom.If we now consider the acids cf a series, we find that as the acidsget weaker their greater absorptive power as compared with theirsalts diminishes, or even vanishes altogether, whilst extremelyfeeble acids absorb much less than their sodium salts.This is duet o the fact that these very weak acids are almost non-ionisable; thatY Y 678 RELATION BE rWEEN Tf1E ABSORPTION SPECTRA, ETC.is, although there are more non-ionised (stressed) molecules present,yet these molecules are so inert that the ionising solvent has littleor no effect on them, and they are therefore stressed t o a very smallextent.O n the other hand, their salts are good electrolytes, andso contain non-ionised molecules capable of ionisation, and there-fore in a state of stress, which readily accounts for their greaterabsorptive power. I f we go still further and deal with acids soweak that their salts are practically completely hydrolysed in solu-tion, we arrive a t a state where the spectra of acid and salt areagain identical. This state seems to have been reached in the casefinal hydrogen atom in arsenious acid and hydrogen sulphide, forin neither case does the replacement of this atom by sodium pro-duce any change in the absorption spectrum.We might represent the relation between the absorptive powerof a completely ionised salt and its acid, which will be consideredcapable of existing in any degree of ionisation, by means of theaccompanying curve (Fig. 3).The horizontal line represents theunvarying absorptivs power of the fully ionised salt, whilst theFIG. 3.-22 J2&-& h$24 Straight line =salt of' comtaitl iomkaliou c6nd abwptioa.Cur red line = acid. -+ Decyeasiny imisation of acid oiily.absorption of the hypothetical acid, varying from complete ionisa-tion a t A to absolute inertness a t B, is given by the curved line.As the ionisation of the acid decreases, the absorption a t firstincreases, and then diminishes, becoming again equal to, and thenless than, that of the salt, the spectra again becoming the samewhen the acid has grown so weak that its salts are completeIyhydrolysed.Summary.-Leaving aside cases of structural change, the resultsseem to indicate that:(1) The nature of the change in absorption produced on theiieutralisation of an acid by a diactinic base is independent of thebase used, but varies slightly in degree with the different bases.(2) Strong acids show the same absorption as their sodium salts.(3) Acids of moderate strength, such as acetic, are more absor-bent than their sodium salts, the difference between the spectrumof the acid and its salt growing less when we deal with weakeracids.(4) Very feebly acid substances are less absorbent than theirsaltsORGANIC DERIVATIVES OF SITATCON. PART XSTT. 679( 6 ) A general explanation of the observed facts can be made byassuming that a non-ionised molecule capable of ionsation exists ina state of stress, and is more absorptive than a similar free ion orthan a molecule incapable of ionisation.UX’IVERGITY COLLEGE,GOWER STREET.QC‘EEN’S UNIVERSITY,B ELFA S1’

ISSN:0368-1645    

DOI:10.1039/CT9140500669

出版商:RSC    

年代:1914

数据来源: RSC

摘要:

ORGANIC DERIVATIVES OF SILICON. PART XSTT. 679LX IX.--Organic Dei-ivatiztes of Silicon.. Pa9-t X X [I.The So-culled Siliconic Acids.By JOHN ARTHUR MEADS and FREDERIC STANLEY KIPPING.IN 1870, Friedel and Ladenburg obtained an amorphous solid byhydrolysing ethylsilicon trichloride, EtSiCl,, with water (Ber., 3,15); they named this product silicopopionic acid and assigned toit the formula C,H,*SiO*OH.Shortly afterwards, by similar methods, Ladenburg preparedother amorphous organic silicon compounds which he described assiliconcetic acid, CH,*SiO*OH, silicob emzoic acid, C,H,*SiO*OH,and rilicotolylic acid, C7H7*SiO*OH, respectively ( A 11 nnlen, 1875,179, 143); these products, which he believed to be analogous t othe carboxylic acids, were classed by him as siliconic acids.I n comparatively recent times, Khotinsky and Seregerikoff pre-pared ‘ I silicobenzoic acid,” ‘( silicoxylylic acid,” and ‘’ silico-naphtltoic acids *’ by hydrolysing eshrs of the type Sift(OEt),with cold concentrated hydriodic acid (Ber., 1908, 41, 2946); andMelzer, a t Ladenburg’s suggestion, prepared a number of aliphaticand aromatic “ siliconic acids ” by hydrolysing trichlorides of theformula RSiC1, with cold water (Ber., 1908, 41, 3390).All these so-called siliconic acids were analysed, with results whichin most cases agreed exceptionally well with those required by therespective formuk ; the percentages of carbon, hydrogen, andsilicon were, in fact, usually given to two decimal places (or evento three decimal places in Melzer’s estimations of hydrogen).Further, except in the cases of those which were infusible, fairlydefinite melting poinh were given t o nearly all these (( acids.”I n view of these facts, a further study of the siliconic acids inorder to ascertain whether or not such compounds really existed,might have been deemed superfluous; nevertheless, such an in-vestigation has been undertaken.The results obtained by one of ua alone, and in conjunction wit680 MEADS AND KIPPING:Robison (T., 1912, 101, 2108, 2125, 2142; this vol., pp.40, 484)have clearly shown thai%ompounds containing the group >Si(OH),are very prone to undergo condensation ; that instead of giving thecorresponding oxides >SiO by the h 6 of the elements of water,such dihydroxy-derivatives give rise by progressive condensatioiito a series of open- and closed-chain products formed from two,three, four, and more molecules of the diol.Consequently, it seemed highly probable that trihydroxy-corn-pounds, obtained by the hydrolysis of trichlorides, RSiCl,, o r ofesters, RSi(OEt),, would behave like the dihydroxy-derivatives, andgive mixtures of condensation product4 the nature of which, as inthe case of those obtained from the silicanediols, would vary withthe conditions of hydrolysis and with the subsequent treatment ofthe product.The validity of these inferences has been tested by an examina-tion of the products of hydrolysis of phenylsilicon trichloride, andthe results have shown that the so-called silicobenzoic acid is amixture, doubtlm a complex one, which varies in character ac-cording t o the method of preparation.When, for example, phenyl-silicon txichloride is hydrolysed with ice-cold water, i t gives aglue-like product which is readily soluble in ether, but when thetrichloride is hydrolysed with steam, it gives a hard, vitreoussubstance which is quite insoluble in ether and in all the commonsolvents.The hydrolysis of the trichloride with a dilute, ice-cold aqueoussolution of ammonium hydroxide gives a glue-like product whichdissolves freely in ether; if, however, hydrolysis is carried out a tthe ordinary temperature, all other conditions remaining un-changed, the product is a powder which is only partly solublein ether.The hydrolysis of the trichloride with an aqueous solution ofpotassium hydroxide gives a solution of potassium salts; when thissolution is treated with carbon dioxide in the cold, the precipitatedacid is completely soluble in ether, whereas, if precipitation occursa t the ordinary temperature, the acid is only partly soluble.All the soluble preparations of silicobenzoic acid may be separatedinto physically different fractions which themselves have the pro-perties of mixtures; these fractions differ very widely in solubility,melting point, and other physical properties, so that it may beinferred that their components differ very considerably in molecularweight.The results of analyses agree with the view that all the fractionsar9 mixtures of condensation products formed from phenyltri-hydroxysilicane by the loss of the elements of water.The morORGANIC DERLVB'I'IVES OF SILICON. PART XXII. 681readily soluble fractions have compositions differing very little fromthat of silicobenzoic acid, C6H,*Si02H ; as the solubility diminishes,the composition of the fraction approaches more and more closelyto that of the anhydride, (C,H,*SiO),O.l\iIolecular-weight determinations in acetic acid solution giveresults which indicate that the average molecular weight of themost soluble components of the glue-like preparations correspondswith that of a condensation product formed from 4 or 5 moleculesof phenyltrihydroxysilicane. The less soluble fractions of thesepreparations give higher results, which correspond approxi-mately with those required for compounds formed from 5 or 6molecules of the trihydroxy-derivative. It may be inferred, there-fore, that the very sparingly soluble and the insoluble fractions,which have approximately the composition (PhSiO),O, containmuch more complex substances.The question arises, what is the nature of the compounds con-tained in these various preparations? Here we enter a region ofspeculation, because the theoretically possible condensation pro-ducts of the trihydroxy-compound are innumerable ; we may, there-fore, defer the consideration of this difficult problem until ourexperiments have been carried further.Although the facts recorded in this paper only relate to theso-called silicobenzoic acid, it seems highly probable that all theother preparations which have been described as siliconic acids arealso complex condensation products of the trihydroxides RSi(013):;.EXPERIMENTAL.Preparation of Phenylsilicon Trich loride.The method employed for the preparation of this compound wasessentially the same as that described by Kipping in the case ofdiphenylsilicon dichloride (T., 1912, 101, 2113), except that onlyabout 1& molecules of magnesium phenyl bromide were used to1 molecule of silicon tetrachloride.The crude trichloride, collectedfrom about l l O o to 120°/50 mm., was purified by repeated distilla-tion under diminished pressure; a long-necked flask, fitted with arod-and-disk column, was employed in these operations, and ulti-mately a liquid boiling at 198-200° was collected.The yield ofthis liquid was about 50 grams from 120 grams of silicon tetra-chloride.The trichloride thus isolated probably contained traces of di-phenyl, an impurity the presence of which was of little importance,as i t could be so easily eliminated in subsequent operations. Asa matter of fact, traces of this compound were actually isolate682 MEADS AND KIPPING:from the products of hydrolysis of some samples of the trichloride.Other possible impurities, which it was most important to get ridof a t this stage, were silicon tetrachloride and more particularlydiphenylsilicon dichloride. As, however, the boiling point of eachof these compounds differs so very greatly from that of the tri-chloride, it can hardly be doubted that both these substances werecompletely removed by careful fractional distillation.It may alsobe added that the absence of silicon tetrachloride was easily provedby direct experiment, and that throughout the whole investigationof the products of hydrolysis of the trichloride no evidence of thepresence of diphenylsilicanediol or of any of its condensatioiiproducts was obtained.Decomposition of Phenylsilicon Trichloride with Water.Phenylsilicon . trichloride is immediately decomposed by waterwith development of heat. I n carrying out this hydrolysis thewater was cooled in ice and vigorously stirred, whilst the oil wasvery slowly dropped in from a tap funnel. A variable quantityof a pasty solid was usually precipitated, but the major portion ofthe product of hydrolysis remained in solution.The aqueoussolution was extracted with ether, and the ethereal extract waswell washed with water; most of the ether was 'then evaporatedwithout previously drying the solution, as the use of calciumchloride or other dehydrating agent seemed to be inadvisable.The syrupy product was transferred to a beaker and placed in adesiccator, which was rapidly exhausted ; this treatment caused theproduction of a very voluminous, frothy mass, which after sometime became brittle, and was then e a d y powdered. I f this processwas not adopted, and the whole of the ether was evaporated underthe ordinary pressure, the product solidified only very slowly, evenwhen subsequently kept in a vacuum, and ultimately became agelatinous mass, which was very difficult to manipulate.-Different samples of the colourless powders prepared in this waywere dried until constant in weight under the ordinary pressure;at first there was a rapid loss, but later on the preparations driedso slowly that, as a rule, they did not become constant in weightuntil after the expiration of about three weeks.The dried samplesshowed no sign of crystalline structure when examined under themicroscope ; they generally melted gradually from about 60° toabout looo. They were all easily and completely soluble in suchsolvents as ether, benzene, or chloroform, a fact which proved thatthey were free from silica; they were only moderately soluble incold alcohol, and sparingly so in cold light petroleum.The air-dried substance was also insoluble in water: the fact that thORGANIC DERIVATIVES OF SILICON. PART XYII. 683initial product of hydrolysis remains for the greater part in solution,seems to indicate the existence of an unstable trihydroxy-derivative,SiPh(OH),. From all solvents the substance was deposited in theform of a glue; it was, in fact, immediately converted into a sticky,gelatinous mass when it was brought into contact with most of theordinary solvents. It dissolved easily in a 20 per cent. solution ofpotassium hydroxide, and it was also soluble, but not so readily, ina 20 per cent. solution of sodium hydroxide, but it seemed to bequite insoluble in a concentrated solution of ammonium hydroxide.A considerable quantity of the above-described preparation wasdissolved in pure ether and fractionally precipitated with lightpetroleum, these operations being continued systematically untilthe original substance had been separated into 5 fractions, each ofwhich had undergone a t least 5 precipitations. All the fractionsthus obtained were glue-like or vitreous masses; when dried untilconstant, they all melted over a range of more than 1 5 O .The mostreadily soluble fraction which might have contained diphenyl wasnot examined; the melting points of the other four (No. I is themost readily soluble fraction) were as follows :I. 65-80O; 11. 75-90'; 111. 80-100°; IV. 100-130°.A portion of each of the above fractions was further separatedinto various sub-fractions with the aid of different solvents; thesesub-fractions showed a graded behaviour, ranging from that of aglue-like mass, melting from 65O upwards, and readily soluble inether, to that of an amorphous powder, which decomposed a t ahigh temperature without melting, and was practically insolublein glacial acetic acid.As it seemed impossible to obtain any crystalline substance, thefour main fractions referred to above were kept in a vacuum untilconstant in weight, and then analysed:111.I.11. <-A-- IV.C ......... 52.7 52.6 52.7 52.8 52.8 per cent.H ........ 4.4 4.4 4-7 - 4.4 9 sSi ......... 20.9 20.9 20.8 20.9 21.0 ,,C,H,*SiO,H requires C = 52.0 ; H = 4-3 ; Si = 20.5 per cent.(C,H, *SiO),O ,, C = 55.6 ; H = 3.9 ; Si = 21.95 ,,Although these results show that all the four fractions haveapproximately the composition C,H,*SiO,H, the whole of the ex-perimental evidence points to the conclusion that the productobtained by decomposing phsnylsilicon trichloride with water is notphenylsiliconic acid, but a complex mixture of condensation pro-ducts of the trihydroxy-compound, PhSi(OH),.The main com-ponents of this mixture have probably the composition PhSiO,H, o684 MEADS AND KIPPING:n.PhSi(OH), - zH20, and they are probably mixed with subst,ancesof the composition zPhSi(OH), - (II: + y)H,O ; the composition ofany such mixture would lie between that of phenylsiliconic acidand its anhydride.Molecular-weight determinations with fractions 11, 111, and IVin acetic acid solution by the cryoscopic method gave: Fraction 11,M.W.630; fraction 111, M.W. 717; fraction IV, M.W. 786. Itseems reasonable to infer from these results that fraction I1 con-sish principally of condensation products formed from 3, 4, 5, and6 molecules of phenyltrihydroxysilicane, and having molecularweights of 415, 553, 692, and 830, approximately, respectively; themore sparingly soluble components of fractions I11 and I V areprobably condensation products of even greater complexity.Actiort of IIent OTL the Product of I-ydrolysis.A sample of the product obtained by hydrolysing the trichloridewith ice-cold water, but which had not been fractionally precipi-t ated, was dried under diminished pressure until constant inweight, and then heated, first a t looo, and later on up to about200O.It first melted and then gradually changed into a hard,brown mass; a t the end of about forty hours it had become con-stant in weight, and the loss then amounted to 6.3 per cent., thetheoretical loss for the change 2PhSi0,H = (PhSiO),O + .H20 being6.4 per cent. The product was almost insoluble in ether andbenzene.Hydrolysis of the Trichloride with Ammonium Hydroxide Soliitioi;.The product obtained by slowly adding phenylsilicon trichlorideto a well-stirred, dilute, ice-cold solution of ammonium hydroxideseems to be identical or nearly so in properties and compositionwith that formed when the trichloride is hydrolysed with ice-coldwater. When, however, phenylsilicon trichloride is dropped intoa dilute solution of ammonium hydroxide which is not kept cold,i t is converted into a hard, white solid, very different in outwardproperties from the glue-like product already described.Thispreparation, having been dried in the air and washed with lightpetroleum, does not melt when it is heated, but gradually chars;when it is treated with benzene or ether, about 25 per cent. passesinto solution, and there remains a colourless powder which ispractically insoluble in these solvents. The soluble matter is aglue-like substance which liquefies gradually from about 90-looo ;the insoluble powder does not melt when it. is heated, and a samplegave on analysis, Si= 21-36, whilst (PhSiO),O requires Si = 21.9per centORGANIC DERIVATIVES OF smcoN.PART XXII. 685The products obtained by hydrolysing the trichloride with asolution of ammonium hydroxide are, theref ore, mixtures, thenature of which varies with the conditions of the experiment; atlow temperatures the preparations probably consist largely of thosecondensation products having the compoaition PhSiO,H, but atthe ordinary temperature, higher condensation products havingapproximately the composition (PhSiO)20 are formed.Precipitates of ‘I Phenylsilicoszic Acid ’’ obtained from AlkalineSOlZl t,ion.As already stated, the products obtained by hydrolysing phenyl-silicon trichloride with water are slowly dissolved by a 20 per cent.solution of potassium hydroxide. When such a solution is neu-tralised with very dilute hydrochloric or acetic acid, it does notgive any immediate precipitate, but in the course of some hours agelatinous separation t a k a place.I f the frmhly-acidified solutionis extracted with ether, it yields a gluelike product, which, whendried, melts below looo, and is readily soluble in ether or benzene;the preparations thus obtained seem, in fact, to be very similar tothose produced by decomposing the trichloride with ice-cold water.The addition of ammonium chloride to the alkaline solutionimmediately produces a flocculent precipitate, which has quite adifferent charactm from that of the glue-like preparation; i t chars,but does not melt, when it is very strongly heated, and whentreated with a considerable quantity of ether or benzene, only asmall proportion of it passes into solution.When carbon dioxide is passed into the clear alkaline solutionobtained by adding phenylsilicon trichloride to a well-stirred ice-cold 20 per cent. solutior, of potmsium hydroxide, a colourlessprecipitate, rather granular than gelatinous, is produced.Thisprecipitate is easily separated by filtration and, when washed anddried, is obtained in the fo;m of a fine, colourlec;~ powder. Thenature of such preparations, however, seemed to vary considerablywith the temperature a t which they had been precipitated; if a tthe ordinary temperature, they contained a considerable proportionof those substances which are practically insoluble in ether,whereas when precipitated from ice-cold solutions they were almostcompletely soluble in that liquid.Although the preparationsobtained by precipitation with carbon dioxide were in such aconvenient, powdery form, it was found to be impossible to freethem completely from occluded potassium salts by washing themwith water; they were, therefore, dissolved in ether, and the filteredethereal solutions were well washed, The products which remaine686 MEADS AND KIPPIXG:when the ethe’r was evaporatied soon became solid, and gave whitepowders when they were crushed.Since the (( phenylsiliconic acid ” obtained by precipitation withcarbon dioxide in the cold differed very considerably in outwardproperties from that prepared by hydrolysing the trichloride withice-cold water, it wits carefully examined.When air-dried, i tmelted indefinitely from about 140° to 160O; i t was readily solublein benzene, chloroform, or ethyl acetate; not easily, but completelysoluble in glacial acetic acid, and only partly so in alcohol. Itdissolved, but only very slowly, in a concentrated aqueous solutionof potaasium hydroxide, much more rapidly in alcoholic potassiumhydroxide.Analyses of two different preparations gave 21.6 and 21.4 percent. of silicon respectively ; these values are considerably higherthan those of the preparations obtained by decomposing the tri-chloride with ice-cold water, and are not far removed from thoserequired by a compound of the composition (PhSiO),O. A con-siderable quantity of the substance was dissolved in ether andfractionally precipitated with light petroleum.These operationswere continued until six main fractions had been obtained, eachof which had undergone some six to eight precipitations. Allthese fractions were colourless powders, which seemed devoid ofcrystalline structure ; they differed considerably in melting point,and in their behaviour towards solvents. The most soluble frac-tion (I) melted from about looo to l l O o , the next (11) from aboutl l O o to 140°, and the next (111) from about 150° upwards; fraction(IV) melted partly from about 210°, but fractions V and VI merelycharred at high temperatures. The most readily soluble fractiondissolved readily in hot light petroleum and was also completelysoluble in alcohol; the most sparingly soluble fraction was prw-tically insoluble in hot light petroleum, and only very partlysoluble in alcohol ; the intermediate fractions showed an inter-mediate behaviour.Silicon determinations were made with fractions I1 to Vinclusive: 11, Si=21*4; 111, Si=21.56; IV, Si=21*84; V, Si=21-56 per cent.The ‘‘ phenylsiliconic acid ” obtained by precipitation withcarbon dioxide in the above-described manner is, therefore, acomplex mixture, and the analytical data agree with the view thatthe main components of this mixture are condensation productsof phenyltrihydroxysilicane, having approximately the composition(PhSiO),O.The method just described is essentially the same as that usedby Ladenburg for the preparation of his ‘( phenylsiliconic acid ORGANIC DERIVATIVES OF SILICON.PART XXII. 687(Ber., 1873, 6, 379; Amnalen, 1874, 173, 143). He precipitatedthe acid with carbon dioxide, dissolved the precipitate in, or ex-tracted it with, alcohol, and evaporated the alcoholic solution. Theresidue, after having been washed with water and dried in a vacuum,melted a t 9Z0, and gave, on analysis, C=52-24; H=4*93 (PhSi0,Hrequires C=52*0, H=4*35 per cent.). When dried at looo, hisacid gave a product which contained C ~ 5 5 . 5 4 ; H=4*37 as themean of two analyses, and Si= 21-29 ; from these results, Ladenburgconcluded that this product was the anhydride (PhSiO),O, whichrequires C = 55.6, H = 3.9, Si = 21-95 per cent.Although these data agree fairly well with the calculated per-centages in each case, it is obvious that they cannot now be accepteda evidence that Ladenburg's preparations were homogeneous oras supporting his views as to the nature of his preparations. Thathis so-called phenylsiliconic acid and its anhydride were both coni-plex mixtures, similar in character to those described above, therecan be but little doubt, and it also seems highly probable that histolylsiliconic acid and tolylsiliconic anhydride were also mixturesof an analogous nature.Finat? Condensation Prohct of Phenylt&hydroxysilicane.It was observed by Hipping and Hackford (T., 1911, 99, 138)that the product obtained by passing steam into phenylsilicon tri-chloride is quite different from that which results when the tri-chloride is decomposed with water, and consists of a hard, brittlesolid, which is practically insoluble in all the common solvents, andhas the composition of phenylsiliconic anhydride.The facts r ecorded above seemed to show that this insoluble anhydride repre-sented the final condensation product, or mixture of condensationproducts, of phenyltrihydroxysilicane ; we therefore prepared someof this substance for further examination.The sample obtained by passing steam into the trichloride,purified as described above (p. 681), was repeatedly extracted withether, which removed a very small proportion of a glue-like sub-stance. As small quantities of diphenyldichlorosilicane mightpossibly have been present in the purified trichloride, and as it wasimportant to prove the absence of the former, the glue-like matterwas carefully examined in order to we whether it contained anycondensation product of diphenylsilicanediol.So far as could beascertained, however, no such compound wils present; the glueseemed to be identical with those preparations of " phenylsiliconicacid " obtained by decomposing the trichloride with water, and nocrystalline substance, except traces of diphenyl, could be obtainedfrom it688 MEADS AND KIPPING:Sodium Salt of Phe?zylsiliconic B cid.If all the above-described preparations are mixtures of condensa-tion products of phenyltrihydroxysilicane, i t seemed probable thatthey would all be hydrolysed by sodium hydroxide, and that finallyall the condensation products would give one and the same sodiumsalt, namely, sodium phenylsiliconate, PhSiO,Na, just as the con-densation products of diphenyl- and dibenzyl-silicanediol are finallyconverted into the sodium derivatives of the simple diols (Kippingand Robison, loc.cit.).As, however, the hydrolysis of the less soluble preparations ofthe acid might require the use of concentrated solutions of alkalihydroxide, it was first necessary to ascertain whether such solutionsdecomposed the silicon compounds into benzene and an alkalisilicate. For this purpose, some of the acid was dissolved in awarm 20 per cent. solution of sodium hydroxide, and the solutionwas boiled during some hours under reflux; a t the end of this timenot a trace of benzene was observed, and the solution had no odourof the hydrocarbon.When a very much more concentrated solutionof potassium hydroxide was used, although a distinct odour ofbenzene could be observed, after the solution had been heatedduring some hours, there was no appearance of oily drops in thecondenser. It may be concluded, therefore, that the decompositionof " phenylsiliconic acid " observed by Ladenburg (Zoc. cit.) onlytakes place to an appreciable extent a t the high temperature whichlie employed.I n order to isolate the sodium salt of phenylsiliconic acid, asample of the product obtained by hydrolysing the trichloride withice-cold water wae dissolved in alcohol and treated with an alcolzolicsolution of sodium ethoxide (approximately 3 mols.) ; the mixturewas then left to evaporate at the ordinary temperature in anexhausted desiccator containing potassium hydroxide.After sometime the solution deposited colourless crystals of a sodium salt,which was only sparingly soluble in alcohol; this product (I) wasrapidly separated by filtration, and washed with absolute alcohol.The alcoholic filtrate was further concentrated (in absence of mois-ture and carbon dioxide), and in this way a second crop of crystals(11) was obtained.These two samples were freed from alcohol under diminishedpressure and analysed; for this purpose the aqueous solution of aweighed quantity of the salt was treated with excess of N/lO-sulph-uric acid, and the filtered solution was titrated with alkali.Fraction (I) gave Na=13.5 and fraction (11) gave Na=14*4;CG€I,-SiO,Na requires Na = 14.3 per centORGANIC DERIVATIVES OP SILICON.PART XXlI. 659These results seem to show that fraction (11) of the salt waspractically pure sodium phenylsiliconate, which had been formed bythe complete hydrolysis of the various condensation products con-tained in the ‘‘ acid ”; fraction (I), which doubtless consistedprincipally of this same compound, probably contained a smallproportion of salts derived from some of the condensation productswhich had not been completely hydrolysed. The behaviour of thetwo fractions towards water accorded with this view; whereasfraction (I) was not completely suluble, and gave solutions whichsoon deposited a considerable precipitate, fraction (11) gave a clearsolution, in which a precipitate was not formed until after thelapse of an hour or so.Sodium Salt from the FinaL Condensation Product.The insoluble solid formed by the action of steam on phenyl-silicon trichloride (see above) was treated with an alcoholic solutionof sodium ethoxide (approximately 2 mols.), and the mixture wasleft in a closed flask.A t the end of some twelve hours, the in-soluble condensation product seemed to have been completelyconverted into a mass of crystalline sodium salt; this preparation(I), which formed about 70 per cent. of the theoretical quantity,was separated by filtration, washed with alcohol, and placed in anexhausted desiccator containing potassium hydroxide. The alcoholicfiltrate, when kept at the ordinary temperature during some days,deposited a crop of colourless needles; this sample (11) wasseparated and washed.These two samples of sodium salt seemed to be very similar tothe corresponding preparations just described ; the one (fraction 11)gave a clear solution with water, but the other (fraction I) was notcompletely soluble. Sodium determinations gave the followingresults :Fraction I, Na = 12.6.C,H,-Si02Na requires Na = 14.3 per cent.It appears, therefore, that even the insoluble mixture of the finalcondensatior) products of phenyltrihydroxysilicane is converted onhydrolysis into a salt which consists essentially of sodium phenyl-siliconate, but a t present there is no evidence that the salt has thesimple formula PhSiOaa.The study of phenylsiliconic acid and of its salts is being con-tinued, and several other siliconic acids are also being investigatedin various directions; it is hoped that the true siliconic acids havingthe simple molecular formula RSiO,H, and also some of theFraction 11, Na = 13.9690 ROSS: THE RATE OF TRANSFORMATION OFcondensation products of the trihydroxy-derivatives may soon beisolated.The so-called inethylstannonic acid, CH,*Sn02H, is also beingexamined in these laboratories on much the same lines; evidencehas already been obtained that the compound in question is acomplex condensation product of CH,*Sn(OH),.The authors desire to express their indebtedness to the RmearchFund Committee of the Royal Society for a grant in aid of thiswork.UNIVERSITY COLLEGE,NOTTINCHAM

ISSN:0368-1645    

DOI:10.1039/CT9140500679

出版商:RSC    

年代:1914

数据来源: RSC