346 SENTER: REACTIVITY OF THEXXX1X.-Reactivity of the Halogens in Organic Corn-pounds. Part I I? I n t e r a c t i o n of Rromoacetic,a-BrornopropiorLic, and a-Bromobutyric Acids andtheir Sodium Sults with Silver Salts in A~UL'OWSSolution. Cutalytic Action of Xilvey* Halides.By GEORGE SENTER.IN previous papers (Trans., 1907, 91, 460; Proc., 1908, 24, 89;Arrhenius Jubelband, 1910, 11, 511; Trans., 1909, 95, 1827) theresults of an experimental investigation of the rate of displacementof the halogens in the lower members of the series of halogen-substituted fatty acids have been communicated, and the mechanismof the respective reactions has been discussed. The present paperdeals mainly with the interaction of the first three members of theseries of a-bromo-fatty acids and their sodium salts with silvernitrate and silver acetate in dilute aqueous solution.It was shownmany years ago by Beckurts and 'Odd0 (Ber., 1881, 14, 576; 1885,18, 222) that the reaction between silver nitrate and the lowermembers of the series of bromo-fatty acids in aqueous solutlion isrepresented quantitatively by the equation :R*CHX-CO,H + AgNO, + H20 =R*CH(OH)*CO,H + AgX + HNO, (I),in which R represents hydrogen or an alkyl group, and X a halogenatom.In the course of the present investigations, the remarkableobservation was made that reactions of the above type are verymarkedly accelerated by silver bromide; even when the solutionis only 1/300 molar with reference to silver bromide (which is, ofcourse, mainly present in the insoluble form), the rate of thereaction may be doubled or even trebled.This action appears toaccount for a number of hitherto unexplained observations madeby previous investigators.The other more important results of the investigation are thatthe rate of reaction increases very considerably with the increasein complexity of the alkyl group R, and that the sodium saltsreact more rapidly than the free acids. Nitric acid exerts a verypowerful retarding effect on the reactions in which silver nitrateis concerned. Silver acetate reacts with the bromo-fatty acids morerapidly than does silver nitrate under equivalent conditionsHALOGENS IN ORGANIC COMPOUNDS. PART IV. 347EXPERIMENTAL.Method of Measurement.-The reacting substances were mixed a tconstant temperature in small tubes of Jena glass, which werethen corked and kept in a thermostat a t a temperature keptconstant to within Omlo. A t definite intervals the contents of atube were poured into a slight excess of a N/50-solution ofammonium thiocyanate, which a t once stopped the reaction, 5 C.C.of a concentrated solution of iron alum and 5 C.C.of concentratednitric acid, free from nitrous fumes, were then added, and theexcess of thiocyanate estimated by titrating with N / 50-silvernitrate.It is known that the afbove method of titration does not giveaccurate results for chlorides, owing t o the solubility of the precipi-tated silver chloride in ammonium thiocyanate. Theoretically, how-ever, there should be no appreciable error in the case of bromides,owing to the much smaller solubility of silver bromide, and thisconclusion has been confirmed experimentally by Rosanoff andHill ( J .Amler. Chem. SOG., 1907, 29, 1467). I n order to ensurethat no error arises in the present case owing to the presence ofprecipitated silver bromide, the method of titration describedabove has been tested in various ways, and has been found to givequite accurate results.The tubes in which the reactions were carried out held about15 c.c., and in all cases the reaction mixture measured 12 C.C.The titration values quoted in the tables are throughout (exceptin table I) the mean of two simultaneous expeTiments. Themeasurements were made at 26'0O.Reaction between a-Bromopropionic Acid and Silver Nitrate.The results of a typical series of experiments with a-bronio-propionic acid and silver nitrate are quoted in table I.The unitof concentration is t,he amount of silver nitrate contained in 1 C.C.of a N J 50-solution of the salt, and the concentration of the halogen-fatty acids (and their sodium salts) is expressed in equivalent units,in accordance with the experimental fact that R*CHBr-C02H (andR*CHBr*CO,Na) is equivalent to AgNO,. The constants in thelast column are calculated according to the equation for a reactionof the second order:I x k=--when the reacting substances are present in equivalent amounts.I n all cases, a- x in the tables represents the concentration of thet 'a(a - x348 SENTER: REACTIVITY OF THEsilver nitrate a t the time t , as this is the substance which is actuallyestimated.I f it is desired to refer k, the velocity-coefficient or velocityconstant of the reaction, to a concentration of 1 mol.per litre, thevalues of E given in the tables must be multiplied by 12 x 50 = 600,since the values quoted are obtained by titrating 12 C.C. of thereaction mixture with 1 /SO molar thiocyanate.Throughout this paper, the concentrations given refer to thereaction mixture; thus, in the experiments quoted in table I, thereaction mixture was initially N l 3 0 with reference to boththe reacting substances.TABLE I.Silver nitrate, AT/ 30. a-Bromopropionic acid, N / 3 0 .t (min.). a-x. k. a - x. k.0 20.0 - 20.0 -10 18’1 0.00050 18.0 0*0005630 14’7 0*00060 14-7 0-0006090 10.2 0.00058 10’0 0’00056These figures show that the results of parallel experiments are ingood agreement, and the fair agreement of the velocity-coefficientsin the course of a reaction is in accord with the assumption thatthe reaction is bimolecular.It will be shown in the sequel, how-ever, that the latter result is only attained owing to il combinationof two factors which influence the reaction in opposite directions.It is well known that one of the best methods of deciding the“order” of a reaction is to measure its velocity with varyinginitial concentrations ; in this way disturbances arising in thecourse of the reaction are to a great extent eliminated. The resultsfor the reaction under discussion are given in table 11, the valuesof k, when the concentrations are not equivalent, being calculatedby the general formula for a bimolecular reaction :where the symbols have the usual significance.TABLE 11.Concentration of Concentration ofsilver nitrate.bromopropionic acid. k.N i l 0 N/40 0.00047N/20 N/40 0-00059N/40 N/40 0’00054N/20 N/20 0-00053N/20 N/10 0*00051The above results appear to show that, in spite of certainirregularities, the rate of the reaction is in the first instancHALOGENS IN ORGANIC COMPOUNDS. PART IV. 349proportional to the initial concentration of the reacting substancesthrough a range of concentration from N/10 to N / 4 0 in each case.It was soon noticed that when the aqueous solution of the acidwas not freshly prepared, the reaction was considerably more rapid.This is illustrated by the results quoted in the accompanying table.TABLE 111.Silver nitrate and bromopropionic acid, each N / 3 0 .(Acid solution, two hours old.)t (min.). a - x. k.0 20 .o -10 18.5 0.0004130 15.4 0 *0005090 10.5 0*00050(Acid solution, kept two days at 26".)t (min.). n-x. k.0 20.0 -10 16.4 0 -00 11 030 12.0 0.001 1190 8'3 0*00080It is clear that the initial velocity is nearly three times as greatwhen the acid solution is kept two days before the rate of reactionis measured.It is well known that a-bromopropionic acid is slowly decomposedby water, according to the equation:CHMeBr*CO,H + H,O = OH-CHMe-CO,H + HBr,and in an earlier paper the results of an experimental investigationof this reaction have been communicated.A little hydrobromicacid is formed during the reaction, and a t once reacts with partof the silver nitrate on mixing the reacting substances. As thespeed of the reaction is deduced from the rate a t which the silvernitrate is used up, the apparent speed in the presence of hydro-bromic acid will be greater than the actual speed with whichsilver nitrate and a-bromopropionic acid react. I n order toestimate the error thus caused, the rate of hydrolytic decompositionof bromopropionic acid a t 2Go has been measured a t 2 6 O with thefollowing results (table IV). The concentrations of bromopropionicacid and of hydrobromic acid are expressed in terms of the numberof C.C.of N/2O-sodium hydroxide required to neutralise 5 C.C. of thesolution, which was approximatelyTABLE IV.Concentration of HBr formed.Five C.C. of acidTime (days). solution titrated.0 19-703 20'106 20.359 20.6012 20'80The reaction at 2 6 O is thereforeI 3 In C.C. ofNJ20-NaOH. Normality.0'00 -0'40 0.0010'65 0-00160.90 0'00231.10 0.0028extremely slow, and the amoun350 SEPU'TER: REACTIVITY OF THEof hydrobromic acid formed is much too small to account directlyfor the results given in table 111.There would appear to be at least two plausible explanations ofthe results in question: (1) that the silver bromide formed inthe course of the reaction exerts a catalytic influence; (2) thatthe bromopropionic acid in aqueous solution undergoes a slowchange into a second more active modification. The first suggestioncan a t once be tested by adding some hydrobromic acid orpotassium bromide to the bromopropionic acid before adding thesilver nitrate. Some of the results obtained in this way are givenin table V.TABLE V.Composition of Mixture I.I Composition of Mixture 11.t (min.). a-2. k.0 20'0 -10 18 5 0*0004130 15.4 0~0005090 10.5 0.00050t (min.). n - x. k.0 20 .o -10 15.7 0.0018720 13 '3 0.0012640 10.0 0'00125The data in-the above table show that, after mixing, the twosolutions are exactly equivalent in concentration ; the only differenceis that in the second mixture a small amount of silver bromide(and potassium nitrate) is present. The mixture is only 1/150molar with respect to this salt, and doubtless the greater p%rt ofit is present in the insoluble form, yet the remarkable result isobtained that this trace of silver bromide more than trebles theinitial speed of the main reaction.Moreover, this by no meansrepresents the maximum catalytic power of the silver bromide, asthe greater part of it coagulates and rises to the top of the solutionsoon after mixing the reagents.Reference t o the data in table I11 shows that the solution ofbromopropionic acid, which had stood two days at 2 6 O , cannothave been more than 1/1000 normal with reference to hydrobromicacid, so that the small amount of silver bromide formed by inter-action of the acid with silver nitrate has a very powerful catalyticaction.The catalytic acceleration of reactions of this type by silverbromide and iodide is further referred to in a later part of thepaper (pp.357, 358)HALOGENS IN ORGANIC COMPOUNDS. PART IV. 351Effect of Acids and of Sodhm Nitrate o n the Reaction Velocity.( a ) Nitric Acid.-Nitric acid, even in very dilute solution, verymarkedly retards the reaction between silver nitrate and a-bromo-propionic acid. Some of the results, typical of an extended seriesof experiments on this point, are given in the accompanying table.TABLE VI.Silver nitrate, N / 2 0 . Bromopropionic acid, N / 2 0 .No nitric acid. Nitric acid, N/30. Nitric acid, iV/15. - - 7t. a-x. k. t. a-32. k. t. a-x. k.- 0 30.0 - 0 30.0 - 0 30.010 26.7 0*00041 20 28'0 0*000120 30 28'45 0.00006030 21'5 0.00045 60 25.0 0~000111 90 25.75 0'000062The figures show that in the presence of N/15-nitric acid therate of the reaction is diminished to about 1/7th of its originalvalue.( b ) Benzenesulphonic Acid.-For comparison with nitric acid,some experiments were made with benzenesulphonic acid under thesame conditions.It was found that the latter acid retards thereaction to a rather greater extent than nitric acid. The initialvalue of the constant in the presence of N / 15-benzenesulphonicacid is 0.000053, as compared with 0*000061 for ~'V/15-nitric acid.It is probable that benzenesulphonic acid is a rather strongeracid than nitric acid, but the difference in the reactivities seemsrather greater than can be accounted for on this basis.(c) Lactic Acid.-As lactic acid is one of the products of thereaction, its influence on the velocity was measured with thefollowing results :TABLE VII.Silver nitrate, N / 20.Bromopropionic acid, N / 20.No lactic acid.t (min.). a - 2. k.0 30.0 -10 26.1 0.0005030 20'4 0-0005380 14'9 0 '00056Lactic acid, NJ5.t (min.). a - x. k.0 30 +O -10 26-9 0 0003830 22.1 0.0004060 18'4 0'00036(d) Sodium Nitrate.-The results of a series of observations withthis salt amre given in the accompanying table; much higher con-centrations were used than in the case of the acids352 SEN'I'ER : REACTIVI'rY OF THETABLE VIII.Silver nitrate, N / 30. Bromopropionic acid, N / 3 0 .No sodium nitrate. Sodium nitrate, ml2.Sodium nitrate, mil. - - - t (min. ). a - x. k. a - 2. k. n - 5. k.0 20 -0 - 20-0 - 20 -0 -10 17'9 0*00060 18.3 0,00045 78.7 0*0003530 14.6 0*00062 15-7 0-00045 16-2 0.00039The facts that nitric acid and benzenesulphonic acid, which areboth highly ionised in solution, retard the reaction to about thesame extent, and that the NO,' ion has only a very slight reta,rdingaction, indicate that the effect in question is mainly due to theH ions. This is confirmed by the fact that lactic acid, which is acomparatively weak acid, has a very slight retarding action. Thebearing of these results on the mechanism of the reaction isdiscussed later (p. 361).Effect of Alcohol and of Acetone on the Rate of Reaction.A few experiments were made in which half the water used LCSsolvent was displaced by alcohol and by acetone respectively; theresults were as follows:TABLE IX.Silver nitrate, ZV/ 30.Bromopropionic acid, N / 3 0 .Solvent.. . Water. Water + alcohol. Water + acetone. - - c-0 20 '0 - 20.0 - 20 -0 -10 18.0 0-00056 17-5 0*00071 17'4 0.0007330 14.6 0.00062 13.9 0.00073 14.3 0-00067It is interesting to note how small an alteration is produced inthe reaction velocity by the displacement of half the water byalcohol or by acetone. It has already been shown by Euler (Ber.,1906, 39, 2726) that the rate of reaction between chloroacetic acidand silver nitrate is approximately the same in water and in 45 percent. alcohol.t (rnin.). a - z. k. a - x, k. a - x. k.Sodium Bronzopropionate and Silver Nitrate.The reaction in this case is represented by the equation :CH,*CHBr*CO,Na+ AgNO, + A,O =CH,=CH(OH)*CO,H + AgBr + NaNO,.The sodium bromopropionate was prepared just before thereaction by careful neutralisation of a solution of bromopropionlcacid with sodium hydroxide.The results of one series of experiHALOGENS IN ORUANIC COMPOUNDS. PART IV. 353ments, in which the relative activities of the acid and its sodiumsalt are compared, are given in the accompanying table:TABLE X.Silver nitrate, N/30. Silver nitrate, N/30.Bromopropionic acid, N/30. 1 Sodium bromopropionate, N/30.t (min.). a-x. k.0 20.0 -10 18-0 0.0005630 14'7 0'0006090 10-2 0 *00054t (min.). a - 5. k.0 20.0 -5 16.6 0'002015 11'0 0.002745 6 ' 4 0.0024The velocity-coefficients, k,, for a bimolecular reaction in thecase of the sodium salt are only in moderate agreement, a resultwhich is doubtless to be anticipated.The initial velocity is aboutfour times that obtained for the free acid. The considerableincrease of th6 coefficient between five and fifteen minutes is nodoubt due to the catalytic influence of silver bromide. A fewmeasurements were also made to determine the influence of theinitial concentration on the reaction velocity, with the followingresults :k. ......... Sodium bromopropionate, N/30 } 0*0020 Silver nitrate, 3/30 ........................Silver nitrate,:N/60 ........................ } 0.0030 Sodinm bromopropionate, N/60 .........showing that the initial velocity is the greater the more dilute thesolution.Bromopropionic Acid and Sz7uer Acetate.I f the conclusion drawn from the experiments already described-that silver bromide exerts a catalytic influence on the rate ofreaction-is valid, the velocity-coefficients calculated for a reactionof the second order ought regularly to increase, owing to theincrease in the amount of silver bromide as the reaction proceeds.However, nitric acid, another product of the reaction, exerts aretarding influence, and the result of these two effects is that inmany cases the reactions follow the law for a bimolecular reactionfairly accurately.In order to eliminate the retarding influence ofnitric acid, it was considered desirable to perform ilr few experimentswith the silver salt of a weak acid, and for this purpose silveracetate was chosen.The reaction in this case is represented by theequation :CH,*CO,Ag + CH,Br*CO,E + H,O =CH,*CO,H + AgBr + OH*CH,*CO,H.Some typioal results are quoted in table XI354 SENTER: REACTIVITY OF THEF?*esli B.r-omopropionic -4 cid.Bromopropionic acid, ~V/30. I Broiriopropionic acid, N/30.Silver acetate, iV/30. Silver acetate, N/lOO.f (mill.). n - x. Jc .0 20.0 -5 18.2 0 '001 0015 14.0 0'0014330 9-5 0.00186t (iiiin.). n -- 2. k.0 6'0 -10 4.75 0.0013030 2-75 0.001450-00156 60 0-70Bromopropionic A c i d (kept two days at 26O).Bromopropionic acid, NJ30. 1 Bromopropioriic acid, iV/60.Silver acetate, N/30. Silver acetate, N/30.t (min.). a - z . k.0 20'0 -5 16.8 0*0020015 10.5 0.0030730 7.6 (0.00272)t (min.).n - x. k.0 20-0 -10 16.5 0'0025030 12.8 0.00327It will be observed that in all these experiments t,he velocity-coefficients calculated for a reaction of the second order increasevery considerably during the reaction, doubtless owing t o thecatalytic influence of the silver bromide. The same effect is seenin the solution which has been kept two days, and in which, there-fore, a little hydrobromic acid has been formed. The very smallamount of silver bromide produced as soon as the two solutionsare mixed is sufficient t o double the initial speed of the reaction.This catalytic effect is also illustrated in the following table; inone case a small amount of potassium bromide is added t o the acidbefore mixing :TABLE XII.Silver acetate, N/25 ............Bromopropionic acid, N16.25.2 ,,Water .............................. 2 ,,t (min ). a-x. k.0 16.00 -5 14.20 OvO015810 12.40 0*0018120 9.65 0*0020630 7-00 0.002688 C.C. Silver acetate, NJ25 ............ 9 c.cRroniopropionic acid, N16.25. 2 ,,Potassium bromide, N/25 ...... 1 ,,t (min.). a-x. k.0 16.00 -6 11.30 0.0052010 9'80 0.0039520 7-00 0.00402These results indicate that 1/300 molar silver bromide morethan trebles the initial speed of the reaction, but that the velocityfalls off somewhat as the reaction proceeds. This is doubtlessconnected with the fact that the greater part of the silver bromidesoon coagulates, and rises to the top of the solution; it can thenexert no catalytic influence.Relative Velocities with Silver Nitrate and Silver Acetate.HALOGENS IN ORGANIC COMPOUNDS.PART IV. 355Simultaneous measurements were made with silver nitrate and silveracetate in equivalent concentration, in order to obtain an accuraterecord of their relative activities with, bromopropionic acid. Theresults are as follows:TABLE XIII.Bromopropionic acid, Nf30. 1 Bromopropionic acid, N/30.Silver nitrate, N/30. Silver acetate, N/30.t (min.). a-x. k.0 20.0 -10 17'8 0 -0006230 14'2 0-0006890 9.5 0 -00 0 6 1t (min. ). a - x. k.0 20'0 -5 17.9 0.0012017 12-6 0*0017045 6-6 0*00220It follows that the initial velocity with the nitrate is about halfThe bearing that with the acetate under corresponding conditions.of this result on the mechanism of the reaction is considered later.Reaction between a-Bromobutyric Acid am? Silver Nitrate.Corresponding measurements to those just described for bromo-propionic acid have been made with bromobutyric acid, but not inthe same detail.CHEtBr*CO,H + AgNO, + H,O = OH*CHEt.CO,H + AgBr + HNO,.The results are similar to those obtained for bromopropionic acid,except that the velocity-coeflicients diminish more rapidly duringthe reaction.The data for experiments in which the initial con-centrations were varied are given in table XIV, the initial valuesof the velocity-coefficients being given in the third column :The reaction is represented by the equation:Concentration ofsilver uitrate.N/60N/40N/20N/10N/40N/40TABLE XIV.Concentration ofbromobutyric acid.k.N/60 0 -0060N/40 0'0034Nj4Q 0.0014A7/20 0.0034N/10 0.0028N/40 0'0021It is clear from these results that when the concentration ofsilver nitrate is kept constant, and that of the bromopropionic acidvaried, there is not much alteration in the magnitude of thevelocity-coefficients; in other words, the rate of the reaction isapproximately proportional to the concentration of the bromo-butyric acid. On the other hand, the coefficients diminish con-siderably as the initial concentration of silver nitrate is increased,which indicates that when the silver nitrate concentration isincreased, the rate of $he reaction does not increase in the sameproportion. This does not correspond with the behaviour of silverVOL.XCVII. BB356 SENTER: REACTIVITY OF THEnitrate and a-bromopropionic acid, where the velocity-coefficientsretain approximately the same value with varying concentrationEfect of Nitric d cid.-The actual observations are quoted inthis case, as they illustrate very clearly the falling off in themagnitude of the velocity-coefficients as the reaction progresses :(p. 348).TABLE XV.Silver nitrate, AT/ 20. Bromobutyric acid, 8 / 2 0 ,No nittic acid. Nitric acid, N/15. Nitric acid, N17.5.6 A - i F * ' ----- t (min.) a - z. k. t (min.). a - z. k. t (ruin.). C6 - X. k.0 30.0 - 0 30.0 - 0 30.0 -5 23'4 0*0019 15 24'6 On0O048 30 25.3 0.0002016 18.8 0'0013 45 21'5 0.00029 90 22'6 0'00012These results show that the retarding effect of nitric acid isconsiderable, and is approximately proportional to the concentrationof the acid.Sodium Brombutyrate and Silver Nitrate.-The magnitude ofthe constant in 1/30 molar solution of each of the reacting sub-stances is 0*0060.The speed is therefore about 2.5 times thatwith the free acid, for which the constant is about 0.0024.Bromoacetic Acid and Silver Nitrate.As this reaction has already been investigated to some extentby Euler [Zoc. cit.), it has been considered sufficient for the presentpurpose to make a few measurements with the object of comparingthis acid with the two higher acids as regards its reactivity withsilver nitrate, and, further, to determine if this reaction, like theothers, is catalytically accelerated by silver bromide.The resultsare given in table XVI:TABLE XVI.Silver nitrate, N/12'5 ......... 5 C.C.Bromoacetic acid, N/5 ......... 2 ),MTeter .............................. 5 ),t (min.), a - x . k.0 20.0 -240 19.5 0 *00000481200 17'6 0 ~00000572760 13'4 0*0000089Silver nitrate, NI12.5 ......... 6 C.C.Bromoacetic acid, N/5 2 $ 9 Potassium bromide, N/25 ...... 2 ,,Water 2 $ 9.......................................t (min.). a-x. k,0 20.0 -180 18.1 0 '0000291200 11-6 0 '0000302760 9 -6 0~000023The data quoted in the table show that &he rate of reactionbetween silver nitrate and bromoacetic acid is very slow at 26O,the rate is only about 1/100th of the corresponding reaction withbromopropionic acid.Further, silver bromide greatly accelerateHALOGENS IN ORGANIC COMPOUNDS. PART IV. 357the reaction-in a solution 1/150 molar with respect to this salt,the initial velocity is about six times that in the absence of thesilver bromide.Sodium Bromoacetate and Silver Nitrate.-The results given intable XVII show that sodium bromoacetate reacts with silvernitrate about three times as rapidly as does the free acid.TABLE XVII.Bromoacetic acid, iV/30.Silver nitrate, N/30.t (min.). n - x. k.0 20'0 -240 19.5 0 '00000541200 17-52 0.00000682760 13'8 0 ~000005 1Sodium bromnacetate, iIT/3O.Silver nitrate, Nj30.t (min.). a - x. k.0 20'0 -240 18.6 0'0000 161200 13'2 0 ,000n212760 7'4 0*000031Methyl Zodide mad Silver Nitrate.As this reaction has been measured by several previous observers,who, however, did not detect the catalytic influence of the silveriodide, it has been considered desirable to repeat the measurements,with special reference to the effect of silver iodide in aqueoussolution. Some typical results are given in the accompanyingtable :TAELE XVIII.Solvent :I.Silver nitrate, N/20 .....Methyl iodide, NJ20 ...5 ,)5 c,.c....................... Water.. 2 Y Yt (min.). n -x. k.0 12.50 -15 11 -60 0.0004230 10.95 0.000:3960 9-70 O*OOOY9120 8 -30 0*00034180 6.80 0.00037TVat er.11.Silver nitrate, N/20 ........Methyl iodide, A720 5 Y9Potassiiim iodide, Nl20.. 1 > )6 C.C. .............t (miii.). a-- x. k.0 12.50 -15 10.70 0 *O 008030 9.65 0'0007660 8 -00 0*00075120 G 7 0 0*00058180 5.10 O'OUOi15S o l vent : Alcohol.I V (composition as 11). I 111 (composition as I).t (min.).a -a. k.0 12 *5 -10 10.75 0 *OO 1 3 030 8.50 0.0012660 6 -73 0.001 16120 4 *go 0.00104t (rnin.). n - x. k.0 12.5 -10 10.4 0*0016230 8 -3 0-0013560 6.4 0*00127120 4 -5 0*00119These results show that the reaction in aqueous solution is con-siderably accelerated by N / 240-silver iodide, and there appears toB B 358 SENTER: REACTIVITY OF TEEbe a corresponding, but much smaller, acceleration in alcoholicsolution. The latter result, however, is of a preliminary nature,as only one series of measurements was made in alcohol.I n all these cases it has to be remembered that the amount ofsilver iodide distributed through the solution is only a smallfraction of the total amount, as the precipitate, especially inalcoholic solution, soon coagulates, and then rises to the top or sinkst o the bottom of the solution, being thus to a great extent removedfrom the sphere of action.I n experiments I11 and IV, besides silver iodide, a, littlepotassium nitrate (1/240 molar) is formed on mixing the solutions.Direct experiment shows, however, that even in 1 / 20 molar solutionpotassium nitrate exerts no appreciable influence on the rate ofthe reaction in aqueous solution, and the same may safely beassumed as to its effect in alcoholic solution in so small con-centration.According to the above table, the rate in ethyl alcohol is to therate in water as 3 : 1, a result not in satisfactory agreement withthe recent measurements of Burke and Donnan (Zeitsch.pltysilcal.Cl~en~., 1909, 69, 148), who find the ratio to be about 6 : 1.DISCUSSION OF RESULTS.(1) Tke Cntalytic ZiLfEuence of Silcer Haloyen Salts.-In theprevious pages it has been shown that silver bromide and silveriodide, even in extremely small concentration, exert a markedaccelerating effect on reactions in which silver salts and halogencompounds are concerned. The conclusions drawn by previousobservers as to the mechanism of such reactions require revisionin the light of this observation.The magnitude of the effect under favourable conditions is shownby the fact that 1/1000 molar silver bromide (about 0.002 gramin 12 C.C.of the reaction mixture) doubles the initial rate ofreaction between silver nitrate and bromopropionic acid. Unfor-tunately, an accurate investigation of this effect is rendered verydifficult by the fact already mentioned, that the precipitate sooncoagulates and is withdrawn from the sphere of action by risingto the top or sinking to the bottom of the solution.I n order to find whether the catalytic influence pertained to thehalogen compound in all forms, about 0.1 gram of freshly pre-cipitated and carefully washed silver iodide was added to a mixtureof silver nitrate and methyl iodide, and the rate of the reactionmeasured. The mean value of the velocity-coefficient in thepresence of the added iodide was 0.00041, in its absence 0*00035,a comparatively small acceleration.The data quoted in tablHALOGENS IN ORGANIC COMPOUNDS, PART IV. 359XVIII show that 1/8th of this amount of iodide, precipitated inthe reaction mixture, produces a much greater acceleration.It seems probable, therefore, that the catalytic power is connectedwith the fineness of division of the silver iodide, possibly with itsoccurrence in the colloidal (hydrosol) form. Lottermoser andRothe (Zeitsch. physikal. Chem., 1908, 62, 359) have shown thatsilver iodide hydrosol is much less stable when the silver nitrateis in excess than when excess of potassium iodide is present. Asthe silver salt is necessarily in excess in the reactions now underinvestigation, the comparatively rapid coagulation of the precipitateis accounted for.The view that the acceleration is connected withthe colloidal form of the silver iodide appears to be supported bythe observation that the catalytic effect is much smaller in alcoholicsolution, in which the hydrosol appears to be less stable.These observations are interesting also in connexion with theexperiments of Miss Burke and Donnan on the reaction betweenthe alkyl halides and silver nitrate in alcoholic solution. Theyfound that whilst the velocity-coefficients calculated for a reactionof the second order remained more or less constant with variationof the initial concenhtion of the alkyl iodide, they increased withincreasing concentration of the silver nitrate. I n other words, ifwe consider a reaction-mixture originally N / 2 0 with regard toboth components, at the moment when the concentration has fallento N / 4 0 , the reaction is found to be proceeding more rapidly thanin a solution in which the reacting substances are originally N / 4 0 :I n spite of a very detailed investigation, the results of which havejust been published (Zeitsch.physilal. Chem., 1909, 69, 148), theauthors have obtained no satisfactory explanation of thisphenomenon, although they favour the suggestion of Wegscheiderand Frank1 (Monatsh., 1907, 28, 91) that it is the non-ionised silvernit.rate which enters into reaction.It is evident, however, that the results could be a t once accountedfor if silver iodide exerts a catalytic action in alcoholic solution,as it has been proved to do in aqueous solution in the presentpaper.The experiments in alcoholic solution quoted in tableXVIII are, as already mentioned, of a preliminary character, andthe matter is now being further investigated by Miss Burke. I f theabove explanation proves tenable, there will no longer be anyexperimental justification for Wegscheider’s suggestion (Zoc. cit.)that it is the non-ionised silver nitrate which reacts.(2) The Mechanism of t72e Reactions.-The mechanism of thesereactions appears to be rather complicated, and the full discussionis postponed until the results of further investigations are available,more particularly the rate of reaction of the halogen-substitute360 SENTER: REACTIVITY OF THEesters with silver nitrate in alcoholic solution. It will be sufficientfor our present purpose to summarise the more important resultscommunicated in this paper which have a bearing on the mechanismof the reactions.Reactions of this type have often been discussedby previous observers, but no very definite conclusions as to theirmechanism have been drawn. All that can be said with certaintyis that the relatively great velocity is conditioned in some wayby the tendency t o formation of the halogen silver salts (compareEuler, Ber., 1906, 39, 2726; Wegscheider, Monatsh., 1907, 28, 79).I n this connexion it is interesting to note that definite compoundsof silver nitrate with certain organic halogen compounds (forexample, AgNO,,CH,I.CN and AgNO,,CH,I,) have been preparedby Scholl and Steinkopf (Ber., 1906, 39, 4393).The fact that silver bromide and iodide exerts a catalytic effecton the reactions renders the interpretation of the results somewhatdifficult, as we are not entitled to assume that the observedvelocities are proportional to the intrinsic velocities.Pendingfurther investigation, however, it may be assumed that the nearestapproach to the relative intrinsic velocities is obtained by takingthe initial velocities of the respective reactions.It has already been pointed out that as regards compounds of thetype R*CHBr*CO,H and their sodium salts, the velocity increasesgreatly with the complexity of the substituting group R. Undercorresponding conditions, the relative reactivities of the first threeacids and their sodium salts with silver nitrate are as follows:CH,Br 'C0,H.CHMeBr'C0,H. CH E tBr'C0,H.0-0000055 0 *00055 0.0025Ell r1001 14501CH,Br'CO,Na. CH McRr'C0,Na. CHEtBr'C0,Na.0 *000016 0-0025 0'0060[41 [4501 [1100]The numbers in brackets give the relative reactivities of thecompounds referred to the slowest as unity. The relative velocitiesdepend to some extent on the concentrations for which the measure-ments are made-the above values are valid for N / 3 0 solutions ofthe reacting substances.The velocities of these reactions have already been compared witht.hose in which the halogen has been displaced in the presence ofwater alone, and when the sodium salts are acted on by alkali, andit has been pointed out that the reactions in which silver salts areemployed are much more rapid.The reactions now under con-sideration also differ from those described in the previous papersas regards the magnitude of the difference in the reactivities. Itis not usually considered that the substitution of a methyl grouHALOQENS IN ORGANIC COMPOUNDS. PART IV, 361for hydrogen makes a very serious difference in the reactivity ofadjacent groups, and yet bromopropionic acid is a hundred timesmore active than bromoacetic acid as regards silver nitrate. It ishoped that measurements with the corresponding esters will throwsome light on the causes of this remarkable difference.The fact that the sodium salts of the bromo-fatty acids reactmore rapidly than the free acids with silver nitrate is doubtlessconnected with the fact that the concentration of silver salt ishigher in the former solutions than in the latter. The equilibriain the case of brornopropionic acid are represented by theequations :CHMeBr*CO,H + AgNO, CHMeBr*CO,Ag + HNO, (1).CHMeBr*CO,Na + AgNO, t CHMeBr*CO,Ag + NaNO, (2).As bromopropionic akid is a relatively weak acid, the equilibriumfor reaction (1) will be displaced towards the right to a smallerextent than in equation (2), and therefore the concentration ofsilver bromopropionate-perhaps the substance which really reacts-will be smaller in the former case than in the latter.Similarconsiderations account for the fact that the initial rate of reactionof silver acetate is greater than that of silver nitrate (p.355). I nthe former case, the equilibrium is represented by the equation :CHMeBr*CO,H + CH,*CO,Ag t CHMeBr*CO,Ag + CH3*C0,H (3).and owing to the fact that acetic acid is a much weaker acid thanbromopropionic acid, the equilibrium will be displaced towardsthe right, and the concentration of silver brornopropionate will berelatively great.Another way of interpreting these results is to assume that it isthe CHMeBr-COO’ ion which reacts with silver nitrate. It caneasily be calculated that the ratio of the CHMeBr-COO ion con-centration in sodium bromopropionate and bromopropionic acid inN/30 solution is about 4.5 : 1, which approximates to the ratio oftheir reactivities with silver nitrate. Similarly, the CH,Br*COO’ion concentration in sodium bromoacetate and the free acid isabout 4 : 1, whilst the ratio of their reactivities with silver nitrateis about 3 : 1.The suggestion that it is mainly the ions of the bromo-fattyacids which react with silver nitrate is further supported by theexperiments with nitric acid (p.351). It can readily be calculatedthat the CHMeBr-COO’ ion concentration in N / 20-bromopropionicacid is reduced t o about 1/7th of its value by the addition ofN/15-nitric acid, which is just the ratio in which the reactivityof bromopropionic acid towards silver nitrate is reduced by thesame proportion of nitric acid.The fact that the rate of reaction is approximately proportiona362 BARNETT AND SMILES : DERIVATIVES OFto the concentration of the bromopropionic acid instead of to thesquare root of the concentration (p. 348) appears at first sight tobe opposed t o the view that the ions are the active agents, butthis may be due to complications arising from the catalytic influenceof the silver bromide.There is no conclusive evidence as to what function of the silvernitrate is concerned in these reactions, but the fact that sodiumnitrate retards the reaction considerably speaks rather for the viewthat the Ag ions are the main aceive components. Should thissuggestion prove, on further investigation, to be well founded, aninteresting explanation of the great reactivity of bromo-f atty acidswith silver salts may be given on the basis of considerationsdeveloped in a previous paper (compare Trans., 1909, 95, 1839).It has been suggested that the relatively slow reaction betweenCH,*CHBr*COO’ ions and OH’ ions is connected with the mutualrepulsion of the negative charges, and it may therefore be antici-pated that the reaction between the CH,*CHBr*COO’ ion and apositively charged ion (in this case the Ago ion) will be relativelyrapid (compare table, Zoc. cit., p. 1835).The interpretation of the results is complicated by the equilibriarepresented by equations (l), (Z), and (3) (p. 361). An attemptwill be made to prepare pure silver bromopropionate and measureits reactivity, but owing to the great instability of the salt it willprobably be difficult to obtain trustworthy results.I n conclusion, I desire to thank Mr. R. W. Davies andMr. T. J. Ward, of St. Mary’s Hospital Medical School, for valuableassistance in the experimental part, of the work.CHEMICAL DEPARTMENT,ST. MARY’S HOSPITAL MEDICAL SCHOOL, W