112 SEGBLLER : THE RELATIVE ACTIVITIES OFSV.--The Relative Activities of Certain O r p n i c todo-comp~urtds with Sodium Plzenoxide. P a ~ t I V.Y’he Injuence of the Solvent.By DAVID SEGALLER.IT is well known that the solvent has an enormous influence onchemical reactions generally, but up to the present comparativelyfew workers have made measurements of the velocity-coefficientsof reactions with a view t o study the effect of the medium. Prac-tically nothing is known of the mechanism of reactions in whichthe solvent plays an important part, and although there are oneor two theories which attempt to account for this influence of thesolvent, there is as yet very little experimental evidence with whichto test these theories. We owe our most complete data probablyto Menschutkin, who studied the rate of formation of tetraethyl:ammonium iodide in about twenty solvents at looo (Zeitsch..physiknl.Chem., 1890, 6, 41). Menschutkin also investigated theinfluence of the solvent on some other reactions, for example, theaction of acetic anhydride on isopropyl and on isobutyl alcohol(ibid., 1887, 1, 611). Walden (ibid., 1908, 61, 633) measured thereaction-velocities between triethylamine and ethyl iodide invarious solvents a t 50°.It has been sought a t various times to connect the change of thevelocity-coefficients with change of the solvent in various reactionswith some physical properties of the solvent, such as viscosity anddielectric constant.Very little evidence has been produced to enable us to connectthe viscosities of the solvents with their influence on the rate ofreaction.On the other hand, i t is probable, as suggested byThomson and Nernst, that substances with the higher dielectricconstant would have a higher ionising power, so that if a givenreaction takes place between ions (and this is by no,means provedfor all organic reactions) the velocity of such a reaction would begreater in the solvent which has the greater dielectric constantCERTAIN ORGANIC IODO-COMPOUNDS, ETC. 113A comparison of the velocity-coefficient obtained by Menschutkinwith the dielectric constants shows that there is a certain q u a l i htive connexion. As shown below, the results of the measurementsdealt with in the present paper do not support the view thatthere is any direct connexion between the influence of the solventand its dielectric constant.Van’t Hoff (Vorlesimgen, 1901, 1, 219) has put forward animportant theory t o account for the influence of the solvent.He supposes that the effect of the solvent is due to two factors:(a) a catalytic action, and ( b ) an action depending on the solu-bility relations between the reacting substances and the solvent.For the latter effect he deduces the equation K’=K*S, so thatK’ is the product of the velocity-coefficients found and the solu-bility- or partition-coefficient of the reacting substances.A number of investigators conclude that the theory of van’tHoff actually does account for the facts observed by them; thusDimroth (Anrulen, 1910, 377, 127) has measured the velocityof transformation of methyl 5-hydroxy-l-phenyl-l : 2 : 3-triazole-4-carboxylate into its neutral isomeride in a number of solvents atloo.He shows that there is no relationship between the influ-ence of the solvent and the dielectric constant, but that there isa direct connexion between the velocity of transformation of oneisomeride into the other and their solubilities. The catalytic effectof the solvent is only minute.Von Halban (Zcitsch. physikal. Chem., 1909, 67, 129) in aninvestigation on the decomposition of triethylsulphonium bromide,and Bugarsky (ibid., 1910, 71, 705) from experiments on thereaction between bromine and ethyl alcohol, both support theconclusions of Dimroth.On the other hand, Freundlich and Richards (ibid., 1912, 79,681) in a recent paper find no connexion between the rate oftransformation and the solubility of chloroamylamine in piperidinehydrochloride.The present communication deals with the change of thereaction velocities between ethyl and propyl iodides with sodiumphenoxide in the following solvents : methyl alcohol, ethyl alcohol,propyl alcohol, isobutyl alcohol, isoamyl alcohol, and acetone.An,experiment was also made in ethyl alcohol, some sodium iodidebeing added to the solution.EXPERIMENTAL.The measurements of the velocity-coefficients were carried outWith exception of the reaction in acetone, all solutions were N / 5 ,VOL. cv. 1as described by the author (T., 1913, 103, 1154, 1421)114 SEGALLER : THE RELATIVE ACTIVITIES OFand the velocity-coefficients were calculated as before.I n the caseof acetone the procedure was somewhat different. Sodium phen-oxide dissolves very readily in boiling anhydrous acetone, and oncooling long, colourless needles crystallise out, coiitaining onemolecule of acetone of crystallisation. The needles gave onanalysis Na = 13.18. C,H,*ONa,(CH,),CO require Na = 13-22 percent.A t the temperature of the atmosphere a N/5-solution of sodiumpheno-xide is almost solid. For this reason and also on account ofthe very great rapidity of the reaction in acetone solution N/10-solutions were employed.The reaction mixture was in this case prepared a t the tempera-ture of the thermostat in a stoppered Jena-glass bottle, and atvarious intervals of time 20 C.C.were withdrawn by means of apipette also warmed to the temperature of the bath. The measurement was then completed as before. The temperature of thethermostat WM 58'6O.(4.t.356992112160(4 *t.506090130147t.6090120Solvent : Methyl Alcohol(Zero = 19 '40 C.C. ).a - x. K.16-27 0'028314.32 0'026513.10 0.026912.02 0 028210.35 0.0281Rmean = 0*0276.Solvent : Propyl Alcohol(Zero=18*15 c.c.).G - x. K.13-39 0.939112-75 0.038911.30 0.03719'65 0-03749.00 0'0361K mean = 0.0381.(6). Solvent : Ethyl Alcohol(Zero= 18.95 c.c.).t. a - x. K.14 16 -21 0-063622 15.03 0.062838 13 12 0.061860 11.19 0.061090 9.22 0'0618K mean = 0.0622.( d ) .Solvent : isol3utyl Alcohol(Zero = 18.25 c. c. ).t. ( I - x. K.30 16.72 0-016770 14.70 0'0189100 13'15 0.0212130 12.92 0.0175146 11.75 0.0191Kmean = 0.0187.(e). Solvent : GoAmy1 Alcohol (Zero= 19-50 c.c.).a - x. 1% t.' a - x. K.13.00 0'012312.32 0.0121l i . 0 0 0'012515.95 0.012615.01 0.0128 Kmean = 0'0124CERTAIN ORGANIC IODO-COMPOUNDS, ETC.Propyl Zodide.115(a). Solvent : Methyl Alcohol(Zero=l9*10 c.c.).t. a - z. K.30 10*00 0'010660 16.85 0-0116100 15.35 0.0126130 14-95 0'0110140 14.56 0.0116Kmean = 0.0115,(c). Solvent : Ethyl Alcoliol(Zero= 18'60 c.c.). N/lO-Sodium Iodide.t. a - x. K.50 15'50 0'021570 14'53 0'021591 13 '59 0.0223115 12'40 0'0242130 12.00 0'0227K mean = 0.0224.(e). Solvent : isoButyl Alcohol(Zero = 19'20 c.c. ).t. n - 2 . . K.70 17-30 0'00818120 16'28 0'00776150 15.80 0.00775247 14'92 0 *O 0 829350 12-52 0.00792Kmean = 0'00798.(6).t.507090102117(d 1.t.4061100120140(f 1.t.116158230366453Solvent : Ethyl Alcohol(Zero=18'30 c.c.).I t - x. K.14.90 0'094913.92 0.02451'2 -82. 0'065912-60 0 024311 *95 0.0248K mean = 0 0249.Solvent : l'ropyl Alcohol(Zero = 18.85 c.c.).a - z. K.16-78 0*016415.80 0'016714.62 0.015313 90 0.016113'35 0'0158Kmean = 0.0161.Solvent : isoAmyl Alcohol(Zero=19.96 c.c.).a - x. hi.17 *52 0'0060116.96 0-0056116 '30 0,0048914'70 0'0039113'25 0.00560K mean = 0.00540.(9). Solvent : Acetone (Zero=10'50 c.c.),t.a - x. K.7 8-40 0.67012 7 -50 0-63417 6.60 0.727t. a - x. K.23 5.80 0.65741 4-70 0.599K mean = 0.657.The results obtained are summarised in the following table,which contains the mean valuea of the velocity-coefficients for ethyland propyl iodides. The third column gives the dielectricconstants of the solvents at 20°:Ethyl .Propyl hielectricSolvent. iodide. iodide. constant.Methyl alcohol ........................... 0.0976 0-0115 31'2Ethyl ,, ........................... 0'0622 0.0249 25.8(Mixture containing N/lO-NaI) 0 '0224 -Propyl alcohol ........................... 0-0381 0.0161 22 2isoButyl ,, ........................... 0 '01 87 0 *OO 7 98 20 0isoAmyl ,, .......................... 0.01245 0'00540 16'0 ........................... 0-657 21 *5 Acetone -The results show that there is no obvious connexion betweenthe influence of the solvent and its dielectric constant. It wasI 116 SEGALLER : THE RELAllVE ACTIVITIES, ETC.to be expected that the reaction in methyl alcohol would be thefastest of those in the alcohols, and Menschutkin finds the reactionhe investigated t o be faster in methyl alcohol than in ethyl alcohol.The values of K obtained here, however, show that the velocityis less in methyl alcohol than in ethyl alcohol.This fact issupported by Eecht and Conrad (Zeitsch. physikal. Chem., 1889,3, 421, who find that the rate of ether formation is less in methylthan ethyl alcohol.The following table compares the relative reactivities in thevarious solvents found in this investigation with the results ofMenschutkin for tetraethylammonium iodide and of Hecht andConrad.The value of K in methyl alcohol is taken as unity ineach case:Sodium phenoxide Hechtand propyl iodide. and Conrad. Menschutkin.Methyl alcohol .................. 1 '00 1 -00 1'00Ethyl !, .................. 2-17 1 *95 0 *71Propyl ,, .................. 1 -40 - -isoButy 1 . , ................. 0.69 - 0 '50isoAniyl , 0 '47Ace tone .................. 57'13 - 1-18. .................. - -The velocity in acetone solution is remarkably high. This highvalue, taken in conjunction with the small velocity in methylalcohol, leads to the conclusion that the reaction considered isprobably not one between ions.This conclusion is, however, difficult to reconcile with the resultobtained when sodium iodide was added to the reaction mixture,the velocity-constant being thus somewhat lowered.An attempt to appIy van% Hoff's equation t o the resultsobtained does not give any connexion between the solubility andthe influence of the solvent; the product of the solubilities ofsodium phenoxide and of ethyl iodide in the alcohols varies onlyto a slight extent, and not a t all in consonance with the largevariations in the velocity-coefficients.The r61e of the latter musttherefore be assumed to be catalytic.I n connexion with the above results a slight modification mightbe suggested in the preparation of phenol ethers.The usual laboratory methods for the preparation of anisole,phenetole, etc., require several hours' heating.The very highreaction-velocity in acetone solution suggested the preparation ofthese substances in acetone solution. Sodium phenoxide (1 mol.)was dissolved in boiliiig acetone, and methyl iodide (li mols.)added. The mixture was boiled under reflux for fifteen minutes,and the oil separated in the usual way. A yield correspondingwith 95 per cent. of the theoretical was obtained. The methoCONSTITUTlON OF THE ORTHO-DIAZOIMINES. PART IV. 1 'rwas also tried for the preparation of phenetde and of phenylhexadecyl ether with the same results.General Conclusions.(1) The solvent medium has a considerable influence on thevelocity of reaction between alkyl iodides and sodium phenoxide.(2) There is no relationship between the influence of the solventand its dielectric constant.(3) The connexion between the solubility relations and theinfluence of the solvent suggested by van't Hoff does not hold inthis reaction, and it is to be msumed that the r81e of the solventis that of a catalyst.(4) The results obtained indicate t.hat the reaction is not" ionic."The author desires to express his thanks to the Research FundCommittee of the Chemical Society for a grant which has helpedto defray the expsnsm of this investigation.THE SOUTH WESTERN POLYTECHNIC, THE Ru'rLIsH SCIIOOI,,CHELSEA, S. W. MEKTON, S.W