Dissociation Constants of Para Substituted Benzoic Acids in Alcoholic Solutions Citation Lewis, William Bradley (1933) Dissociation Constants of Para Substituted Benzoic Acids in Alcoholic Solutions. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/N64X-D109. https://resolver.caltech.edu/CaltechTHESIS:02282017-141954872 Abstract The dissociation constants for benzoic and p-chlorobenzoic acids were measured in aqueous solution, and the constants for benzoic, p-chlorobenzoic, p-bromobenzoic, p-iodobenzoic, p-toluic, p-ethyl benzoic, and p-cuminic acids were measured in 80% alcoholic solutions by means of a quinhydrone electrode. A description of the electrode, and the vacuum tube amplifying circuit used in connection with the potantiometer, is given. The synthesis of p-toluic, p-ethyl benzoic, and p-cmninic acids is described. Item Type: Thesis (Dissertation (Ph.D.)) Subject Keywords: Chemistry Degree Grantor: California Institute of Technology Division: Chemistry and Chemical Engineering Major Option: Chemistry Thesis Availability: Public (worldwide access) Research Advisor(s): Unknown, Unknown Thesis Committee: Unknown, Unknown Defense Date: 1 January 1933 Record Number: CaltechTHESIS:02282017-141954872 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:02282017-141954872 DOI: 10.7907/N64X-D109 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 10074 Collection: CaltechTHESIS Deposited By: Benjamin Perez Deposited On: 01 Mar 2017 15:52 Last Modified: 16 Aug 2023 23:08 Thesis Files Preview PDF - Final Version See Usage Policy. 5MB Repository Staff Only: item control page

DISSOCIATION CONSTANTS OF PA...T\A SUBSTITUTED BENZOIC ACIDS IN ALCOHOLIC SOLUTIONS by William Bradley Le\vis In partial fulfillment of the requirements for the degree of Doctor of Philosophy California Institute of Technology Pasadena, California 1933 1. ABSTRACT 1he dissociation constants for benzoic and p-chlorobenzoic acids were measured in aqueous solution, and the constants for benzoic, p-chlorobenzoic, p-bromobenzoic, p-iodobenzoic, p-toluic, p-ethyl benzoic, and p-cuminic acids were measured in 80% alcoholic solutions by means of a quinhydrone electrode. A description of the electrode, and the vacuum tube runplifying circuit used in connection with the potantiometer,is given. The synthesis of p-toluic, p-ethyl benzoic, and p-cmninic acids is described. INTHODUCTION The present research was undertaken in hope that the dissociation constants of a series of pars. substituted benzoic acids could be determined vri.th sufficient accuracy to draw definite conclusions as to the relative inductive effects of the several substituent groups. Early investigators were co:ni'ronted with the fact that monosubstituted benzene derivatives fell into two classes, those that form meta disubstituted ~eriva.tives, and those fonning ortho or para disubstituted derivatives. Groups such as the nitro, the sulfonic, and t..~e carboxyl group., were found to be strongly meta-directing to the exclusion of the other oosi tions whereas groups such as the hydroxyl and the a~ino groups directed ortho and para, but not meta. The orientating effect of poles has shown that a positive pole is strongly meta-directing, even when removed from the benzene nucleus by a carbon atom in a side chain , cf. + J 8~~ C H5• CH - 2 0 -.rMe lUV.l 3 tyU meta. + C~ 5 .NMe 3 , ~ 10010 meta; On the other hand, negative poles are strongly ortho, para directing; witness, the rapid bromination of phenol in alkaline solution to symmetric tribromophenol. For other groups, that a.re not definitely polar, a dipole is fonned. 'Ihese can be ortho,para-directing, or meta-directing de- pending on the electron-attraction of the group. Groups that attract electrons strongly a.re meta-directing, those that attract them weakly, or repell them., are ortho,pa.ra-directing to a lesser of greater extent. The groups studied in the present research a.re weakly ortho, para-directing groups. They have not enough electron attraction to be meta-directing, nor enough electron repulsion to be strongly ortho,,para- 3. directing. However, their effects are noticeable in the de gree of dissociation of the substituted benzoic acids. A group that attracts electrons, tends to shift the electron pair between the hydrogen and the oxygen atom in the carboxyl, away from the position they would assume in the unsubstituted acid, towards the oxygen, thus leaving the proton freer to pass into solution as an ion. Similarly, those that repell electrons, shift them away from the oxygen toward the proton, making it less free to ionize. The halogen groups e:ir.hibit a weak electron attraction,, and the alkyl groups a weak electron repulsion. On the above ba sis, it wa s thought that the increasing kernel charge in the series of halogen-substituted benzoic acids, might manifest itself in a distinguishable increase in the dissociation constants, passing f r om chlorine to iodine. A simila r change might be exhibited in t he alkyl-substituted acids passing from methyl,, to ethyl,, to isopropyl. 4. METHODS Several general methods have been used in investigating the dissociation constants of weak acids: and potentiometric measurements. conductivity, colorometric For acids that are weak as benzoic acid, and as insoluble as the halogen substituted benzoic acids, conductivity measurements would be attended with such errors, due to the difficulty in technique required to measure so feebly conducting solutions, that accurately detennined acid constants seemed unattainable. Similar difficulties attend colorometric determinations. '.!he insol- bbili ty of the acids would rule out potentiometric titrations. TI1e quinhydrone electrode had been used with success on benzoic acid, and ~ra.s adopted as the most practicable means available. 5. ELECTRODE REACTION The quinhydrone electrode furnishes a convenient oxidationreduction potential., and may be used as a hydrogen-ion concentration cell: (1) Pt., quinhydrone, H+ ( c 1 ) l\ W (c 2 )., quinhydrone., Pt Considering the electrode reaction, QH = Q + 2H+ ( c ) + 2 e the electrode potential is E and = E - RT ln i9l_ ( c t o 2F (QH) since quinhydrone dissociates in solution to give equal parts of quinone and hydroquinone, the ratio fil_ = 1 and (QH) (2) E = E0 - RT F ln ( c ) • Thus the cell (1) has a potential given by (3) -E = ~ ln ~ F c2 c is written instead of a and it will be shovm that the method is con- 6. cerned with the concentration rather than with the activity. The acid whose strength is to be me a.sured is dissolved at a definite mola.lity in a. 0.09 molal solution of KCl. This forms a half cell, which is measured against a standard half cell containing HCl 0.01 (All solutions were ma.de 0.09 molal in KCl molal., KCl 0.09 molal. to eliminate liquid junction potentials.) ion activities in the two half cells. - log '( and y is proportional to the Consider now the hydrogen To a first approximation, 1e- where t'" is the ionic strength is the activity coefficient: = (Y..~, log y2.. = ol.~-i. log Thus Y, o<. ::. f J I(.._ (Z14'r\) 1 i . ~oi, ( :DXT/ l. =o. ~-c) '- c. '0 <t y,- :: ei.(~e~-0,) and }' 2.. So rewriting (3) substituting activities for concentrations -E a, = 59.• 15 log_ millivolts ~ c, + log _}! = 59.15 (1 og -- = ) r~ C2 c, 59.15 log_ + 59 .15 a ot. ~.re. L Cz fe 1 ) c, = 59.15 log Cz + /!). E here t'' = 0.1000 ~t whence = 0.0900 A. E "'~. = o. 3162 "~?. = o. 3000 = -59.l5x0.5066xO.Ol62 =-0. 48mv. Since this value is of the order of magnitude of the reproducibility 7. of the experiments, it can be neglected, and concentrations may be used in place of activities. From equation (3) we find EF + RT E log C2 = - =Ex 0.010907 +log c 1 But in all our runs E is negative and log c 1 = -2.0000; so taking the absolute value of E, (4) -log c 2 =Ex 0.016907 + 2.0000 where c2 is the hydrogen ion concentration from the organic acid. The Acid Dissociation Constant To calculate the dissociation constant of the acid, we take a.s our definition ka -- C - C2 where c is the molal concentration of the acid and c 2 is the hydrogen ion concentration determined by equation (4). The dissociation constant of an acid is the equilibrium constant for the reaction ( 5) HA = H+ + A- Before the concept of activities, the constant was written according to the usual mass action calculation rn-t-)(A-) ( 6) (HA) (H+) =(A-), and (HA) and since = = molality of acid assuming no dissociation cH+ = (H+) = molality of H+ ions where °HA 8. expression (6) may be written With the advent of the concept of activity, the acid dissociation constant was written ( 8) (cH )2 y~ = (cHA- cH+) where aH and aA are respectively the activities of the hydrogen ion and the anion, and aHA is the activity of the undissociated portion of the acid. If y is taken as the mean ion activity coefficient, and the activity of the undissociated acid is taken equal to its concentration, expression (8) holds. where In the case of aqueous solutions of benzoic acid Y' can be determined, KA is calculated as well as ka• In alco- holic solutions where y is not known, only ka is given. In order to facilitate computation, curves were plotted of ka against E. In all the final runs, the acid concentration was 0.02 molal. Pre- liminary work showed that too dilute solutions were attended by unsatisfactory results due, possibly, to adsorption of hydrogen ions on the glass of the containers. Because of the insolubility of the halogen substituted benzoic acids, it was necessary to change from aqueous solutions to alcoholic solutionso These will be discussed later. Apparatus The cell used consisted of an inner and an outer half connected through 9. a ground glass joint, which was wetted with saturated KCl solution when the cell was being assembled. the standard HCl, KCl solution; ic acid, KCl solution. The inner cell held 3 cc. of the outer cell, 17 cc. of the organ- These amounts gave the same liquid level in both containers, thus reducing diffusion across the ground glass joint to a mininmm. The inner electrode was a rectangular platinum foil 8 mm. by 5 mm. The two outer electrodes were coiled platinum wires o These proved more reproducible than platinum foil. Due to the relatively high resistance of the cell, it was necessary to use an amplifying circuit in conjunction wi th the potentiometer. The circuit used consisted of two vacuum tubes balanced one against the other by varying the grid bias. The description of the electri- cal circuit is given below. Potentiometric Circuit The diagram shows the insertion of the amplifier into the potentiometer circuit. This was accomplished by removing the wires ~ E_, from the switch posts ~' .£', and soldering them together. The switch was then connected to the plates of the amplifying circuit. After the potentiometer is adjusted to the standard cell in the ordinary manner, the single -pole double - throw switch, s, is turned so as to short out the potentiometer-cell circuit. The two tubes are then brought into balance by means of the subsidiary potential (tapped off resistances R and R ) applied to the grid of the float- 3 ing tube. 4 The switch is then thrown, applying to the grid of the other tube the electromotive force of the cell, which is counterbalanced by the potentiometer. This method has the advantage of 10. having the D.C. resistance of the vacuum tube always in series with the electrolytic cell. The high resistance reduces the current drawn dur- ing a reading to a negligible value, thus preventing polarization errors. Solutions The alcoholic solutions were made up to 80% alcohol by volume. This value was chosen in order that their molality in KCl might be 0.09 and still permit the organic acids to be sufficiently soluble. If the original volume percent alcohol is ~' and the desired volume percent is !_, dilute !!. cc. of the original to 100 cc. by adding E_ cc. of water, thus obtaining 100 cc. of alcohol !_percent by volume. be the density of the original alcohol; Let d1 d2, of the final alcohol; da, of absolute alcohol. Then a s - 100 t 100 t a s The amount of water in 100 cc. of final alcohol : 100 d 2 - t da II If " II !!. cc. of original " II " " "to be added~bd : lOO(d25d1) grms. . grms. t where d is the density of water. If the final solution is to be !!!_molal in some constituent, an aqueous solution is made lm molal in that constituent, where l -= ~ • Then b cc. of this solution diluted to 100 cc. with alcohol will be m molal 11. in the constituent, and .i percent aleohol by volume. So in making up alcoholic solutions, tbree solutions are necessary: Solution I alcohol ~percent by volume Solution II HCl (J; x 0.01 molal), KCl O. x 0.09 molal) Solution III KCl (1 x 0.09 molal) To make standard reference solution, dilute b cc. of II to 100 cc. with I. To make standard organic acid solution , dissolve sample of acid in less than a cc. of I, add b cc. of III, and dilute to 100 cc. with I. Preparation of Materials The alkyl substituted benzoic acids, with the exception of the toluic acids, are not obtainable from any chemical supply house. of synthesis presented themselves: Two methods 1. the introduction of a CHO group para to an alkyl group by the Gatterman reaction, followed by oxidation to the acid; 2. the replacement of an amino group by a CN group by the Sandmeyer reaction, followed by hydrolysis to the acid. The preparation of p. toluic aldehyde is described in Organic Syntheses Vol. XII, p. 80. The procedure was followed, using carefully dried toluene, freshly prepared Cu 2 01 2 (Org.Syn.), and freshly opened AlC1 • 3 The absorption of CO was good, and after 4.5 hours, the reaction mixture had thickened to the consistency of butter. The yield of alde- hyde was 51. 5% of the theoretical. This was converted to the acid by the Ca.n..~izzaro reaction, with a 72% 12. yield. The same method was employed with other hydrocarbons. Two runs with n. butyl benzene, two with curaene, and one with ethyl benzene. In all these cases, the absorption of CO was poor, and the yields of aldehyde were nil or very small. Two more runs were made with ethyl benzene, each yielding a low percentage of an aldehyde. The aldehyde from the first was oxidized with alkaline permanganate and converted entirely to phthalic acid. The aldehyde from the second was converted by a Cannizzaro reaction to an impure ethyl benzoic acid. Repeated recrystalizations failed to raise the melting point above 87°. The contaminant was probably o. ethyl benzollic acid. This dii'ference in behavior between toluene and ethyl benzene, viz: that the toluene directs the entering co entirely to the para position, whereas the ethyl benzene seems to favor the ortho position as well, might prove worthy of future investigation• The failure of this method, except in the case of toluic acid, (which was available from Eastman), led to the adoption of alternatives. For Curninic acid, p. cuminic aldehyde was obtainable from Eimer and .Amend. This was oxidized by alkaline permanganate to p. cuminic acid, in almost quantitative yield. Para aminoethylbezene was prepared by a molecular rearrangement of ethyl aniline in the presence of zinc chloride. (Hickenbottom J.c.s. The method used was a slight modification of the one there described. 90 g!'lllS• of freshly distilled ethyl aniline were heated with 45 grms. 13. of i'used zinc chloride in a 300 cc. Erlenmeyer flask on a sand bath. The flask was fitted with a short air reflux condenser (14 inches long), and heating was adjusted to lreep the mixture boiling gently. After l0.5 hours, the hot, red mixture was poured into a porcelain dish, where it solidified. The solid zincichloride salt of the pri- mary amine was triturated in a mortar with ether to remove any of the unchanged ethyl aniline. The ether was filtered off; the product pressed between sheets of filter paper, and decomposed with 200 cc. of 50% NaOH solution to the free amine, which was separated by steam distillation. This was extracted with ether, dried over CaC12 , and the ether was evaporated of't~ The yield of p. amino ethyl benzene was 50 grms, 54.5% of the theoreticalo In converting this to the nitrile, standard procedure was follo wed; except that after forming the cuprous cyanide complex, enough sodium carbonate was added to neutllal.ize the acid from the diazo-mixture. This prevented the evolution of HCN, which being undesirable from the standpoint of general comfort, tends also to reduce the yield of the nitrile. 37.5 grams of the neutral sulfate of the amine was diazotized, by dissolving it in a mixture of 40 ~ams of suli'uric acid in 100 cc. of water. This was cooled in a pan of ice, and to it was added, with constant stirring, 18 grms. of sodium nitrite in 50 cc. of water. The diazo solution was added to the warm cuprous cyanide solution slowly ove~ a period of about 10 minutes. This solution had been prepared previously by 65 grms. of copper sulfate dissolved in 200 cc. 1 .4 . of water, heated on a water bath. the fume hood, To this was added, cautiously in 55 grms. of NaCN in 100 cc. of water. After the evo- lution of cyanogen had ceased, and the entire amount of cyanide had been added, 60 grms. of sodium carbonate was added. The reaction was done in a 3 litre balloon flask. The mixture was heated 15 minutes after the diazo solution had been added, and was then steam distilled. The nitrile was taken up in ether, the ether evaporated and the product was shaken with 4 grms. of stannous chloride in 10 cc. of HCl to remove isocyanide. It was then poured into water, extracted with ether, dried over Cac1 2 , and ether evaporated off. Yield of purified nitrile, 13.4 grrns; 59.5% of the theoretical. This was hydrolysed with 40 grms. of sulfuric acid in 20 grms. of water, heating for 10 hours in an oil bath at 155°. acid 7.5 grms ., 43.5% of the theoretical. Yield of crud~ This was purified by dissolving in sodium carbonate, reprecipitating with HC l, twice recrystallizing from 50%alcohol, and finally subliming . fied acid, Yield of puri- 5.5 grms. The rema ining acids, p. chloro benzoic, p. bromo benzoic, and p.iodo benzoic, were from Eastman. The first t wo were purified by twice re- crystallizing from 50%alcohol, while the p. iodo benzoic was sublimed just before usingo 15 . Table I m.p. 0 c. Acid * Benzoic 121 p. Chloro benzoic 2)(-238 p. Bromobenzoic 251-252 * p. Iodobenzoic * p. Toluic * p. Ethylbenzoic 265-268 177-178.5 * p. Cuminic 115.5-116.5 112-112.5 *Sublimed. Preparation of Alcohol The alcohol for the final runs was refluxed with KOH and granular aluminum for two hours to render it aldehyde free. I. & E. Ch. Vol. 5, No.2, p. 100 (1933) ). (Stout and Schuette It was distilled into a nitro gen filled container, an d forced out by nitro gen when used. Its density at 25° gave no t est for aldehyde with Schiff's reagent. was determined pyknometrically. Density 25° 0.80064 Vol. percent alcohol 97.114 Density 80.00%alcohol 25° 0.83904 s "' 97.11% b t ~ 80.00% 1 whence Solution I Solution II Solution III - It grms./cc. grms./cc. 18.oo cc. 5.5556 alcohol 97.114% HCl 0.055556 molal KCl 0.50000 " KCl 0.50000 16. The HCl solutions were made up by weighing out constant boiling HCl prepared after the method of Faulk and Hollingsworth (F & H, J.A.c.s • .45. 1220 (1923) ). 1 The KCl used was twice recrystallized, fused, and powdered. Technique of E.M.F. Measurements Solutions of the acids under investigation were made up to 0.02000 molal strength in 100 ac. samples. These, with the reference HCl solution, were kept in a thermostated oil bath at 25.00° ± 0.01°. In making a run, the ground glass connection between the inner and outer half cell was wetted with saturated KCl solution; a pinch of quinhy- drone (prepared method Bijlmann Bulo Soc. Ch. ~' 245 ) was added to each half from a knife blade; the electrodes were inserted in the outer half cell, and the cell was mounted on a shaker in the thermostatic bath. The solutions were then added by means of pipettes so built that the liquid level in the two half cells was the same. The electrode for the reference solution was then inserted, and the shaker started. Pot entiometric measurements were made on the average of every two minutes over a period of ten minutes . Between runs , the cell was carefully washed and dried. The electro des were kept in the same solution that they were surrounded by in the cell. This treatment gave, by far, the most reproducible and consistent results. It might be noted that treating the electrodes with cleaning solution is disastrous; subsequent runs fail to agree, by wide varia- tions, with those previous to such treatment, and new electrodes I!Illst be made. 17. In the case of runs with aqueous solutions, shaking was not necessary; identical results were obtained with or without. But for the alcoholic solutions, wide dmscrepancies occurred which were eliminated by shaking continuously throughout a run. Experimental Re sults Preliminary runs in aqueous solutions 0.09 molal in KCl of various molalities of HCl against the standard reference concentration of 0.01 molal HCl, showed that for not too dilute solutions, the hydrogen ion concentration could be determined to a few percent. When the acid con- centration was less than 0.0001 molal, the determinations gave i nvar iably too low a concentration. This could be explained by assuming that. hydrogen ions were adsorbed on the walls of the cell and containing vessel. A very slight molal adsorption per unit area of surfac·e would account satisfactorily for the discrepancy. It was due to this that alcoholic solutions were employed o Table II HCl solutions in 0.09 molal KCl against standard reference solution HCl 0.01 molal, KCl 0.09 molal. -E observed CH calculated Run molality HCl -E calculated 1 l.OOOxl0-3 59.15 mv. 59.30 0.9949 x 10-3 2 " n 59.10 10003 tt It 59.41 0.9905 tt tt 59.40 0.9909 r; tt 58.60 10022 3 4 5 18 .. Table II - Continued Run molality HCl -E calculated -E observed 6 l.ooo x lo-3 59.15 58.94 1.009 7 tt " 58.90 1.010 8 tt ti 59.23 0.9975 9 tt ti 57.70 1.059 10 " It 57.05 1.086 ll l.ooo x io-4 118030 120.52 0.9185 x 10-4 12 tt " 120.45 0.9210 13 " " 124.62 0.6157 14 It ti 119.41 0.9591 15 " " 126.62 0.7244 CH calculated Measurements were made with aqueous solutions of benzoic acid at 2.000 x 10-2 molal and 2.000 x 10-3 molal concentrations. The results appear in Table III. Table III Dissociation Constant of Benzoic Acid in Aqueous Solution Eun Molality -.E ka x 105 KA x 105 1 2.000 x 10- 2 52.35 9.10 5.96 2 " 52.05 9.31 6.16 3 lt 52.10 9,08 5.95 4 52.60 8.94 5.85 5 " " 52.50 8.99 5.88 6 " 52.45 9.02 5.91 mean: 9.11 5.98 19. Table III - Continued ka x 105 KA x 105 84.00 8.92 5084 tt 84.20 8,~4 5.72 9 " 84.50 8.52 5.58 10 tt 84.. 30 8.68 5.68 11 tt 84.65 8.40 5.50 12 tt 8~.4'i 8.56 5.61 mean: 8.64 5.66 Run Molality -E 7 8 2.000 x 10-3 note: y =- 0.81 KA ::::: ka y'). (Kilpatrick and Chase J .A.C .s • .5}, 1732 (1931) ) ' Runs were also made with aqueous solutions of p. chlorobenzoic acid. Due to its low solubility, no solution could be stronger than 4 x 10-4 molal. Table IT Dissociation Constant of p. Chlorobenzoic Acid in Aqueous Solutions ka x 105 KA x 105 106.29 10.8 7.10 Run Molality -E 1 3.9540 x lo-4 2 tt 106.36 10.7 7.01 3 tt 106.18 10.9 4 " 106.89 8.05 7.14 5.28 5 6 " 107.31 II 105.26 9.72 12.0 7.85 7 II 106.15 11.0 z.21 mean: 10.5 6.86 6.45 20,, Table IV - Continued x 105 Run Mola1ity -E 8 1.9770 x 10-4 120.44 8.19 9 tt 122.43 6.43 mean: 7.31 k a 4.81 The final runs in 80%alcoholic solutions are given in Table V. Table V Dissociation Constants of 0.02000 Molal Acids in 80% .Alcohol 0.09 Molal in KOL -E Acid ka x 107 Acid mv. p. Iodobenzoic mean: p. Chloro benzo ic mean: -E ka x io7 mv. 92.41 38.0 35.5 97.01 26.5 94.40 32.5 94.24 33.0 93.42 91.30 41.5 97.23 35.2 26.1 93.18 36.0 94.36 33.2 93.63 35.0 92.77 37.2 105.29 13.5 96.64 27.5 105.91 13.0 96.06 29.0 106.14 12.7 95.30 3005 107.08 l2o0 96.50 27.8 95.45 30.4 92.99 36.5 93.41 Po Bromobenzoic Benzoic 106.11 . 1208 .2 1. Table V - Continued Acid -E ka x io7 Acid mv. p. Toluic p. -E ka x io7 mv. 115.67 6.1 lll.68 p. Ethylbenzoic 109.59 9.9 8.4 113°45 7.4 113.60 7.2 111.77 8.4 112.42 8.o 113.28 7.5 116.60 5·.9. 110.z8 9.0 mean: 113.99 7.1 111.77 8.4 Cuminic 108.66 10.5 108.41 10.7 109.49 10.0 108.77 10.5 ~11.65 8.5 mean: 109.39 10.0 22.• Cle c tro l ytic Ge l l ---- 1 1 . 23". Elect:tic Circuit 1.5 Description of Amplifying Circuit R1 Super Da.vohrn wire-wound resistance, 100,000 ohms. ~ Ordinary filament rheostat. R3 20 oluns. R4 50 oluns. Tubes: Western Electric, 102 D. Potentiometer~ Leeds and Northrup, Type K, Standard Cell: Weston, Galvanometer: Leeds and Northrup, No. 200334. No. 6915 Cat. 2420 A EM F = 1 0 01870 Sens. 31 meg. Res. 3_40 ohms. SUMMARY I Dissociation Constants in Aqueous Solutions Acid Observer Benzoic Ostwald (1) Euler(2) Schaller (3) Ko rtwrigh t ( 4) White and Jones ( 5) Kirschman, Wingfield and Lucas ( 6) Present pa.per p-chloro- Ostvrald (1) benzoic Present paper II Molality KA x 10 0.02 5.98 ti t,.\, i II L. :ro II \, . I '\ ti 5 b-10 (.. lo 3 " II 5.98 3. 954x10- 4 7.3 Dissociation Constants in Alcoholic ~olutions Acid Mola.lity k a x 10 0.02 12.8 p. chlorobenzoic II 30.4 p. bromobenzoic II 35.0 p-iodobenzoic II 33.2 p.toluic " 7.1 p.ethylbenzoic ti 8.4 p.cuminic II 10.0 benzoic 7 (1) Osti..va.ld, Z.physik.Chem., 3, 241 (1889). (2) Euler, ibid., 21, 257 (1896). (3) Schaller, ibid:-;- 25, 497 (1898). (4) Kortwright, ibid.-,-21, 311 (1896). (5) White and Jones, .Am:Ch.J., 44, 159 (1910). (6) Kirshman,, Wingfield and Lucas, J.A.c.s., 52, 23 (1930). 26. .. Conclusions Runs with 0.02000 molal benzoic acid in a queous solutions give values for KA with a maximi.un deviation of 2% from the mean; and at the dilution, 0. 002000 molal , a maximum de vi at io,n of ?Jfo. In the case of p. chlorobenzoic acid, the aqueous solutions gave values for KA v~ying 8% from the mean. This was thought excessive, especially when the determinations at half the original concentration fell considerably lower. The low value for ka in both cases, raises doubt as to its validity. The runs in alcoholic solutions, although consistent only to 20% of the mean values, show a much better spread from the halogen- substituted benzoic acids to benzoic, and from benzoic to the alkyl substituted benzoic acids., In the case of the halogen• substituted acids, no distinction can be drawn between the strengths of the three. The saroo can be said for the alkyl- substituted acids, although there is a slight indication that p. cuminic acid is stronger than p. ethyl benzoic is stronger than Po toluico Since the research hoped to show definite distinctions between the three halogen- substituted acids, and between the three alkyl- substituted acids , the results are negative. However, it was shown that there is a definite distinction between the influence of an halogen group, and an alkyl group; and that the differences in influence between the members of these two classes are of a much lower order of magnitude. The research was undertaken under the direction of Professor H. J. Lucas, to whom thanks are due for his many helpful suggestions and timely criticisms.