The Reaction in Acid Solution between

By Wilson J. Broach, Robert W. Rowden and Edward S. Amis. Contribution from, the Department of Chemistry, University of Arkansas, Fayetteville, Arkans...
0 downloads 0 Views 531KB Size
Feb. , 1952

TETRABROMOPHENOLSULFONPHTHALEIN AND SILVER NITRATE IN ACID

189

THE REACTION IN ACID SOLUTION BETWEEN TETRABROMOPHENOLSULFONPHTHALEIN AND SILVER KITRATE BY WILSON J. BROACH, ROBERT W. ROWDEN AND EDWARD S.AMIS Contributionfrom the Department of C h i s t r y , University of Arkansas, Fayettevillc, Arkansas Receined Nooembsr 87, 1860

The reaction of silver nitrate with tetrabromo henolsulfonphthalein was studied at 25.1' and 30.1'. The reaction n~echanism was found to be consecutive first order. *he specific velocity constants, the ener 'e8 of activation, tbe Arrlienius frequency factors, and the entropies of activation were determined for the removal of t f e first and second bmmine atoms. The specific velocity constant for step 1 was found to be from 8 to 10 times as great as that for ste 2. The rates of removals of the third and fourth bromine atoms from the dye molecules were shown to be negligibly slow. %e negative entropy values were explained as due either to a rigid activated as compared to a reactant state or to a small transmipsioil coefficient. Transmissions of the reaction mixtures were found to increase with time. This phenomenon was discussed at some length.

The reaction of silver and mercuric nitrates with brominated alkyl and aryl compounds has been studied by various authors.' The kinetics of the reaction between silver nitrate and tetrabromophenolsuIfonphthalein, however, have not been studied. It was thought of interest to make a study of the kinetics of this reaction, since it gave promise of being a reaction, the mechanism of which would involve the consecutive removal of bromine atoms from the dye molecule by the silver ion. Since the dye is only slightly soluble in water solution and the silver nitrate could be present in much higher concentrations, it was thought the mechanism might be pseudounimolecular.

sented the mechanism as being the two simultaneous reactions RI AgNOa EtOH --+AgI HNOa ROEt (A) RI AgNOa --+ AgI R*O.NO* (B) They found that the bimolecular velocit,y constant was increased by increasing the initial concentration of silver nitrate and to a leMer extent by decreasing the initial concentration of alkyl halide. Baker' found t,hat the rcact,ion of p-methylbenzyl bromide and p-nitrobenzyl bromide followed the mechanism represented by reaction ( A ) and (B) above (7670 in accordance with (A)), but that the reactions were unimolecular throughout. Roberts and Hammett' studying the reaction of benzyl chloride and mercuric nitrate in water and dioxane solution concluded that the mechanisni was that of the formatmionof ionic intermediate which t,hey called a carbonium ion, benzyl ion, CsH&H?+. They found the reaction to be bimolecular with one mole of mercuric ion reacting with two moles of benzyl chloride. The equation for bimolecular reactions was modified by Roberta Experimental and Hammett to account for the latter observation. The The brom phenol blue was Eastman Kodak No. 752 final products of the reactions were benzyl nitrate and benzyl tetrabromophenolsulfonphthalein.' Other chemicals used alcohol, which corresponds to the mechanism in equations were of C.P.grade. The thermostat at 25.1' was constant (A) and (€3) with water substituted for ethyl alcohol in to =kOi05' and the one at 30.1' was constant to within equation (A). fO.O1 Brown bottles were used to prevent possible catalTo determine the mechanism of the silver nitrate and tet.ysis of the reaction or decomposition of the precipitated rabromophenolsulfonphthalein reaction it was thought' silver bromide by light. Each run was set up by i tting necessary to test for the presence of nitrate in the organic 200 ml. of 6.97 X lo-' M tetrabromophenolsulfonpEtralein, product, since it is known that ring organic nitrates arc 50 ml. of 0.9974 M nitric acid and 50 ml. of 0.08611 M readily hydrolyzed in water solution. The absence of orsilver nitrate into a bottle. The silver nitrate RW added ganic nitrate was proven by makine a specific test for nitrolast and the time of the beginning of the run was taken as the gen. To make this test, the organic product was extractcd last drop of silver nitrate entered the reaction bottle. The from the reaction mixture with diethyl ether and the ether bottle was sto pered immediately and the mixture shaken was evaporated at room temperature. The dry residue was and placed in t i e thermostat. At intervals the bottles were again extracted with ether and the ether again evaporated removed from the thermostat, quickly and vigorously at room tern erature. The dry residue was taken up in shaken and returned to the thermostat. The run was alcoholic Naz808solution so that any free nitric acid carried analyzed by filtering through a previously prepared and over in the ether extractions would be converted to NaNO,. weighed porcelain crucible contafnin a n asbestos mat. The The alcohol was evaporated and the dry residue was extime of completion of the run was tafen when the last of the tracted with dry ether and the ether evaporated. The dry solution passed through the filter. Timing was not critical residue was a ain extracted with dry ether and the ether since the shortest run, except the first, involved at. least evaporated. t h i s product was used in making the nitrogen 17 hours and the longer runs involved several hundred hours. test. The nitrogen test was that given by Shriner and FuAfter filtering the reaction mixture t,he bottle and crucible son.6 Repeated tests gave nL indication of t.he presence of were t,horoughly washed with distilled water containing nitrate: therefore, from comideration of prcvious work approximately 2 ml. of nitric acid per liter of water, then and from the fact thcre was no organic nit,rate in the prodwith 95% ethyl alcohol and finally with pure distilled water. ucts, we believe the over-all mechanism of thc reaction is If the crucible tended to become clogged with adsorbed dye, RBr, Ag+ 2H20 --+ RBh(0H) AgBr HIO+ alcohol wash was used to desorb the dye. The cruciblc waa brought to constant weight in an oven and the weight RBrr(0H) Ag+ 2H20--+ (C) of silver bromide determined from the difference in weight RBro (0H)t AgBr HaO+ of the crucible and the weight of the crucible plus the silver bromide. From the time rate of production of silver bro- Ifany organic nitrate is formed it is subsequently hydrolyzed. mide the rate of reaction of silver nitrate with tetrabromoI n the reactions above, one silver ion reacts with one dye phenolsulfonphthalein was calculated. or substituted dye molecule to produce one molecule of Mechanism.-Burke and Donnan studied the interaction AgBr. Since each step requires a collision with a silver ion of alcoholic silver nitrate with alkyl halides.' They repre- i t would Beem reasonable that the step involving the original removal of bromine atom would be much faater than any of

+

+ +

+ +

+

.

+

(1) Euler, Bsr., 89 2726 (1908). Chaw and Kilpatrick. J . Am. Chem. Soc.. 64, 2284 (1932). Baker, J. C h m . Sw., 987 (1934). Roberta and Hammett. J . Am. Chem. Sw., 69,1063 (1937). Panepintoand Kilpatriok. ibid., 69, 1871 (1937). Amis and La Mer. ibid.. 61, 905 (1939). (2) Burke and Donnan,

J . Chon. Soc., 88, 666 (1904)

+

+

+

+ +

+ +

(3) Baker, ibid., 11, 987 (1934). (4) Roberta and Hammett, J . Am. Cham. Soc.. 69. 1063 (1937). (5) Shriner and Fusoo, "The Systematic Identification of Organic Compoundo." John Wiley and &M. Inc., New York. N. Y.. 1948. pp. 52-54

WILSON J. BROACH, ROBERT W. ROWDEN AND EDWARD S. AMIS

190

Vol. 56

In Tables I and 11, column 3 contains precipitated AgBr expressed in moles per liter at 25.1' and 30.1°,respectively. Column 4 of Tables I and I1 contains precipitated silver bromide expressed in moles per liter at 25.1' and 30.1°, respectively, coming from step 1 of reaction (D) and calculated from equation 3. Columns 5 of Tables I and I1 contain precipitated silver bromide expressed in moles per liter at 25.1' and 30.1", respectively', coming from step 2 of reaction (D) and calculated by equation (4). The reaction velocity constants IC1 and IC2 used in calculating z and y were obtained using the following procedure. In early stages of the run it was assumed that step 1 of reaction (D) predominated and therefore the the quantity ( a - x) represented the concentration of the dye corresponding to time t. Therefore a plot of precipitated silver bromide against time was made and its slope found at a time For details see Amis.6 sufficiently early so that y was not significant. This slope corresponded to dx/dt in equation (1) Data and when divided by the corresponding (a - x) In Table I are recorded the data taken at 25.1 gave preliminary values of kl. The method of obtaining an approximate value of kz was to take the TABLE I slope of the silver bromide-time curve after sufKINETIC RUNS BETWEENTETRABROMOPHENOLSULFONficient time had elapsed so that x was constant. PHTHALEIN A N D SILVER NITRATE AT 25.1 This slope was divided by (x - y) and gave kz ackl = 2.95 X b = 3.80 X 10-3 cording to equation (2). (x - y) was obtained in AgBr, X. I/. (5 + u), Run Time, (nioles/l.) (moles/l.) (inoles/l.) (tnoles/l.) the following way no. hr. x 103 x 103 x 108 x 103 the steps. The removal of the second bromine atom would be second in reaction velocity and the removal of third and fourth bromines would be comparatively very slow. These conclusions are reasonable statistically. Therefore, we shall write the over-all reaction for the purpose of calculating experimental results as RBr, --+ RBrs(0H) --+ RBrs (OH)z (a) (5) (Y) (D) where a, z and y represent the over-all concentrations of original dye and first and second substitution products, respectively. z also represents the silver bromide produced in step 1 and y represents silver bromide from step 2. kl is the velocity constant for step 1 and k~ is the velocity constant for step 2. Mathematically dx/dl = kl(a x) (1) dy/dt = k z ( ~- y) (2) Solution of equabion ( 1 ) and (2)yields x = a - ae-w (3) akt (1 - e--kll) akl ( 1 - e--kzi) y = kz - kl (4)

1 2 3 4 5 6 7 8 9 10

0.17 24.8 24.9 52.6 73:o 74.2 99.9 144.9 244.0 461.0

O.OO0 ,378 .376 .580 .756 .776 .816 .962 1.070 1.206

0.003 ,362 .362 .550 .618 .619 .661 .687 .696 .697

0.000 ,018 .019 .064 .lo4 ,105 .155 .237 .381 ,558

0.003 .380 .381 .614 .722 .724 ,816 .924 1.077 1.255

x + y = c

(5)

where C is total concentration of AgBr precipitated. Then y = c - x

and 5

-y

=

z - (C

-5)

(6) = 22

-c

(7)

Finally kl and lcz were adjusted by trial and error and have the values represented in Tables I and 11. From these tables it is observable that kl is approximately 8 times kz at 25.1" and 10 times at 30.1'; Table I1 contains data taken at 30.1 O. in the third columns of the tables the actually observed AgBr precipitate is represented. Columns 4 TABLE I1 contain values of x, columns 5 calculated values of KINETIC RUNS BETWEENTETRABROMOPHENOLSULFONy, and columns 6 calculated values of x y, All PHTHALEIN AND SIIJVFARNITRATE .4T 30.1 these substances are expressed in moles per liter. ki = 5.30 X lo-': kt = 5.70 X 10-8 y should correspond The calculated values of x 21. x+l/ to observed concentration of siIver bromide. ObRun Time, Crkof%i.) (mo?&l.) (inoles/l.) (moles/i.) no. hr. x 10s x 103 x 103 x 103 servations of columns 3 and 6 in Tables I and I1 1 0.11 0.000 o.oO0 0.000 01.oO0 demonstrate that this expectation is fully realized. .423 2 17.6 .436 .024 .447 In Fig. 1 are plotted the concentrations of AgBr .024 3 17.6 .428 ,423 .447 (x y) versus times at the two different tempera.038 .497 4 23.6 ,520 ,535 tures. From the figure the agreement of the two is .038 .497 5 23.6 ,545 .535 obvibus. In Fig. 2 is plotted the concentrations of 6 51.5 .764 .120 ,652 .773 a - 2, z - y, y and x on curves A, B, C and D, 7 95.9 .996 .938 .245 .693 respectively. The curve for a - x versus time rep8 148.0 1.098 .361 .697 1 .058 resents the rate a t which original dye is being con9 168.0 1.138 ,697 ,397 1 ,094 sumed by step one of the reaction. Curve B rep1U 168.0 1.162 .697 .397 1 .094 resents the rate of change of concentration of the 11 244.0 1.192 .697 .503 1 ,200 dye intermediate due to a combination of steps 1 12 263.0 1.208 .523 1 .220 ,697 and 2 of the reaction. Curve C represents the rate 13 430.0 1.280 .697 .630 1 ,327 at which final dye product is produced by step 2 of ,697 14 527.0 1.330 ,658 1 ,355 the reaction. This is likewise the rate at which 1.5 530.1 1.308 ,697 ,659 1 ,356 AgBr is produced in this'step of the reaction. The 16 530.8 1.385 :697 .659 1 .356 curve D represents the over-all rate of production of both intermediate and final dye product. This cor(6) Arnis, "Kinetics of Chemical Change in Solution," The Macmilresponds to the over-all rate of production of AgBr Ian Company, New York. N. Y..1949, pp. 14, 15.

+

+

+

191

TETRABROMOPHENOLSULFONPHTEALEIN AND SILVER NITR.4TE IN ACID

Feb., 1952

7

I

0 100 200 300 400 500 600 700 800 Time in hours. Fig. 1.-Plot of (z y) and of AgBr versus time. The curve represents the calculated values of (z y) and the circles represeont experimental data for 4gBr. I, T = 30.1*; 11, T = 25.1 ; time in hours a t 25.1 = time in hours at 30.1 "-200 hours.

+

+

from step 1. The curves have forms expected for consecutive first order reactions.' The removals of the third and fourth bromine atoms from the dye molecule mere negligibly slow compared to removal of the first and second bromine atoms. Even after 1,558 hours a t 25.1 ' only 54% of the bromine had been ,removed from the dye. Another run stayed at 30.1' for 488 hours and then remained a t 70"for 16 hours and only 54% of the bromine had been removed inthis run. From these data and when it is taken into consideration from Tables I and I1 that the specific reaction rate of step 1 is from 8 t o 10 times as great as the specific rate of step 2 depending on temperature and when it is also observed that the specific rate of step 1 for a 5" rise in temperature is roughly 1.8 times as great and that the specific rate of step 2 is increased by 1.5 times for the same temperature interval, then it seems entirely justifiable to neglect the rate of removal of the third and fourth bromine atoms over the time and temperature ranges listed in the Tables I and 11. Statistically the rates of removal of the third and fourth bromine atoms would be extremely slow, since there would be only half as many bromine atoms available for removal per dye moleculeand the chance of a third and fourth collisions of the same dye molecule with a silver ion while the molecule remains sufficiently deformed, correctly orientated, and having necessary energy for reaction is very slight. The precision of the data is observable from runs 2 and 3 in Table I and runs 2 and 3 , 4 and 5 , and 9 and 10 of Table 11. It is observable that in no case is the precision of duplicate runs less than 5% and in some cases it is less than 1%. The accuracy of the measurements compared to theory from columns 3 and 6 of Tables I and I1 is never less than 6.2% and averages 2.8%. The accuracy therefore compares favorably with the precision. These can be considered fair degrees of both precision and accuracy since in the early stages of the run the dye goes over to a colloidal state in these solutions of high electrolyte concentrations as is shown by the Tyndall cone effect. This effect is (7) Daniels, "Chemical Kinetics," Cornel1 Cniyersity Press, Ithaca, N. P.,1938, p. 29.

I

I

AI

I

f n r z

100

200 300 400 500 Time, hours. Fig. 2.-Piot of (a - z),(z - p), y and r in molesfliter versus time in hours; k, = 2.95 X 10-2, kz = 3.80 X temperature = 25.1 : curve A = a - x; curve B = z y; curve C = y; curve D = z.

-

not dependent on silver nitrate but to the electrolyte concentration as was shown by replacing the silver nitrate by an equal concentration of sodium nitrate. Furthermore the dye is somewhat adsorbed by the precipitated silver bromide and thus n small fraction of the dye is removed from the effective zone of action with silver nitrate even though the solutions are shaken regularly. The fact that the dye does become somewhat colloidal and is t o a small extent adsorbed on the silver bromide makes filtration slow and requires careful washing with both water and organic solvent. It is obvious that these effects did not greatly influence the over-all rates, but their influence on precision and accuracy were, no doubt, important. In Table I11 are listed the calculated values of the energies of activation, Arrhenius frequency factors and entropies of activation for the specific reaction rates kl and kz. The energies of activation were calculated using the equation the Arrhenius frequency factors mere calculated using equation A( T log k) log2 = aT and the entropies of activation were obtained from Eyring's equations which is (10)

In this equation K is the transmission coefficient, taken as unity, k 1 is Boltzmann gas constant, h the Planck constant, As the entropy of activation TABLE 111 ENERQIES OF ACTIVATION.ARRHENIUSFREQUENCY FACTORS AND 'ENTROPIES OF ACTIVATION FOR THE 'SILVER NITRATETETRABROMOPHENOLSULFONPHTHALEIN REACTION FOR THE TEMPERATURE RANGE 25.1 TO 30.1 ' O

Step

1

2

Temp. interval, OC.

Energy of activation A E , cal.

25.1-30.1 25.1-30.1

21,050 14,570

(8) Ref. 6, p. 151.

Arrhenius frequency log Z factor

Entropy of activation cal./degree A8,

13.90 8.26

-11.2 -36.6

192

WIMONJ. BROACH, ROBERT W.ROWDEN AND EDWARD S.AMIS

Vol. 56

and the rest of the symbols have their usual signif- since the curves are similar for silver nitrate and soicance. dium nitrate solutions although the latter salt The negative values of activation entropies ob- causes no substitution of bromine by hydroxyl. served in Table I11 may be due to an activated The fact that the curves in both salt solutions at the complex which is more rigid than the reactant dye beginning of the runs have different slopes from molecule! those of the curves at longer time intervals may be However, these low entropies of activation could due to the early conversion of the dye to a colloidal result from low Arrhenius frequency factors arising state. Some of the increase of transmission with from small transmission coefficients. In other time is no doubt due to the adsorption of the dye by words, a K value less than unity! precipitated silver bromide; however, this must be It is observed that the values of both AE and log a minor cause, since again sodium nitrate produces Z are much greater for step 1 than for step 2. no silver bromide. Hence the larger part of the inThus the Arrhenius frequency factor in step-.l as crease must be due to some change with time in the compared to step 2 more than compensates for the absorptive capacity of the dye molecules for the higher energy of activation required by step 1, per- wave lengths of light studied. This change could mitting this step to be the much faster of the two. be due to a fading reaction of the dye in strong acid Absorption data on the reaction mixture were solution of salts similar to the acid fading of tritaken at various times, using a Beckman model DU phenylmethyl dyes.’O Whatever the change in the quartz spectrophotometer. The absorption runs dye molecule, it does not affect the rate with which were set up exactly as the kinetic runs and all reac- silver ion replaces bromine in the dye molecule. tions were at the same concentrations in both type.s As to whether the reaction is heterogeneous can of measurements. A t specified times samples be investigated from the following considerations. were removed from the reaction vessel, centrifuged First suppose the reaction to take place only a t for 10 minutes, and liquid sample siphoned from the surfaces of colloidal silver bromide or assume above the solid sediment. These were measured the reaction to be catalyzed by this salt. These against a standard made up freshly each time ex- assumptions would be illogical since it would be actly as the original run was made. In Fig. 3 are impossible to see how the reaction would be initiplots at various time intervals of the wave length of ated because originally there is no silver bromide light in millimicrons versus percentage transmission present. Then too the reaction would, in the early of sample compared to the standard. stages of the run, speed up with time since the amount of silver bromide would increase. I n the case of absorption this would be true until the dye present had all been absorbed on the surface of the bromide and from then on the rate would decrease due to the depletion of the dye. spm Actually the rate was found always to depend on athe concentration of the dye remaining unreacted, .o 60 according to our theory, and the observed time I rate of production of silver bromide decreased in 40 agreement with these calculations. 9 Second assume that the reaction took place only i; 20 on the surface of colloidal dye particles. Then a molecule would have to be absorbed on the colloidal 0 dye surface only long enough for two bromine atoms to be removed, and then give place to, and not at 400 450 500 650 600 FIJ 400 450 500 550 600 any future time interfere with, the absorption of an x. unreacted dye molecule. This would seem improbFi . 3. Plot of the per cent. transmission versus the wave able. Why should an unreacted dye molecule be l e n d , T i n mp for tetrabromophenolsulfonphthalein in absorbed and a reacted one be desorbed and not inAgNO, + HNOI and .in NaN08 + “08 solutions: A, AgNO, + “0, solution of dye: 0 daye, 4 days, 9 days, terfere by being reabsorbed? How would such a 21 days; B, NaNOa + HNOS solution of dye: 0 days, 4 mechanism give a rate proportional to unreacted dye concentration. days, 9 days, 18 days. It seems more probable that the dye became colIt is observed for a given wave length the percent- loidal to a certain extent almost immediately and age transmission increases with time in both silver that these colloidal dye particles which by observanitrate-nitric acid solution of dye and the sodium tion remained in suspension were loosely connitrate-nitric acid solution of dye. This is true structed and permeable to silver ion. The extent except for the very earliest period of time for which of the colloidality was included in the rate constants the absorption wave has a little different slope than so that the calculated rate based on the concentrafor the other time intervals and actually crosses one tion of unreacted dye, including the initial concenof the shorter interval curves in both types of solu- tration of dye, agreed with the observed rate of tions. The increase of transmission with time can- production of silver bromide. The reaction is too complex and the theory and not be attributed to the replacement of bromine atoms in the dye molecules by hydroxyl radicals observation are in too close agreement for the interpretation to be merely fortuitous. (9) Glasstonc. Laidlcr and Eyring. “The Thoory of Rate Prorewcs,” McCraw--Hill Book Co., Inc., New York, N. Y.,1941, p. 297

(10) Diddle and Porter, J . Am. Chem. Soc., 37 1571 (1015).