THE REACTIOS B E T K E E N BROhIOSCCCIn’ATE IOX AND THIOSULFATE ION1 F. A. LONG
AND
A. R. OLSON
Department of Chemistry, University of California, Berkeley, California Received September 26,1956
About five years ago several articles appeared containing experimental data which the authors of the articles felt were at variance with the Bronsted theory of salt effects. Thus Conant and Peterson (2) stated that they were unable to correlate their data on some coupling reactions with the theory. La Mer and Kamner (10) found that although increasing the ionic strength had the theoretical effect on the reaction between a-bromopropionate ion and thiosulfate ion, it had the opposite effect on the corresponding reaction between P-bromopropionate ion and thiosulfate ion. To explain this anomalous salt effect, La Mer and Kamner advanced their theory of oriented collisions, according to which a thiosulfate ion repels the carboxyl group of the acid ion and thus produces a favorable orientation for replacement of the bromine by the thiosulfate ion. The amount of this favorable orientation is decreased by the presence of other ions, thus overshadowing the primary salt effect. Sturtevant (14) showed from theoretical grounds that such a theory could not account for the phenomenon. The P-bromopropionate reaction showed several other anomalies. Thus the beta-substituted compound reacted faster than the alpha-substituted compound, n-hich is contrary to a widely accepted rule of organic chemistry. La Mer (9) also reported that the heat of activation of this reaction had a temperature coefficient of over 100 calories per degree. Bedford (1) and his colleagues, studying the reaction between thiosulfate ion and bromosuccinate ion, in which the bromine is alpha to one carboxyl group and beta to the second, not only confirmed the peculiar salt effect of La Mer and Kamner, but found a temperature coefficient of the heat of activation fully three times as large as that reported by La Mer. These results were the more surprising since Olson and Long (13) have shown that reactions between bromo- and chloro-succinic acids and chloride and 1 Presented a t the Symposium on Molecular Structure, held a t Princeton University, Princeton, New Jersey, December 31,1936 to January 2,1937, under the auspices of the Division of Physical and Inorganic Chemistry of the American Chemical Society. 267
268
F. 4 . LONG AND A . R . OLSON
bromide ions exhibited heats of activation that in every case were independent of the temperature. Apparently the only essential difference between this latter work and the investigations of Bedford, was that Olson and Long kept their solutions distinctly acid in order to exclude side reactions. Quite recently Xielsen (11) studied the alkaline saponifications of various ester ions in which the charge was progressively farther removed from the reaction position. No effect due to the position of the charge was observed. Nielsen also performed a few simple experiments on the reactions of thiosulfate ion with bromosuccinate and p-bromopropionate ions. His procedure was simply to follow simultaneously both the rate of disappearance of thiosulfate and the rate of appearance of hydrogen ion for varying ratios of the reactants. He showed that the second-order constants for the bromosuccinate reaction exhibited a pronounced drift with time, as is also apparent in the original investigations of Bedford and coworkers. A few calculations given by Nielsen show that a combination of first- and second-order reactions, when treated as pure second order, could give salt effects of a sign opposite to that predicted by the Bronsted theory. He decided that an interpretation of the bromosuccinate and P-bromopropionate results necessitated an accurate restudy of these reactions, and concluded, “At present, therefore, it cannot be said that these reactions are clear-cut or a definite contradiction to the Bronsted theory.” As long ago as 1912 Holmberg (3) and Johansson (7) showed that bromosuccinate ion reacted in neutral solution to form a p-lactone,
Br H
I I -0OC-C-C-COOI I H
H
H H
I
-+ -0OC-C-C-CO
/
+ Br-
I H I
Lo-J
Further work by Johansson and Hagman (8) sho1Ted that the formation of lactones is probably a common reaction for beta-substituted acids and is often concealed only by the high reactivity of such lactones. It was to avoid this lactone formation that Olson and Long, in the work mentioned above, maintained a high hydrogen-ion concentration. It therefore suggested itself to us that the anomalous results reported for the reaction of thiosulfate ion with bromosuccinate ion might be connected with this phenomenon. This could be very simply tested by a method involving the use of optically active materials. Thus, from Z-aspartic acid and from Z-asparagine we can prepare Z-bromosuccinic acid and its Z-monoamide, HOOCCBrHCHZCONHz, respectively, by similar reactions. The properties of the
BROMOSUCCISATE 10s AKD THIOSULFA4TE IOh’
269
amide are remarkably like those of the acid. They are both levorotatory and to about the same extent, I n acid solution, they react bimolecularly with substituting agents at nearly the same rate t o give optically active products having dextrorotation. T h e n the solutions are neutralized and
1
-
1.5
-
F
v)
z 0 V +
9 I 8
:1: 2
-
I 0
200
1
400 600 TINE IN MINUTES
! I 800
FIG. 1. Showing unimolecular character of reaction. Concentrations are 0.100 molal bromosuccinate ion and 0.1554 molal thiosulfate ion. For curve 1, which was obtained by titration of thiosulfate, logarithms of unreacted bromosuccinate have been plotted against the time. Curve2gives theplot of log (ao- C Y ) / ( C Y-~ C Y - ) against the time where C Y O and CY, are the initial and final polarimetric readings and CY is the reading a t time t .
TABLE 1 Reaction of I-bromosuccinamidate and thiosulfate at 40°C. S203--= 0.200 molal; amide = 0.200 molal T I M E IN MIXUTES
DEGREES ROTATION
0 17 64 232 358 716 1312 1597 2202
-4.80 -4,58 -4.01 -2.80 -2.17 -1.10 -0.18 $0.01 +0.41
BIMOLECULAR RATE CONSTANT
0.0107 0.0101 0.0102 0,0102 0.0101 0,0109 0.0106 0.0115 0.0107 0.0104 (Av.)
* Calculated treated with thiosulfate ion, big differences appear. K i t h the acid, the resulting thiosulfate compound shows a strong levorotation. The rate of reaction, measured either polarimetrically or by titration with iodine, is unimolecular. Figure 1 gives the results of a determination of this rate for concentrations of 0.100 molal bromosuccinate ion and 0.1554 molal
270
F. A. LONG AND A. R. OLSOS
thiosulfate ion. The amide, on the other hand, when reacting with thiosulfate ion, yields a dextrorotatory compound. The reaction is bimolecular as is shown in table 1,where we give the results for a polarimetric determination of the rate of this reaction at a temperature of 40°C. I n addition, under comparable conditions, the reaction of the amide is many times slower than the reaction involving bromosuccinate ion. The important distinction between the bromosuccinic acid and its beta amide that might lead to a different mechanism is the fact that, as has previously been shown by Holmberg (4),the amide is incapable of forming a lactone. Thus the reaction of this amide with thiosulfate ion is bimolecular and leads to a product whose rotation is of the opposite sign from that of the original acid, as would be expected from the ordinary process of direct substitution accompanied by inversion. The unimolecular character of the reaction involving the bromosuccinate ion and the fact that the rotation of the final product is of the same sign as that of the original bromosuccinic acid suggests very forcibly that the reaction takes place through the production of the lactonic ion as a n intermediate. For the Z-bromosuccinamide the reaction would then be LCOOCH2CHBrCONHz
+ S2O3--
+ Br-
.+d-COOCHzCHSzOaCONHz
while for the Z-bromosuccinate ion the reaction is very probably l-COOCHzCHBrC00 *%
d-COCHzCHCOE
I
I
+ Br-
io-i Here the order of the reaction would depend upon m-hich is the ratedetermining step, but the final product would always have a levorotation. Still further evidence that the above mechanism is the correct one may be obtained by considering each of the two steps in detail. Step 1, the production of the lactone from the bromosuccinate, has been extensively investigated by Johansson (7, 8) and by Holmberg ( 5 ) . Johansson followed the rate of production of the lactone by titrating the bromide ion with silver nitrate solution. For a 0.02 molal solution at 25°C. he found a unimolecular rate constant of 0.00240, a value which is in striking agreement with the rate constant given by the results shown in figure 1, which is 0.00245. The d-P-propionolactonic acid which is formed from the Z-bromosuccinate ion has never been isolated in a pure state, but in neutral or slightly acid
27 1
BROMOSVCCISSTE 10s AND THIOSCLFATE 10s
solutions it is fairly stable. A concentrated solution of the lactomi iiiay be obtained by treating a solution of bromosuccinate ion with silver nitrate solution over a period of several hours. Precipitation with excess silver nitrate frees this solution from any malate or broinosuccinate ions. Finally, acidification, extraction with ether, and a reextraction with water gives an almost pure solution of the lactonic acid. An exact determination of the concentration of this .qomewhat unstable material is difficult, but it can lie done with fair accuracy by titrating to neutrality, adding excess hydroxide ion to effect hydroly.sis, and then back-titrating after the hy-
If a solution of the optically active lactone is neutralized and treated with thiosulfate ion a rapid reaction takes place. Knowing, however, the initial rotation of the lactone, it is possible to follov the reaction polariTABLE 2 Deterininatl'on o j rate f o r Iactonate a n d thiosulfate ions at 25°C. d-lactonate ion = 0.0822 molal; thiosulfate ion = 0.1036 molal TIME I V MINCTES
O
2 33 3 16 4 58 6 34 10 42 14 18 26 34 co
II
I
i
a --a
DEGREES ROFATIOS
1
BIMOLECULAR RATE
comr.th-r
.-
$1 32
0 0 0 0 -0 0 0 0
79 66 41 17 19 40 62 8.5 1 01 1 40
I
I
0 0 1 1 1 1 2 2 2
I
I
I
66 91 15 51 72 94 17 31 72
I
'
0 99 0 91
1 02 1 02 1 09
107 1 06 1 03 (Av )
metrically. Table 2 give.; the data for such a determination. The initial concentration of lactone was 0 0822 molal and that of the t h i o d f a t e ion 0.1036 molal. I t is obvioui from these data that the reaction is bimolecular. Another run with the ianie concentration of lactone but with a thiosulfate concentration of 0.2 inolal gave the slightly higher bimolecular constant of 1.12 as coiitrasted t o 1.03. I t iq important to notice that the optical rotation of the product is qtrongly levorotatory just as n as given by the mixture of bromowcciiiate ion and thiosulfate ion. 111 addition the rate of the bimolecular reaction between the lactone and thiosulfate ion is fast enough 10 that strp 1 in the postulated mechanism would be the ratedetermining Step, step 2 being .imply a rapid follow reaction. T h ~ qall the knoirii factq are in accord with the poqtulated mechanism. Since the qualitative inveytigations outlined above showed that the T H E JOURNAL O F PHYSICAL C H E M I S T R Y , V O L .
41,
NO.
2
27 2
F. A . LOKG AKD A. R. OLSOS
previously postulated bimolecular mechanism was wrong, we attempted to recalculate the rate constants assuming a unimolecular reaction, but published data were too fragmentary. We have, therefore, investigated the reaction between bromosuccinate ion and thiosulfate ion for various concentrations of the reactants and various ionic strengths at several temperatures. The reaction has been followed by a titration method and by a polarimetric method where feasible. PREPARATION O F MATERIALS
The Z-bromosuccinic acid was prepared from Z-aspartic acid according to the method described by Holmberg (6). The l-bromosuccin monamide was made by a similar method from l-asparagine. Since we could prepare the optically active bromosuccinic acid in a very pure state, we used this material for the titrations as well as for the polarimetric determinations. The standard solutions were made up with the usual precautions. Twice recrystallized potassium iodate was used as the primary standard for determining the concentration of the thiosulfate solutions. The usual buffer solution was 1/30 molal disodium phosphate and 1/30 molal potassium dihydrogen phosphate, which had a p H of 6.8. For a few runs, however, a mixture of these two substances giving a pH of 6.4 was used. EXPERIMENTAL METHODS
I n making a run the bromosuccinic acid was weighed into a volumetric flask, dissolved in water, and enough sodium hydroxide solution pipetted in to just neutralize it. Then the proper amount of sodium thiosulfate solution was pipetted in, buffer solution, if used, mas added, and the solution brought to the proper volume. For the work at 25°C. the time of addition of sodium hydroxide was taken as zero time. For runs a t other temperatures, however, the solutions were allowed to come to temperature before the final volume adjustment was made, and the time of the first sample was taken as zero time. Since the reaction is unimolecular, this procedure was satisfactory. For all runs the unit of time is the minute. When titrating thiosulfate ion, samples were removed from the mixture as the reaction progressed and pipetted onto crushed ice to stop the reaction. For the bromide ion, which was determined by the Volhard method, the samples were pipetted into known amounts of silver nitrate solution containing excess nitric acid. The reaction has been followed in all cases up t o a t least 65 per cent completion. Titrations were performed with calibrated volumetric burets. For all of the dilute solution titrations corrections were made for the indicator constants. The thermostat temperatures, which are accurate to =tO.Ol"C., were measured by thermometers that had been calibrated by the Bureau of Standards.
273
BROMOSUCCINATE IOT A T D THIOSULFATE ION EXPERIMCNTSL RESULTS
I n table 3 and figure 2 we have collected the data for the results at 25°C. for 0.1 molal concentrations of bromosuccinate ion. The straight lines given by the logarithmic plots show immediately that the reaction is unimolecular; this is confirmed by the small changes in the slopes shon-n for the different concentrations of thiosulfate. The rate of disappearance of thiosulfate ion is thus determined solely by the rate of production of the
RUN
1
TABLE 3 Data at 25'C. for 0.1 molal bromosuccinate ion
I
I CONCENTRATION OF THIOSULFATE
A D D E D SALT
VNIMOLECULAR RATE CONSTANT
X IO3
molal
0.1513 0.1513 0.1513 0.1554 0.1554 0.400 0.1513
Buffer (Polarimetric) (Polarimetric) 0 9 .tf KNO,
TIME
0.754 0.754 0.754 0,808 0.766 1.500 1.684
2.471 2.474 2.481 2.454 2.425 2.58 3.389
IN MINUTES
FIG.2. Logarithmic plots for 0.1 molal bromosuccinate ion and varying concentrations of thiosulfate ion a t 25OC. The nunihers of the curves correspond t o the n u m bers of the runs as given in table 3.
lactonic ion. Fortunately the rate of reaction of the lactone with thiosulfate ion is so fast that in these concentrated solutions we can neglect the amount of lactone which disappears, either by hydrolysis of the lactone or by the re-formation of bromosuccinat,e ion through reaction with bromide ion. This has been corifirnied by a quantitative determination of the rates of t h e three reactions which are competing for the lactone. The specific rate constant for the bimolecular reaction between thiosulfate and lactone
274
R. OLSOX
F. A . L O S G A S D A .
hag previou-ly been shon n to lie 1.03; for the himolecular reaction hetv-cm lactone and bromide ion the bpecific rate ic approxiniately 0,001. At a pH of 6.4 the uiiiniolecular rate of reaction of t h c lactone with water as determined polarimetrically in a buffered solution iy 0.000413. The latter two rates are from unpuhli~hcdn-orli in tlii3 laboratory. TABLE -4 Polarimrtric c.icfei.viinaiion of rate Bromosuccinate ioii = 0.1 molal; thiosulfate ioii
=
I
TIME IN DIISVTTES
0 11 37 212 287 390 61 1 655 m
-3 2 2 2 2 2 2 2 1
k =
S E T READISG
DEGREES ROT.\TIOS
00 97 91 33 44 32 143 11 89*
-1 11 1 08 1 02 0 66 0 55 0 43 0.2t55 0 22
I I
, I
0.1554 iiiolal 2.303
a >-
1
CYo-a
CY
- log __
0 0 0 0 0 0
00238 00215 00243 00243 00241 002425
(AV,)
* Calculated.
I
0
200
400
600
800
TIM1 IN PIhUTES
FIG.3. Logarithmic plots for 0.01 molal hroiiiosuccinate ion a t 25°C. The numbers refer to the runs given in table 5 , Curves 1, 4,and 6 shoJv a slight induction effect which curve 7, obtained h y titrating the liromide ion, does not show.
I n table 4 we give detailed data for the polarimetric determination of the rate of this reaction, which is listed a. run 5 in table 3. In thiy rim lye measure not only the appearance of the 1-thio-uccinatc ion but alqo the disappearance of the 1-bromowccinate ion. Severthele+ the cpecific rate constant which we obtain from thew data agrees within the limits of experimental error with the results obtained from the titration method, which measures the disappearance of thio-ulfatc ion only. T h i ~agrecment is
2i5
B R O M O S T C C I S I T E 10s A S D THIOSL7LFhTE IO?;
,&oi\-n in figure 1, which gave tlit' logaritlimic plot; for tn-o runs having the sanie concmtration. of 1,roiuosiicciiintc ion and t1iio;ulfate ion but, follon-ed polarinietrically in 0110 cn.c and by titration in the other. The rate constant from the polarinirtric niet!iotl is less accurate than that for the titration method, hecause of ,;onie uncertainty in t h e value of thc molal rotation of th(>l-thiosuccinate ion. Runs in n-liich the hroniomccinate ion ih 0.04 niolal are listed in table 5 and plotted in figure 3. A discussion of tlie changes exhibited hy the rate eonstant; a,I. H., .~SD COWORKERS: J. Am. Chem. SOC.66, 280 (1934); 67, 1408 (1935). (2) COXANT,J. B., A N D PETERSON, 15'. A . : J. Am. Chem. SOC.62, 1220 (1930). (3) HOLMBERG, BROR:Ber. 46, 1713 (1912). (4) HOLMBERG, BROR:J. prakt. Chem. 87,466 (1913). J. prakt. Chem. 87,456 (1913); 88,553 (1913). (6) H O L M B E RBROR: G, (6) HOLMBERG, B R O R :Ber. 60, 2198 (1927). (7) J o ~ a ~ s s oHJ.: s , 2 . physik. Chem. 81, 573 (1912).
BROMOSUCCISATE IOK A S D THIOSULFATE IOK
28 1
(8) JOHANSSON, HJ., A N D HAGYAN, S.11.:Ber. 66, 647 (1922). (9) LA MER,5’. K.: J. Am. Chem. SOC.66, 1739 (1933). (10) LAMER,V. K., A N D KAMNER, MILDRED E.: J. Am. Chem. SOC.63,2832 (1931). (11) NIELSEN,RALPH F.: J . Am. Chem. SOC.68, 206 (1936). (12) OLASDER,ARNE: Z. physik. Chem. 144, 49 (1929). (13) OLSOS, A. R., AXD LOSG,F. A.: J. Am. Chem. SOC.68,393 (1936). (14) STURTEVANT, J. LI.: J. Chem. Physics 3, 295 (1935).
