iringence form a more satisfactory basis for quantitative study. The wave length of isatropy (at which e = w ) is an important fixed point on the birefringence-wave length curve. Crossed axial plane dispersion in the near ultraviolet can be detected and memared photographically provided the wave length is longer than that of the ultraviolet cut-off of the microscope system used. Thiourea becomes uniaxial a t 3780 8.and hence is a good experimental subject. Most organic substances have absorption bands in the ultraviolet and all types of dispersion tend to be intensified as these bands are approached from the high wave length side. For this reason many substances may have strong measurable dispersion in the visible violet or ultraviolet but very little in the balance of the visible spectrum. In such cases the dispersion may acquire sudden prominence over a relatively short wave length interval as illustrated by the sudden appearance of crossed dispersion in stilbene.
[COMMUNICATION NO.878 FROM THE
Summary 1. The dispersion of the optic axid angle in trckns-stilbene has been measured. This substance is optically positive for wave lengths above 4070 k. and optically negative for shorter wave s. The rare type of dispersion giving rise to this effect has tentatively been called a x b l dispwswn with change of sign. Strong crossed dispersion is also present. 2 . Dispersion of the refractive indices a, fl, and y in stilbene has been determined by Merwin’s method. 3 . The refractive indices e and w of benzil were measured over a wide range of wave lengths. As shown by Jelley, this substance is optically isotropic a t 4205 8., optically positive above, and negative below this wave length. 1. The optical properties of thiourea have been measured. This compound exhibits crossed axial plane dispersion near 3780 k. in the ultraviolet . WILMINGTON, DELAWARE
RECEIVED JULY 6, 194
KODAK RESEARCH LABORATORIES]
The Inversion of Menthone with Trichloroacetic Acid in Aprotic Solvents BY A. WEISSBERGER
The reaction of diazoacetic ester with trichloroacetic acid in hexane yielding the trichloroaceto-glycolic ester, CC18C02CH2C02C2Hs, proceeds through a reactive complex composed of one molecule of the ester and two molecules of the acid.lI2 The reaction is slowed down considerably by the addition of certain substances like ethers, ketones and alcohols.2 This effect was attributed to the formation of compounds of the addenda with the acid, and a spectroscopic study of the salt formation of dimethylaminoazobenzene arid trichloroacetic acid in benzene3 corroborated this interpretation. Parallel with these investigations was studied the inversion of nienthone.4 This reaction also can be measured in aprotic solvents, but it differs from the decomposition of diazoacetic ester in some respects. In the latter reaction, carried out in aprotic solvents, the acid becomes part of the reaction product, .i. e., it (1) Bransted and Bell, TRISJOURNAL, 68, 2475 (1931). (2) Weissberger and H6gen, Z . $hysik. Chern., 11166, 321 (1931). (:1) Weissberger and Fasold, i b i d . , 11117, 65 (1931). {A) Wonneberger. Dissertation, Leipzig, 1930; Darken, Dissertat i o n , Lripzig, 1034
“catalyzes its own reaction,”6 while in the menthone inversion the reaction product is a stereoisomer of the starting material. It was, therefore, of interest to see whether the reactive complex in the menthone inversion is simpler than in the diazoester decomposition, and to investigate whether addenda have the same or a different effect in both processes. The investigation of the menthone inversion itself is reported in the following. Menthone, 1-methyl-4-isopropyl-cyclohexane-3one, has two asymmetric carbon atoms and exists in four optically active isomers. The two pairs of antipodes differ from one another by geometrical isomerism ^I
k!
Menthone
H
H
Isomenthone
The trans formulas are attributed to menthone, ( 5 ) Bell, “Acid-Base Catalysis,” Oxford University Press, N e a York, N.Y.,1941.
Jan., 1943
INVERSION OF
MENTHONE WITH TRICHLOROACETIC ACID I N
k R O T I C SOLVENTS
103
the cis formulas to isomenthone.6 The isomeri- mization through the formation of diastereoisozation of 1- and d-menthone under the influence mers. Identity of the final equilibria under varyof acids or bases was discovered by Beckmann,’ ing conditions is therefore not to be expected. The inversion proceeds as a first-order reactiong and the first rate determinations were made by Vorlander.8 Tubandtg showed that the same dx/dt = kmx - k’m(1 - x ) (1) equilibrium is reached from a menthone of ( a ) ” ~ where m is the total concentration of menthone; -27.75’ and from a “Rechtsmenthon” of (cx)~ODx and 1 - x are the fractions of the isomers, and +27.79’, and that the reaction velocity is iden- k and k’, the velocity coefficients of the opposite tical in both cases. Tubandt assumed that isomerizations. The sum k k‘ is determined “Rechtsmenthon” is a mixture of I-menthone and from the change in rotation of the reacting soluof d-isomenthone, while Grossman and Brauer‘O tion took a different view because they observed that k + k’ = f log,, ’Uo (2) P m - P1 LLRechtsrnenthon)) was not inverted in formic acid. We found that under identical conditions pa, p , , and p t are, respectively, the rotation angles the reaction velocity and the final rotatory power a t the time of the first reading, at equilibrium and of the equilibrium mixture are identical not only a t a time t after the first reading. in the inversion of 2-menthone, and of “RechtsThe rate of the alkali-catalyzed menthone inrnenthon” but also of a d-isomenthone of (CY)~OD version is independent of the menthone concen85.6’. In formic acid, this d-isomenthone was tration and proportional to the alkali concentrainverted a t the same rate as I-menthone, and tion, the final equilibrium being identical for all reached the same equilibrium angle as the latter solutions.9 The inversion with acids is more comcompound. These observations confirm Tu- plicated. bandt’s findings. The stability of “RechtsmenTubandtg found, with hydrochloric and sulfothon” in Grossman and Brauer’s experiment was salicylic acid in ethanol, that the inversion rate caused by the accident that this menthone was a falls when the menthone concentration is inmixture of d-isomenthone and I-menthone of, or creased. He suggested that menthone forms nearly of, the composition of the equilibrium in compounds with the acids and showed that esteriformic acid as a solvent. fications with acids as catalysts arelslowed down The rearrangement takes place a t the asym- by the addition of menthone. At constant menmetric carbon atom in the a-position to the car- thone concentration, the inversion velocity in most bonyl group (b). It is accompanied by a proto- solvents increases more rapidly than the concentropic change a t the other carbon atom adjacent tration of the acid. The composition of the final to this group1’while the arrangement on carbon equilibrium between I-menthone and d-isomenatom (a) is untouched; the oxidation of both part- thone is, according to Tubandt, independent of ners of an equilibrium leads to the same ketocar- the nature and concentration of the inverting bonic acid.12 The two diastereoisomers with the agent and of the menthone concentration, dehigher rotatory power are partners in one equilib- pendent, however, on solvent and temperature. rium, those with the small values partners in the Grossman and Brauer inverted I-menthone by other. This agrees with the theory13 that the addition t o various acids without a solvent and carbon atom adjacent to the carbonyl group makes found that, after evaporation of the acid, mixthe larger contribution to the activity of the tures of different activities remained. After the completion of the present experimolecule. The activity of the equilibrium mixture is caused by the asymmetry of carbon atom ments, Bell and CaldinI4 investigated the men(a), which, in turn, keeps (b) from complete race- thone inversion in chlorobenzene a t 100’ and used it for the comparison of different acids. The reac(6) Sugden and Whittaker, J . Chem. SOL,117, 1868 (1925); Zeitschel and Schmidt, Be?., 59, 2301 (1926); Carter, J . Chem. Soc., tion rate was found to vary linearly with the 1278 (1927);Read, Chcnz. Rev., VII, 1 (1930). menthone concentration. At a constant concen(7) Beckrnann, Ann., 150, 322 (1889). tration of the menthone, i t increased faster or (8) VorlHnder, Bn.,86,273 (1903). (9) Tubandt, Ann., 889, 41 (1905); 854, 259 (1907); 877, 284 slower than the acid concentration, depending (1910). on the conditions. Cryoscopic and polarimetric (10) Grossmanand Brauer, J . praRt. Chem.. 98,37,38,62(1918). (11) Bartlett and Vincent, THISJOURNAL, 56,4992 (1933). measurements showed that trichloroacetic acid (12) Beckmann and Mehrliinder, Ann., 969,367 (1896).
+
+
(13) Freudenberg and Ruhn, Bn.,64,703 (1981).
(14) Bell and Caldin, J . Chcm. S o c . , 382 (1938).
A. WEISSBERGER
104
and menthone form a stable complex, and that the complexes formed with other acids are the less stable, the lower the ionization constant of the acid.
Experimental l-Menthone was prepared from l-menthol and distilled in vacuo. Four different samples were made. Two of them underwent a further rectification over boric acid.” Kinetically, these samples did not behave differently from the others. The rotatory power of the equilibrium mixtures, however, differed owing to the presence of some menthol in the unrectified samples. d-Isomenthone was prepared from d-isomenthol, for which we thank Messrs. Schimmel and Co.. Miltitz. I t was rectified over boric acid. Trichloroacetic acid, solvents and other materials were purified as described by Weissberger and Fasold.3 The observations were carried out with a Lippich halfshade polarimeter, allowixig readings of 0.01 ’, in jacketed 200-mm. polarimeter tubes. The temperature was held at 20.0 =t0.1O by means of a thermostat, unless stated otherwise. The calculations were made according to ( 2 ) . Some typical experiments are given in detail to show the reproducibility of the rate coeficients and their constancy ill a given ruii TABLEI Menthone, 0.5 mole/liter; trichloroacetic acid, 0.5 mole, liter: benzene D x - p’i = 5.81 pr‘ - po = 5.77 l.3
rniri 0 60 120
180 210 240 1440 m
9,
- 1 31 -0.83 - .35 .06
+ +
,25 .45 +3.81 +4.50
+
!k
+ k’)lor 6.2 6.5 li.5 6.5 ri.5
6.4
..
;AI
pL
0 60 120 180 240 360 1440
-1.28 -0.78 -- . 3 3 * .09 -+ . 4 6 + I . 12 f3.81
(k 4 % ’ ) l o ‘
0 -10 60 90 120 I80
.-0 . 7.3 +0.15 +0.92 +1.58 4-2.42
6 . .i (1 . 5 ti.5 f j. 5 6.4
t3.09 +4.W +5 83
(1
”U
31.3 31.4 31.6 31 ti :(I 7
:3fl 60
120 180 30(1
-0.64 10.25 T0.64 +1.66 +3.16 t4.11 i-5.15 75.87
31.9 31.7 31.6 31.7
31.6 31.9
With constant concentration of menthone and varying concentrations of trichloroacetic acid in benzene as a solvent, the average values of Table I1 and Fig. 2 are obtained. Starting with low concentrations of the acid, the reaction velocity increases faster than proportional to the acid concentration. The increase, however, reaches a maximum and falls off with high concentrations of the acid. The maximum lies a t a concentration of about two moles of acid per mole of menthone. Table I11 and Fig. 1 1.5, Zeitschel and Schmidt. Ner., 69, 2298 (1986)
Trichloroacetic acid, mole/litcr
0 . 10
.25 .30 .40 50 .75 .93
.oo
1 1.04
1 .10
1.23 1.50 2.00 3.00
!k
(k f k’)lO~/cooc. acid
+ k’)IO‘ 0.31 1.5 2.1 3.9 6.5 13.3 17.0 18.5 19.3 20.4 22.5 26.4 31.6
3.1 6.0 7.0 9.8 13.0 17.8 18.3 18.5 18.6 18.6 18.3 17.6 15.8 13.3
39.9
show that this relation is maintained a t other menthone concentrations and in other solvents. TABLEI11 Menthone, 0.25 mole/liter : benzene Trichloroacetic acid, moleiliter
U.25 .40
50 00 .70 .80 L .(IO
rk
+ k’I101
(k
+ k’)!04/conc. acid
4.4 9.75 12.7 15.5 17.5 19.5 23.7
17.6 21.4 25.5 25.8 25.0 24.1 23.7
hlenthone, 0.1 mole/liter; benzene 20 ,25
II,
.30 40
.50
Ti.49
31 3
TABLE I1 Menthone, 0.5 mole/liter; benzene
t j . ti
Menthone, 0.5 mole/lit.er; trichloroacetic acid, 8.0 mole, liter; benzene p a - p” = (i..i(i o.< - po = ti.5i 30
Vol. 65
L . oo
34.0 35.2 38.3 36.5 33.8 26.2
11.8
8.8 11.5 14.6: 16.9 26.2
Menthone, 0.5 mole/liter; hexane 2.8 11.2 .50 10.2 20.4 I . 00 24.2 24.2
11.28
Meut hone, 0.25 mole/liter ; hexane 18.7 37.3 823 36.8 32.6
11.5
The rate of inversion varies with the solvent. Under identical conditions-menthone 0.5 mole/ liter, acid 0.5 mole/liter-(k k’)104 is in hexane 10.2, in benzene, 6.5, and in chloroform, 4.6. The reaction rate falls with the dielectric constant of the solvents. This may be caused by the dependence of the formation of the reactive complex (see below) on the dielectric constant of the surrounding medium. However, other factors may also contribute toward this solvent effect.I6
+
(16) Dimroth, Ann., 371, 127 (1910); H. v. Halban, “Habilitationsschrift,” Wiirzborg, 1909: Moelwyn-Hughes. “The Kinetics of Reactions ia Solution,” Oxford. 1933.
Jan., 1943
INVERSION OF MENTHONE WITH TRICHLOROACETIC ACIDIN
105
h R O T I C SOLVENTS
TABLE IV Benzene Acid 0.25 mole/liter Menthone, mole/liter
0.1 .25 .5
(k
+ k')104
Benzene Acid 0.5 mole/liter Menthone mole/Iite;
0.1
8.8 4.4 1.5
.25 .5 1.0
(k 4- k')lO4
16.9 12.7 6.5 2.3
Benzene Acid 1 mole/liter Menthone, mole/liter
(k
0.1 .25
.6 1.0
Hexane Acid 0.5 mole/liter Menthone, mole/liter
+ k')10'
(k
0.1 .25 .5
26.2 23.7 18.5 9.2
+ k')lO4 23.8 18.7 10.2
final rotation of a 0.5 molar menthone solution with increasing concentration of the acid is rather steep up to an acid concentration of 1 mole/liter, i. e., an acid menthone ratio of 2 . In this region, the curve becomes more flat. The reason for the dependence might be (1) an effect of the acid on the rotation of the components of the equilibrium, without a shift in the concentrations of the latter, and (2) a change in the position of the equilibrium itself. The following experiments give further information. A 0.5 molar solution of a mixture of Z-menthone and d-isomenthone in benzene had a rotation of f0.37" in a 100-mm. tube. This rotation increased immediately to +1.4" when the solution was made to contain 2 mole/liter trichloroacetic acid. The value of +1.4' was obtained by extrapolation to the time of mixing because the angle shifted owing to the progressing isomerization. A completely inverted solution of 0.5 mole/ liter of menthone and 0.25 mole/liter of trichloroacetic acid in benzene in a 100-mm. tube rotates +3.02'. When the same volume of 2 molar trichloroacetic acid in benzene is added, the rotation jumps to f4.58' in a 200-nim. tube. From there it shifts a t a rate characteristic of the new conditions to the value appropriate for them. These TABLE V experiments show that both factors mentioned Temp., OC. Pm (k + k')10' above are responsible for the dependence shown 19.95 2.56 13.5 29.90 2.44 30.6 in Table VI. 40.01 2.34 65.8 An idea of the shift in the concentrations of the 49.32 2.27 126.8 isomers in the equilibrium mixtures with the acid The rotation of the equilibrium mixture, pm, concentration is given by Table VII. In these rises with the concentration of the trichloroacetic experiments the acid was extracted with aqueous acid, as shown by Table VI. The increase of the sodium bicarbonate solution after equilibrium was reached. It should be noted that the reTABLEVI sults are not very accurate; small amounts of Menthone, 0.5 mole/liter in benzene menthone are taken up by the sodium bicarbonAcid, mole/liter
According to Tubandt, the rate of inversion of menthone in 0.01 N alcoholic hydrochloric acid falls by 6% if the menthone concentration is raised from 0.26 to 1.04 moles per liter. Table IV shows that the influence of the menthone concentration on the rate of the reaction with trichloroacetic acid as a catalyst in benzene and in hexane is much more pronounced. When the menthone concentration rises from 0.1 to 0.5 M , the reaction rate in benzene falls to 70.5% if the trichloroacetic acid is 1 N, to 38,4y0 if the trichloroacetic acid is 0.5 N , and to 17.0% if the trichloroacetic acid is 0.25 N . The (k k') values plotted against the menthone concentration lie on a straight line as long as the menthone concentration is lower than the concentration of the acid. If, however, the menthone concentration becomes higher than that of the acid, the retarding effect of the menthone is diminished. The temperature coefficient of the reaction was determined between 20 and 50" in solutions containing 0.25 mole/liter of menthone and 0.5 mole/liter of trichloroacetic acid as shown in Table V. The activation energy calculated in the usual way is 6,200 cal.
+
I
PCO
0.1 .25 .5 .75 1.0 2.0 3.0
$0.95' +1.51 $2.22 +2.51 +2.68 +2.93 $3.09
TABLEVI1 In 100-mm. tube; menthone, 0.5 mole/liter in benzene Trichloroacetic acid, mole/liter
Pa,
tion of the acid
0 . 1 normal .25 normal 1.0 normal
$0.95 +1.50 +2.64
+0.75 .93 +1.12
p m after extrac-
+
ate solution, and the extraction of the different amounts of trichloroacetic acid changes the concentrations of the remaining solutions.
The total concentrations of menthone (m) and of acid (a) are m = [MA] a = [MA]
+ [MI + 2/14?]+ [+4]
(6) (7)
Discussion of Results From ( 5 ) and (6) follows It is obvious that the reaction rate is not simply 1 [mi]= I)Z K S [A (8) proportional to the concentration of hydrogen i[-41 ions or of the undissociated acid either in its Elimination of [A,] through (4) and elimination dimeric or in its monomeric state. The situation of [MA] through (8) in (7) gives of singularities in regions where the ratio of trichloroacetic acid to xnenthone is 2 :1 suggested that the reactive complex might have this composition. The result that the rate of inversion Equations (1) and (3’1 can he written as ax k with hydrogen chloride in benzene is proportional = m(k + k’) FFk7 x - k---__ +k ’ k’ (1 to the square of the acid concentration* emphasized this suggestion. A number of assumptions were and tried for the formation of the ternary complex, for instance, that it is formed from nienthone and the (3’) dimeric acid. These deductions were not in acand, since the reaction velocity in the equilibrium cordance with the experiments. The evaluation, li is zero, i. e. however, of assumptions which resemble those I { =k ’- = X = k ’ (10) made about the niechanisni of the diazoacetic k I-X iester decomposition2 gave results in good agree- as ment with the experimental values. dx 1 _. Stobbe and Haertell* and Kendall and Gib- at hons19 showed that ketones and acids form coni- and 1 1 plexes. Tubandt’s evidence for acid complexes dn - = [MA] [ A ] ( k+ W‘) with menthone was quoted above. The existence d 1 (3”) of a complex between menthone and trichloroDivision of (1”)and ( 3 ” )gives acetic acid in hydrocarbon solution follows from cryoscopic measurements,14 from the instantaneous effect of the acid on the rotation of nienthone, To find out whether (k k ’ ) / a has a maximum and from the action of menthone in the diazoin agreement with the experiment, it is differenacetic ester decomposition.2 We assume that the tiated with respect to a inversion takes place by interaction of a binary acid-menthone complex with a further molecule of the monomeric acid. Equivalent to this assumption would be the spontaneous rearrange- The second factor on the right-hand side of (12) ment of an intermediary complex of one molecule is always positive, which can be shown by the of menthone with two molecules ol monomeric formation of the reciprocal value da/d[A] from acid. If A is the monomeric acid, Az, the dimeric iU), and it car1 therefore be neglected in the disdcid, M, the nienthone, x, its unreacted fraction, cussion of the maximum. Using the values for k’ from (11) and for a from (9), we differenMA, the menthone-acid complex, and [] indicates k tiate concentrations, the reaction velocity is
(
(
+
+
d x j d t = k[MAl[Ajx
- k‘[MA][AI(l - x!
(3)
We write 1Al’/[A?l = R I [A][M]/[MA] = Iiz
( ;j
I
(17) I n these evaluations, I enjoyed the most valuable assistance of Dr. C. Wagner, to w h o m I express sincere thanks. I should also like to thank Mr. D. S . Thomas of these Laboratories for his help.-A. W. (IS) Stobbe and Haertel, A i m . , 370, 106 (1909). ’ 10’1 Rendall aud G i l ~ L w r i Trrrs ~, ~ O I ; K S A L , 37. 14Sl (19131
Jan., 1943
INVERSION OF
MENTHONE WITH TRICHLOROACETIC ACIDZN APROTICSOLVENTS
The term in { ) must equal zero a t the maximum, and, solving for [A],,,, we obtain
107
Considering the fact that in aqueous solution the ionization constant of benzoic acid lies between that of acetic and monochloroacetic acid, one might surmise that trichloracetic acid is associated [AI- = (Kz m ) (141 to an appreciably lesser degree than benzoic acid. Since (14) has one real positive root, a maximum However, in view of these uncertainties, the calexists and amaxis given by (9). culations were carried through with several values K1, the association constant of trichloroacetic of K1,viz., 0.06, 0.04, 0.01, and even 0.001. acid was calculated from freezing-point deterThe equilibrium constant Kt of the menthone minations of G. v. Frank,*O and E. Strasser.21 acid complex-(5)-for benzene solutions a t the These authors give the association factor f, i. e., freezing point, can be evaluated as follows from molecules calcd./molecules exp. for a concen- freezing-point determinations of benzene with tration range of 0.023 mole/liter to 0.241 mole/ trichloroacetic acid and menthone and with menliter. From their values K 1was calculated accord- thone alone, and the association constant of the ing to acid: [M][A] and [MA] in (5) are calculated with m, KI and A T. u is the concentration of -1 y that part of the acid which is not bound to menK1 = a (15) thone, 2, the sum over all concentrations on which (l depends the freezing point, and C, the molecular with an average of 0.06. For a 0.25 molar solu- freezing-point depression of benzene 5.13. Action, we found f = 1.58 and K 1 = 0.048. A higher cording to (7) and (4) degree of association was found by Bell and [MA] = a - ( ~ [ A zf ] [A]) = a - u (16) Arnold,22who state that dry trichloroacetic acid and in benzene consists essentially of double moleZ = [AI [Ail [MA1 [MI = A T / C (17) cules over the whole concentration range investigated, 0.01-1.5 mole/liter. From measurements We call of LeF&vre and Vine,23 on the other hand, a [AI [Ail = [AI ([A12/Ki) = (P (18) value K1 = 0.04 is obtained. According to these and consider p and u functions of A-according to authors, trichloroacetic acid in benzene is less (18)-with the result associated than the dichloroacetic acid ; this, u = 1/2(K1 4~ - ~ K I ( K I 4 ~ ) ) (19) in turn, less than the monochloro compound, and only acetic acid itself is present almost com- Since, however, pis known according to (18), (17), pletely in the dimeric form. It appears that the and (6), and because association of the acid in benzene is the higher, the = (AT/C) - m (20) smaller its ionization in aqueous solution. This [MA] can be calculated according to (18). The suggestion was made by A. HantzschZ4and is very values for [MI and [A] now follow from (6) and plausible indeed. The ionization constant of a (18). Table VI11 shows the results of the freezcarboxylic acid is the higher the less tightly the ing-point determinations in benzene and the valproton is bound in the carboxylic group. The ues for Kz obtained with several values for K 1 . association of the acid, on the other hand, is caused According to (14) and (9) and with the values by the sharing of two protons between two acid for K I and Kz, data for amaxare obtained which molecules. If, in a first approximation, we neglect are compared with the experimental results in other factors, we may expect that an acid mole- Table IX. cule with a low tendency to keep its proton will The agreement is best with K 1 = 0.06 and 0.04, also have a relatively low tendency to share the but fair even with much lower values. While proton of another acid molecule. A high, but the position of the maximum depends strongly on limited, association of benzoic acid in benzene, the menthone concentration, it varies relatively K = 0.0015, was found by Wahl and Rouse.25 little with K 1 and K Zas long as these values are (20) G . v. Frank, Dissertation, Leipzig, 1922, p. 79. within the results of the experimental evidences (21) E. Strasser, Dissertation, Leipzig, 1929, p. 18. mentioned above. Hence, the theoretical treat(22) Bell and Arnold, J . Chem. Soc., 1432 (1935). (23) LeFSvre and Vine, ibid., 1795 (1938). ment of the assumption that the'menthone in(24) Private communication version takes place in a ternary reactive complex JOURNAL, 63,3002 (1941) (25) Wahl and Rouse, THIS
d$ +
($
9
+
+
+
+
+
+
+
os
A. WEISSBERGER
1
ASSOCIATIONCONSTANTSOF 0.100 .l5 2 .i . I 11 25; 15 .25 273
0 752 1 183 1 641 I 176 1 731 I 496 I 849 2 403
I
1)
7
u
...> 2,; 2i i 3 >
TABLE VI11 MENTHONE-ACID COMPLEXIN BENZENE
THE
K:
AT, C
17
I
I
0 147 28 7
0 038
320 229 338 292 :? 50 (7 1
070 118 081 142 110
.50
K2
( i r n n ~C X ~ '
amahcair
0.04 .04 .04 .04 .04 .04
0.32 .32 .32 55 . 65 .55 " 55 1.05 1.05 1 05
0.2: .26 .21 36 32 .45
.OI4 .01
.0-1
)-I 01
(Kt = 001)
= 004)
I' K I
0 027 034 045 0-16 027 070 038 036
040 OfjO
046 097
K?
= 0.001!
O 015 020 02 1 ,036
,0143 052 ,013 0181
For the calculation, values are given to [A]
R1
0-+
K:
Kz
0.032 037 057 037
035 060 030 0-11 053 048 03#5
195
0.0ti .04 ,01 .06 .01 ,001 .OR
(KI
n 032
081
TABLEIX
. 25
IRI= OOG)
cp
POSITION OF MAXIMAL ACID REACTIVITY
0.1
VOl. 65
IO4
.3Y 1.05 1.00
0.87
"
Interpolated from Figs. 1 and 2. experimental curves.
composed of one molecule of the menthone and two molecules of the monomeric acid yields maxima for the rate-acid concentration curves for the various concentrations of menthone which agree well with the experimental results.
a s parameter. According to (11) kL+L U
=
(k
+ k')
~
[MAIIAI ma
(21)
u and [MA] are determined according to 4, 7 and 8. In this comparison it has to be kept in mind that the factor (k k') in (21) displaces each calculated curve by a value log (k k') which is constant for each curve. Only the shape of the curve-and the position of the maximum in respect to the abscissa-can, therefore, be considered. The calculated and the experimental values to be compared are given in Figs. 1 and 2.
+
+
T -1 1.6 I+
1.9
X
&,
-+I
1.4
1.2 'j
1.0
SI
, e
I & 1.4
EXPERIMENTAL
&'+
W
g
0.8
c
f
0.6
$1
0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 1 log a. Fig. 2.-Menthone 0.1 mole/liter : xxx, experimental values: --- rnenthone 0.5 mole/liter: 000, experimental values. 0 0.2
w 0.8
bc.
+
06
51
0 4 11 8 0.0
-
/-\
8,-00 I .6, 0 07.
K,- 0 01 , K f 0 0 4
I
0.4
02
:
T
OR
08
1.0
1.2
log n
Fig. l.-&ienthone 0.25 niole/liter . xxx, experimental log [(k k ' ) / a ( I / @ k'))]. values; *, ordinate = 3
+
+
+
In the following, the shape of the dependence of the reaction rate acid concentration is studied over the whole range of the latter. The course of (k k ' ) / a is investigated as a function of a, using values of Kz of the order of magnitude just determined, and the curves are compared with the experimental values.
+
+
For rn = 0.25, curves were calculated with each of the values mentioned above and the corresponding values for Kz. At higher concentrations of the acid, these curves show good agreement with the observations. The calculated curves are too flat, however, in the region of lower acid concentrations. With Kz values which are 50% smaller, agreement of the calculated curves with the experimental is obtained over the whole range. If one considers the fact that in the calculation of KP the change of the association of the acid by the addition of menthone is neglected, and that small alterations in AT have a big effect on
Jan., 1943
I N V E R S I O N OF
MENTHONE WITH TRICHLOROACETIC ACIDIN AF’ROTIC S O L V E N T S
Kz-a
1% alteration of AT changes K2 by almost 15Y0-the agreement between theory and experiment must be considered as good. Since the curves for m = 0.25 show in principle the same behavior with the three values of 0.04, 0.01 and 0.001 for K1, the curves for m = 0.1 and m = 0.5 are given for K1 = 0.04 only. The situation with the menthone concentration of m = 0.5, a t which most of the experiments were made, is similar to that with m = 0.25. The calculated value for KZ shows good agreement with the experimental curve a t higher concentrations of the acid. With smaller values for Kz, satisfactory agreement over the whole range is obtained. For a menthone concentration of 0.1 rnole/liter, a satisfactory agreement, even for the higher acid concentrations, is obtained with smaller values for K Zonly. The agreement shown in Figs. 1 and 2 confirms that the racemization of menthone by trichloroacetic acid in hydrocarbons or halogenated hydrocarbons as solvents proceeds in a complex formed by one molecule of the menthone and two molecules of the monomeric acid. This mechanism explains not only the dependence on the acid concentration, but also the observations on the falling of the reaction rate with rising concentration of the menthone (Table 111). From Equation (11) it is evident that an increase in m will, by formation of [MA], cause a decrease in [A] and k k’, and furthermore that in the region of our experiments the effect on [A] will be greater for m a. When Weissberger and Hogen2 analyzed the reaction rate of diazoacetic ester with trichloroacetic acid, it was found that the assumption of a reactive complex composed of one mole of diazo ester and two moles of acid gave the best agreement with the experiment. The agreement was very good in the region of excess diazo ester but not good in the region of excess acid. This behavior is now understood because the authors neglected the association of the trichloroacetic acid. The effect of this omission is negligible as long as the diazo ester is in excess, but it becomes large when the reverse is the case. In the absence of a more active proton acceptor, the acid molecules share their protons with each other and associate to dimers. In the presence of another proton acceptor, the latter accepts the protons of the monomeric acid molecules, and the equilibrium between dimeric and monomeric acid is shifted. If the new acceptor is more avid for protons or present in
+
109
sufficient excess, the dimeric acid practically disappears; e. g., on addition of one mole of water to a benzene solution of trichloroacetic acid no dimeric acid could be found cryoscopically. l 4 Hence, as long as the diazo ester, a good proton acceptor, is present in a t least equimolar concentration to the acid, the latter can be considered as unassociated without detriment. If, however, the acid is in excess over the ester, this simplification is no longer permissible. The observation that the reactive complex in the menthone inversion is similar to that of the reactive complex in the diazoacetic ester decomposition is particularly interesting in view of- the fact that in the latter reaction the acid enters the final product. This circumstance obviously does not affect the kinetics of the process as compared with the menthone inversion. The information about the reactive complex in the menthone inversion given above agrees with and supplements the accepted views on acidcatalyzed enolizations.26 It is assumed that the relevant part of the binary complex has the constitution C
HOiCCCla
I ,“t C=O
dy ‘ H
In this complex the protons on the carbon atoms in the a position to the carbonyl group have become more mobile, the acid proton causing a shift of electrons through electrostatic attraction and through the resonance effect indicated by the arrows. Th&e protons are therefore capable of an exchange reaction with the proton of a second acid molecule, and, if this exchange involves a replacement of the proton on the reverse side of the asymmetric carbon atom, the I-menthone goes over into d-isomenthone and vice versa.
Summary 1. The inversion of I-menthone to d-isomen-
thone by trichloroacetic acid in benzene, hexane, and chloroform was investigated a t 20’ with various concentrations of acid and menthone. 2. The catalytic activity per mole of acid has a maximum when the molar ratio of acid to menthone is about 2. This relation holds in different solvents and at all concentrations of the menthone which were investigated (0.1 to 1 mole/liter). (26) L. P. Hammett, “Physical Organic Chemistry.” McGrawHill Book Co., N e w York, ii. Y . . 1940.
110
M. G. SEVAG
3 . With a constant concentration of the acid, the reaction rate sinks with rising concentration of the menthone. If the menthone concentration is lower than the acid concentration, the retarding effect of the menthone is more pronounced than with menthone concentrations higher than those of the acid. 4. The temperature coefficient of the reaction was determined at intervals between 20 and 50’. The heat of activation is 6200 cal. 3 . The rotation of the equilibrium mixture depends on the concentration of the trichloroacetic acid for two reasons. The acid affects the specific rotation of the components of the equilibrium mixture, and, with higher concentration of the acid, the equilibrium shifts in favor of the d-isomenthone.
[CONTRIBUTION FROM THE
Vol. 65
6. The analysis of the kinetic results shows that the inversion takes place by interaction of a binary acid-menthone complex with a further molecule of the monomeric acid. Equivalent to this assumption would be the spontaneous rearrangement of a complex of one molecule of menthone with two molecules of monomeric acid. 7. Freezing-point determinations were carried out on benzene solutions of menthone and trichloroacetic acid to determine the constant of the complex formation between menthone and acid, and the association constant of the acid. 8. By comparison of the reaction rate of “Rechtsmenthon,” I-menthone and d-isomenthone, it was confirmed that “Rechtsmenthon” is a mixture of E-menthone and d-isomenthone. ROCHESTER, N. Y .
DEPARTMENT OF BACTERIOLOGY, THESCHOOL OF MEDICINE,
RECEIVED AUGUST14, 1942
UNIVERSITY O F PENNSYLVANIA]
Properties of p-Hydroxylaminobenzenesulfonamide and a Related Molecular Complex’ BY M. G. SEVAC Mayer2 reported a crystalline substance melting a t 161’ but did not give either its preparation or analysis. Bratton, White and Marshall3 obtained a product melting a t 139.5-141.5’ and having the composition CsHsNzSOs which they regarded as p-hydroxylaminobenzenesulfonamide. Thorpe and Williams4confirmed the work of Bratton, el al. Burton, et a L 6 obtained a substance melting at 163-164’, the analysis of which did not fully agree with that required by CeH8N2SOs. Burton6 reported two substances, one melting a t 139-140’ and the other a t 160-161’. He did not give their analyses but considered them to be dimorphic forms. Sevag and Shelburne’ obtained two forms of hydroxylaminobenzenesulfonamide, one melting a t 139-141’ and the other a t 181.5’. In view of the above conflicting results, a detailed study of the properties and the (1) This work started under a grant from The Commonwealth Fund, and was continued under grants from The Josiah Macy, Jr. Foundation, and Merck and Co. (2) R. L. Mayer, Bull. acad. med., 117, 727 (1937); Biol. mcd. Paris (Suppl.), 27, 45,75 (1937). (3) A. C. Bratton, H. J. White and E. K. Marshall, Jr., Proc. SOC. E x # . B i d . Mcd., 42,847 (1939). ( 4 ) W.V. ThorpeandR. T. Williams, Biochm. J . , 86,61 (1941). (5) H.Burton, J. W. McLeod, T . S. McLeod and A. Mayr-Hartiug, Bud. J . E*912. Path., 21, 292 (1940). (6) H. Burton, Chemistry €9Induslry, 60,449 (1941). (7) M. G . Sevag and M Shelbume. J . Ract.. 49, 411, 421. 447 t10421
nature of these substances was considered necessary. The present study confirms the findings of Bratton, White and Marshall regarding the nature of the substance melting at 141.5’. On the other hand, the substance melting a t 161.5’ is found to differ from it in C , H and N analyses, salt-forming and solubility properties and in the volume of oxygen they consume. According to the findings of Bambergefl hydroxylamine on oxidation yields one mole of nitrosobenzene and one mole of hydrogen peroxide. One mole of nitrosobenzene is then combined with one mole of hydroxylamine forming azoxybenzene and water. The oxidation of p-hydroxy1aminobenzenesulfonamide seems to follow the same course. The substance melting a t 141.5’ consumed a volume of oxygen corresponding to 92.4%, and that consumed by the substance melting a t 161.5’ to only 64% of the theoretical value (Fig. 1). It was evident that the latter substance contained 36% a non-oxidizable substance. The analysis of the products in the reaction mixture revealed the presence of 35% sulfanilamide (Table I, compare columns 1. 5 and 8). These results show that the substance melt(#>
F’ Ramberger, He? , SS, 113 (1 900).
