March, 1938
AMINECATALYSIS OF DIMETHYLACET ACETOACETIC ACID DECOMPOSITION 595
A d . b l e d . for CJ%OJG: C, 66.73; H, 8.27. Feuod: C, 58.62; H,3.66. 6-Aminoacridioe-2-sdfonic Acid.-To the hot wlution of 6-aminoacridone-Z-sulf~nic acid (6 g.) in s o d i w hydroxide (lo%, 9 9.) and water (120 CC.),sodium a m a l g w (5%, 80 g . ) was added. The mixture was shaken gently a t 80' for two hours. The filtered hot solution on acldlfying with hydrochloric acid deposited prismatic needles of pink color; yield 5.5 g. It forms brownish-orange rods from water, being unaltered a t 360" and soluble in hot water and hot alcohol with green fluorescence. The almost colorless solution in concentrated sulfuric acid shows intense greenish-blue fluorescence. This compound is rather sensitive to heat especially in alkaline solution. Anal. Calcd. for ClSH&aNa: C, 66.93; H, 3.65. Found: C, 56.95; H, 3.77. The sodium salt gives yellow thick plates from water. It is soluble in water and alcohol with green fluorescence. Anal. Calcd. for C1aH90fi2SNa: Na, 7.77. Found: Na, 7.15. b C l r b o x y m e t h y l a m i n o a c r i ~ e - Z ~ oAcid.-A ~e SOlution of 6-aminoacridine-2-sulfonicacid (0.55 g.), monochloroacetic acid (0.2 g.) and hydrated sodium carbonate (0.6 g.) in water (8 w.) was refluxed for two hours and gave, on acidifying with hydrochloric acid, aggregates of violetred fibrous needles (0.4 g . ) which turned to tar a t 345-360'. Anal. Calcd. for ClbH120bN2S: C, 54.22; H, 3.61. Found: C, 54.32; H, 4.08. pChloroaeetykminophenyl-arslnie Acid.-A mixture of p-aminophenyl-arsinic acid (1.1 g.) and rhloroacetyl
_--_ ...._
c M d e (0.9 9.) was warmed a t 70' for a few mlnutes and pawed on cooling into w.atar. The resctiep prodwt w w dissolved in wdium carbonate and precipitated with hydrochloric acid in colorless plates, m. p. 295-296" (dec.). I t is readily soluble in methanol but insoluble in water. A d . Calcd. for CsHo0,NClAs: C, 32.72; H, 3.07. Found: C, 33.00; H, 3.38. f5-Arseeetinylparieoacridiue-2-oJfonic Acid.-A wlution of 6-amiwacridine~2-sulfonicacid (0.5480g.) and pchloroacetylyninophenyl-arsinicacid (0.5720 g.) in sodium hydroxide (lo%, 1.6 g.) was warmed a t 70-80' for three hours and acidified with hydrochloric acid on cooling. The solid wag digested with warm water, washed with water and then with methanol; yield 0.75 g. The substance crystallizes in dark red prisms, t u r n i q to tar a t 340-360". It is soluble in dilute sodium carbonate and insoluble in alcohol and acetone. The yellow solution in concd. sulfudc acid exhibits green fluoreacenee. A d . Calcd. for GiHisGNsSAs: C, 47.48; H,3.39. FOWI: c, 47.83; H, 4.02.
I hereby thank Professor Hata for his interest in the work.
summary Fuming sulfuric acid acts upon &nitroacridone as well as 6-nitrodiphenylamine-2-carboxylicacid to give 6-nitroacridone-2-sulfonic arid. TOKYO, JAPAN
RECEIVRD OCTOBER 5, 1937
-
[ C O N T R ~ B U T I O N FROM T H B C H E M I C A L LABORATORY OF T H E ROYAL vETERlti.$RY
AND
AGXICULTURAL COLLBPE 1
Amine Catalysis of the Ketonic Decompositios of a,a-Dimethylacetoacetic Acid BY KAI JULIUS PEDERSEN
I t has been found by Widmark and Jeppsson' and by Ljunggren' that the ketonic decomposition of acetoacetic acid CHsCOCH2COOH +CHsCOCHs f COI
is catalyzed by primary amines when the pH of the solution is kept within certain limits. In more acid and in more alkaline solution the amines do not catalyze. The influence of the acidity of the medium is explained by assuming that the reaction takes place either between the neutral amine and the undissociated acetoaretic acid, nr between the positive aminium ion and the acetoacetate ion. I t is the object of this paper to examine more closely the amine catalysis of the decmqmsition (1) E. N.P. Widmark and C.A. Jeppmsn. Skand. P i c k . Phvsiol , 41, 43 (1922). (2) C. Ljuaggren, "Kntalytisk Kolryruvapjllluiag UT Ketokarbonsyror," Dissertation, Lund, 1925.
of @-keto carboxylic acids. Instead of acetnacetie acid there was used a,a-dimethyIracetoacetic acid, whose uncatalyzed decompoeition has been studied in earlier papers."' The two amines aniline and o-chloroaniline were chosen as catalysts. Experiments were carried out at 24.97 and 34.93' in aniline-aniliniurn chjoride buffer solutions and in ochloroaniline-o-chloroanilinium chloride b d e r solutions. To all the solutions had been added enough sodium chloride to make the total salt concentration 0.150 N . We denote the undissociated dimethylacetoacetic acid by HA, the uncharged amine by 3, and the total concentration of t h e reacting substance by x (HA) (A-). Amofding to the Swedish investigators',2 we may interpret the decomposition in the amhe buffer solutions as
=
+
(3) K.J. Pedulocl. Tar Jm.r4,Il. -8 (0 K.J Pedersen. ibid , II,240 (1986).
(L9W.
596
&I
JULIUS PEDERSEN
the result of three simultaneous reactions, namely, the unimolecular decompositions of HA and A-, and a bimolecular reaction between B and HA -dx/d MHA) 4- ki(A-1 Kz(B)(HA) (1) K O and kl have been determined separately in the earlier paper.’ Expression 1 holds only very roughly for the experiments in this paper. For the aniline buffers we must add a term of the form ks(B)2(HA), and for the o-chloroaniline buffers one of the form kd(BH+)(HA). Thus the velocity in these buffer solutions may be expressed
+
as -h/dt
= ko(HA)
+
+
(2)
For the experiments in aniline buffers the term containing kr, and for those in o-chloroaniline buffers the terms containing kl and ka are negligibly small. The solutions of sodium dimethylacetoacetate were prepared from the same sample of purified ethyl ester (“fraction 2”) as was used in the earlier paper,4 and in the same way as described there. The decomposition was followed by measuring the carbon dioxide pressure above the solution in an apparatus described earlier.6 Before an experiment was begun the apparatus was evacuated with the water pump, filled with carbon dioxide, and evacuated again. Experiments in o-Chloroaniline BuffersThe o-chloroaniline buffers were prepared from o-chloroanilinium chloride and sodium hydroxide solution. The o-chloroanilinium chloride used was prepared by dissolving o-chloroaniline (British Drug Houses, Ltd.) in boiling 10% hydrochloric acid. The crystals obtained by cooling in ice were recrystallized from 10% hydrochloric acid. The preparation was analyzed by titration with sodium hydroxide and by determination of the amount of bromine which it could take up.6 One mole of pure o-chloroanilinium chloride needs four equivalents of bromine. The results of both titrations ageed with those calculated from the formula within less than 0.5%. We use the following symbols. Initial and final values are indicated by the subscripts 0 and a, respectively; P = p , difference between the final pressure readings and the reading t minutes after the start; (o = Po/xo, proportionality factor of the equation P = qx; a = (A-)/x, degree of dissociation of dimethylacetoacetic
-e,
-
K. J. Pedusen, Tms JOUBNAL, 68, 18 (1931). (6) K. J. Pedersen, ibid., 66, 2616 (1934). (6)
acid. Analytical concentration of the solution at the start of the experiment: a molar o-chloroaniliniunl chloride, b molar o-chloroaniline, xo molar dimethylacetoacetic acid, (0.150 - a) molar sodium chloride. L = KBK+/KHA, ratio between the dissociation constants of the acids BH+ and HA. The values of (B), (BHf), a,and (H+), corresponding to each of the readings are first calculated by successive approximations from the four equations
+
-
(3)
(B) b (H+> d / ( ~ (BH’) (H+) d / q a = (B)/[(B) 4- UBH+)I (H+) %a+(BH+)/(B)
+
4-ki(A-1 &@)(HA) ka(B)*(HA) kr(BH+)(HA)
Vol. 60
-
+
(4) (5) (6)
For the dissociation constants of dimethylacetoacetic acid and the o-chloroanilinium ion are used the values found in 0.150 N sodium chloride solution in earlier publication^.^^' They are given a t the top of Tables I and 11. In order to show that the rate of decomposition found in the experiments may be expressed by equation 2 we transform this equation as explained below. For ko and kl we use the values found earlier4 and given a t the top of Tables I and 11. The solutions are so acid that kl(A-) is always negligibly small. As we shall see we get agreement with the experiments when we leave out also the term with k3. Equation 2 may then be written -&/dt
= [ko
+ kz(B) + kABH+)I(1 -
a)%
(7)
When x is sufficiently small the reaction will follow the unimolecular law with the constant k = [Ro h(B)m kdBH+)ml(l - a m ) (8) From equations 3 and 4 we find
+
(B),
+
- (B) = (BH+) - (BH’), d / q
- (HC) t (H+),
By means of this equation and P = ax we transform equation 7 into -dp/Pdt = [KO kdB), 4-h(BH+),] (1 a) ( h - h ) ( l a ) ( d / q - (H’) 4-(H+)-) If we use the abbreviation (8) it may be written
+
-
-
By integration we find -K*t=logP+
A-IogPo
(9)
where the small correction A is A - I [ K * -k a-
(h*- k4*)(1
-
- (H+)
~ ) ( d / p
+ (H+),)df
(7) K. J. Pedmen. Kgl. Danskc Vid. Setsk.. Ma1b.-fys. Mcdd., 16 No. 3 (1937).
AMINECATALYSIS OF (Y,C~-DINLETHYLACETOACETIC ACID DECOMPOSITION
hlarch, 1938
597
TABLE I KETONICDECOMPOSITION OF a,a-DIMETHYLACETOACETIC ACID IN O-CHLOROANILINE BUFFERSOLUTIONS AT 24.97 KBH+= 2.12 X 10-3. KEA 5 5.11 X 10-4. L = 4.15. KO* = 1.822 X 10-8. k1* = 3.99 X lo*. xo was always about 0.019. (H+)m
(BH+)m
x 108 b X 102 X loa x 108 13.56 36.41 12.83 0.73 21.58 28.40 20.15 1.43 36.54 38.55 36.44 2.01 43.72 46.68 28.3 2.96 63.65 36.31 60.41 3.24 67.21 71.56 28.30 4.35 84.31 88.7 36.49 4.39 108.1 113.6 36.41 5.47 From the experiments are found: kn* =
(B)m X 103
a
01
first read.
ko* X
loa X
k* X 10: UDD
37.14 0.3074 0.4109 ,2629 29.83 ,2122 .2023 38.45 ,1792 31.27 .1334 .1470 .1362 39.54 .1268 32.74 .lo51 ,0988 .0994 .lo44 40.79 41.88 .0822 .OS54 0.03244, kd* = 0.00316.
Exptl.
Calcd.
1.808 2.103 2.534 2.540 2.832 2.785 3.071 3.198
1.807 2.103 2.540 2.537 2.847 2.771 3.056 3.221
(1
- am)
1.073 1.343 1.453 1.554 1.574 1.631 1.632 1.666
kr* X 10' kd* X 10' X(l-am) X(l-am) X(B)m X(BH+)m
0.710 .713 .995 .865
0.024 .047 ,092
1.108
.165 .190 .239 .312
0.950 1.185 1.243
,118
TABLB I1 KETONICDECOMPOSITION OF a,a-DI?AETHYLACETOACETIC ACID IN O-CHLOROANILINB BUFFERSOLUTIONS AT 34.93 L = 6.34. ko* = 6.36 X lo-*. kl* = 17.5 X lo-'. 1c0 was always K3Hf = 2.93 X IOea. K H A= 4.62 X IO-'. about 0.019. kr* X 10' kd* X 10' k* X Calcd. 10' 10') X(l-am) X(1-am) Exptl. Xko* ( l -X am X (B)m X(BH+)m
(I
a X 100
b X
loa
(H+)m X loa
(BH+)m
x
(B)m
x
103
109
first read.
a-
0.2446 0.3048 14.79 41.12 15.84 40.07 1.05 ,1847 42.18 .1656 29.37 31.43 40.12 2.06 .1327 44.45 43.10 ,1239 47.51 40.04 3.06 .0871 44.95 .OS41 4.91 74.3 79.2 40.04 104.1 46.69 .0645 .0661 110.7 40.06 6.63 From the experiments are found: &,* = 0.0643, &a* = 0.0068.
6.33 7.52 8.22 8.95 9.36
6.33 7.56 8.18 8.91 9.40
4.42 5.19 5.52 5.81 5.94
1.84 2.21 2.40 2.64 2.80
0.07 .I6 .26 .46 .66
and k* = 0.4343 k. A is found by rough graphi- much each of the three simultaneous reactions concal integration, preliminary values of the con- tributes to the total reaction. A is plotted Experiments in Solutions of u-Chloroanilhium stants being used. When log P the exagainst t the points fall along a straight line whose Chloride and Hydrochloric Acid.-For slope determines k* of the experiment. The planation of the results of the kinetic measureagreement with equation 9 was always very satis- ments in o-chloroaniline buffer solutions it has factory. The values of the velocity constants been necessary to assume that a bimolecular rek* computed in this way are given in the eighth action between the undissociated keto acid and the aminium ion coitributed to the catalysis. column of Tables I and 11. In order to find k2* and ka* we introduce the This effect is expressed in equation 7 by the term known values of ko* and a minto equation 8, and containing kd, and its importance is seen clearly from the last column of Tables I and 11. We calculate &z* (B)a, k4* (BH+) for each of the experiments. From the values of this expres- may expect that this reaction also will take place sion for the whole series of experiments at the in strongly acid solution, and, consequently, that same temperature we compute k2* and kr* by the catalysis by o-chloroaniline is not confined to the method of least squares. The results are given a limited pH interval. In order to test this dia t the bottom of the tables. The agreement is rectly experiments were carried out at 24.97' in tested by calculation of k* for each of the experi- solutions containing hydrochloric acid (c molar) ments from the values of k2* and kr*. The results u-chloroanilinium chloride (a molar), and sodium of this calculation are given in the ninth column chloride (s-c-a molar). Within each series of of Tables I and 11. A comparison of the values experiments c and the total salt concentration s in the eighth and ninth columns will show that the were kept constant, while a was varied. agreement is good. This good agreement confirms If equation 7 holds the unimolecular velocity that it was pemissible to leave out the term with constant is given by the equation k, in equation 2. The last three columns of Tables h [ko M B ) h(BH+)1(1 a) (10) I and I1 have been computed in order to show how o was always smaller than 0.01 and was cal-
+
+
-
+
+
-
598
KAI JULIUS I?EDBRSEN
culated with sufficient accuracy from a rough estimate of KxA. By means of equation 6 and the abbreviations x = x2 + A4 (11) A2
A4 3
+
W B H + / [ ( H + ) KBH'] k4(H+)/[(He) -k K B H ~ ]
(12) (13)
equation 10 may be written as k/(l
- a) = ko + xa
(14)
Vol. 60
variation of k4* with (H+) agrees with the assumption that the form stable in strongly acid solution has the acid strength 0.040 and the catalytic constant 0.25 X while the form stable in weakly acid solution has the catalytic constant 3.40 X This is seen from the last column of Table 111, which gives k4* calculated from the formula kr*
- 0.25 X lo-'
(H+) 3.40 X 10-8 - k4* Within a series of experiments a t constant hyThe formula contains three arbitrary constants. drogen ion concentration X is constant. It was The agreement may therefore be accidental. It determined graphically by plotting k / ( l - CY) is not surprising that a constant value of k4* was against a, and drawing a straight line through the found for the buffer solutions (Table I). It folpoints. In order to separate the catalytic effects 15 that the relative variation lows from formula of BH+ and B, X2* was calculated by means of of k4* in the buffer solutions examined is not conequation 12 from the known values of kz* and KBHt. k2* = 0.03244 was taken from Table I. siderable. If the results are recalculated using For the solutions where s = 0.15 was used KBH+ kI* from formula 15 instead of the constant value exactly the same value of kz* is ob= 2.12 X lod3. For the more concentrated so- 3.16 X tained. lutions it was considered sufficiently accurate to It has not been possible to give a satisfactory use the value KBHt = 1.5 X obtained by explanation of the variation of k4. It is very extrapolation to s = 0.75 of the formula given in improbable that the o-chloroanilinium ion takes an earlier paper.' up another proton in the solutions of hydrochloric x4* was found by subtracting Az* from X*. acid used for the experiments, and if this were the Finally, k4* was calculated from &* by means of case the decrease of the hydrogen ion concentraequation 13. The results are summarized in tion would be so great that it, especially in the Table 111. The last column but one gives the solutions where a > c, could hardly escape detecvalues of k4* determined from the experiments as is that the o-chloroanition. Another possibility just explained. For comparison k4* found for the buffer solutions is given in the last line of the linium chloride used for the experiments has contained a catalyzing basic impurity whose corretable. sponding acid has the strength about 0.04. The TABLEI11 assumed impurity can only be present in small KETONIC DECOMPOSITION OF DIMETHYLACET ACETOACETIC ACIDAT 24.97" IN SOLUTIONS CONTAINING a MOLAR0 - amount because it has not been detected by the analyses of the substance, and it must therefore CHLOROANILINIUM CHLORIDE, .C MOLARHYDROCHLORIC ACID, AND (s-a-c) MOLARSODIUMCHLORIDE be a very powerful catalyst. However, t h i s exEach line gives a summary of n experiments where c planation also is very improbable, because the was kept constant while u was varied from 0 t o urnax. experiments have been carried out with preparaPrep. ki*lOS c S am-. ?I no. X*lO* Xz*103 X d * l O a Exptl. Calcd. tions of different origin, and which have been 1.98 2.15 0.150 4 2 0.32 0.02 0 . 3 0 0.30 0.31 0.478 0.75 ,256 4 2 .60 .10 -50 .50 .49 purified in different ways. In Table 111 prepn. 1 .228 .75 .256 4 2 .95 .21 .74 .76 .72 refers to the preparation which was also used for .121 .75 .256 4 2 1.42 .40 1.02 1.03 1.03 the buffer solutions. It was made from o-chloro.057 .15 .075 4 1 2.58 1.19 1.39 1.44 1.55 ,057 .15 .075 4 2 2.65 1.19 1.49 1.55 1.55 aniline (British Drug Houses) as already de.057 .15 .075 2 3a 2.65 1.19 1.46 1.51 1.55 scribed. Preparation 2 was from o-chloroaniline .057 .15 ,075 2 3b 2.60 1.19 1.41 1.46 1.55 ,004" .15 8 1 3.16 3.11 (Kahlbaum, "reinst"). The chloride was recrysExperiments in buffer solutions (Table I), (H+) tallized five times from 10% hydrochloric acid -0.004. and, finally, once from water. Preparations It is seen from the table that k4* is far from being 3a and 3b were different fractions obtained by independent of the hydrogen ion concentratlon . distilling Kahlbaum o-chloroaniline in (KZCW, in k4* decreases with increasing hydrogen ion conan all-glass still fitted with a Widmer column. centration as if the catalyzing molecule were As seen especially from the experiments in 0.057 gradually being transiormed into a less active molar hydrochloric acid the four preparations all form by taking up a hydrogen ion. Thus, the gave the same catalytic effect.
-
= 0.040
(15)
March, 1938
AMINECATALYSIS OF DIMETHYLACET ACETOACETIC ACIDDECOMPOSITION 599
Experimentsin Aniline Buffer Solutions.-The solutions were prepared from aniline and hydrochloric acid. The aniline used was the middle fraction obtained by distillation in vacuo of Kahlbaum aniline for analytical purposes. In the aniline buffers the hydrogen ion concentration was so small that it was necessary to make a correction for the hydrocarbonate ion present in the solution. If the carbon dioxide pressure is pc& cm. above a solution with the hydrogen ion concentration (H+) the hydrocarbonate concentration is where K ~ ois, the ordinary dissociation constant of carbonic acid, f the Ostwald absorption coefficient for carbon dioxide, R the gas constant in liter atm./degree mole, and T the absolute temperature. For the buffer solutions the correction will therefore be
For KBH+, the dissociation constant of the anilinium ion, was used the values found earlier* for 0.150 N sodium chloride solution. They are given a t the top of Tables IV and V. For Kco2 the values 7.8 x 10-7 a t 250 and 8.5 x 10-7 a t 35' were estimated from the measurements of Shedlovsky and M a c I n n e ~and , ~ Giintelberg and Schiijdt.lo Forf we have used the values 0.821 at 25' and 0.59 a t 35' from the measurements of Geffcken" (the latter value by extrapolation). P is here the difference between the corrected pressure readings a t the times m and t P a P m A P m - P - AP (17) The meaning of the other symbols is the same as or analogous to that in the treatment of the o-chloroaniline buffers. Instead of equations 3 and 4 we have (B) =, b 4- (H+) - A P / v - d / v (18) (BH+) = a - ("3 Ap/v d / v (19) From the equations 16, 17, 18, 19, 5, and 6, we find by successive approximations the values of A$, P, (B), (BHf), a, and (H+), corresponding to each of the readings. We shall now show that expression 2 holds for
+
+
(8) K. J. Pedersen, Kgl. Danskc Vid. Sclsk., Math.-fys. Mcdd., 1& no. 9 (1937). JOURNAL, 67, 1705 (9) T. Shedlovsky and D. A. MacInnes, THIS
(1935). (10) E. Gantelberg and E. Schibdt, 2. physik. Chcm., 186, 893 (1928). (11) G. Geffcken, ibid., 49, 257 (1904).
the aniline buffers when we set kq = 0. The expression is transformed into
+
-dx/dt
+ +
+
ko(HA) ki(A-) kaL(BH+)(A-) ksL(B)(BH+)(A-) = [ko(l - a ) kia kzLor(BH+) ~ s L ~ ( B ) ( B H + ) ] x( 2 0 )
+
+
When x is sufficiently small the reaction will follow the unimolecular law with the constant k
ko(1.-
a,)
+ kia, + k2Lam(BH+), + kaLa,(B),
(BH+),
(21)
We further use the abbreviation 8
( ~ z Lt (B)mksL)a'm/vk
(22)
By means of the equations 18, 19, 21, and 22, equation 20 may be written as
where
-1
L
a[kL
- ksL((BH+), PP
- (B),
($ + P )
[APm
+
- A P +dHf)
v(H+),]
By integration of equation 23 we find
The small correction l f * ( P ) d l , where * denotes
that all velocity constants should be expressed by means of decadic logarithms, is found by rough graphical integration. Before we can carry out the computation we must know approximate values of the constants in order to estimate 6 and Sdf*(P)dt. We therefore make a preliminary
l
computation of k* setting j3 = 0 and p ( P ) d t = 0. From the values of K* thus determined we find by means of equation 21 preliminary values of ka*L and ks*L which we use for the final computation. For ko* and kl* we have used the values found in an earlier paperj4and given a t the top of Tables IV and V. When
is plotted against t the points fall along a straight line whose slope determines k* of the experiment. The agreement with equation 24 was always very
KAIJULIUS PEDERSBN
600
Vol. 60
TABLE IV
-
DECOMPOSITION OF a,a-DIMETHYLACETOACETIC ACID I N ANILINE BUFFERSOLUTIONSAT 24.97 KBH+= 2.35 x lo-'. KEA = 5.11 X L 0.0460. ko* = 1.822 x 10-8. kl* = 3.99 x 10-6. KETONIC
q was always about 0.019.
(I
( B H '9m
(B) m
first read.
k* X 109
Exptl.
am
0.0525 0.0368 0.872 0.9385 1.242 .0528 .0867 .951 .9727 1.276 .0532 .1890 .982 ,9872 1.386 .0536 .2884 .987 .9915 1.493 .1279 .0646 .871 ,9165 2.889 .1281 .922 .0930 .9403 2.985 .1282 .2412 .967 .9761 3.396 ,1284 .2950 .974 ,9803 a. 557 From the experiments are found: kn*L= 0.02209, kr*L = 0.0197.
Cslcd.
+x -
[ko* (1 u m ) kiam] 108
1.240 1.276 1.383 1.495 2.894 2.994 3.406 3.552
0.116 .054 .027 .OlQ *
156
.113 .047
.om
kz*Lpm X (BH:)m
x
10
1.088 1.134 1.160 1.174 2.589 2.661 2.764 2.780
kr*Lum X (B)m(BH+)m x 108
0.036 .OB .196 .302 .149 .220 .595 .732
TABLE V
KETONIC DECOMPOSITION OF K B H += 3.48 X 10-6.
always about 0.019.
ACID I N ANILINE BUFFER SOLUTIONS AT 34.93 L = 0.0754. ko* = 6.36 X 10-8. kl* = 17.5 X 10-0. xo was
a,@-DIMETHYLACETOACETIC
KEA = 4.62 X 10-4. a
(BH +)m
first read.
(B)m
Um
k* X 10' Exptl. Calcd.
0.05453 0.0425 0.858 0.9118 3.89 .lo16 .945 ,9616 ,05377 3.89 ,9794 4.12 .05430 .1949 .972 .980 .9%9 4.31 ,05423 .2670 .0790 .0474 .829 .8884 5.46 ,0794 .0467 .826 .8866 5.47 .0792 .1420 .946 .9597 5.73 .9597 5.72 .0792 .1420 .948 .0788 .2455 ,971 .9770 6.12 .0792 .2417 .970 .9759 6.14 .8165 8.26 .1293 ,0436 .756 8.82 .888 .9111 .1304 .loo9 .942 .9520 9.52 .1295 .1941 10.11 .959 .9638 .1311 .2633 From the experiments are found: k3.L = 0.0647, kX*L = 0.0512.
satisfactory. The values of the velocity constants computed in this way are given in the fifth column of Tables IV and V. In order to find kz*L and ks*L we introduce the known values of ko*, kl*, and a minto equation 21, and calculate k $ * L ( B H + ) k3*L(B) ( B H + ) for each of the experiments. From the values of this expression for the whole series of experiments a t the same temperature we compute kz*L and ks*L by the method of least squares. The results are given a t the bottom of the tables. The four last columns have been computed from the constants kz*L and k3*L in order to show the agreement with the formula, and to give an impression of how much the uncatalyzed and each of the two catalyzed reactions contribute to the total reaction. The good agreement confirms that it was permissible to leave out the term with &, corresponding to a reaction between the anilinium ion
+
'
-
[ko* (1 a m ) k t * L a m kt*Lam + k ~ * ~ m ]X (BH+)m X (B)m(BHt)m x 103 x 10: x 108
0.58 .26 .15
3.90 3.88 4.12 4.30 5.44 5.46 5.74 5.74 6.11 6.13 8.25 8.88 9.52 10.12
.I1 .73 .74 .27 .27 .16 .17 1.18 0.58 .32 .25
3.21 3.35 3.44 3.46 4.54 4.55 4.92 4.92 4.98 5.00 6.83 7.69 7.97 8.17
0.11 .27 .53 .73 .17 .17 .55 .55 .97 .96 .24 .61 1.23 1.70
and the undissociated dimethylacetoacetic acid. In the anilinium buffers 1- CY is small, which naturally makes the part played by this reaction less important. In order to see if the reaction takes place to an appreciable extent in more acid solution some experiments were carried out in 0.018 and 0.056 M hydrochloric acid containing anilinium chloride. For these solutions (B) and the terms with K1 and K3 are negligibly small. We therefore apply the equation k = ko(1-a) k&a(BH+) kd(l--a)(BH+) to the unimolecular velocity constant K found in the experiments (Table VI). All the constants except ka are known. We find k4* = 0.00056 independently of the hydrogen ion concentration. We see that the effect exists, but it is too small to be detected in the aniline buffer solutions. In an earlier paper' it was shown that aniline, in agreement with formula 2, has no appreciable effect in sufficiently alkaline solution.
+
+
AMINECATALYSIS
March, 1938
OF
THYLACETOACETICACIDDECOMPOSITION 60 1 CY,(Y-DIME
TABLE VI or that BH+ reacts with A- (corresponding to KETONICDECOMPOSITION OF a,a-DIMETHYL.ACETOACeTIC In Table the velocity constants kd and k&). ACID AT 24.97' IN SOLUTIONSCONTAININQ c MOLAR VI1 the velocity constants have therefore been HYDROCHLORIC ACID,a MOLAR ANILINIUM CHLORIDE, AND given in both ways. If we assume that the (0.150-c-a) MOLAR SODIUM CHLORIDE. THECONSTANTS GIVENAT THE TOPAND BOTTOMOF TABLE IV HAVEBEEN krhenius equation k = Ae-Q'RT holds for the velocity constants, we may calculate the conUSED cx
1na
'15
56.5 18.26 55.7 18.40
37.7 56.4 75.4 112.7
P
0.0091 .0271 ,0091 .0268
ka X ki*L kr X 108 k* X 108 103 X 108 (BH+) Exptl. Calcd. ( 1 P) (BH+)a (1 - a )
-
1.833
1.835 1.807
1.844 1.860 1.901
1.837 1.864 1.901
1.772 1.807 1.773
0.007 .034 ,015 .067
0.021 .031 .042 .061
From the experiments is found: k,* = 0.00056.
stants A and Q for each of the simultaneous reactions. Table VI11 gives the values of A and Q, when the time unit is the second. TABLE VI11 THECONSTANTS A AND Q OF THE FORMULA k = Ae-4/RT TIMEIN SECONDS
The results of the experiments both in o-chloroaniline and in aniline buffer solutions have been summarized in Table VII. It is impossible on the basis of the kinetic experiments to distinguish between the two possibilities mentioned in the introduction, namely, that B reacts with HA (corresponding to the velocity constants kz and &), Aniline TABLE VI1 T H E DECOMPOSITION OF a,a-DIMETHYLACETOACETIC ACID
0-CHLOROANILINE AND ANILINE BUFFER SOLUTIONS The experimental data agree with the following expressions for the velocity, where HA denotes the undissociated keto acid, B the amine, and L the ratio between the dissociation constants of the acids BHf and HA: (1) ko HA) kdA-1 k2(B)(HA) k&3)2(HA) kd(BH+)[HA) and (2) ko(HA) kdA-) kzL(BH+)(A-) kL(B) (BH+)(A-) ka(BH+)(HA). The velocity congtants have the following values (k* = 0.4343 k, time in minutes) IN
+
+ + +
++
24.97'
ko* ki* (kz*
o-Chloroaniline
Aniline
0.001822 3.99 X lo-' 0.03244 0 ,00316 .1346 ,480 .429
+ +
34.930
0.00636 17.5 X 100.0643 0
,0068 .408 .858 ,680
0
.02209 .0197
.OM7 .0512
Q (cal./mole)
A
ko k1 k2
k, k2L ks
1
KC
k& kd
4.76 X 1.09 X 1.921 X 2.24 X 4.044 X 1.165 X 2.503 X 2.291 X 5.06 X
10" lOI3
108 10' 1012
10' 10' 10"
lo9
23.0 X 27.1 X 12.53 x 14.0 X 20.31 X 10.64 X 8.43 x 19.68 X 17.49 X
10s lo3 103 lo9 lo3
loa 103 103
lo3
I wish to thank the head of the laboratory, Professor Niels Bjerrum, for his kind interest in my work. summary The ketonic decomposition of a,a-dimethylacetoacetic acid has been investigated a t the temperatures 24.97 and 34.93' in o-chloroaniline and aniline buffer solutions. Several reactions take place simultaneously. Tables VI1 and VI11 give a summary of the experimental results. The decomposition was also studied in solutions of o-chloroanilinium chloride (and anilinium chloride) containing hydrochloric acid. The influence of the hydrogen ion concentration seems to be more complicated than it appears from the experiments in the buffer solutions. COPENHAGEN, DENMARK RECEIVED DECEMBER 15, 1937
