46 $8
PETERE. YANKWICH AND A L R E R r L. PROMISLOW [CONTRIRUTIOS FROM
NOYES
CHEMICAL LABORATORY, UNIVERSITY O F
Vol. 76
ILLINOIS]
Intramolecular Carbon Isotope Effect in the Decarboxylation of Liquid Malonic Acid Near the Melting Point BY PETERE. YANKWICH AND ALBERTL. PROMISLOW RECEIVEDAPRIL24, 1954 The intramolecular isotope effect in the decarboxylation of liquid malonic acid has been investigated at 140'. The product acetic acid was degraded via the Schmidt reaction in order to ascertain and permit correction for carbon isotope inhomogeneity in three-carbon malonic acid skeleton. The best results lead to k,/kl = 1.0292 & 0.0007, a value in agreement lvith the prediction of Yankwich and Belford (1.0284 i O . O n l Z ) , but not with that of Bigeleisen (1.0198).
There have been published the results of several investigations designed to determine the magnitude of the intramolecular carbon isotope effect in the decarboxylation of malonic acid a t temperatures slightly above its melting point. The rate constarits may be defined' :LS k.3
C'W0I-I /,
+ C'"*
'C'2001-I
+ C"0z
H,?,
ka
+ C'2I-I:!C1200II + C1*HaC'300H
(I)
(2)
I n Table I are collected average values of the isotope effect expressed as 100 [ ( k 4 / k 3 ) - 11; the appended errors are average deviations and were computed by the present authors.
The present study was conceived in such a manner as t o : (a) make possible the determination of both the carboxyl and methyl carbon isotope ratios for such carbon in acetic acid product; (b) eliminate any possible but unrecognized effect on the results due to the use of a continuous collectioti (sweep) procedure; (c) obtain data for carbon dioxide and acetic acid products in each run. T o these ends, we have carried out the decarboxylation of malonic acid just above its melting point in sealed reactors (batch process), and have employed a degradation procedure for acetic acid based on the Schmidt reaction5 in order to establish the carbon isotope distribution in the original malonic acid.
Experimental
TABLE I
The analyses were divided into three series. Series A way a test of the batch decarboxylation procedure; the evolved SUMMARY O F ISTRAMOLECULAR ISOTOPE EFFECTRESULTS:acetic acid was burned over copper oxide. The acetic acid lOO[(k,/k.i) - 11 was degraded in all experiments of series B ; but the results 7', Calcd. based on L l h and Av. of Oxidn. indicated t h a t further improvement of the degradation OC. coz HAC all results method R e f . procedure was desirable. I n series A and B the malonic 138 1 . 9 zt 0 . 1 ( ' 2 ) O s b , ,, w c 1 acid employed was Eastman Kodak Co. white label grade; 138 2 . 9 3z 0 . 4 ( 4 ) 2 . 2 & 0 . 2 (3) 2 . 5 i 0.4 Dd 2 preliminary tests showed that the purity of the material 2.4 (1) 3.1 (1) 2 . 8 =t0 . 3 LV (99.9 f O.lyoby acidimetric titration, m.p. 135') was not 137 2 . 9 i 0 . 3 (3) 2.5 (1) 2 . 8 =t0 . 3 D 3 increased significantly by recrystallization, and it was used a (n) = number of experiments. * Only one introduction without further purification other than careful drying. I n series C the improved degradation procedure was emof CO, to mass spectrometer in one expt. W = oxidation ployed. The malonic acid used was a synthetic material of M A or HAC by Van Slyke-Folch or similar wet combustion prepared by the method described by Yankwich and Stivers.3 technique. D = oxidation of MA or HAC over CuO a t The product was twice recrystallized from acetonebenzene 700-800 O . mixtures, then sublimed in vacuo ( a t 90-100°). Decarboxylation Apparatus and Procedure .-The batch The agreement among the various sets of results summarized in Table I is not too good; further- reactor is shown in Fig. 1. It consists essentially of a chamber B, provided with a cold finger F , a break-off seal H , more, the internal consistency of any given set (re- and tubes A and E for vacuum system connections. The sults calculated on the basis of carbon dioxide or sample size was chosen so that the final partial pressure of acetic acid analyses should be the same) is little acetic acid vapor at the temperature of the decomposition better. It is difficult to find likely specific causes of was much less than the vapor pressure of the liquid acetic acid. the observed discrepancies. The reactor is connected to a high vacuum system an inert gas sweep through A and warmed with a Bunsen flame during evacuaI n some of the was employed to remove decarboxylation products tion. After the reactor is cool, dry COz-free air is admitted A and the vessel removed from the vacuum system. from the reaction vessel; acetic acid (HAC) and through weighed sample of dry malonic acid is shaken through carbon dioxide were later separated and purified A the constriction C into the body of the reactor B; the bull) by distillation. (Bigeleisen and Friedman' give no is then connected t o the vacuum system and evacuated for details of their method of product collection.) In 2-3 hours, after which period it is sealed a t C while under all cases a malonic acid was employed which had vacuum. By means of a clamp placed a t E, the whole bulb is imbeen shown to be isotopically homogeneous in car- mersed for ld-18 hours in a n oil-bath thermostated a t 14O..i bon by means of the soda lime--acetate pyrolytic f 0.5'. (The reaction was found to be more than 98% degradation which yields the methylene carbon complete in 12 hours. From the data of Hinshelwood. infra, one would expect 98% decarboxylation in about atom of malonic acid as methane. The wet oxida- vide 4 hours.) The bulb is then taken from the bath, the suptions were quantitative, presumably; one could porting tube cut a t G and a standard taper joint attached then ignore an isotope effect of the type noted by there. The breaking magnets are inserted through this joint and the reactor attached to the high vacuum system. Evans and Huston.4 When the connections are evacuated, a cold bath (Dry Ice(1) J . Riyeleisen a n d L. Friedman, J . Chcm. P h y s . , 17, 998 (1919). trichloroethylene) is raised about F to condense the acetic (2) J. 0 . Liuilsay, A . P.;. Dourus a n d IT. G. Thode, Can. J . Chcm., 29, acid product, the seal H is broken, and the carbon dioxide 192 (liZS1). product distilled from B to a liquid nitrogen-coolrd finger' (3) 1.' E. YaukwicL and E. C. Stivers, J . Chcm. Phrs., 21, 61 (1953). (4) B.A . Rvnas and f . L. Kuaton, ibid., 19, 1214 (19511.
(5)
K.P.Schmidt, B w . , 678, 704 (lQ24).
Sept. 20, 1%51
DRCARIIOXYL.\TION OF LIQUID~ T A L O N I C XCID
4fi49
H
I CM.
I
ABSORBER (ASCARITEANHYDRONE)
I
REACTION v E S SEL I
Fig. 1.-Batch-type
\
decarboxylation vessel (letters refer t o text).
on the vacuum line. The CO, is freed from traces of acetic acid by a second distillation; it is transferred finally to a mass spectrometer sample bulb. (In series C a n additional purification consisting of solution of the CO? in carefully degassed distilled water was carried out.) In the experiments in series A the cold finger F was broken off from B and acetic acid samples removed with a micropipet for combustion. I n all other runs the acetic acid was vacuum transferred into a n appropriate reaction vessel for degradation. Degradation Apparatus and Procedure.-Acetic acid was degraded by procedures based upon Phares' applications of the Schmidt reaction,s which, in this case, may be written HN3
I
1
I
I
SCRUBBER
I
LG&l C U fUBE,425%
ABSORBER IANHYDRONE)
-
PRESSURE
CO, HANDLING
100% HzSOd
+ CHaCOOH + HzS04 Nz + COz -I- CH3NHn.HzSO4
(3) The reaction trains used in this research are shown schematically in Fig. 2; the "B" train was employed in series B experiments, the '"2" train in series C. As will be seen below, the isotope ratios obtained for carboxyl carbon when the "B" train was used were low and erratic. While the ratios obtained from methylamine combustion seem very reproducible, there is no guarantee that they are not in error also. A sweep of the mass spectrum of the samples showed t h a t erratic behavior was always accompanied by a large (20-300 millivolts) background peak at mlq 30; it seemed likely t h a t another impurity was present at m / q 44 (established a t 20 volts for routine analyses) which affected the measured ion current ratio ( m / q 45)/(m/p 44), assumed t o be due to isotopic forms of carbon dioxide only. Since these difficulties were removed, and internal consistency of the results established, by the introduction of the heated copper tube, they are presumed to have been due t o oxides of nitrogen produced during the degradation and methylamine c o m b ~ s t i o n . ~ The reaction vessel is shown in Fig. 3. Into the pearshaped flask J is placed ca. 50 mg. of reagent grade sodium azide and 0.3 ml. of ca. 100yOsulfuric acid (3 parts concentrated reagent acid and 1 part 2070 fuming acid); the flask is cooled in flaked ice until reaction ceases, then allowed to warm to room temperature. The flask J is then attached to the vacuum system and cooled to D r y Ice temperature while being evacuated (a small amount of gas evolution takes place before the solution freezes). The acetic acid in F is then distilled into J, which is then filled with d r y COz(6) E. F. Phares, Arch. Biochem. Biophys., 3 3 , 173 ( 1 9 5 1 ) . (7) The background at m / q 30 was likely due to N14O'a. The low carbon ratios were due probably to a contaminant at m / q 44, most
likely N"Nl'O18.
Fig. 2.-Block
diagram of trains employed in acetic acid degradation.
free air, removed from the vacuum system, and put into place in the degradation train. (While these operations are being carried out, the train is swept with a slow stream of helium .) A water-bath at 30" is placed about J; over a period of 10-15 minutes the temperature of the bath is increased to 70°, after which time the sweep is continued for one hour. The carbon dioxide collected, which is from the carboxyl carbon atom of the acetic acid, is trapped in the gas-handling system and removed for mass analysis (minimum yield 90%). Flask J is cooled to room temperature, with the waterbath in place, and 5 ml. of .5 k f carbonate-free sodium hydroxide introduced through L . The reactor is resealed and the water-bath temperature raised to 100'; the system is then swept for 40 minutes. The carbon dioxide collected, which is from the methyl carbon atom of the acetic acid, is trapped in the gas-handling system and removed for mass analysis. (The yielddepends on the sweep time, exerts little effect on the carbon isotope ratio, and can be made equal t o t h a t in the previous step at will, by prolonging the sweep.) As a test for occurrence of cross contamination of the carboxyl and methylene carbon atom samples, C14-carboxyllabeled acetic acid was degraded by the procedure used with C series samples. The carboxyl activity appearing in the methyl carbon sample depended upon the sweep time between collection of those samples and was of the order of 0.0770 for 1-3 hours sweep. Isotope Analyses.-The carbon isotope ratios of the various samples were obtained from measurements with a Consolidated-Nier Isotope-Ratio Mass Spectrometer. T h e carbon ratios of each sample were determined once on each
-1 n5n
that of carbon dioxide obtained from complete combustion of the original malonic acid; R n , that of carbon dioxide obtained from complete combustion of product acetic acid. Any of these R values may be used to compute the mole fraction of CI3, X ,in the particular saniple, through the relation X = R / ( 1 Hj. If it is assumed that the concentration of spccies with carbon skeletons 12-13-13 and 13-12-13 111:iy be rieglected (;.e., insignificant double labeliiig 1, then
-+
k ? / k j = .TS.i/Sc
M
(41'0
LVhere XAis not determined directly, the c:ilcul:ition of the isotope effect may be based on the observed ratios of the effluent carbon dioxitle and the original malonic acid, in which case k,/k3
=
(2-7" - .Yc)/.?.-c
(5)
or on those of acetic acid and rnnlonic acid, i n which case kr/k:, = ( ~ X B Su)/(X.Yn - 2 S i r )
in both these relations the assumption of carbon isotope homogeneity in the three skeletal carbon positions is made. Results The corrected carbon isotope ratios, their averages and computed values for each X, arc collcctetl TABLE I1
Fig. 3.-Reaction
vessel used in acetic acid degradation (letters refer to text).
of at least three introductions of sample carbon dioxide into the manifold of the spectrometer. A triad of ratio determinations was considered satisfactory only if the average deviation from the mean ratio was 2 X 10-8 or less. T w o working standards, one consisting of tank carbon dioxide, the other of carbon dioxide obtained from complete combustion of a small sample of malonic acid, were employed during the course of the work. Frequent cross checks were made among the effluent-gas samples and the working standards. I n order to eliminate the effects of small variations in the response of the mass spectrometer, all measured ion current ratios were expressed in terms of t h a t of the tank carbon dioxide standard. The mass spectrum of each sample was obtained over the range m / q 26 to 48. The sample was rejected if this spectrum deviated appreciably from t h a t of tank carbon dioxide which had been subjected t o the same high vacuum mariipu1,itions. The ratio of ion currents ( m / q 45) /(m/q 44) was corrected for the contribution t o the ion current a t m / q 45 of the ipecics C12O16O17by application of the value -0.000800 to the measured ratio of ion currents after the latter was corrected for incomplete resolution.
Calculations The decarboxylation of liquid malonic acid has been shown t o be first order with respect to the ncid.8.9 Rate constants for the reactions of interest here are defined in equations 1 and 2. I n this paper five corrected isotope ratios, R = (CL30i6)/ (C'20:6), are employed in the evaluation of k 4 / k 3 : Rc, that of the effluent carbon dioxide; RA, that o f the carboxyl carbon atom of the product acetic , of carbon dioxide obtained from the :wid; R A ~that methyl carbon atom of product acetic acid; R I I , ( 8 ) C . N. Ilinshelwood, J . Chcm. Soc., 117, 150 (1920). (0, .I I.askin, T ~ a n s .Siberiun Acud. A g r . Forestry. 6 , 7 (I'J").
CORRECTED T O N CURREKTRATIOSASD CAI.CI:I,ATRD CI3 MOLEFRACTIONS All I< and X vulucs I1:ive been rnultiplietl 1)y I O e Series A IZUU No.
RC
RI,
1
10734
2
10731
3
10734
4
10736
5
1073G
10971 10974 101168 10985 10974 10963 10970 in97 I 10974 10967
Av.
R
s Series I3
R i i n h7t)
ij "
i
8 Av. R
.Y
inm
10735 Z t 2 10621 i 2
rt I'
108.73 f 1
I? C in745 10744 10744 10744 zt 2 10650 2
Ri
1020 1043 1002 1023 15 0902 f 1 5
*
*
Series C Run So.
RC
RA
Rn
0
10724
11047
10813
10
10722
11035
lOS10
11
10722
11037
10834
Sept. 20, 19.54
R.\DIATION CHEMISTRY
in Table 11; the appended errors are average deviations from the mean value. Internal Consistency of the Results.-Procedural deficiencies are frequently reflected in the several X values for a given series of experiments being in improper relationship to each other. Where the acetic acid is burned, as in series A, the X values should be related as 2xB XC = 3 x D (7) for that series, with the X ' s multiplied by lo6 for (10821 convenience, we calculate: (21706 + 8) f 2) = (32327 S), while the value observed is (32328 f 6). Where the acetic acid is degraded, the X values are related as
+
+
*
XA
+
X M
f XC
=3xD
(8)
For the results obtained in the series B experiments we compute: (10902 =t 15) 4- (10741 f 2 ) (10630 f 2) = (32273 f l5), the observed value being (32328 f 6). This discrepancy is believed to be due largely to the erratic results for carboxyl carbon in the acetic acid degradation. The data from the series C experiments are much more consistent; we calculate: (10919 + 5) (10723 f 3) (10609 + 2) = (32251 f G ) , and observed (32259 f 15). Average Values for k4/k3.-Average values of 100[(k4/k3) - 11 calculated with equations 5 and 6 are given in Table 111; the results from our series A are compared with those of Lindsay, Bourns and Thode2 in which dry combustion procedures were employed. The mean value obtained from the
+
+
+
TABLE I11
AVERAGE\'ALLIES
OF
100[(k4/k3) - I] CATLULATED ON Two B 4SES H A C basis
COYb??ls eq .)
L, B and T Series A
2 87 2 92
* 0 46
+
[RADIATION1,ARORA
0 4
rORY A N D
e q fi
2 20 i 0 21 2 90 =k .13
OF
ALIPHATIC ALCOHOLS
465 1
former is 2.91 f 0.0s; that from the latter is 2.53 f 0.27. The most accurate results for the isotope effect should come from the series C experiments. In Table IV are given values of 100[(k4/k3) - 11 calculated with equation 4. The mean value obtained is 2.92 0.05. If the calculation is based on average X values, a procedure which emphasizes the experimental error in the degradation, a value 2.92 + 0.07 is found.
*
TABLE IV VALUES FOR
100[(k4/kj)
- 11 CALCULATED B Y En. 4 FOR SERIESC
Run No.
xc
XA
9 10 11
IO610 10608 10608
10926 10914 10917
100[(k,/kd)
Av.
2.98 2.88 2.91 2.92
- 11
f 0.03 f .03 f .03 f .05
Discussion Bigeleisenll has concluded that k4/k3 is temperature independent, which requires that its value for the CI3case a t hand be 1.0198. The mean of all results obtained a t 138" bv Lindsay, Bourns and Thode2 is 1.0253 + 0.0027, carbon isotope homogeneity of the malonic acid being demonstrated. Our results a t 140°, with the effect of carbon isotope inhomogeneity eliminated through the use of acetic acid degradation, are k4/k3 = 1.0292 + 0.0007. It is interestin? to note that an interpretation of related experiments in quinoline medium'? leads to a prediction of k 4 'k3 = 1.02S-l f 0.0012 a t 140'. Acknowledgments.-This research was supported by the A. E. C. We are indebted to Professor Robert F. Nystrom for many helpful suggestions, and to Mrs. R. IV.Hill for the mass spectronieter analyses. (11) J. Bigeleicen. J. P h y s . Chein.. 66, 823 (19.52). (12) P. E. Yankwich and R. I,. Belf CzHj > CHI in the alcohols studied. The mechanism of formation of some of the products has been discussed
Introduction Knowledge of the influeTce of structural and f u n c t i o d factors on the radiolytic behavior of organic compounds is necessary both for the developmerit of a theoretical basis for the radiation (1) Much of the work presented here is from the University of California Radiation Laboratory Report IJCRL-1378 (June 1951) by W.R. McDonell (declassified Ph D. thesis) (2) Presented in part a t the Lob Angrles Meeting of t h e American Chemical Society (March 1853).
istry of organic compounds and the uses of radiation chemistry in organic synthesis. Few systematic studies of the radiation chemistry of generic groups of Pure compounds have been made. L i d and Bardwel13 using a-particles, studied the saturated hydrocarbon gases methane to butane, and Honig and Sheppard4using both a-particles and deuterons (3) S. C. Lind and D. C. Bardwell, THIS JOURNAL, 48, 2335 (1926). (4) K. E. Honig and C. W. Sheppard, J. Phys. Chcm., 60, 119 (1948).
