526
SOTES
TABLE I1 CHEMICAL SHIFTSO B SUBSTITCTED ANILISICMIODIDES Compound
($L8&+f
N(CHa)s
C-CHs
1.11
2.14
C-H
-2.66
Br
0.88
-2, very broad
ence in magnetic anisotropy of the methyl and bromo groups. Therefore it is very unlikely that in both compounds non-equivalent methyl groups on the same nitrogen atom could have the same magnetic environment. Although radiochemical studies have not been reported for o-bromo-K-trimethylanilinium iodide, if the supposed non-equivalence of the methyl groups on nitrogen is due to the steric effect of an ortho substituent as postulated,6 a bromo group should be as effective as a methyl group in causing this non-equivalence. Conclusion With the possible exception of o-nitroaniline, it has not been possible to demonstrate hindrance to rotation about the carbon-nitrogen bond in any of the aromatic amines which were studied. The possibility of another explanation for the radiochemical results with o-methylK-trimethylanilinium iodide should be considered. Experimental The proton resonance spectra were obtained with a Varian Associates nuclear magnetic resonance spectrometer operating a t 56.44 Mc./sec. The spectra were calibrated by audiofrequency modulation of the transmitter which determined peak positions to an accuracy of 0.01 p.p.m. The thermostated probe assembly has been described previously.6 All compounds except those described below were obtained commercially. Methylation of obromoaniline gave the N,N-dimethyl derivative.? Replacement of a nitro group in o-dinitrobenzene by dimethylamine produced o-nitro-N,S-dimethylaniline.8 Methylation of o-methyl and obromo-N,N-dimethylaniline with methyl iodide gave the corresponding X-trimethyl derivatives (6) S. Brownstein, Can. J . Chem., 87, 1119 (1959). (7) K. Auwers, Ber., 45,2530 (1907). (8) P. Friedlaender, Monatsh. Chem., 19, 627 (1898). (9) J. von Braun, Ber., 49, 1101 (1916).
VOI.
The absorbance of 2.00 and 3.00 mM solutions of copper(I1) perchlorate was determined as a function of added lithium or hydrogen bromide over the range 1.5 to 7 mM in a medium of 1.00 M lithium perchlorate in 90% ethanol. Preliminary measurements were done on a Gary Model 11 recording spectrophotometer; the values used in calculations were obtained with a Beckman Model DU spectrophotometer and stoppered, 1.000-em. silica cells. Reagent grade chemicals were utilized throughout and standard analytical methods used for determining concentrations. The calculation method used was that of AIcConnell and Davidson4 in the form advocated by Data were collected at 290, 296, and 300 mp spanning the band for a run in hydrobromic acid, and at the peak value for a run in lithium bromide. Table I shows the average formation constants calculated from least squares fitting of slopes to the straight lines obtained in the McConnell-Davidson plots, together with the uncertainties in the constants due to scatter of points in the individual runs. The best value for K1 in the medium, 90% ethanol and molar lithium perchlorate, may be taken as 10.0 f 0.8. The indicated precision, based on run to run variability, is in gocd agreement with that derived by assuming a plausible uncertainty (=!=0.001unit) in the absorbance measurements, and suggests that the trend in K , with wave length is apparent rather than real. Our results are in reasonable agreement with the constants reported by Farrington (K1 = 2.1) and Sasanen (K1 between 3 and 4) at the same ionic strength in water, allowing for the change in the dielectric constant of the solvent. An alternative method of calculation proposed by Newton and Arcand6 also was used, as this method would reveal the presence of higher complexes if any were to be found in these solutions. Good agreement on the value of KI was obtained which further supports the assumption that over the concentration range studied, only the first complex is formed. TABLE I FORMATION CONSTANT OF CuHr * IN 90% ETHANOL;CONSTANT ENVIRONYEXT 1.00 M LiClO4 IONIC Wave length,
T H E FORMATION COXSTANT OF THE CuBrf ION I N 90% ETHXSOL BY JOHN C. BARNESA N D DAVIDS. HUME Department of Chemistry and Laboratory far A’zlclear Science. Massachusetts Instztute of Technclogy, Cambridge 60, Nasu. Received August 16, 1068
I n connection with a study of complex formation between copper(I1) and bromide ions in various solvents we have had occasion to examine the formation of the CuBrf ion in 90% ethanol. Formation constants for this complex in aqueous media have been reported by Farringtonl and Nasanen2 from spectrophotometric measurements on the absorption band which appears at 283 mp when small additions of bromide are made t o copper(I1) perchlorate solutions. Kosower, Martin, and Meloche3 found the band maximum shifted to 310 mp in absolute ethanol, and in 90% ethanol we have found it to be centered at 296 mp. (1) P. S. Farrington, J . Am. Chem. Soc., 74, 966 (1952). (2) R. WWnen, Acta Chem. Scand., 4, 816 (1960). (3) E. M. Kosower, R. L. Martin, and V. W. Meloche. J . A m . Chem, Sac. ‘19, 1.509 (1957j1
67
mp
290 230 300
HBr data
10.51 f 0.04 10.14 f .03 9.17 f .06
LiBr data
.......... 10.00 f 0.03
.
......
Acknowledgment.-This work was supported in part through funds provided by the U. S. Atomic Energy Commission under Contract hT(30-1)-905. (4) H. McConnell and N. Davidson, zbid., 72, 3164 (1950). (5) W. D. Bale, E. W. Davies, and C. B.Monk, Trans. Faraday Soc., 52, 816 (1956).
(6) T. W Newton and G 1 4 Aroand, J A m . Chem Soc , 75, 2449 (1953).
T H E EFFECT OF CRESOLS ON MALONIC ACID BY LOUISWATTSCLARK Department o f Chemistry, Western Caralzna College, Culloiuhee. .V. C. Received August .B3 1068
Kinetic studies have been carried out in the past on the decarboxylation of malonic acid in more than 50 different non-aqueous solvents selected from a wide range of different homologous series. These include aromatic amines, aromatic and aliphatic alcohols, manocarboxylic
XOTES
Feb., 1963 acids, aromatic nitro compounds, ethers, thiols, aldehydes, phosphates, sulfoxides, and polyhydroxy compounds. The present paper reports the results of kinetic studies carried out in this Laboratory on the decarboxylation of malonic acid in three additional solvents, o-cresol, mcresol, and p-cresol. A comparison is made of the reaction in other hydroxylic solvents. Experimental Reagents .-( 1) The malonic acid used was analytical reagent grade, 100.07, assay, m.p. 134'. To ensure perfect dryness it was stored in a debiccator containing sulfuric acid (2) The cresols used in this investigation mere highest purity chemicals. Immediately before the beginning of each decarboxylation experiment each sample of each liquid was distilled at atmospheric pressure directly into the dried reaction flask. Apparatus and Technique.-The details of the apparatus and technique used in this research have been described previously.2 I n these experiments a sample of malonic acid weighing 0.1806 g. (corresponding to 40.0 ml. of COZ a t STP-weights based upon the actual molar volume of COZ,22,263 ml. a t STP) was weighed into a fragile glass capsule weighing approximately 0.1 g. and blown from 7-mm. soft glass tubing. Approximately 50 ml. of solvent (saturated with dry COS gas) was placed in the 100-ml. 3-neck, standard taper reaction flask immersed in the thermostat oil bath. The temperature of the oil bath was controlled t o within 1 0 . 0 5 " or better, and was measured by means of a thermometer graduated in tenths of a degree and calibrated by the U. S. Bureau of Standards. Steam point corrections and stem corrections were applied carefully to ensure the utmost precision in temperature readings.
Results Decarboxylation experiments were carried out in each solvent a t three different temperatures over a 20' temperature range. The experiments were performed two or three times at each temperature. In the case of each of the solvents used in this investigation log (Vm V,) was a linear function of time over about the first 70y0of the reaction. The average rate constants, calculated in the usual manner from the slopes of the experimental logarithmic plots, are brought together in Table I. The parameters of the Eyring equation, based upon the data in Table I, are shown in Table TI, along with corresponding data for the decarboxylation of malonic acid in several other acidic type solvents reported previously. Data for the reaction in several alkanols are shown in Table 111. TABLE I APPARESTFIRST-ORDER RATECONSTANTS FOR THE DECARBOXYLATION OF MALONIC ACIDIN THE CRESOLS Solvent
o-Cresol
m-Cresol
p-Cresol
Temp., (cor.)
x 104 (seo. -1)
Av. dev.
140.63 149.33 160.08 140.63 149.33 160.OS 139.33 149.33 160.08
3.87 7.24 15.16 3.84 8.73 23.4 4.10 IO. 23 24.4
rtro.01 f .02 =I=.04 =I=.02 f .02 z?c . 1 f .01 f .05 f .1
OC.
(1) (a) G. Fraenkrl, It. L. Belford, and P. E. Yankwich, J . Am. Chem. Soc., 76, 15 (1954); (b) L, W. Clark, J. Phys. Chem., 60, 825 (1956); (0) 60, 1340 (1956); (d) 60, 1583 (1956); (e) 61, 1009 (1956); (f) 61, 1575 (1957); (9)62, 79 (1958); (h) 62,368 (1958); (1) 62, 500 (1958); (j) 62, 1468 (1958): (k) 64,41 (19130); (1) 64,508(1960); (m) 64,677 (1960); (n) 64,892 (IQBD): (0)66,
(2)
2271 (1901), (p) 66, 125 (1962).
L, \ \ . Clark, zbndL,EO, 1150 (1956).
527
TABLE I1 KINETICDATAFOR DECARBOXYLATION OF MALONIC ACID I N THE CRESOLS A N D IN OTHERACIDIC TYPESOLVENTS AH* (kcel./ mole)
Solvent
Molten malonic acids Propionic acid'k Hexanoic acid'" rn-Cresol Valeric acid'" P-Mercaptopropionic acidIk p-Cresol Nitrobenzene'h PhenollL Decanoic acid'" o-Cresol o-NitrotolueneIb
35.8 33.6 32.5 32.3 32.2 30.3 29.8 28.1 27.3 26.6 24.2 23.5
*
AS (e.u./ mole)
AF*iuo (koa]./ mole)
+11.9
30.9 31.1 31.2 31 . O 31.2 31.1 30.8 31 .O 30.9 31.1 30.9 31.0
+ 6.1 -I- 3 . 2 + 3.2 + 2.4 - 1.9 - 2.4 - 7.2 - 8.9 -11.0 -16.5 -17.9
TABLE 111 KINETICDATAFOR
DECARBOXYLATION OF MALONIC ACID SEVERAL ALKANOLS
THE
IN
Solvent
AH* (koa14 mole)
Benzyl alcohol1m n-Butyl alcohol'n Isoamyl %-Hexyl alcohol'n 2-Ethy lhexanol-l In Diisobutylcarhinolln Cyclohexanollm
29.7 27.2 27.1 26.0 24.8 24.8 23.0
AS* (e.u./ mole)
+ 1.0 -
4.4 - 4.5 - 7.6 -10.4 -10.7 -15.0
AF*ira (kcal./ mole)
29.3 29.0 29.0 29,l 29.0 29.1 29.1
Discussion of Results Abundant proof of the validity of the proposed mechanism for the decarboxylation of malonic acidla is furnished by the data shown in Table 11. Malonic acid is an electrophilic agent, the polarized carbonyl carbon atom having a deficiency of electrons. The solvent is a nucleophilic agent, the hydroxyl oxygen atom furnishing electrons. The malonic acid attacks the solvent molecule, forming an intermediate complex which suff ers cleavage. The rate-determining step is the formation of the transition state. The ease of formation of the complex is determined by two factors: (1) the effective negative charge on the nucleophilic atom, and (2) the accessibility of the nucleophilic atom. The activation energy or enthalpy diminishes as the effective negative charge on the nucleophilic atom i n ~ r e a s e s . ~ The entropy of activation decreases as the steric hindrance increases.6 It is interesting t o note in Table I1 that the largest enthalpy as well as the largest entropy of reaction obtains in the case of the decomposition of molten malonic acid. I n every case the addition of solvent lowers both the AH" and the AS*of the reaction. A very close similarity will be noticed in the parameters for the reaction in the two solvents hexanoic acid and m-cresol, indicating a close correspondence in the properties of these two solvents. As the strength of the acid decreases (in other words as it becomes more basic) the enthalpy of reaction decreases. The +I effect of the methyl group in p-cresol increases the electron density on the phenolic oyygen. (3) L. W. Clark, ibid., 67, 138 (1963). (4) K. J. Laidler, "Chemical Xinetlcs," McGraw-Hill Book do., Inc., New York, N. Y., 19.50, p, 138. (5) S. Glaeatone, K. J. Laidler, and H. Eyring, "The Theory of Rate Prooease~,"McGraw-Hill Book Cot, Inm, New Yatki XsY., i941, pi 22.
NOTES
528
- 15
-10
-5
0
+5
+10
A S * (e.u./mole).
Fig. 1.-Enthalpy-entropy plot for the decarboxylation of malonic acid alone and in solution: I, monocarboxylic acids, nitro compounds, and phenols; 11, alcohols.
This leads to a lowering of AH* on going from m-cresol to p-cresol. When the methyl group is in the o-position the positive inductive effect is further increased as shown by the low value of AH* for the reaction in ocresol. The ortho effect is clearly revealed by the very large decrease in the entropy of activation on going from p-cresol to o-cresol. An enthalpy-entropy plot for the reactions listed in Table I1 is shown in line I of Fig. 1. Line I1 of Fig. 1 is a plot of the parameters for the reactions in the seven alcohols listed in Table 111. Let us consider first of all line I of Fig. 1. It will be observed that we are dealing here with the reaction of an acid in acidic solvents comprising several homologous types. The points on this line represent results obtained from different sources over a period of several years. We have a total of twelve points covering a wide range of enthalpies and entropies of reaction. The fact that these points form a straight line indicates that the reaction takes place in each solvent by the same mechanism.6 Malonic acid itself and all the solvents included on line I of Fig. 1 with the exception of the nitro compounds contain the hydroxyl group. The best straight line that can be drawn through these points has a slope of 407'K., which corresponds to 134'. This is the so-called isokinetic temperature, that is, the temperature a t which the rate of reaction is the same in all solventslBand it is also precisely the melting point of malonic acid. This means that a t the melting point of malonic acid its rate of reaction is not affected by the presence of acidic solvents. The intercept on the zero axis of line I of Fig. 1 is 30.9 kcal. This is the free energy of the reaction in any solvent a t the isokinetic temperature. I n column 4 of Table I1 the experimental free energies of the reaction in all the solvents a t the melting point of malonic acid are shown and are seen to agree with the theoretical value within plus or minus one or two tenths of a kcal./ mole. Returning to line 11 of Fig. 1 we see that this line is parallel to line I and lies below line I. Since it is parallel to line I it has the same slope. It often is noted in kinetic studies that a change in type of solvent results in the formation of a new line parallel to the original line.6 It means in this case that the decarboxylation of malonic acid in alcohols has the same mechanism and the same isokinetic temperature as it does in acids, phenols, and nitro compounds, but that its rate of reaction a t the isokinetic temperature is greater in alco( 6 ) J.
E. Leffler, J. Or& Chem., 20,
1202 (1985).
Vol. 67
hols than in acids. The intercept of line I1 of Fig. 1 on the zero axis is 29.1 kcal./mole. Again, the experimental values for the free energy of activation of the reaction a t the isokinetic temperature in the case of the seven alcohols listed agree within plus or minus one or two tenths of a kcal./mole with the theoretical, as shown in column 4 of Table 111. It is not believed that this coincidence between the melting point and the isokinetic temperature in the case of the decarboxylation of malonic acid is purely fortuitous but rather that it represents a general and fundamental kinetic relationship. Support for this belief has been furnished by the results of a long and intensive study of the decarboxylation of oxanilic acid in the molten state and in s ~ l u t i o n . ~The enthalpyentropy plot of the reaction in twelve solvents including aliphatic and aromatic ethers and aromatic amines over a wide range of enthalpies and entropies of activation yielded a straight line having a slope of 423'K. or 150°, which corresponds to the melting point of oxanilic acid. It also is considered of some significance that the enthalpy and entropy of activation for the decomposition of molten malonic acid are both larger than those for the reaction in any solvent. A similar circumstance was noted in the study of the oxanilic acid reaction in the presence of non-ionizing solvents. In the case of solvents showing increases in the enthalpy and entropy of activation it was indicated that probably ionization had occurred and we were no longer dealing with the un-ionized acid. The fact that the data for nitrobenzene and o-nitrotoluene fall on line I of Fig. 1 indicates that the nucleophilicities of the aromatic nitro compounds are comparable to those of the monocarboxylic acids and phenols. The ortho effect again is strikingly demonstrated by the low value of AS* for the reaction in o-nitrotoluene. Acknowledgments.-The support of this research by the National Science Foundation, Washington, D. C., is gratefully acknowledged. (7) L. W. Clark, J. Phys. Chem., 66, 1843 (1962).
T H E RELAXATION SPECTRUM OF NICKEL-TRIGLYCISE COMPLEXES' BY JEFFREY I. STEINFELD A N D GORDON G. HAMYES Department of Chemistry and Research Laboratory of Electronics, Massachusetts Institute of Technology, Cambridge, M aasaohusetts Received August $8, 1068
Kinetic studies of the successive steps of formation of multi-ligand complexes of Co(I1) and Ni(I1) with imidazole, glycine, and diglycine have been carried out using the temperature-jump method.2 The measured rate constants could be successfully correlated with a mechanism involving the formation of an ionpair between the metal species and the ligand followed by dissociation of a water molecule from the inner hydration shell of the metal ion. The equilibrium constant, KO, for ion-pair formation was calculated using a Debye-Huckel interaction potential. The first-order rate constant, ICo, describing the water dissociation was calculated by dividing the experi(I) This work was supported in part by the U. S. Army Signal Corps, Air Force, Office of Scientific Research, and Offioe of Naval Research and in part b y the National Institutes of Health (RG 7803). (2) G.G. Hammes and J. I. Steinfeld, J . A m . Chem. Soc., 84, 4639 (1962).
