NOTES
180
Vol. 65
and solubility data6ssdemonstrates that for activity coefficients of inert uncharged molecules, electrostatic interackions between solute and solvent are second-order effects in most cases. Even with iona, electrostatic effects are only part of the story. In a forthcoming paper on the activity coefficients of ions, it will be shown that it is only with the smallest of ions t,hat electrostatic effects predominantly govern the activity coefficients. In the eq., log f = AMP,k M is a measure of the difference in internal pressure between solvent and reference solvent and P is the parachor. From the derivatim of this eq., P is used solely as a measure of the molecular volume of solute and is comput'ed from the available atomic and structural increments.' For a reaction in which reactants and transition state are all non-polar, such as many free radical reactions, the Bronsted eq. combined with the McGowan eq. leads to the eq. log k/ko = AM(PA PB - P*). From studies on the variation of rates with external pressure, it has been found that the volume of activation, AV*, is nearly zero for non-polar reactions.839 Thus P A PB = P* andlogk/lco = 0. Electrostatic effects and kxlP terms are independent variables so that when electrostatic effects are present
k E t O H = 5.5. In 80% ethanol, Av* = -23 ml./mole and the values in water and ethanol must P would be (2.0)(0.0120) be similar. The ~ M term (23) = 0.55. This represents 10% of the total log kH&/kEtOH and so to a first approximation it is proper to interpret Y values as due to electrostatic interactions between solvent and solutes.
(7) 0. R. Quayle, Chcm. Revs., 63, 484 (1953). (8) J. Buchanan and S. D. Hamann. Trans. Faraday Soc., 49, 1425 (1953); H.G.David and S. D. Hamann, ibdd., 60, 1188 (1954). (9) K..I. Laidler and C. Chen, ibid., 64, 1026 (1958); K.J. Laidler, Faraday SOC.Disc.., 22, 88 (1956).
decarboxylation experiment each sample of each solvent was distilled a t atmospheric pressure directly into the dried reaction flask. Apparatus and Technique.-The kinetic experiments were conducted in a constant-temperature oil-bath by meaauring the volume of Cot evolved at constant pressure, aa described
THE DECARBOXYLATION OF OXAMIC ACID I N ANILINE AND IN O-TOLUIDINE BY LOUISWAws CLARE' Department of Chemistry, Saint Mary a/ the Plains College, Dodge City, Kansas Received July 6 , 1960
In the decarboxylation of malonic acid in nonionizing, polar solvents, the rate-determining step appears to be the formation of a transition complex involving coordination between the electrophilic, polarized carbonyl carbon atom of the un-ionized diacid and an unshared pair of electrons on a nucleophilic atom of the solvent molecule.2 In kinetic studies of this reaction in more than fifty non-aqueous solvents inductive and steric effects in harmony with this proposed mechanism invariably were observed.s It appears highly probable, also, that the decarboxylation of oxalic acid in polar solvents log k/ko = ~ Mterms P + electrostatic terms involves the same sort of mechanism as does that of It is thus the approximate cancelling of the k M P terms malonic acid.4 that explazns why log k/ko depends primarily on Although oxamic acid, apparently, never has electrostatic e.flecects. However, it is in these cases been the subject of any kind of kinetic study, the where electrostatic effects are important that analogy between it and oxalic acid suggested that AV* differs significantly from zer0,8.9 and the it, too, should be capable of undergoing decarboxyP is not exact. Values lation in polar solvents by essentially the same cancelling of the ~ M terms of AV* have about &30 ml./mole as an upper mechanism as that of oxalic acid and malonic acid. limit. For .water-hydrocarbons, AM = 0.0130, Preliminary experiments in this Laboratory re~ vealed that smooth decarboxylation of oxamic acid and this is the largest k~ value yet found ( k = 0.0120 for wa,ter-ethanol). Thus an upper limit to does take place a t moderate temperatures in polar the contribution of kxlP terms to log k/ko is ~ M A V * solvents. In order to ascertain whether or not this = (2.0)(0.0130)(30) = 0.8.lo decarboxylation follows the same mechanism as The eq. log k/Ao = mY has been found to that of oxalic acid and malonic acid kinetic studies correlate the rates of solvolysis reactions with of the reaction were carried out in two aromatic solvent." The term Y depended on the solvent amines, aniline and o-toluidine. Results of this and was called the "ionizing power" of the solvent. investigation are reported herein. This interpretation seemed anomalous because in Experimental the many examples studied, log k/ko was as much Reagents.-(1) The oxamic acid used in this investigation dependent om changes in log f of the non-ionic was analytical reagent grade, 100.0% assay. Two methods reactant as on changes in log f of the ionic or polar of assay mere employed: volumetric and gasometric. In transition state. By calculating the contribution the volumetric analysis a weighed sample wm digested for three hours with excess standard sodium hydroxide until all of the ~ M t,erms P to log k/ka, the contribution of ammonia waa driven off, the solution cooled and the excess the e1ectrost)atic effect can be determined by base determined by titration w t h standard hydrochloric difference. One example will demonstrate the acid. In the gasometric analysis, each decarboxylation ex eriment, when properly carried out, yielded the theoretiapproach. carvolume of carbon dioxide at STP. (2) The solvents were For the solvolysis of t-butyl chloride, log AH&/ reagent grade chemicals. Before the beginning of each
+
+
(10) The factor of about 2.0 arises because Vb = 0.49P and Vm = 0.42P where 'vb and Vm are the molecular volumes a t the b.p. and m.p. (J. C. McGowan, Roc. trau. chim., '76, 199 (1956)). (11) E. Grunwald and S. Winstein, J . A m . Chem. SOC.,7 0 , 846 (1948); S. Winstein, E. Grunwald and H. W. Jones. ibid.. 73, 2700 (1951); A. H. Fainberg and S. Winstein. ibid., '78, 2770 (1956).and '79, 5937 (1957): S. Winstein, A. H. Fainberg and E. Grunwald, ibid.. '79. 4146 (1967).
(1) Western Carolina College, Cnllawhee, N. C. (2) G. Fraenkel, R. L. Belfard and P. E. Yankwich, J . Am. Chsm. Soe. 16, 15 (1954). (3) L. W . Clark, THIEJOURNAL, 64, 917 (1960).and previous papers in this seriea. (4) L. W.Clark, ibrd., 61,699 (1957).
NOTES
Jan., 1961 in a previous paper.’ In each experiment a 0.1585-g. sample of oxamic acid (the amount required to produce 40.0 ml. of C02 at STP on complete reaction) was introduced in the usual manner into the reaction flask containing a weighed sample of Eolvent saturated with dry COZgas.
Results In some experiments at the beginning of this study some peculiar experimental difficulties were experienced: (1) the experiment would halt before the entire nmoun t of COzwas collected, and (2) the evolved COz would then start to be resorbed, the entraining liquid. rising up into the buret, finally coming to complete rest a t zero volume. In seeking for an explanation of this behavior some white crystals were observed adhering to the inside walls of the condenser. These crystals were collected and analyzed and found to be oxamic acid. Apparently the formamide produced concomitantly with the COP in the decarboxylation experiment tended to combine with the COz by a reverse reaction to give the original compound. It was found that, by employing a rather large volume of solvent and using a gentle stirring action instead of a more vigorous one, this reverse reaction could be circumvented almost completely until practically all the original oxamic acid sample had decomposed. (Upon standing, however, resorption of the evolved COPwould begin to take place, and after several hours the entraining liquid would again rise to the :zero line in the buret. Crystals of oxamic acid, corresponding to the stoichiometric weight used originally, would then be found adhering to the condenser walls.) Inasmuch as oxamic acid apparently had not been investigated previously from a standpoint of kinetics a careful study was made of the effect of concentration on the reaction velocity constant. In this study seven decarboxylation experiments were carried out in aniline at 160.54’ corr., using, respectively, 50, 71, 75, 90, 120, 130 and 135 g. of solvent. The results of this study indicated that there was no detectable change in the specific reaction velocity constant at this temperature with change in concentration. However, inasmuch as better results were obtained when larger volumes of solvent were used, the practice was adopted of running future experiments in 110-135 g. of solvent. The rate of decarboxylation of oxamic acid in aniline and in o-toluidine was then measured a t four different temperatures over a range of about 20”. The plot of log ( V , - Vt) was linear over nearly the entire course of the experiment indicating that the reaction is first order. 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 11. Discussion The effective positive charges on the two carbonyl carbon atoms of oxalic acid are partially E effect) neutralized by a shift of electrons (a from the two hydroxyl oxygen atoms. In oxamic acid these charges are partially neutralized by a E effect of one hydroxyl oxygen atom and one amide ( 5 ) L. W . Clark i b i d , 60, 1160 (1956).
+
+
181 TABLE I
APPARENTFIRST-ORDER RATE CONSTANTS FOR THE DECARBOXYLATION OF OXAMIC ACID IN ANILINEAND IN 0TOLUIDINE Solvent
Aniline
+Toluidine
Av.
Temp.,
No. of
k X 104, nec.-1
deviation
142.73 146.31 151.90 160.50 137.36 142.44 149.30 156.22
3 2 3 7 2 3 3 2
0.859 1.66 4.55 20.00 1.26 2.86 8.25 23.73
10.003 i .02 i .02 f .04 f .005 f .02 f .02 == ! .03
‘C. cor.
runs
TABLE 11 KINETIC DATA FOR THE DECARBOXYLATION OF OXAMICACID IN ANILINEAND IN +TOLUIDINE Solvent
AH*, kcal.
Aniline 0-Toluidine
59.7 53.7
AS, e.u.
+68.0 +57.1
k d x 104 sec. -1
3.17 8.66
nitrogen atom. Since nitrogen is less electronegative than oxygen, the nitrogen will release electrons more readily than will the oxygen. This means that the carbonyl carbon atom of oxalic acid will have a higher effective positive charge than will that of oxamic acid. If, as analogy suggests, the decarboxylation of oxamic acid proceeds in polar solvents by the same mechanism as that which has been established for oxalic acid-an electrophilic carbonyl carbon atom of the acid coordinating with an unshared pair of electrons on a nucleophilic atom of the solvent molecule to form a transition complex-then, in a given solvent, it would be expected that the AH* of the reaction would be lower for the acid whose carbonyl carbon atom has the higher effective positive charge-namely, oxalic acid.6 In Table I11 are shown the values of the Eyring parameters for the decarboxylation of oxalic acid, oxamic acid, and malonic acid in aniline solution. It will be observed (lines 1 and 2 of Table 111) that, in aniline, the AH* for the decarboxylation of oxalic acid is actually 19 kcal. less than that for the decarboxylation of oxamic acid. These data are strong evidence for the validity of the postulated mechanism. TABLE I11 KINETICDATAFOR THE DECARBOXYLATION OF OXALICACID, OXAMICACIDAND MALONIC ACIDIN ANILINESOLUTION AX*, Reactant
Oxalic acid4 Oxamic acid Malonic acid’
kcal.
40.3 59.7 26.9
AS*, e.u.
+16.2 +68.0
- 4.5
kxo’ X 104, 8ec. -1
1.45 0.52 50.0
Further confirmation for the proposal may be obtained by a study of the data in Table 11. If oxamic acid coordinates with a pair of unshared electrons on the nitrogen atom of aniline in a manner similar to that of oxalic acid and malonic acid, then the presence of a methyl group ortho to the amino (6) E. J. Laidler, “Chemical Kinetics,” McGraw-Hill Book Co., Inc.. New York, N. Y..1950. p. 135. (7) L. W. Clark, THISJOURNAL,62, 79 (1958).
182
NOTES
group would be expected to have two effects: (1) it would release electrons due to a positive inductive effect, thus increasing the effective negative charge on the nitrogen atom and giving rise to a decrease in AH*, and, (2) it would sterically hinder the approach of oxamic acid to the nitrogen atom, resulting in a decrease in AS*. In Table I1 (lines 1 and 2 ) it will be observed, in fact, that both AH* and AS* for the decarboxylation of oxamic acid decrease on going from aniline to o-toluidine. At 150” the decarboxylation of oxamic acid takes place about three times as fast in o-toluidine as in aniline in spite of the increased steric hindrance, due to the compensating effect of the lowered activation energy. In view of the fact that all the experimental results obtained are in complete accord with the theoretical predictions, there remains little doubt but that the original assumption is correct, namely, that oxamic acid decomposes in polar solvents by the same mechanism as do oxalic acid and malonic acid. In Table I11 it will be noticed that the AS* for the decarboxylation of oxamic acid in aniline is considerably larger than that for oxalic acid. This indicates that, in all probability, discrete molecules of oxamic acid are involved in the rate-determining, coordinating step, whereas, in the case of oxalic acid, an association cluster or so-called “supermolecule” composed of perhaps three single molecules is involved.* If,as appears to be the case, the mechanism of the decomposition of oxalic acid is similar to that of malonic acid, the fact that, in aniline, AH* for the malonic acid. reaction is lower than that for oxalic acid (see Table 111,lines 1 and 3) indicates that the effective positive charge on the polarized carbonyl carbon atom of malonic acid is greater than it is on that of oxalic acid. This difference in charge between the two acids is probably attributable in great measure to the presence of the methylene group in the malonic acid molecule which tends to prevent the transmission of inductive effects between the two terminal carboxyl group^.^ Oxalic acid and oxamic acid may be regarded as types of a-keto acids, whereas malonic acid may be regarded as an example of a 0-keto acid. On the basis of these results it may be presumed that, in general, other a-keto and 0-keto acids will undergo decarboxylation in polar solvents by the same mechanism as that which has been proposed for oxalic acid, oxarnic acid and malonic acid. If this is correct, it would be anticipated that, in view of the large difference in activation energies, @-keto acids will be more unstable than a-keto acids, that is, they will suj?fer decarboxylation more easily under similar conditions. As a matter of fact this anticipation is found to be correct. For example, in the decarboxylation of oxalacetic acid in the presence of nucleophilic catalysts, pyruvic acid and carbon dioxrde, not formylacetic acid and carbon dioxide, are obtained.10 Oxalacetic acid may be regarded both as an a-keto acid and a @-ketoacid. (8) W . HUckel, “Theoretical Principles of Organic Chemistry,” VoI. 11, Elseviet. PubIishiag Co., New York, N. Y., 1968, P. 329 el scq. (9) R. Q. Bfea ste?, ”Otganic Chemistry.” Second Edition, Prentic6Ball, Ino., Nub ‘York, k,Y , , 1968, p, 691,
Vol. 65
In the decarboxylation reaction it is the p-keto moiety which suffers cleavage first, the a-keto portion proving to be the more stable. Acknowledgments.-( 1) Valuable assistance was rendered by Donald McCoy in carrying out the analyses. (2) This research was supported by the Nat,ional Science Foundation, Washington, D. C. (IO) F. Haumwits, “Bioohemiatry,” John Wiley and Sons, Ino.. New York, N. Y., 1955, p. 54.
CRYSTAL STRUCTURE OF POTASSIUM SILYL BY M. A. RINGAND D. M. RITTER Departmnt of Chemistry, University of Wahshinpton, Seattle 6 , Washington Received July 6 , 1960
Potassium silyl, KSiHa, recent,ly prepared,’ has been examined by X-ray powder diffraction, the first such measurement on a metal derivative of a Group IV hydride. Eight reflections were observed using Cu K a radiation (X 1.5418 k.) on a Philips powder camera of 56.78 mm. effective radius, determined using six NaCl reflections with correction for absorption error. The calibration was necessary because insuflicient back reflection was obtained to apply the Straumanis technique. The structure is cubic of the NaCl type with the K positions at OOO, %%.io, O%%, the silicon positions at OO%, %W,OXO,XxX and a0 = 7.15 f 0.02 A. giving No effort was an X-ray density of 1.29 g. made to locate the hydrogen atoms, owing to the use of so few reflections and the probability that the silyl anions rotate. The observed intensities (listed in Table I) were obtained by graphical integration of the microphotometer curve. The scattering factors of Viervoll and Ogrima for silicon, and for potassium those of Berghuiss were used to obtain the calculated intensities.
%Os,
TABLEI RELATIVE INTENSITIES
hkl
Cslculaded
Observed
111 200 220 311 222 400 420 422
0.07 1 .oo 0.73 .03 .26 .12 .32 .24
0.02 1 .oo 0.78 .02 .30 .09 .33 .16
We are indebted to Mr. Henry Montgomery and to Professor E. C. Lingafelter for their assistance and advice. This research was supported by the United States Air Force through the Air Force Office of Scientific Research of the Air Research and Development Command under contract No. AF (1) M. A. Ring and D. M. Ritter. J . Am. Chem. SOC., 83, in press. ( 2 ) H. t’iervoll and 0. Ogrim, Acta Crust., 2, 277 (1949). (3) J. Berghuis, Ij. M. Haanappel, M. Potters, B. 0. Loopatra, Ci at hlaoGillavry and A, La Vconsndaaf, Auto Cryeta,8 , 478 (1866)d
