Proton Nuclear Magnetic Resonance Effects in ... - ACS Publications

R. E. Glick, and D. F. Kates. J. Phys. Chem. , 1958, 62 (11), pp 1469–1470. DOI: 10.1021/j150569a038. Publication Date: November 1958. ACS Legacy Ar...
0 downloads 0 Views 281KB Size
Nov., 1958

1469

NOTES

TABLE I1 KINETIC DATAFOR THE DECARBOXYLATION OF MALONIC ACID I N METHYLANILINE, ANISOLE, PHENOL AND THIOPHENOL AH AS AF1400 klt00 x los E* A Solvent

(1) (2) (3) (4)

Methylanilinel Anisole Phenol Thiophenol

(cal.)

27,400 31,020 28,110 35 ,140

*

(cal.)

(sec. -1)

1.54 X 4.00 X 2.80 X 8.75 x

10l2 lOl2 10" 1014

Arrhenius and Eyring equations are listed in Table I1 with data for methylanilinel included for comparison. If we examine the data for anisole and methylaniline (lines 1 and 2 of Table 11) we see that H is higher for the reaction in anisole than for that in methylaniline. This we would predict since oxygen is more strongly electronegative than nitrogen and hence the unshared pair of electrons on the oxygen atom are less available than are those on the nitrogen. (It has been shown previously that an increase in the effective negative charge on the nucleophilic atom of the solvent molecule, causing a n increase in the attraction between the two reagents, lowers the enthalpy of activation for this reaction.1) The entropy of activation AS* is slightly larger for anisole than for methylaniline. This could be attributed to the fact that in methylaniline the nitrogen has attached to it a methyl group and a hydrogen atom as well as the phenyl group, whereas in anisole the oxygen has a methyl group and a phenol group but no hydrogen. I n anisole one would expect that the malonic acid would have a greater probability of approaching and coordinating with the unshared electrons than it would in methylaniline. The higher value of AS* in the case of anisole is in accordance with this expectation. The mechanism for the decomposition of malonic acid in anisole appears, therefore, to be essentially the same as it is in methylaniline. At 140" malonic acid decomposes 32 times as fast in methylaniline as it does in anisole as a result of the higher basicity of the amine. Lines 2 and 3 of Table I1 reveals that both AH* and AS* are lower for phenol than for anisole. If phenol remained undissociated one would expect just the opposite of this, since the +I effect of the methyl group in anisole should increase the availability of the electrons on the oxygen atom. Furthermore, a methyl group should offer greater steric hindrance to the approach of the malonic acid than would a hydrogen atom, resulting in a smaller value of AS*. Now of all possible molecular species present in molten phenol near its boiling point, the only one which would have a greater effective negative charge on the nucleophilic atom than anisoig is the phenolate ion. It appears, therefore, that in the reaction in molten phenol the malonic acid coordinates with the phenolate ion. The lowering of AX* is undoubtedly attributable to the fact that extensive association obtains between phenolate ions and undissociated molecules of phenol. I n thiophenol hydrogen-bonding does not obtain and therefore association does not take place. Furthermore, in the absence of base, dissociation into thiophenolate ion cannot take place. Thiols are stronger acids than phenols, and are conse-

*

26 ,600 30,170 27,260 34,300

*

(e.u.1

-

5.33 3.67 8.90 +19.17

*

(cal.)

28 ,800 31,690 30 ,930 26,370

(sec. -1)

440.0 13.5 33.7 19.0

quently weaker bases. One would predict, therefore, that AH* would be larger for thiophenol than for phenol or anisole. The results are entirely in accordance with expectation as shown in Line 4 of Table 11. The higher value of AS* as compared with anisole corresponds with the fact that the sulfur atom is larger than the oxygen atom, and a hydrogen atom is smaller than a methyl group, hence there would be less steric hindrance in the formation of the activated complex. Acknowledgment.-The support of this research by the National Science Foundation, Washington, D. C., is gratefully acknowledged. The thiophenol used in this research was furnished by the Pitt-Consol Chemical Company, Newark, N. J.

PROTON NUCLEAR MAGNETIC RESONANCE EFFECTS IN AROMATIC SOLVENTS BY R. E. GLICKAND D. F. KATES Department of Chemistry, The Pennsylvania State University, Universify Park, Pennsylvania Received June $3,1968

When aromatic substances are used as solvents in nuclear magnetic resonance studies, an anomalous shift in the position of the solute resonance frequency is observed as compared to that predicted by the magnetostatic modela2 This effect has been interpreted as due t o the field-induced magnetic anisotropy of the aromatic 7r-electrons8 and is the intermolecular counterpart of the intramolecular contribution to the chemical shift for hydrogen attached to aromatic rings as postulated by P ~ p l e . ~ Previously, Bothner-By and Glick3 and Reeves and Schneider,6 examining the behavior of various substances in aromatic solvents, concluded, this intermolecular shift being relatively small for acetone and dioxane3 and large for chloroform,a-6that chloroform's hydrogen displayed an acid-like inter(1) This work was supported in part by the Office of Naval Research, Project NR055-328. Reproduction in whole or in part is permitted for any purpose of the United States Government. (2) The magnetostatic model is given by equation 1. Hi = 8iHv (1 - ad (1). For ( l ) , Ha is a reference resonant field (for a proton, Ho is the resonant field for the bare nucleus i n vacuo), si is a specific shielding factor and is a function of the electron environment of the ith nucleus, a is a shape factor, K is the magnetic susceptibility of the media, and Hi is the observed field for the ith nucleus. Using the Lorenz-Lorentz approximation for the molecular cavity and a cylindrical sample cell oriented transversely t o the magnetic field, a has the theoretical value of - 2 n / 3 (-2.09). The variations of proton resonance frequency with solvent have been found to follow equation 1 if a is given the value of -2.60.8 (3) A. A. Bothner-By and R. E. Glick, J . Chem. Phys., 3 6 , 1651 (1957). (4) J. A. Pople, ibid., 84, 1111 (1956). (5) L. W. Reeves and W. G . Schneider, Can. J . Chem., 36, 251 (1957).

1470

NOTES

action with the a-electrons of the aromatic system. Further, the intermolecular effect was observed when chloroform was dissolved in benzenes substituted with electron donating substituents and not observed when electron withdrawing substituents were present. I n an attempt to extend these prior studies to additional haloforms we have examined the proton n.m.r. frequency shifts for various substances dissolved in benzene. Experimental The proton solvent shifts for the substances listed in Table I are the differences in resonant frequency between that for the pure liquid and that of the solute extrapolated to infinite dilution in benzene divided by 40 (solvent shifts in p.p.m.). I n the haloform cases the extrapolation was made by the least square method from points over the range 100 .to 10 volume % a t intervals of 10 volume % and for the ' except remaining materials a t intervals of 20 volume % for the last dilution. All measurements were made on a Varian High Resolution Nuclear Magnetic Resonance spectrometer at 40 mc. using standard sampling and frequency measuring techniques. The various chemicals were obtained as gifts from the Dow Chemical Company or the Shell Chemical ComDanv or were reagent grade chemicals from commercial siurces.

TABLE I 1

Cornpd. in m diln. in benzene

AH,b

p.p.m.

3

2 A.",C

p.p.rn.

AH",d

p.p.rn.

4 Molecular e vol. (25') (cc.)

CHzClz 1.80 0.67 1 . 1 3 64.3 67.0 CHzClBr (D)" 2.05 0.92 1.13 CHzBrz 2.00 1.18 0 . 8 2 70.0 62.3 CHaI 1.88 1.12 0.76 80.9 CHCla 1.95 0 . 7 2 1.23 83.1 CHC1,Br (D)" 2.12 0 . 9 5 1.17 CHCIBrz(D)" 2.32 1.15 1.17 85.4 83.7 CHBra 2.25 1.30 0 . 9 5 Dioxane 0.55' 0.25 .30 85.5 Acetone .30' - .12 .42 73.3 .OO 182 &Butyl peroxide@)" .30 .30 a D, gift from the Dow Chemical Co.; S, gift from the Shell Chemical Co. b Frequency difference between the proton resonance for the solute infinitely dilute in benzene and in the pure liquid divided by 40 (in parts per million). Frequency difference for proton resonance for t-butyl peroxide infinitely dilute in benzene and compounds listed divided by 40; a diamagnetic susceptibility correction. R. Glick and S. J. Ehrenson, to be published. AH AH 6 Halomethane deV4ties from "Products of the Dow Chemical Company, The Dow Chemical Company, Midland, Michi an, 1955. Others from C. P. Hodgman, "Handbook of Ehemistry and Physics," Chemical Rubber Publishing Co., Cleveland, Ohio, 31st Edition, 1949. Molecular volume of benzene is 88.8 cc. Ref. 3.

F. .

-

f

The observed solute proton resonance shift, in parts per million, from that of the pure liquid to that at infinite dilution in benzene ( A H ) are listed in Table I, column 1.

Vol. 62

A correction for the magnetic susceptibility difference between the pure liquid and that of benzene ( A H ' ) is found in column 2, Table I. This is essentially the solvent shift redicted from equation ( 1 ) (a equal to -2.60) with K for %enzenea equal to -0.47 X 10-6 rather than the macroscopically observed value of -0.62 X 10-8. The shift is actually the difference between that found for t-butyl poroxide extrapolated to infinite dilution in benzene and in the compounds listed6 in Table I. The intermolecular aromatic contribution ( A H " ) is listed in column 3. Column 4 contains the molecular volume of the solute. It may be noted that A H is no higher on the average for the haloforms than for the other halomethanes. The protons in the latter substances are a t least three orders of magnitude less acidic than those in the former.' Following Pauling's treatment of axial differences in magnetic susceptibility for aromatic systems,8 the 7-electrons are taken to be equivalent to a conducting wire giving rise to an opposing magnetic field when an external field is applied. Using a free electron model, Waugh and FessendenO evaluated an average effective field in the region external to the aromatic ring, AHerf (in p.p.m.), in terms of equation 2.

For (2) n, e, and m refer to number, charge and mass of the electron, a is the radius of the aromatic ring, and BO is a field distribution function tabulated in terms of the cylindrical coordinates p and z . If this function is averaged over a circular plane of radius 2.5 A. parallel to and a t 8 distance of 2.8 A. from the plane of the aromatic ring the mean field at this plane will be 0.62 p.p.m. lower than that calculated from medium diamagnetic considerations alone. The conditions specified above appear as reasonable mean distance for a solute molecule interstitially spaced between two aromatic rings. If a second ring is considered, an even lower average field would be calculated. An op osing but lesser contribution (due to the essential l/r8 beRavior of equation 2 ) would be found for a position on the periphery of an aromatic ring. Thus, from the AH'' values 1isted.h Table I, together with the quantitative computation, it appears that, If a specific interaction exists between the haloform proton and the 7-electrons as is indicated bv infrared evidence,'O this intermolecular n.m.r. effect need not be diagnostic for such interaction. This deviation from the magnetostatic model may manifest itself as a mean positioning of the solute molecule between aromatic planes; a statistical rather than specific interaction. The controlling feature is most likely a molecular volume relationship between the solute and solvent molecules. We gratefully acknowledge gifts from the Dow Chemical Company and the Shell Chemical Company of the chemicals designated in Table I. (6) R. E. Glick and S. J. Ehrenson, t o be published. (7) J. Hine, J. Am. Chem. Soc., 73, 2438 (1950); J. Hine, A. M. Dowell, Jr., and J. E. Siligley, Jr., ibid., T S , 479 (1956). ( 8 ) L. Pauling, J . Chewa. Phys., 4, 673 (1936). (9) J. S. Waugh and R. W. Fessanden, J . Am. Chem. Soc., 79,846 (1957). (10) C. M. Huggins and J. C. Pirnentel, J. Chem. Phys., 23, 896 (1955).