Oct. 5 , 1962
RELATIVE SELECTIVITY OF CF,
concentration, Fig. 3 . I t is seen t h a t over a wide concentration range the points fall on a straight line. The intercept a t zero concentration gives the portion of the reaction which is catalyzed by the hydronium ions and the water (the uncatalyzed reaction). A mechanism where the association reaction in equilibrium 2 becomes rate determining a t lower pH values, agrees with the observed data of general acid catalysis. The fact that such a proton-trans-
[COSTRIBUTION FROM
RADICALS I N ADDITION
REACTIONS
3661
fer reaction is expected to occur in a much shorter time suggests that a t lower p H values the rate of the fading reaction is determined by reaction 4 and t h a t this latter reaction is being catalyzed by the buffer . Acknowledgments.-The author wishes t o express his sincere thanks t o Dr. S. D. Bailey, Chief of this Laboratory, who helped make this work possible and to John Soma and Dr. J. n‘einstein who helped out on some of the experimental work.
THE DEPARTMENT OF CHEMISTRY, STATEUNIVERSITYCOLLEGEOF SYRACUSE UNIVERSITY, SYRACUSE 10, XEW Y O R K ]
Addition of CF, Radicals to Aromatic Hydrocarbons.
FORESTRY AT
The Relative Selectivity of CF,,
BY A. P. STEFANI AND -11.SZWARC RECEIVED FEBRUARY 2, 1962 The relative rate constants ( k 2 ) of CF3 addition to a series of unsubstituted aromatic hydrocarbons were determined. I t was shown t h a t a linear relation exists between log ( k s / n ) and the atom localization energy of the most reactive center ( i f the hydrocarbon, n being the number of such centers. This indicates the formation of an incipient C-CFa bond in t h e transition state, i.e., in spite of the strongly electrophilic character of CF3 such a transition state resembles a u-complex and not a *-complex. The linear plot of log(ke)cpatwsus log(k2)CHa has a slope 0.8for the series of investigated hydrocarbons, i.e., benzene, naphthalene, anthracene, etc. This slope is taken as a measure of the intrinsic sclectivity of CF3 radical; its value, lower t h a n unity, shows that the intrinsic reactivity of CF8 is larger than t h a t of CH3. The results obtained in this series of unsubstituted aromatic hydrocarbons are compared with those obtained in the series ethylene, propylene and isobutene. T h e latter reflect the high electrophilic nature of the CFSradical and must not be used in determining its intrinsic selectivity. T h e problem of x - and u-complexes formed in the addition of a radical to a suitable substrate is discussed. It is suggested t h a t a *-complex is a charge-transfer complex, and its formation does not require activation energy. The formation of u complex does require activation energy, this species being the classical adduct radical in which a new u-bond is formed. Conditions favoring the stability of *-complexes are considered. The significance of such complexes is discussed with reference to the process of formation of g-complexes.
A method allowing one t o determine the relative rate constant of CF3 radical addition t o a series of suitable substrates was described in a preceding paper.‘ The radicals were generated by photolysis of hexafluoroazomethane in iso-octane, their reaction with solvent yielding CF3H CFa
+ iso-octane -+- CFIH + iso-octyl radical
(1)
The addition of a suitable substrate t o the solution leads to the reaction CFs
+ substrate -+
CF3.substrate
(2)
which competes with (1) for CF3 radicals. The experimental conditions were such that all the CF3 radicals which escaped “cage” recombination underwent reaction 1 or 2, and none interacted with the radicals formed in the system. Analysis of the products allows calculation of the ratio of the rate constants k 2 / ; k l , and since kl remains constant the data provide information about the relative reactivities of the substrates toward CFa radical addition. The results obtained for a series of non-substituted aromatic hydrocarbons are reported in this paper. Experimental The experimental technique developed for these studies was described fully in the preceding paper’ to which the . high-pressure mercury reader is referred for details. % lamp (General Electric, XH-6) was used for the irradiation, and since Pyrex glass was employed in the experiments, all light corresponding to X i , S o r . . 81, 1185 I 1 939). ( I S ! G. A Russell. i b i < i ,80, 4 Y S i (195S). (19) P I . ..\bell and L H. Piette, F i f t h I n t e r n . S y m p on Free Radicals 11961)
!
J
1 I -complex
______ no
V 6
polar stuctures
- w i t h polar
structures
-complex Fig. 5 .
resonance (e.s.r.) spectra of radicals produced by the addition of Br atoms to some olefins a t - 8 O O . The structure of the spectra proves t h a t the Br atom adds to the middle of the C=C bond indicating, in our opinion, the formation of a a-complex. I n our interpretation, the low temperature and low C-Br bond dissociation energy prevents (or slows down) the process .rr-complex
+u-complex
and hence the a-complex accumulates in the system. Moreover, since the recombination of a-complexes is probably slower than t h a t of ordinary radicals ( i e . ) u-complexes) , their stationary concentration is higher, providing convenient conditions for e.s.r. studies. The formation of a-complexes resulting from the addition of 0 atoms to olefins is discussed by Cvetanovic.'o H e assumes t h a t such an addition requires a substantial activation energy (see Fig. 6) and t h a t the resulting a-complex is separated from the ultimate u-complex by a lower potential energy barrier. Thus, the formation of the a-complex is, in his opinion, the rate determining step in the u-complex formation.
t;
V
G -Complex Fig. 6.
Li-e raise two objections to this representation : (1) There is no valid reason to expect a high potential energy barrier in a-coinplex forniatioii. (2) If such a barrier does exist, then i t is superfluous to consider a lower hump separating the a- and u-complexes. The system would acquire the (20) R. J. Cvetanovic, C a n J . Chein., 38, 1678 ( 1 9 6 0 ) ; see also J . C h e w P h y s . , SO, 19 (19.59): a n d G. Boocock and R J Cvetanovic. Cail. J . Chew., 39, 2436 (1961).
3666
MAXT. ROGERS AXU
necessary excess of energy in passing the first hump, and conceptually the second hump would only be of significance if a mechanism existed to dissipate the internal energy of the complex before i t reached its final u-state. I n the radical or free atom addition i t is doubtful whether such a mechanism does exist. One point, however, needs further discussion. If both a - and u-complexes are feasible, does the transition state resemble a a- or a u-complex? The transition state under consideration is the one described by the hump separating the x - and u-complexes and not by a hump preceding the a complex (if such a hump exists a t all). Our studies of the addition of CF3 radicals to unsubstituted aromatic hydrocarbons indicate t h a t the transition state for this reaction resembles a u-complex, i.e. even for this strongly electrophilic radical an incipient C-CF3 bond is formed in the transition state. The addition of 0-atoms may follow a different course, and, if this is the case, we suggest that the di-radical nature of 0-atoms, and not their electrophilic character, may be responsible for such behavior. The following point needs, however, further discussion. If both a- and u-complexes are feasible, does the transition state resemble a a- or a a-complex? The transition state under consideration is the one described by the hump separating the a - and u-complexes and not by a hump preceding the a-complex (if such a hump exists a t all). The addition of CF3 radicals to benzene and some of its derivatives was studied recently by Whittle, rt aLZ1and their findings suggest that a covalent C-C bond is formed in this process. The results of our studies permit us to go further and claim that the incipient C-C bond is formed in the (21) S. W. Chailes and E. U’hittle, Tuans. F a r a d a y Soc., 66, 794 ( 1 ~ 1 c i O ) ; S. W. Chailes, J. T. Pearson a n d E . n’hiffle. ibid., 57, 1336 (1961).
[COSTRIBCTIOS FROX K E D Z I E CHEXICAL
JOHN
D. GRAHAM
Vol. 84
transition state since the respective rate constants per reactive center were found to be related to the respective localization energies. This means t h a t even for this strongly electrophilic radical the transition state resembles a u- and not a a-complex. The addition of 0-atoms may follow a different course, and, if this is the case, we suggest that the diradical nature of 0-atoms, and not their electrophilic character, may be responsible for such a behavior. I n conclusion we wish to emphasize that the electrophilic nature of a radical is reflected in its response to the presence of electron-donating or electron-withdrawing substituents in the substrate. Such a response is observed whether the addition leads directly to a u-complex or proceeds via a a-complex. The electrophilic character of a radical might favor the formation of a a-complex, although such a complex need not necessarily be formed in every addition involving an electrophilic radical. Whenever a a- and a-complex are formed in the same process, the formation of the former cannot be the rate determining step in the formation of the latter. The transition state of the addition of a strongly electrophilic CF3 radical t o aromatic or olefinic substrates resembles a a-complex. Relatively stable a-complexes probably are formed with extremely electrophilic radicals or free atoms. They are favored by reaction9 leading to relatively weak a-bonds in the final adduct radical and by low temperature. The adducts observed by Abell and Piette are probably a-complexes. IVe wish to acknowledge the financial support of this work by Wright Air Development Division, Contract XF33(GlG)iGG:! and by the National Science Foundation. \Ye also wish to thank Dr. F. R . M a y o for reading the manuscript and for his constructive suggestions.
LABORATORY, h’lICHIGAN
STATE UNIVERSITY,
EASTLAXSING, htICHIGAS]
High-resolution Fluorine Magnetic Resonance Spectra of Some Perfluoroalkyl Derivatives of Sulfur Hexafluoride1 BY
hIAX
T. ROGERS .4ND
JOHN
D.
GR.iHAh12~3
RECEIVED MARCH19, 1962 T h c high-resolution fluorine magnetic resonance spectra of a group of perfluoroalkyl deri\ratives of sulfur licxailuorid been analyzed and t h e chemical shifts and spin-spin coupling constants determined. T h e fluorine chemical shifts are to vary with the electronegativity of the substituents, Fluorine-fluorine coupling constants are observed between nuclci separated by three, four and five bonds, but the coupling constants d o not vary in a coherent manner with a change in tl1c number of bonds separating interacting nuclei. Our results are iri agreement with the idea t h a t coupling between fluorine nuclei separated by three or four single bonds is due t o both through-space and through-bond interactions b u t t h a t the i n portance of the through-bond interaction falls off rapidly with number of intervening bonds and the coupling is largel\. a through-space interaction when the nuclei are separated by five single bonds. T h e very small coupling constants observed for interaction betvcen nuclei separated by three bonds in t h e CFs-CF2 and -CF2--CF2- fragments w ~ u l dstill appear t o be anomalies possibly t o be explained as resulting from rotatioiid averaging of traits and gauche coupling constants of 01,posite signs.
Introduction Althougll ~~~~i~~~ lllaglletic reSOllaIlce for a number of fluorocarbon derivatives have been (1) This rusearch was sugportcd by a grant froin t h e .‘itoniic Eilcrgy Comniissiun.
studied, much less systematic data for FI9cheiiiical shifts and spin-spin coupling constants exist than ( 2 ) Sterling Chemistry Laboratory, Yale University, S e w H a v e n , Connecticut. [,1) T a k e n in part i r < ~ ttnh e 1’11.U. t h e gan State University, 1 Y G l .
