217 ”F

---

217

COMMUNICATIONS TO THE EDITOR

CH3CF3 undergoes unimolecular decomposition via concerted H F elimination.6 To provide an independent check on the CH3CF3parameters used in our calculations, we have also carried out a parallel study of activated 14CH3CF3produced via radical recombination. The recoil experiments involved 19F(n, 2n) 18F and (15) A fundamental part of the theory of regular solutions is the 3He(n, p) 3H nuclear a c t i v a t i ~ n . ~ ~ Scavengers -~~ inabsence of “solvation” of the molecules of a solution.16 Thus in a cluding HzS, 0 2 , fluorinated olefins, and mixtures of regular solution, specific chemical interactions are considered absent.*e Christian and Tucker have included both physical and these substances assured the efficient removal of chemical interactions under their definition of “solvation.” thermalized recoil atoms and organic radicals. Identi(16) J. H. Hildebrand and R. L. Scott, “Regular Solutions,” Prenticecal results were obtained with these scavengers, except Hall, Englewood Cliffs, N. J., 1962, p 3. that HzS permitted experimental determination of RESEARCH DEPARTMENT AARON X. FLETCHER ‘8F-labeled radicals.jb The 14CH3CF3 experiments CHEMISTRY DIVISION involved co-photolysis of 14CH3COCH3-CF3COCF3 NAVALWEAPONS CENTER mixtures with large excesses of CHaCF3 present as CHINALAKE,CALIFORNIA93555 “bath” species. All the experiments involved radiogas RECEIVED SEPTEMBER 17, 1969 chromatographic a n a l y ~ i s . ~ ~ - ~ ~ The CH3CF3 results and other relevent data are given in Table I.’ The ratio of Ic,(E) values for (CHICFzlsF):(CH2TCF3)was found to be 2200 f 300, The Chemistry of Nuclear Recoil Fluorine-18 which indicates unambiguously that Dib (CH3CFz18F) Atoms. 111. The Average Energy and >> EV’b(CH2TCF3).2r5&*8 Calculated P b ( R j X ) values

the broadlSgl6 definition of “solvation” used by Christian and Tucker allows them to place a label of “solvated” on all molecules in solution, the label does not necessarily mean that “solvation” is always going to play a significant part in the chemistry of the molecules.

Mechanism for F-for-F Substitution in CH3CF3

Sir: Energetic substitution (1) is one of the novel chemical reactions that are uniquely characteristic of high-energy chemical systems. l v 2

X*

+ RjY -+(RjX’) + Y X-for-Y

(1) An important fundamental question about such reactions has to do with the energy ranges in which they occur. Experimental studies of the levels of vibrational excitation for the (R,X’) product species have provided the available information about reaction energy ranges for T-for-H (2) and F-for-F (3) reactions3

+ CH3CF3 -+(CHZTCF3+)+ H 18F+ CH3CF3 -+(CH3CFz”F’) + ”F T*

T-for-H

(2)

F-for-F

(3) Average residual (RjTt) energies from (2) have gen1 eV.4 erally been found to lie within the range 5 We have sought similar information for F-for-F reactions based upon experiments with nuclear recoil IsF atoms. The results from our first experiments, which involved pressure-dependence studies with perfluorocyclanes, indicated extensive unimolecular F~ decomposition of the products from (3) in C - C ~ and (Rjl*F) c-CeF8 arising from remarkably large Pb value^.^,^^,^^ In an attempt t o provide support for these unusual results, further experiments have been carried out involving activation of a single chemical species by both reactions 2 and 3. The primary goal of this work was the direct comparison of experimental unimolecular rate constants for CH3CF2l8F from (3) and CHzTCF3 from (2), with the aim of minimizing errors in the calculated Evib(RjX)values. Activated

*

(1) (a) A. P. Wolf, Advan. Phys. Org. Chem., 2, 201 (1964); (b) R . Wolfgang, Ann. Rev. Phys. Chem., 16, 15 (1965); (e) R. Wolfgang, Accounts Chem. Res., 2, 248 (1969). (2) Syzbols include (*) and (t),translational and vibrational excitation; EVlb(R,X),mean residual energy from (1); ka, unimolecular rate constant; A , S, and E,, Kassel parameters (ref 8b, 8c); ZAAand n A , the binary collision frequency and molecular density at pressure P; Z, the gas compressibility factor; and d = s refers to the fugacity at which half the energined (R,Xt) species undergo collisional stabilization, (3) Center-of-mass X-for-Y reaction energies are related through energy conservation to the sum of reaction enthalpy changes, product internal excitation energies, and displaced atom kinetic energies. Since the latter are unknown, the present results provide minimum estimates for the average F-for-F and T-for-H reaction energies. (4) (a) Available Evlb(R,T) results include: CHsT, 3.5 =$ 1.0 eV (ref 4b, 4c) ; c-CaHaT, c - C ~ H ~ and T , c - C ~ D ~5T ,i 1 eV (ref 4d-4f) ; (b) D. Seewald and R. Wolfgang, J . Chem. Phys., 47, 143 (1967); (c) Y. N. Tang and F. S. Rowland, J . Phys. Chem., 72, 707 (1968); (d) J. K. Lee, B. Musgrave, and F. S.Rowland, J . Amer. Chem. SOC., 81, 3803 (1959); (e) E. K. C. Lee, and F. S. Rowland, ibid., 85, 897 (1963); (f) C. T. Ting and F. 5. Rowland, J . Phys. Chem., 72, 763 (1968) (5) (a) N. J. Parks, C. F. McKnight, and J. W.Root, Chem. Eng. News, Sept 16, 42 (1969); (b) C. F. McKnight and J. W. Root, J . Phys. Chem., 73, 4430 (1969); (c) J. W. Root, ibid., 73, 3174 (1969) ; (d) R. R. Pettijohn, K. A. Krohn, N. J. Parks, C. F. McKnight, and J. W. Root, manuscripts in preparation. (6) (a) 8 . T.V. Benson and G. Haugen, J . Phys. Chem., 69, 3898 (1966); (b) R. D. Giles and E. Whittle Trans. Faraday Soc., 61, 1426 (1965); (0) A. Maccoll, Chem. Rea., 69, 33 (1969); (d) M. J. Berry and G. C. Pimentel, J . Chem. Phys., 49,5190 (1968); (e) D. C. Phillips and A. F. Trotman-Dickenson, J . Chem SOC.,A, 1144 (1968); (f) S.H. Bauer, J . Amer. Chem. Soc., 91, 3688 (1969). (7) Y. N. Tang, T . Smail, and F. S. Rowland, ibid., 91, 2130 (1969). (8) (a) The basic Kassel relation is not applicable t o situations that involve wide “excitation spectra.” We define an average ka asZs8bsc I

in which (ZAA/nA)d--a is calculated directly from the experimental quantity (P/Z)d,# (note 2). (b) N. B. Slater, “Theory of Unimclecular Reactions,” Cornel1 University Press, Ithaca, N. Y., 1969, Chapter VIII, p 164 ff; (c) N. B. Slater, Trans. Rov. SOC.(London), 246, 57 (1963); (d) Fugacity and collision diameter calculations have already been described (ref 6b,8e). Diameters estimated for CHaCFa and CHzFCFa were 4.98 and 5.0 A. (e) J. W. Root, Ph.D. Dissertation, University of Kansas, 1964. Volume 74, Number 1

January 8, 1970

COMMUNICATIONS TO THE EDITOR

218

Table I : Experimental k,o and Calculated Evib(RjX) Values for Energetic Substitution Reactions Product species

Reaction

+

“CH&Fs CH2TCF3 CH3CF218F CH2”FCFs

I4CH3 CF3 T-for-H F-for-F F-for-H

c-C~F,‘~F

F-for-F

CFa’8F

F-for-F

Exptl half-fugacity, atrn

Exptl ka(B),

0.038 f 0,001 0.13 =I= 0.03 280 f 50 7 . 3 Zk 0 . 1

1.32 X 4.6 X 1.0 x 2.3 X

12.0 f 0 . 5

...

18F*

+ CH3CF3 -+(CH2lSFCF3+)+ H F-for-H

(4)

but our results are consistent with the earlier report of extensive decomposition associated with (4) in cyclanes. A concerted mechanism for F-for-F substitution follows. On the basis of the present results, we therefore conclude that the EVib(Rjl8F)values from (3) in CH3CF3, c-C4Fs, and CF4-and also by implication the 10 eV. Although mean reaction energies-exceed “direct” microscopic mechanisms have often been favored for X-for-Y reactions,’ Bunker has demonstrated through an extensive classical trajectory study that T-for-H reactions involve the concerted participation of several atoms in the reactant molecules.12 In agreement with his findings, our results support a “concerted” mechanism for F-for-F reactions, since Evib(Rj18F)values invariably exceed the various C-F bond dissociation energies by factors of 2 or more. F-for-F reactions simply could not deposit residual energies of this magnitude via any process involving only one C-F chemical bond. The concerted participation of neighboring bonds serves to effect at least partial intramolecular distribution of the excess energy.

The Journal of Physical Chemistry

108 lo8 10’2 1010

3.6 X 1O1O

...

were obtained via a modified Kassel treatment,*vg for which several sources of error can be anticipated.sb>l0 Facts which support these calculated values include : (i) the 14CH3CF3 results agree with available kinetic (CH2TCF3) background data for CH3CF3;6r9(ii) the Evib value is consistent with earlier T-for-H result^;^ (iii) the Pib (CH3CF2’8F) value is consistent with our previous Evib(c-C4F718F) value;5a#5b and (iv) thermodynamic evidence favoring large Evib(RISF)values from (3) is provided by the observation of extensive stepwise decomposition of CF3I8F t o :CFI8F. The minimum endoergicity for this process is 9.3 eV.5a--5c,7Unfortunately, no data are available for comparison with our Dib(CH2l8FCF3) value from (4)

sec-1

Calcd aVib(RjX), eV

4.33 f 0.09 4.7 =I= 1.0 11.5 f 2 . 0 6 . 7 rt 1 . 0

1 : E :;: > 9 . 3 f 0.1

Ref

This work This work This work This work 5b

5b, 7

Acknowledgment. The authors express appreciation to Dr. J. R. Martin and Dr. R. J. Mattson of DuPont for gifts of fluorine chemicals; to the Crocker Laboratory staff and the University of Calfornia, Berkeley, Nuclear Reactor staff for cooperation; to the U. S. Atomic Energy Commission for support;13 and to Professor D. L. Bunker, University of California, Irvine, for advance communication of his results. (9) (a) Kassel S parameters within a fluorohydrocarbon series increase with fluorine substitution (ref 5b,9b). e - c s F , H ~ values ~ range from 13.0 t o 21.0 (1: = 0 to z = 4), and C ~ F , H Lvalues ~ range from 9.0 to 18.0 (z = 0 to z = e), corresponding to active participation of all vibrational modes for both fully fluorinated species. For CHaCFs and CH2FCFa log ( A ) = 13.5, and S = 13.5 and 15.0, respectively (ref 6a-6c). Our ‘4CHsCFa experiments support E, = 2.73 f 0.06 eV based upon Do(CHa-CF3) = 4.33 k 0.09 eV (ref 90) in agreement with earlier work (ref 6, 9b). This value was also assumed for CHzFCFa (ref 6a-Bo). (b) F. P. Herbert, J. A. Kerr, and A. F. Trotman-Dickenson, J. Chem. Soc., 5710 (1966); (c) J. W. Coomber and E. Whittle, Trans. Faraday Soe., 63, 1394 (1967). (10) Several sources of error can be anticipated: (i) large error limits were cited for (P/Z)&s for CHaCF218F, because its determination required extrapolation over a 20-fold pressure range. Our measure ments spanned a 500-fold pressure range from 0.025 to 12.5 atm; other experimental errors are negligible in comparison; (ii) errors in the Kassel parameters (note 9) including their possible energy dependence (ref 5b, lob) ; (iii) errors in the strong collision assumption; (iv) errors in our implicit assumption of identically shaped excitation spectra from (2) and (3) in CHsCFs. I n partial support of this assumption available excitation spectra from (2) have been symmetric about zvlb(R,T) for both Kassel and RiceMarcus kinetic models (ref 4c-4f, 1Oc). (b) D. W. Placzek, B. S. Rabinovitch, and G. Z. Whitten, J. Chem. Phys., 43, 4071 (1965); (c) E. K. C. Lee, Ph.D. Diasertation, University of Kansas, 1963. (11) Y. N. Tang and F. S. Rowland, J. Phys. Chem., 71,4576 (1967). (12) D. L. Bunker and M. Pattengill, Chem. Phys. Lett., 4,315 (1969). (13) This research was supported under A.E.C. Contract AT-(11-1)34, agreement 158. (14) Address correspondence to this author.

DEPARTMENT OF CHEMISTRY UNIVERSITY OF CALIFORNIA DAVIS,CALIFORNIA 95616

CHARLES F. MCKNIGHT NORRISJ. PARKS

CROCKER NUCLEAR LABORATORY AND DEPARTMENT OF CHEMISTRY UNIVERSITY OF CALIFORNIA DAVIS,CALIFORNIA95616 RECEIVED OCTOBER 10, 1969

JOHNW. R o o ~ l ~