1614
NOTEB
It was originally hoped that the fluorine-fluorine coupling constants which are a function of the geometry of the chain would indicate micelle formation. However, no changes in the coupling constants were observed as a function of concentration in aqueous solution. It is of interest to note that Ja-B is somewhat less in acetone solution than in aqueous solution. Acknowledgments. The nmr spectra were acquired by B. J. Nist. This work was performed in part under contract with the Office of Naval Research.
Confirmation of the Inertness of Sulfur Hexafluoride toward Attack by
CF3 and CH3 Radicals I
'40 2250
2710
160
2270 27SO
I
I
180SCALE FOR 'y,O 240 2290 SCALE FOR Ci,O 2350 2750SCALEFOR @,OZelO
260
2310 2830
280 2390 2650
Chemical ehifi,Hr Figure 1. Relationship between concentration and '9.F chemioal shift of a,p-, and y-fluorine atoms of heptafluorobutyric acid in aqueous solution. Trifluoroacetic acid used as external standard.
In pure heptafluorobutyric acid the chemical shift at 26" for the a-fluorine atoms with respect to trifluoroacetic acid as an external standard was 2459 Hz. For the 0- and y-fluorine atoms the chemical shifts were 2863 and 281 Hz, respectively. A t the cmc the chemical shift for the a-fluorine atoms was, therefore, 187 HZ less than in pure heptafluorobutyric acid. Corresponding values for the p- and y-fluorines were 89 and 129 Hz. This behavior suggests that the a- and y-fluorines protect the p-fluorines from solvent effect of water. Pure heptafluorobutyric acid probably consists almost entirely of undissociated molecules polymerized through hydrogen bonding, In dilute solutions below the cmc the acid is largely dissociated into ions as shown by a freezing point depression of 3.3" for a 1 rn solution.¶ Since the change in charge of the carboxyl group upon dissociation should influence chemical shifts of the fluorine atoms, it is not surprising that the change in chemical shift upon dissolving in water is greatest for the a-fluorines. The nonlinearity of the curves in Figure 1 at high concentrations may be related both to a change in the fraction of acid ionized and to a change in structure of the solutions. Somewhere between 2.5 and 100 mol yo the acid becomes a continuous body rather than existing as micelles. Acetone solutions of heptafluorobutyric acid showed only small and equal changes in the chemical shifts of the different fluorine atoms. These were about equal to those expected from bulk susceptibility and general solvent effects. The Journal of Physical Chemistry
by Henry F. LeFevre, Jayavant D. Kale, and Richard B. Timmons Department of Chemistru, The Catholic Unitersity of America, Washington, D . C. (Received IL'ouember 8.5, 1 9 6 8 )
Sulfur hexafluoride has been used extensively as an added inert gas in chemical kinetic studies. For example, it has been employed as an inert energy-transfer agent in the decomposition of inorganic oxideslP2 and in the deactivation of chemically activated species.3 Recently the inertness of SF, with respect to chemical reactivity has been q~estioned,~ with the authors reporting the rate constant for attack of methyl radicals on SF, to yield CHHF and SF5. From their observed reaction rate these workers have concluded that SF, should not be employed as an inert energy transfer agent in the presence of organic free radicals above 140'. We attempted to measure the kinetics of the reaction of CF3 radicals with SF, over a wider temperature range than the 30' interval studied in the CH3 plus SF, reaction.4 The reactivities of CH3 and CF3 radicals are roughly comparable, with CF3 usually demonstrating a lower activation energy for H-atom abstraction from hydrocarbons.6 However, it is possible that the reaction of CF3 with SF, might proceed via a very polar transition state and thus exhibit a higher activation energy than the corresponding CHa reactions. In our experiments we were looking for reactivity via the formation of CF4 resulting from an (1) M . Volpe and H . 9. Johnston, J . Amer. Chem. Soc., 78, 3903 (1956). (2) D.J. Wilson and H . S. Johnston, ibid., 75, 5763 (1953). (3) G. H. Kohlmaier and B. S. Rabinovitch, J. Chem. Phys., 38, 1709 (1963). (4) L. Batt and F. R. Cruickshank, J. Phys. Chem., 70, 723 (1966). (5) Cf. "Tables of Bimolecular Gas Reactions," National Standard Reference Data Series, National Bureau of Standards, Washington, D. C., 1967.
NOTES
1615
abstraction reaction by CFB radicals. We wish to report that no trace of CF4 formation could be detected up to temperatures of 365'.
Experimental Section Reactions were carried out in a conventional highvacuum apparatus. The reaction section had a total volume of 600 cc and contained an all-glass gas circulation pump a.nd a cylindrical quartz reaction cell of volume 124 cc. The reaction cell was placed in an electrically heated furnace. The temperature of the furnace could be controlled to a t least f 2 ' and showed a temperature profile of less than 1' across the reaction cell. The photolysis of hexafluoracetone (HFA) was employed as the source of CF3 radicals. The HFA was obtained from Allied Chemical and the gas was freed of C2F, and CF3H impurities by fractional distillation a t -150'. The SF, was obtained from the Matheson Co. and had a stated minimum purity of 99%. This gas contained a small impurity of CF4 which was reduced (but not completely eliminated) by fractional distillation a t -186'. Photolysis of the HFA was achieved using a General Electric UA-2 medium pressure mercury lamp. Analysis of reaction products was carried out using a dual column-dual detector gas chromatograph. Flame ionization and thermal conductivity detectors were employed. V i a analysis of synthetic mixtures it was demonstrated that 2-m, 0.25-in. 0.d. silica gel columns could effectively separate CF4 and CzF,. Under the conditions we employed, CZF, was not completely resolved from SF,. However, SF, gives no response on the flame ionization detector; therefore CzF, could be determined quantitatively. Using synthetic mixtures of CF4, CzF,, and SF,, we observed that it was impossible to separate completely the CF, and CzFa from the SF, via fractional distillation. Apparently some of the CF4 and CZF, remains trapped in the SF, matrix. I n view of this difficulty, we simply condensed the entire reaction mixture at the end of a run and collected the noncondensable gas at - 196'. This noncondensable gas was shown to be entirely CO via mass spectrometric analysis. The condensable reaction mixture was then transferred to a loop trap of the gas chromatographic gas inlet system and subsequently injected onto the silica gel column. Results and Discussion From the proposed mechanism of Batt and Cruickshank4for the CHBplus SF, reaction we had anticipated the following mechanism for the CFa reactions HFA
+ h~
42CFa
2CFa + CzF,
CF3
+ SF,
4
CF,
+ CO + SFS
(1)
followed possibly by further CF4 production via the disproportionation reaction of CF3 radicals with SFs. Our initial experiments carried out a t 150' revealed no CF4 formation. Therefore, we increased the reaction temperature to 365O, which is approximately the limiting temperature for the thermal stability of HFA. However, even a t this higher temperature, we were unable to detect the formation of any CF4 in this reaction mixture. From our gas chromatographic calibration curves, we determined that the minimum quantity of CF4 which we could detect was 0.025 pmol. In a typical experiment a t 365', we observed 20 Fmol of CZF, formed during a photolysis time of 7200 see for an SF, conmol/cc. centration of 1.5 X Since we did not observe any CF, production, we are not able to determine the rate constant for reaction 3. However, from the observed CzF, formation we are able to calculate a maximum rate constant and from this estimate a minimum activation energy of this reaction, From the relationship
R C F ~ / R C ~=F ~ ~ [' ~S F , ] / ~ Z ' ' ~ and using the value of 0.025 pmol (the minimum detectable amount) for total CF4 and 2.3 X 10la cc mol-' sec-l for k2,6we calculate a value for k3 equal to 2 X lo4 cc mol-' see-l. If we assume a value for the preexponential factor A3 of 10" cc mol-' sec-l, we calculate a value for E3 of 20 kcal/mol. I n all probability A3 is greater than 10" and, since we observed no CF4 formation, we emphasize that the above value of EQrepresents an absolute minimum. For example, a more reasonable value for A3 of l O I 3 cc mo1-I sec-l gives an activation energy of 25 kcal/mol. In view of the lack of CF, formation, we carried out several photolytic experiments with acetone as a source for CH3 radicals. Even a t temperatures of 365' we were unable to detect CH3F even though our detection limit for CHaF was considerably more sensitive than CF4, as CH3F gives a good response on the flame ionization detector. We are unable to account for the CHaF formation observed by Batt and Cruickshank4 other than to suggest that it may have resulted from a surface-catalyzed reaction.
Conclusion In view of these findings we feel the suggestion that SF, be used as an energy-transfer agent only below 140' for reaction systems containing organic radicals is unjustified. Actually, we have demonstrated that SF, is impervious to CF3 and CH3 radical attack up t o temperatures of 365'. Therefore, we feel that SF, can safely be employed as an inert gas a t least to temperatures of 365'. The extreme inertness of SF, remains somewhat of a mystery, especially when one considers
(2)
(3)
(8) P.B. Ayscough, J. Chem. Phys., 3 3 , 1566 (1955). Volume 78,Number 6 M a y 1969
1616 that the alkyl radical abstraction reactions from SF, are exothermic. Acknowledgment. Acknowledgment is made to the donors of the Petroleum Research Fund, administered by the American Chemical Society, for support of this research.
NOTES Table I: Effect of Electron Acceptors upon Singlet- and TriplebExcited T M P D Produced on Irradiating 10-2 T M P D in Benzene (a) Fluorescence Quenching.
Ill0
Nz0
SFa
0 0 2
Air
0.0095
0.064
0.065
0.42
(b) Effect of Additives on Absorption and Decay a t 565 i 5 mp (Pulse: 5 to 7 krads) (i) Chloracetic Acid
Reactions of Radiation-Excited States of
Concn. of additive, mol L-1
Tetramethyl-p-phenylenediaminewith
...
Electron Acceptors
4.66 x 1.88 x 2.84 x 5.76 x 9.40 x 1.58 X
b y D. Greatorex, T. J. Kemp, and J. P. Roberts School of Molecular Sciences, University of Warwick, Coventry, United Kingdom (Received July 18, 1 9 6 8 )
10-4
10-3 10-3 10-3 10-3
10-6 k ,
sec-1
10-4~e
2.51 1.72 1.92 2.16 2.15 2.21 ~ .
2
3.07 f 0 . 2 0.208 0.173 0.192 0.131 0.260 ~ 0.122
(ii) Benzyl Acetate
Tetramethyl-p-phenylenediamine(TMPD) engages 5.05 X 10-8 2.ig 1.90 in several different kinds of interaction with electron 8.16 x 10-3 2.05 2.07 1.65 X 2.05 2.84 acceptors depending upon the' particular acceptor in2.97 X 10-2 1.45 3.35 volved and the electronic state of the TMPD molecule. The charge-transfer (C-T) and electron-transfer reAll solutions saturated; lorefers to Nz-flushed solution. actions between ground-state TMPD and ?r-acceptors in nonpolar and polar media are well documented.' In nonpolar and polar rigid media, TMPD undergoes identical conditions (k,/kr for ClCH2C02Hand benzyl two-quantum ionization via the triplet state but this acetate are 8.2 and 3.0 M-I, respectively), indicating does not depend on the proximity of an acceptor precursors of the solute singlet state, e.g., excited molecule.2J Donor-acceptor collision is important, singlet benzene, to be relatively unreactive toward however, in fluid solutions; for example, the fluorescence these acceptors. However, the triplet-state absorption spectrum of TMPD in a hydrocarbon solvent at 20' is of TMPD at 565 mp in benzene produced by pulse progressively weakened by systematic addition of radiolysis is affected by three of these acceptors in a-methylnaphthalene and a new emission band appears quite different ways: (i) 0.14M NzO eradicates the at longer wavelength which is assigned to a transient spectrum;6 (ii) benzyl acetate in moderate concenC-T complex on the basis of its lack of structure and trations (5X 10-aM to 3 X 10-2M) only partly its variation with temperature, viscosity, and incident reduces the absorption and barely affects its kinetic light intensity.a decay; (iii) addition of chloracetic acid to only Flash photolysis of these solutions produces triplet 4.7 X lo4 M concentration reduces the intensity of the state a-methylnaphthalene but no triplet TMPD or spectral absorption by 30%, leaves it unchanged in TMPD*+and the sequence of steps remains o b s ~ u r e . ~ character, but increases its half-life from 2.3 to 33 psec, Analogous experiments with dimethylnaphthylamine indicating conversion to a spectrally identical species of (donor) and m-dimethylphthalate (acceptor) indicate greater longevity. Collected data are given in Table I. that C-T emission is appreciable only in nonpolar Addition of either benzyl acetate or chloracetic acid solvents but that in polar solvents it is absent and the produces a new, very intense but short-lived absorption donor-acceptor interaction is one of electron transfer to at 435 mp (Ge = 1.3 =t 0.5 X lo6,kl = 7.5 f 0.4 X lo6 produce radi~al-ions.~ We report here that the radio-excited fluorescence of (1) R. Foster and T. J. Thompson, Trans. Faraday SOC.,59, 296 TMPD in benzene solution is strongly quenched by (1963). (2) K. D. Cadogan and A. 0. Albrecht, J . Phys. Chem., 7 2 , 929 electron acceptors including NZO, SF6,COz, chloracetic (1968). acid, and benzyl acetate (Table I and Figures 1 and (3) N. Yamamoto, Y. Nakato, and H. Tsubomura, Bull. Chem. 2). k,/kt for chloracetic acid and benzyl acetate are, SOC.J a p . , 4 0 , 451 (1967). (4) M. Koizumi and H. Yamashita, 2. Phys. Chem. (Frankfurt), 5 7 , respectively, 244 M-1 (4') and 75 M-1 (9'). Quench103 (1968). ing of fluorescence of 2,5-diphenyloxazole in benzene by (5) T. J. Kemp, J. P. Roberts, G. A. Salmon, and G. F. Thompson, these electron acceptors is very much less marked under J. Phys. Chem., 7 2 , 1464 (1968). 0
The Journal of Phzlaical Chemistry
