3351
COMMUNICATIONS TO THE EDITOR tain striking effects on the decay rates of radicals in glassy matrices, the viscosity of perdeuterated 3MP was determined. Four measurements on a 1-g sample in a 0.55 cm in dialmeter viscometer tube were compared with results on an identical sample of the nondeuterated compound (Figure 3). The dotted line in Figure 3 is from measuremients4 with the more accurate 0.9-cm tube. The results indicate that there is no major difference between the viscosities of 3 M p - h ~and 3MPd14 glasses. The viscosity of trans-l,3-dimethylcyclohexaneglass a t 112°K is 8.3 :X lo7P.
terminations of the viscosities of MCHx in the region 135-205°K and of 3MP in the region 95-120°K, using a commercial rotating cylinder viscometer. The former values are above the glass softening temperature, and the latter span it, so that comparison with extrapolated values of the present work is not meaningful. It appears, however, that their value for pure 3MP a t 95°K (lo5 P) is about one order of magnitude higher than that found in our earlier work4at this temperature, although our investigations4 of other highly viscous glasses have agreed well with earlier results extending from lower vi~cosities.~
Related Measurements von Salis and Labharts have recently reported de-
(6) G. A. von Salis and H. Labhart, J . Phys. Chem., 72, 762 (1968).
C O M M U N I C A T I O N S TO T H E E D I T O R
Intramolecular Elimination Reactions in the Photolysis of Fluoroaldehydes
Sir: Recently, perfluoroalkyl radicals have been generated by the photolysis of various fluoroaldehydes,' and data have been obtained for the abstraction of the aldehydic hydrogen atom by the radical Rf
+ HCORr
RfH
+ CORf
The aldehyde has also been used2 as a radical source to investigate the removal of hydrogen from various substrate molecules, i.e.
Ftf + R H A- RtH
+R
This technique is satisfactory if there is no other source of R f H in the reaction system. I n a study of the photolysis of CF&OCFs-HCOCzFs mixtures it beca,me apparent that although the fluoroform formation could be adequately accounted for by the reactions CF3
+ HCOCzFs A CF3H + COCzFs 2CF3 -% CzFa
the pentafluoroethane formation was not similarly expressed by the reactions CzFs
+ ECCOCZF~--% CzFsH + COCZFS 6
2CzFs +C4Fio Our evidence for this conclusion was that although a plot of the ratio R c F ~ H / R ' /us. ~ c aldehyde ~F~ concentra-
tion gave, within experimental error, zero intercept, a ~ ~ ~ l o a markcorresponding plot for R C 2 F b ~ / R 1 / a yielded edly positive intercept. Analysis of the data published for the HCOC2Fh system yielded essentially the same conclusion. When cyclopropanecarboxaldehyde is p h ~ t o l y z e d , ~ ~ ~ its decomposition has been shown to involve production of free radicals and also the formation of propylene by an intramolecular elimination reaction (8), HCOa
A 8
HCO 3. CO
+
a
CHFCH-CH~
It seemed likely that such an intramolecular elimination reaction was also contributing to pentafluoroethane formation when the fluoroaldehyde was photolyzed; Le., the primary processes were
+ CZF5 -% CO + CzFsH
HCOCZFS
HCO
We have photolyzed under similar conditions the aldehyde alone and also aldehyde-nitric oxide mixtures. In the latter cases perfluorobutane formation was completely inhibited although extensive pentafluoroethane formation occurred, the yield decreasing by only -75% (1) G. 0.Pritchard, G. H. Miller, and J. K. Foote, Can. J . Chem., 40, 1830 (1962). (2) G.0. Pritohard and J. K. Foote, J . Phys. Chem., 68, 1016 (1964). (3) G. Greig and J. C. J. Thynne, Trans. Faraday Soc., 63, 1369 (1967). (4) J. J. I. Overwater, H. J. Herman, and H. Cerfontain, Rec. Trav. Chim., 83, 637 (1964). Volume 78, Number 9
September 1968
COMMUNICATIONS TO THE EDITOR
3352 at 400°K compared with the experiments performed in the absence of the inhibitor. We therefore conclude that reaction 10 contributes substantially to pentafluoroethane formation. The fact that only such a relatively small degree of CzFEH inhibition was observed, particularly with regard to the fact that, in the unscavenged experiments there will be substantial contributions to the CzFs radical concentration by the decarbonylation of the COCzFs radical produced in reaction 5 , suggests that photodecomposition of the aldehyde by reactions 9 and 10 must be comparable. Similar examination of the photolyses of HCOCFI and HCOC3F7 showed that the following intramolecular elimination reactions occur appreciably in these systems.
+ CFsH HCOC3F7 +CO + CgF7H HCOCF3 +CO
We therefore conclude that fluoroaldehydes are not suitable for use as photochemical sources of fluoroalkyl radicals in connection with hydrogen atom abstraction reactions and that the kinetic data reported's2 for such reactions are likely to be significantly in error.
RECEIVED MAY10, 1968
A Reply to "Intramolecular Elimination Reactions in the Photolysis of Fluoroaldehydes"
Sir: The point made by Morris and Thynne in the preceding communication' is valid, and an examinat'ion of the data on CzFsCHOphotolysis2 at about room temperature is in accord with their observation. The intramolecular elimination of R f H (where Rr = CF3, CzFs, or C3Fd RrCHO hv +RrH CO (1) us. the cleavage into radicals RrCHO
+
+ hv *Rt + CHO
(2)
is presumably temperature independent, so that the formation of R f H via reaction 1 will be relatively more important than formation by abstraction Rf
+ RtCHO +R f H + CO + Rt
+ Rr +Rr,
(4) will therefore lead to low values (-4 kcal mol-') of EB,assuming that Ed = 0. This is actually the case, as shown by the tabulation of the data4 and the value of E3 = 8.2 kcal mo1-I obtained by Dodd and Smitha in The Journal of Physical Chemistry
+ R H +R f H + R
(5) using the aldehydes as sources are significantly in error. The data on H2 and D2' compare favorably with similar data obtained using perfluoroketones and perfluoroazo compounds as radical s o u r ~ e s , ~and ~ ~the J collected data on CH4 and isobutane are similarly selfc o n s i ~ t e n t . At ~ ~ ~elevated ~~ temperatures the main mode of decomposition of the aldehyde is via reaction 3 and a similar chain propagating step involving the atom or radical produced in reaction 5.1° A reinvestigation of the photodissociation of these aldehydes is warranted, particularly with regard to the possible difference in multiplicities of the electronic states which lead to the two dissociative modes.
(1) E.R.Morris and J. C. J. Thynne, J . Phys. Chem., 72,3351 (1968). ( 2 ) G. 0.Pritchard, G. H. Miller, and J. K. Foote, Can. J . Chem., 40, 1830 (1962). (3) Reaction 3 assumes that the fluoroacyl radicals decompose, which may not be completely true at room temperature. However, the relative instability of CFsCO has been noted; J. C. Amphlett and E. Whittle, Trans. Faraday SOC.,6 3 , 80 (1967). (4) G. 0.Pritchard, J. R. Dacey, W. C. Kent, and C. R. Simonds, Can. J . Chem., 4 4 , 171 (1966). (5) R. E. Dodd and J. W. Smith, J . Chem. SOC.,1465 (1957); this is further verified by the value of E3 = 9.7 kcal mol-' obtained for CeFs CFaCHO using perfluoroazoethane as the radical 8ource.4 (6) Our data2 are not extensive enough a t high temperatures to make this test. (7) G. 0. Pritchard and J. K. Foote, J . Phys. Chem., 68, 1016 (1964). (8) With D2, RR*Dwas, in any case, determined directly.
+
(9) G. 0 . Pritchard and G. H. Miller, J . Chem. Phys., 35, 1135
(1961). (10) I n the C3F7 Da experiments we observed 60 to 100% HD formation based on CIFTDproduced.?
+
DEPARTMENT OF CHEMISTRY OF CALIFORNIA UNIVERSITY SANTA BARBARA, CALIFORNIA93106
G. 0. PRITCHARD
M. J. P E R O N A
RECEIVED JUNE11, 1968
(3)
at lower temperature^.^ The Arrhenius plots used2 to yield values of E3 - ' / 2 E 4 Rr
Rr
Acknowledgment. We thank the National Science E. R. MORRIS J. C. J. THYNNE Foundation for financial assistance.
CHEMISTRY DEPARTMENT EDINBURGH UNIVERSITY EDINBVRGH 9, SCOTLAND
+
CFaCHO photolysis is now substantiated. These authors made a correction for CFsH formation via mode 1, but found it to be unimportant above 150"; and a plot of their results, RCF~H/R~/'C~F~ us. aldehyde pressure, obtained at 209", passes through the origin.0 It does not necessarily follow therefore that the data reported on the abstraction reactions with hydrogen and hydrocarbons
Charged Square-Well Model for Ionic Solutions'
Sir: The simplest physical model for electrolyte solutions seems to be the charged hard-sphere model (1) Grateful acknowledgment is made for the support of this re search by the Office of Saline Water, U. S. Department of the Interior
