COMMUNICATIONS TO THE EDITOR
Nov. 20, 1964
sensitizer behavior suggests t h a t this reaction pathway is not available. l 2 In summary, these data suggest t h a t electronically excited states of the pyrylium oxide I1 are not intermediates in the photochemical reaction I + 11, nor are electronically excited states of the indenone oxide I intermediates in the reverse p r o ~ e s s . ' ~Thus by elimination, both reactions probably proceed b y w a y of vibra-
tionally excited ground electronic states formed b y isoenergetic crossing f r o m T I , T I [ ,a n d possibly also SI a n d
5051
k 2 5 " 1 1 = 5 x 10-6 set.-', k2501V(R=CaH6) = 8 x 10-7 set.-', and k z s 0 1 v ( R = ~ )= 0, are in accord with this prediction Acknowledgment.-We are greatly indebted to Dr. T. F. Ziegler for aid in the preparation and counting of radioactive samples
CYANAMID COMPANY CESTRALRESEARCH DIVISION CHEMICAL DEPARTMEST STAMFORD, COSNECTICUT
EDWINF . ULLMAN WM A HENDERSON, JR.
.%MERICAS
RECEIVED JULY24, 1964
s11.14
Similar though less complete data have been obtained for the photointerconversions I11 $ IV. The effects of sensitizers are parallel to those described above, while direct irradiation of 111 b u t not IV causes rearrangement to the pyrone VIII.'b,c,15Arguments similar to those given again point to reaction by way of a vibrationally excited ground state
0 VI11 \ R=H or C,H,
)
Of particular interest is the observation t h a t lowering the temperature reduces the quantum yields for reactions I1 I and IV + 111." An attractive but not unique interpretation of these results may be made in terms of the relative energies of the transition states in the thermal isomerizations and of the vibrationally excited ground states of I1 and IV formed by isoenergetic crossing from the electronically excited states. Thus the energies of the electronically excited states and transition states are probably sufficiently close to each other t h a t collisional deactivation of the vibrationally excited states may compete with rearrangement to the corresponding epoxides. At increased temperatures more successful competition with collisional deactivation should occur since more energetic vibrationally excited ground states would be produced by internal conversion from thermally populated higher vibrational levels of the electronically excited states. This reasoning suggests that successful competition with collisional deactivation should also become more likely on lowering the thermal transition state energies. A rough parallel between the quantum yields and thermal rates of reactions I1 --+ I and IV + I11 would be expected, since the energies of a t least the singlets of I1 and IV are very similar. The observed order of quantum yields a t 7 i ° K . , I1 >> IV (R = C&) > IV(R = H ) , and the thermal fading rates
+.
-
(12) Additional evidence t h a t nonvertical energy transfer does not play a role in sensitized reaction I I1 is found in the absence of energy transfer from sensitizer triplets of energies substantially below the maximum possible triplet energy of I (ETI < E a = 74 kcal ) , By contrast relatively large apparent uphill energy transfers are frequently associated with the nonvertical energy transfer process; cf. ref. 10 and 11. (13) R;arrangement of SII or T I I to SI or TI is energetically forbidden, for E S I L(50 kcal.)5 is a t least 12 kcal. less than ET[ (-68 kcal.) when related to the energy difference of the ground states (41:= 6 kcal.). T h u s Esr > E S Tand ET, > ET?, both b y more than 12 kcal. (14) Since the preparation of this manuscript it has been found t h a t quenching of reac!ion I + I1 approaches a limit (quantum yield reduced by 115%) with increasing naphthalene concentrations. This strongly suggests t h a t 89?% of I 1 is formed directly from SI without the intermediacy of TI. (15) W e are indebted t o Professor P. Yates for providing a generous supply of 111 ( R = CaHd.
