PHOTOCHEMISTRY 33 



trans compound demonstrates that some (or all) of these excited mole- 

 cules are in a state which is peculiar to the /rans-configuration, possibly- 

 state R. If czs-stilbene is excited to state R it must go, by internal con- 

 version, to state 1' (or possibly to the ground state with a high excess of 

 oscillational energy) in a time much less than the natural half life of the 

 excited state. The fluorescent yield of /rans-stilbene was not measured. 

 Lewis and his coworkers (1940) assumed that the ratio of the nonradia- 

 tive return to the cis- and /ra/ts-forms was independent of whether the 

 excited state was formed by the irradiation of normal cis- or trans-siW- 

 bene. This assumption leads to a value of about 0.5 for the fluorescent 

 yields. Although this latter assumption is consistent with the available 

 data, it is by no means the only reasonable interpretation. Olson's con- 

 clusion (1931) that the compound formed from the excited state will be 

 predominantly the isomer of lower stability should not be expected to 

 apply to a reaction which takes place by way of internal conversion. 

 The probability of such a process will depend on the relative forms of the 

 several potential-energy surfaces and on their points of intersection and 

 not merely on relative times spent in the two rotational configurations. 

 Photochemical reactions of this type deserve much more attention than 

 they have received. They are intrinsically interesting, and an under- 

 standing of them should prove helpful in the interpretation of more com- 

 plex photochemical processes (Pinckard et al., 1948; Stearns, 1942). 



The photoisomerization of o-nitrobenzaldehyde to o-nitrosobenzoic 

 acid involves the breaking of two bonds and the formation of two new 

 ones, but the reaction appears to be strictly an intramolecular process. 

 The course of the reaction is independent of whether the compound is 

 present as crystals, is in solution in a solvent, e.g., acetone, or is in the 

 vapor phase. No detectable oxygen is liberated. In the condensed 

 systems (Leighton and Lucy, 1934) at room temperature the quantum 

 yield is 0.50 ± 0.03. In the vapor phase at 90°C, where the vapor pres- 

 sure is about 4 mm of Hg, the quantum yield is 0.70 + 0.05, but the 

 yield is reduced by the addition of molecular nitrogen, reaching a value 

 of about 0.5 at a nitrogen pressure of 700 mm of Hg (Ktichler and Patat, 

 1936). The yield in solution for the corresponding reaction of 2,4-dini- 

 trobenzaldehyde is also about 0.5 but is approximately 0.7 for 2,4,G-tri- 

 nitrobenzaldehyde. It is plausible that the reaction involves the hydro- 

 gen-bonded quasi six-membered ring and that it takes place by way of an 

 act of internal conversion. Why the yield reaches a limiting value of 0.5 

 in condensed systems or in the presence of an inert gas is not obvious. 

 The theoretical predictions of Leighton and Lucy are incompatible with 

 the subsecjuent experiments of Ktichler and Patat, and therefore this 

 detailed theory apparently must be rejected. 



The photochemical denaturation of proteins and inactivation of 

 enzymes can be classed, somewhat arbitrarily, as rearrangements of the 



