REVERSIBLE PHOTOCHEMICAL PROPERTIES OF DYES 9 



reactions (Hendricks and Borthwick, 1954). To press this analogy 

 farther, I should like to hypothesize that the red-absorbing pigment in 

 photoperiodic systems is a reduced form of a highly conjugated mole- 

 cule which, on absorbing red light, results in the production of a far- 

 red-absorbing species (having lost its labile hydrogens and therefore 

 absorbing at longer wavelengths) and reduction of its substrate. This 

 picture requires that the reduced substrate (AH-) be a promoter of the 

 eventual physiological responses, whereas the oxidized substrate (A) 

 produced with far red is an inhibitor of the first stages of a chain of 

 reactions leading to these responses. 



In a series of papers dealing with the kinetics of photoreduction of 

 dyes we have ascertained that it is not the initially light-excited dye 

 (to its first electronically excited state, lifetime about 10~^ sec) that 

 is chemically reactive, but rather its metastable state (triplet state, 

 lifetime in water about 10"^ sec) that reacts with the reducing agent 

 (see, for example, Oster and Adelman, 1956; Adelman and Oster, 

 1956). Trace amounts of substances such as nitrobenzene will retard, 

 i.e., slow down the photoreduction. This is one indication that long- 

 lived excited states are involved since diffusion calculations show that, 

 in order for the retarding molecules (at low concentrations) to en- 

 counter a dye molecule while it is in an excited state, the reactive state 

 must be of long life. It was further found that the quantum yield of 

 photoreduction decreases with increasing dye concentration. Appar- 

 ently dye molecules in the ground state quench the triplet state of the 

 excited dye molecules. 



In highly viscous media the number of collisions between excited 

 dye and reducing agent is decreased and the photoreduction (and 

 photorecovery) is suppressed (Oster and Wotherspoon, 1954). We 

 have recently found, however, that dyes in a rigid glucose glass are 

 readily photoreduced with visible light and reformed with ultraviolet 

 light (Oster and Broyde, to be published). Apparently the medium it- 

 self acts as the hydrogen transfer agent so that reactions take place 

 with the dye and its immediate surroundings, and no diffusion is re- 

 quired. Incidentally, by this method we have been able to trap certain 

 intermediate reduction states (semiquinones) of the dyes. 



At dye concentrations above about 10~"M, photoreduction practi- 



