1824 SPECTROSCOPY AND FLUORESCENCE OF PIGMENTS CHAP. 37C 



that crystals of these pigments behave, in the first approximation, Hke 

 stacks of monomolecular layers, with interactions lestricted in the first 

 approximation to molecules in a single layer. However, the final stage 

 of the approach to saturation seems to be slower than predicted for a 

 monolayer, and the saturation position of the band in crystals is somewhat 

 beyond that found in monolayers (table 37C.II). This indicates that a 

 final contribution to the total interaction energy comes from molecules 

 belonging to different layers. 



The band shifts observed in the two chlorophylls and the two pheo- 

 phy tins follow, as expected, the order of their oscillator strengths ; but the 

 band shift observed in bacteriochlorophyll is much smaller than could 

 be expected from its higher oscillator strength. An explanation could be 

 sought in a different orientation of the transition dipole in respect to the 

 crystal plane, or the superposition of several electronic transitions. 



It will be noted that the band shift was interpreted above without 

 recourse to overlapping of electron clouds of the individual molecules, 

 i. e. without the assumption of any quantum mechanical resonance phenom- 

 ena. The shift therefore gives no information as to the piobability of 

 energy migration between molecules in the crystal or monolayer (as was 

 first suggested by Rabino^vitch et al. 1953). Information about the latter 

 point must be derived from observations of fluorescence and band width, 

 the first of which indicates the actual lifetime of excitation, and the second 

 the extent of its coupling with molecular vibrations. 



No fluorescence of the chlorophyllide crystals could be detected by 

 Jacobs et al, indicating that it is either extremely weak (<C0.1%), or 

 located beyond 1 mn (where the sensitivity of the detector used drops 

 rapidly). In the first case, the actual life time of the excited state is 

 T^ = T<p = 1.2 X 10~V -C 1.2 X 10-11 sg(j. (cf. section 1 of this chapter). 

 The red band in crystals has about twice the half-width of the correspond- 

 ing band in solutions; this shows that the coupling of electronic excitation 

 with intramolecular vibrations is not destroyed by crystallization, and 

 probably supplemented by coupling with lattice vibrations. The excita- 

 tion must therefore stay in the absorbing molecule for a time >10-'^ sec. 

 (the period of an intermolecular vibiation). This permits •<C120 excita- 

 tion transfers during the life time. In the (less likely) alternative case 

 of an as yet undiscovered fluorescence band above 1 mn, the latter does 

 not overlap with the absorption band at 740 m^, and no resonance excita- 

 tion energy transfer is possible at all. We therefore conclude that, 

 despite the closeness of identical pigment molecules in a crystal lattice, 

 the quantum is "trapped" in the absorbing molecule and only a few (if 

 any) resonance transfers occur before it is dissipated into vibrations. 

 The situation is quite different from that found by Scheibe in one-dimen- 



