Spontaneous emission from reaction centers of photosynthetic bacteria has been recorded with a time resolution of 50 fs. Excitation was made directly into both the special-pair band (850 nm) and the Qx band of bacteriochlorophylls (608 am). Rhodobacter sphaeroides R26, Rhodobacter capsuleas wild type, and four mutants of Rb. capsulatus were studied. In all cases the fluorescence decay was not single exponential and was well fit as a sum of two exponential decay components. The short components are in excellent agreement with the single component detected by measurements of stimulated emission. The origin of the nonexponential decay is discussed in terms of heterogeneity, the kinetic scheme, and the possibility of slow vibrational relaxation.The mechanism of the initial electron transfer step in the reaction center (RC) of photosynthetic bacteria has been the subject of intense study over the past 10 years. This initial step is ultrafast, occurring in about 3 ps at room temperature (1). As the understanding of the RC improves the need arises for more precise kinetic data. In particular, questions arise as to the exponentiality of the observed kinetic signals (2-8), the possibility of differing behavior at different wavelengths (3, 4), the existence of oscillatory components (5), and the existence of spectral shifts (3, 6) accompanying the excitation and subsequent electron transfer processes.The primary method used for ultrafast studies of the primary charge separation step has been time-resolved absorption spectroscopy, generally with low-repetition-rate (10-30 Hz) relatively high-power (excitation pulse energies in the range 1-30 pJ) laser systems. In addition to the limited dynamic range and signal/noise ratios of such measurements, precise determination of kinetics requires that accurate account be taken of all the competing absorptions and bleachings at the detection wavelength. In measurements of the decay of the excited state of the special pair (P*) by stimulated emission, most workers have made measurements at or near the isosbestic point in the spectrum consisting of ground state (P) bleaching and absorption of the radical cation of P (P+) and P*. However, such a procedure makes it difficult to observe longer decay components in the stimulated emission and to look for the presence of spectral evolution or wavelength-dependent kinetics. Zinth and coworkers (7) could not rule out the presence of a 10-to 20-ps component within their experimental accuracy. More recently Vos et al. (5), after significantly improving their signal/noise ratio, reported that the stimulated emission in Rhodobacter sphaeroides R26 (R26) did not decay exponentially but was well described by two decay times (2.9 and 12 ps) with relative amplitudes of 65% and 35%. This observation is very significant for kinetic analyses of absorption changes in other portions of the spectrum, in particular for discussion of whether the primary process should be described by a one-step superexchange or two-step sequential mechanism (2-15).An...