Volume 28, number 1 OPTICS COMMUNICATIONS January 1979 A METHOD OF AVOIDING WAVELENGTH-DEPENDENT ERRORS IN DECAY-TIME MEASUREMENTS R.K. BAUER and A. BALTER Institute of Physics, N. Copernicus University, 87-100 Toru~, Poland Received 5 October 1978 The influence of wavelength-dependent photoelectron transit time on decay-time measurements for different photomul- tipliers is discussed. A new correction method is proposed. From the experimental results two conclusionsare drawn: (a) a discrepancy exists between the calculated and measured correction factors and (b) not only wavelength-dependent errors but also errors caused by different spatial distributions of fluorescence and scattered light must be taken into account. 1. Introduction There are two methods of decay-time measure- ments: phase and pulse fluorometry. In both methods one compares the outputs of a photomultiplier (PM) in response to fluorescence and scattered exciting light. The excitation and emission are usually characterised not only by different wavelengths but also by different band-shapes. Due to these differences the photoelec- trons emitted from the photocathode must have diffe- rent initial kinetic energies (assuming a constant work function) leading to a wavelength dependence of the transit time between the photocathode and the first dynode of the PM. This wavelength dependence is of- ten strong enough to cause considerable errors in de- cay-time measurements in the nanosecond range if it is not taken into account. Several authors have drawn attention to this prob- lem and proposed correction methods [1-5]. Miiller et al. [ 1] illuminate the scatterer with "white" light and observe through a light filter used to select the emission wavelength of the fluorescence sample, which itself is excited through another filter selecting the ap- propriate excitation wavelengths from the "white" light. This method assumes that no phase shift exists for different wavelengths of modulated excitation light (or, in the case of a pulse fluorometer, that the shape and timing of the excitation flash is invariant with wavelength), which is not generally true. It is also necessary to take into account the fact that the spec- tral distributions of scattered and emitted light viewed by the PM may be quite different (especially when using broad-band filters) because the emission and "white" excitation light may have anormously differ- ent intensity distributions. For these reasons, the meth. od proposed in ref. [1] is seen to be of limited appli- cation. These effects are particularly important when mea- surements of lifetimes as a function of excitation and/ or emission wavelength are being undertaken. It is then necessary to correct for the wavelength dependence of the transit time of photoelectrons very carefully, with- out the assumptions mentioned above. The elabora- tion of a new method of avoiding wavelength-depen- dent errors in decay time measurements is reported in the following. In this method we excite the sample and illuminate the scatterer with light of precisely the same wavelength and spectral distribution. 2. Methodological considerations The transit time t of a photoelectron between the photocathode and the first dynode of a PM depends on its initial kinetic energy AE, the acceleration a and the electrode distance d (in our considerations, .we as- 91