Phase fluorimetry, or frequency domain time-resolved fluorescence spectroscopy is, perhaps, the oldest among time-resolved fluorescence techniques. Despite its age and drawbacks, the method is very useful, when large batches of samples need to be characterized quickly. It does not require pulsed lasers. An incoherent light source, a couple of monochromators and a detector is all you need, therefore, this method is – at least, in principle – the cheapest. It is based on the following brilliant insight: fluorescence light leaves the sample slightly later (in fact, on average fluorescence lifetime later) than the excitation light arrives. If we excite the sample using modulated light source (e.g. the light from the lamp will be passed through a rotating disk with holes in it), the fluorescence light will be modulated at the frequency of the excitation light, but will experience a phase shift with respect to the excitation light (fig. 18). The phase shift, loosely speaking, will correspond to the delay between the excitation and fluorescence. Sensitive electronic devices called lock-in amplifiers allow very precise detection of phase shift between two signals. Assuming some shape of fluorescence decay in the sample (usually exponential decay), we use the phase shift to deduce the fluorescence lifetime (the exponent). Tricks can be done to enhance time resolution: one can sweep the modulation frequencies, or use modulation depth of fluorescence as an additional input deducing the time course of fluorescence decay. However, at the end of the day, this is a purely electronic method and picosecond-femtosecond time resolution is a challenge for it.
Fig. 1) Phase fluorimetry: the sample is excited using fast-modulated light (black line). Because the fluorescence takes some time to occur, the detected fluorescence light is modulated at the same frequency, and exhibits a phase shift with respect to the excitation light. From the phase shift, an average fluorescence lifetime can be calculated.
As already mentioned, this method is the simplest, and hence the cheapest: it does not require pulsed lasers, nonlinear crystals and other fancy laserline components. Its accuracy and reliability when pushed to the limits (picoseconds) is inferior to the direct measurement methods, because in order to determine the fluorescence decay times, we must postulate its kinetic shape. Direct methods do not need such assumptions; they simply measure the fluorescence decay. However, since the method is quick and accessible, it has its uses.