This method is essentially a different version of fluorescence upconversion: the general idea of the experiment is the same, with the exception that a different nonlinear optical effect is used for time slicing. The effect in question, as name suggests, is optical Kerr effect. This is a phenomenon, when isotropic material (water or fused silica) becomes birefringent when placed in a strong linearly polarized light field. Birefringence means that the polarization of linearly polarized light traversing the material is partly rotated.
The optical layout of such experiment is shown in fig. 17. Fluorescence light is allowed to pass the shutter consisting of crossed polarizers with nonlinear medium (fused silica window, water cell or carbon disulphide cell ar the usual choices). When the gate pulse is absent, the medium is isotropic, there is no birefringence and the fluorescence is blocked by the second polarizer. When the gate pulse arrives, polarization of the fluorescence light, created by the first polarizer, is rotated due to optical Kerr effect, and some of the light is transmitted through the second polarizer to be dispersed in a spectrograph and detected by a sensitive CCD. Similarly to fluorescence upconversion, the signal arises only from the part of fluorescence pulse overlapping in time with the gate pulse.
The advantages and drawbacks of this method are similar to those of fluorescence upconversion. However, the added advantage of Kerr shutter is the fact that the frequency of fluorescence light is unaffected – we directly measure the intensity of sample fluorescence. As a disadvantage, we note that optical Kerr effect is the third order nonlinearity. This is both good and bad: good because phase matching is not important. Bad, because the gate pulse intensities required for it to occur are enormous. The liquids with relatively high Kerr nonlinearities (water and CS2) feature a non-instantaneous (dipolar) response, limiting their time resolution to several picoseconds.
It must be noted that the time-resolved fluorescence methods based on the tricks of nonlinear optics are not easy: one needs high gate pulse intensities, sensitive detectors and very high-quality optical components for rejecting unwanted light and transmitting the light to be measured.