A time resolved resonance Raman experimental set-up is described. It is based on the use of two independent lasers : a Q-switched Nd-YAG pumped dye laser and an excimer laser. A synchronization electronic device has been made to fire the two lasers with variable delays in the range .0 ns -+ 10 ms with a 10 ns step variation. A drift control unit is used to prevent any slow drifts of the excimer laser. The overall jitter between the two laser pulses is 4.5 ns. INTRODUCTION Time resolved resonance Raman spectroscopy is becoming a very powerful technique for the caracterisation of reactive excited states and transient intermediates in fast photochemical reactions (1,2). It provides the means of identifying the transient species, analysing their structures and bonding, and also of monitoring their kinetics of formation and decay. The two-color pump probe method is generally applied. The pump beam is used to start the photochemical reaction by providing energy to the system and promoting the molecules to high excited states. The probe, which is delayed with respect to the pump, is used to excite resonance Rarnan spectra of the transient species. Both have to be tuned to an electric transition : pump to the absorption wavelength of the molecules in the ground state and probe to the transient absorption of the species being analysed. In recent experiments, we have recorded the resonance Raman spectra of excited states and radicals of various molecules using a Q-switched Nd-YAG pumped dye laser (3,4,5). In these cases, the third or the fourth harmonic of the YAG laser was used as the pump beam and the resonance Raman spectra were excited by the dye laser pumped by either the second or the third harmonic of the YAG. The probe beam was delayed by an optical delay line (3 ns/m). Unfortunately, the low quality of the laser beams makes it impossible to use large delays and most experiments were performed in the range 20-50 ns. These experimental conditions did not allow kinetic measurements to be made. In order to overcome this problem, one has to use two independent lasers and synchronize their firing to increase the delays time between the probe and pump beams. TIME-RESOLVED EXPERIMENTS WITH TWO LASERS We have developed a new time-resolved Rarnan experimental set-up using the same Q-switched Nd-YAG pumped dye laser for the probe beam. An excimer laser was preferred for the pump beam because it is less expensive than a second Nd-YAG pumped dye laser and it emits many wavelengths in the UV region (193, 222, 248, Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jp4:19917123 C 7 4 6 4 JOURNAL DE PHYSIQUE IV 308, 351 nm) that are generally suitable to fit the absorption spectra of most organic compounds. We have used an YG 581 C QUANTEL YAG laser pumping a TDL 50 QUANTEL dye laser. The output of the dye laser can be continuously tuned from 220 nm to 750 nm by direct pumping of red dyes or coumarin using the second and the third harmonics of the YAG respectively, or by frequency-doubling and/or frequency mixing. The Excimer laser is a QUESTEK model 2420. The problem was to synchronize the two lasers with as low a jitter as possible. In all cases, the jitter time must be lower than the duration of the laser pulses, i.e. 7-9 ns for the dye and 20-30 ns for the Excimer. The repetition rate was fixed at 10 Hz because of the YAG laser. The YAG laser needs three trigger orders to be fired ; one to command the charge of the capacitors, the second to fire the flashlamps and the third to trig the pockels cell. The capacitors require about 80 ms to be charged. The excimer laser needs two trigger orders. One to command the charging of the capacitors and the second to fire the thyratron. For this laser, about 30 ms are needed to charge the capacitors. A digital pulse and delay generator has been designed to generate all the trigger orders, all the delays, and also to synchronize a gated multichannel detector with the YAG laser. The schematic diagram and the chronogram of this system are given in figures 1 and 2.