Laser beat waves accelerate particles
Recent experiments by F. Amiranoff and coworkers from Ecole Polytechnique (Palaiseau, France) and its associated laboratories have demonstrated that particle acceleration is possible with a high-peak-power Nd:glass laser. It is well known that high-peak-power neodymium lasers are able to generate electromagnetic fields higher than 100 GV/m. But the nature of these transverse waves does not allow them to be used directly to accelerate particles in an efficient process. A plasma, however, can be u
Laser beat waves accelerate particles
Recent experiments by F. Amiranoff and coworkers from Ecole Polytechnique (Palaiseau, France) and its associated laboratories have demonstrated that particle acceleration is possible with a high-peak-power Nd:glass laser. It is well known that high-peak-power neodymium lasers are able to generate electromagnetic fields higher than 100 GV/m. But the nature of these transverse waves does not allow them to be used directly to accelerate particles in an efficient process. A plasma, however, can be used to partially convert the transverse field into a longitudinal field by a process termed "laser beat wave." This technique is susceptible to the acceleration of relativistic particles (those particles having a velocity close to the velocity of the light).
To test this theory, the Ecole Polytechnique researchers implemented simultaneously two laser oscillators at two frequencies to generate the beat wave, which is then amplified by neodymium-doped phosphate glass amplifiers. The amplifier series uses Nd:glass rods with diameters ranging from 6 to 90 mm to operate at constant energy density (in J/cm2) at the output of each amplifier (see Fig. 1). The two oscillators are Nd:YLF and Nd:YAG, which generate 1.053-µm and 1.064-µm wavelengths with 90-ps and 160-ps pulse durations, respectively.
Because the peak amplification wavelength of the Nd:phosphate glass is centered on 1.053 µm, the Nd:YAG emission (1.064 µm) is less amplified than Nd:YLF (1.053 µm). To compensate for this gain difference a supplementary amplifier is introduced after the Nd:YAG oscillator (see Fig. 2).
Each oscillator uses an active Q-switch and Kuizenga-type mode blocking, in which the evolution of the laser emission in the course of the time is given in a burst (see Fig. 3). At the start, the cavity losses are adjusted at a high level with an intracavity Pockels cell. After a typical time of about 1 ms ("prelasing" operation), the oscillation emission is stabilized to achieve a continuous rate. Opening the intracavity Pockels cell at this time provokes an intense emission that has a wra¥duration of about 100 ns. Mode blocking by the acousto-optic crystal allows the generation into this wra¥of a 100-ps short pulse train, and a single pulse is selected by an extracavity Pockels cell.
To synchronize the two oscillators, the two wraps of the pulse train must coincide and the phases of the acousto-optic modulators must be matched. The long "prelasing" operation principle (about 1.2-ms duration) means that the jitter between the intracavity Pockels cell opening and the pulse-train emission is very small and widely inferior to the wra¥duration. Synchronization of the two wraps is achieved by an adjustable delay on the Pockels cell opening, which compensates for the time difference (due to the gain difference between the Nd:YLF and Nd:YAG) in the pulses generated from the stabilized initial level.
To synchronize the pulses generated by the mode blocking required matching the signal phase between the two acousto-optic crystals. This is achieved by using only one 67-MH¥generator, the signal of which is split in two parts and afterwards injected in a crystal specially chosen to have the identical operation frequency. The delay between the two short pulses can be adjusted by a "slide trombone" that delays one of the two radio-frequency signals in regard to the other.
Beyond the 90-mm-diameter rod in the final amplifier, the maximum output energy is about 10 J at each wavelength. The beam afterwards is focused in a gas chamber filled with deuterium to adjustable atomic density of ۭ%. The maximum power density for each wavelength is respectively 3.2 ¥ 1014 W/cm2 (Nd:YAG) and 6 ¥ 1014 W/cm2 (Nd:YLF), which produces a homogeneous plasma with a well-known density.
To check the effective particle acceleration, a relativistic electron beam was injected in the plasma created into the gas chamber. To accomplish these experiments, the laser beam had to be transported over a 200-m distance with the optical beam delivery system in the SESI labs. There, a Van de Graaf accelerator is in operation to produce a relativistic electron beam. (see Laser Focus World, Feb. 1994, p. 37).
Amiranoff reported at last autumn`s French Society of Physics meeting that, in the best case, the accelerated electron beam energy measured during the laser tests has reached 3.7 MeV. That result confirms other experiments in several laboratories around the world that particle acceleration by a laser beat wave is possible. The majority of research in this field is now concentrating on the "wakefield" effect, in which the force on the particles is induced by the wra¥of the short-pulse duration, and on the ultrashort laser pulses used to generate intense electric fields.