System for nanosecond and picosecond laser ...

0 downloads 0 Views 2MB Size Report
The pulsed voltage was applied to the discharge gap from a high-voltage thyratron generator through a charging resistor =i-I0 k~. The thyratron generator output ...
SYSTEM FOR NANOSECOND AND PICOSECOND LASER SYNCHRONIZATION USING HIGH-VOLTAGE SEMICONDUCTOR SWITCHES N. Mo G. S.

G. Basov, B. L. Vasin, M. P. Kalashnikov, G. Korn,* Yu. Mazur, A. M. Maksimchuk, Yu. A. Mikhailov, V. Sklizkov, V. N. Puzyrev, S. I. Fedotov, and A. Chaushanskii

Results are presented on the development and investigation of a system for the synchronization of the emission of a nanosecond laser, used for plasma heating, with the emission of a diagnostic picosecond laser. The system is based on the use of the fast-semiconductor-switch technology, and ensures a temporal synchronization instability not higher than i00 psec.

I.

METHODS OF SYNCHRONIZING LASERS WITH DIFFERENT PULSE DURATIONS

To determine the dynamics of target ablation inexperiments on laser-driven thermonuclear fusion (LTF) by using x-ray [i] and optical [2, 31 sounding, it is necessary to design and construct lasers operating at nanosecond and picosecond durations and rigidly synchronized with one another.

The main requirements that must be met by such lasers and synchronization

methods are the following: I.

Possibility of varying the pulse durations, ViZo, from I to 5 nsec in the nanosecond

band and from i0 to i00 psec in the picosecond band. 2.

Maximum laser-output irradiance and possibility of subsequent effective amplifica-

tion of the signals to several GW/cm 2. 3.

Good reproducibility of the temporal and energy characteristics.

4.

Minimum temporal instability, not exceeding the picosecond-pulse duration, of gen-

eration of the two pulses. 5.

Reliability of the entire synchronization system.

The synchronization problem can be solved by various methods, which can be arbitrarily divided into the three groups of schemes shown in Fig. io In the first case, the heating and brightening pulses are generated by the same ultrashort-pulse laser (Fig. la) by splitting the laser output radiation beam and guiding the resultant pulses along two independent channels, in each of which the necessary conversion of the radiation duration and frequency are effected; the pulses are synchronized with the aid *Central Institute for Optics and Spectroscopy, Academy of Sciences of the German Democratic Republic. Laser-Plasma Laboratory and Special Design Office. Lebedev Physics Institutes, Academy of Sciences of the USSR. Translated from Preprint No. 85 of the Lebedev Physics Institute, Academy of Sciences of the USSR, Moscow, 1988.

0270-2010/89/1002-0087

$12.50

9 1989 Plenum Publishing Corporation

87

8.

b.

F ~ I -A- ~

,r~-r -'~ C.

Fig. l. Classification of systems for formation and synchronization of laser-emission brightening and heating pulses: PG) ultrashort-pulse generator; LC) pulse-length converter; ODL) optical delay line; FC) radiation-frequency converter; AC) amplification channel; AOG) acoustooptical generator; DL) dye laser; PL) pump laser; SSB) stabilization and synchronization block; IS) isolating system; ML) picosecond master laser~ of the optical delay line ODL.

The simplest duration converters DC for the heating pulse

are passive optical devices of the "stacker" type [4].

The duration is increased in these

devices by adding together a large number of pulses produced by multiple division of the intensity of the initial pulse with duration m0 and shifted relative to one another by a time ~m0.

The synchronization instability is limited in this case only by the thermal expansion

coefficients of the employed "stacker" optical elements and can reach =I psec.

Other ad-

vantages of the system are the possibility of shaping the heating pulse by adjusting the amplitudes of the added signal.

A shortcoming of this scheme, however, as well as of other

schemes in which a nanosecond pulse is formed from picosecond ones, is that the resultant output pulse is highly coherent.

In high-power laser installations this can lead to a lower-

ing of the drawn energy, to an increase of the linear perturbations in the amplification process, and to a lowering of the target-irradiation homogeneity by interference pips.

Among

other shortcomings of schemes with devices of the "stacker" type are the large number of elements employed, the adjustment difficulties, and the concomitant reliability problems encountered when pulses on the order of several nanoseconds are formed. The difficulties of simultaneously obtaining short (=10-11-10-I~ sec) pulses for plasma diagnostics and long (=i0-8-i0 -9 sec) for plasma production can be partly overcome by simultaneously using"stacker" type devices and regenerative amplifiers [5].

This approach was

used, e.g., in the Livermore laboratory [6], where pulse shortening from m0 = 120 psec to 15 psee was accomplished by time compression in the feedback amplifier, while the nanosecond

88

pulse was shaped by a stacker system. filled cells~

The mode locking was effected in this system by dye-

The shortened laser pulse was extracted from the feedback-amplifier cavity

using a high-speed Pockels cell, and the resultant temporal instability • pulse energy spread AE/E = •

psec and the laser-

were achieved by stabilizing the amplification system, thermo-

static control of the regenerative-amplifier cavity, and careful monitoring of a large number of system parameters that govern the processes of amplification and contraction of the pulse injected into the cavity.

Since the optimal ablation regime was realized after approximately

200 passes along the cavity, the pulse used for injection had an energy ~10-i~

-6 J, and

was isolated from the start of the pulse train. The possibility of obtaining nanosecond pulses by directly broadening an ultrashort laser pulse in a regenerative amplifier having a Fabry-Perot etalon mounted in its cavity was investigated at the University of Rochester [7].

In this laser the duration of the shap-

ing pulse was determined by the number of passes of the radiation in the cavity Tout ~ ~in '/~'An electrooptic

Pockels

cell

controlled by high-speed

semiconductor

(gallium arsenide

and silicon) switches [8] was used for Q switching and extraction of the radiation from the cavity.

A single pulse of ~30 psec duration injected in a regenerative-amplifier cavity

having an optical length 30 cm grew after 150 passes to 1.5 nsec, and the reproducibility of the energy characteristics was •

It is impossible, however, to reach the lon'g pulse

durations needed for contemporary LTF research, because it is difficult to effect many passes in the regenerative-amplifier cavity and time delay increases. In the second group of schemes (Fig. ib), both pulses (diagnostic and heating) are formed by two synchronized ultrashort-pulse lasers of unequal duration, with active mode locking effected by one and the same acoustic oscillator.

Synchronization of the diagnostic

pulse is ensured by synchronous optical pumping of the amplification stages in the channel in which the brightening pulse is formed.

This pumping can be effected by an auxiliary pump

laser PL of appropriate lasing frequency, triggered by a single pulse from the amplification channel of the high-power laser facility.

This technique is quite capable of producing

ultrashort (pico- and subpicosecond) brightening pulses from dye lasers synchronously pumped by nitrogen [9], neodymium [I0], or argon [II] lasers.

Among the shortcomings of this method

are the low energies of the produced pulses (10-4-10 -2 J), which are insufficient, e.g., for x-ray brightening of a plasma, the stringent requirements on the thermal stability of the laser cavities and of a number of auxiliary optical elements, and the appreciable fluctuation of the energy of the pulses.

The synchronization temporal instability of this system

is ~i0-!00 psec. Finally, in the third case (Fig. !c), the heating and the brightening pulses are produced by synchronized Q-switched mode-locked lasers.

Notwithstanding the obvious superior-

ity of this method, which permits independent control of the required parameters of the produced pulses, its use was held back by the low synchronization accuracy.

The best synchron-

ization accuracy in such schemes,