Next: Model and parameters of
Up: Multipulse Operation and Limits
Previous: Multipulse Operation and Limits
Introduction
Ultrashort-pulse laser oscillators demonstrate a tendency to
destabilization and pulse splitting when the pulse duration
approaches minimal values (see, for example, [1]). As
the net group-delay dispersion (GDD) in the resonator approaches
zero, a nonregular pulsing or a stable multiple pulse generation
may occur. Both the unstable as well as stable multipulse
oscillation can be generally considered as the obstacle for the
pulse shortening. At the same time, it is possible to use the
pulse splitting for an additional duration decrease as well as for
an increase of the laser repetition rate. The analysis of this
phenomenon is not straightforward because the ultrashort-pulse
laser is a complex system with different interacting
nonlinearities.
At the moment, the following mechanisms of the multipulse
operation have been considered: the appearance of a negative
feedback due to the pulse chirp growth [2]; the
radiation scattering growth accompanying the decrease of the
soliton period [1]; the higher-order dispersions
contribution to laser dynamics [3]; the laser continuum
growth owing to the positive net-gain of the background radiation
outside the pulse [4]; the dynamical gain saturation
and recovery [5]. The multipulse operation can also be
considered as a result of the higher-order soliton dissociation
[6].
In the framework of the soliton approach [7,8]
it was shown, that the multipulse oscillation can be caused by the
formation of the bounded multisoliton complexes in resonators with
both negative and positive net-GDD. The self-interaction of such
bounded soliton-like complexes is governed by the relative phase
of the constituent pulses [9]. The presence of the
continuous local (i.e. through the pulse wings) interaction
between the pulses distinguishes such bounded soliton-like
complexes from the regimes created as a result of the
repetition-rate multiplication due to the change of the gain and
loss saturation balance in the system [10,11,12].
One should also distinguish the stable multipulse operation from
the generation of the double pulses colliding in an active medium
[13,14,15,16]. It should be noted, that the
transition to the multipulse operation is not the sole scenario of
the stability loss in the contiuous-wave (CW) mode-locked
solid-state lasers: the automodulational instability can produce
regular as well as nonregular oscillations of the single pulse or
the so-called picosecond collapse with the abrupt transition from
the femtosecond to the picosecond generation
[17,18].
The stable multipulse operation in the negative GDD region was
experimentally observed in Ti:sapphire
[1,19,20,21], Cr:LiSGaF [22] and
Yb:KYW [23] Kerr-lens mode-locked lasers as well as in
Nd:glass [24], Ti:sapphire [4] and Cr
:YAG [25] lasers mode-locked by semiconductor
saturable absorber mirror. In Ref. [26] the tendency to
the multipulse generation was reported for the positive-GDD regime
in the Cr
:ZnSe laser with passive mode locking
initiated by acousto-optical modulation.
The latter medium allowing diode pumping and possessing excellent
lasing characteristics (see [27,28]) is of interest
as tunable ultrashort pulsed mid-IR source. At this moment, a
variety of generation regimes has been demonstrated: the efficient
pulsed [29] and continuous-wave [30,31]
diode-pumped operation, active mode locking [32], and the
active modulator-assisted passive mode locking [26].
Above all that, Cr
:ZnSe is of interest as the model object
for the study of the multiple pulse operation. This interest is
explained by the combination of the unique characteristics: very
large nonlinear index of refraction [33] as well as
emission and absorption cross-sections. As it will be shown later
on, these factors play a crucial role in the pulse destabilization
and the transition to the multiple pulse operation.
In order to explain the variety of the observed multipulse regimes
one needs to have a basic understanding of the nature of the
multiple pulse generation in Kerr-lens mode-locked lasers. Our
model takes into account the strong saturation of the Kerr-lens
induced fast absorber, the gain saturation and recovery dynamics,
the GDD and the self-phase modulation, making it valid for the
different femto- and picosecond Kerr-lens mode-locked lasers and
sufficiently simple to obtain the physically significant
conclusions.
Our analysis reveal the existence of the two basic mechanisms
causing the multiple pulse operation: the background amplification
due to the gain saturation decrease and the increase of the
"bounded" perturbation in the presence of the strong saturation of
the fast absorber. The basic laser factors defining the pulse
stability were found to be the gain and loss saturation in the
combination with the spectral loss. As it will be demonstrated,
the inter-pulse interaction is strong enough to produce the
correlations of the inter-pulse distances and phases, which are
governed by the GDD and the fast absorber parameters. The
statistical and multistable properties of the bounded multiple
pulse complexes will be considered as well. Finally, the limits of
the single pulse operation and the methods of the stability
enhancement will be analyzed.
Next: Model and parameters of
Up: Multipulse Operation and Limits
Previous: Multipulse Operation and Limits
V.L. Kalashnikov
2002-12-28