Page 373 - High Power Laser Handbook

342 So l i d - S t at e La s e r s Ultrafast Lasers in Thin-Disk Geometry 343

spectral filtering due to the gain bandwidth, stabilizes additional
87
pulse generation against Q-switching instabilities. The basic idea
for the additional stabilization in the femtosecond region is as fol-
lows: If the energy of a pulse increases by relaxation oscillations, the
spectrum of the pulse is broadened by SPM. A broader spectrum,
however, will experience a smaller average gain due to the laser
material’s finite bandwidth. This effect has a much smaller influence
on picosecond lasers, because SPM is much weaker. In addition,
inverse saturable absorption of the absorber further reduces the QML
threshold. 88,89 Again, the basic idea is simple: An inverse saturable
absorption causes a rollover in the nonlinear reflectivity, which
increases the losses for pulses with higher energy, thus damping
relaxation oscillations. Therefore, SESAMs are ideal for mode-locking
diode-pumped solid-state lasers, because semiconductor saturable
absorbers inherently have a large absorption cross section (a low
absorber saturation energy), and their nonlinear reflectivity dynam-
ics can be designed over a wide parameter range. In contrast, the low-
gain saturation energy of VECSELs makes them immune to such
Q-switching instabilities, moreover their short upper-state lifetime
(typically in the nanosecond regime) tends to restrict the pulse repeti-
tion rate to the gigahertz range for stable CW mode locking. High-
power TDLs typically operate in the 1- to 100-MHz regime, because
at much higher pulse repetition rates, QML instabilities become more
severe. This is, however, not a problem for most applications, because
the larger pulse energy at lower pulse repetition rates is advanta-
geous for many applications, such as precision micromachining.
Stable mode-locked operation with dynamic gain saturation
requires the absorber to saturate at lower intracavity energy than the
gain in order to obtain stable pulse formation. For mode-locked VEC-
SELs, the active region typically consists of the same material as the
absorber. Therefore, a large ratio between the mode areas on the gain
and on the absorber material is necessary to achieve saturation of the
absorber before the gain saturates. In this case (see Fig. 13.7a), geo-
metrical issues for the laser cavity give an upper limit for the repeti-
tion rate of a fundamentally mode-locked VECSEL. One way to
overcome these issues is mode locking with similar mode areas on
the gain and the absorber (see Fig. 13.7b), which is referred to as 1:1
mode locking. In this case, SESAMs with low saturation fluence are
required, which can be achieved using quantum dot (QD) SESAMs
instead of conventional quantum well (QW) SESAMs. In QW-SES-
AMs, the product of the saturation fluence F and the modulation
sat
depth DR is proportional to the energy needed to completely saturate
the absorber. This means that these two parameters cannot be adapted
independently. However, for 1:1 mode locking with ultrahigh repeti-
tion rates and very low pulse energies, a low F and DR are required,
sat
according to the QML criterion. This problem could be overcome by