Even though the pulses coming from the oscillator are well suited for some time-resolved spectroscopic experiments, a single Ti:Sapphire oscillator without any amplifiers has several drawbacks:
Very high repetition rate. The light induced processes in biological systems sometimes take milliseconds, or longer to complete (for example, a triplet state of chlorophyll lives for about 3 ms). If the time intervals between the pulses are shorter than the duration of the investigated processes, every pulse coming to the sample encounters the ‘footprints’ of previous pulses still present. That renders the experiment incorrect. One way to solve this problem is moving the sample to investigate a fresh spot each time; however with only 12 ns of time between pulses, it is physically impossible to move the sample far enough for the laser beam to interrogate a fresh spot.
Narrow tuning range. Even though the Ti:Sapphire oscillator is tunable in the range from 700 to 1000 nm, many research tasks require different spectral range – the samples simply do not absorb in the accessible wavelength range. Tuning the wavelength of light is made possible by nonlinear optical devices called optical parametric amplifiers (OPAs), which will be discussed later. However, to operate properly, these usually require higher energies than available from the oscillators. Additionally, some types of time-resolved spectroscopy measure inherently nonlinear signals (e.g. two-photon absorption, photon echo spectroscopy), only observable at high laser field intensities.
To obtain higher pulse energies, a technique called chirped pulse amplification is employed. The laser implementing this technique is called regenerative amplifier. The amplification sequence is illustrated in fig. 7.
Fig. 1) Operation principle of chirped pulse regenerative amplifier. Weak pulses are stretched in time, one out of ~80000 pulses is injected into the amplifier cavity, amplified and compressed in time again. The inset shows the shape of the electric field in the chirped pulse: lower frequencies (redder photons) arrive first, followed by the higher frequencies (bluer photons).
- Oscillator pulses are stretched in time (i.e. their duration is increased) up to several hundreds of picoseconds.
- One out of roughly 80000 pulses is injected into the cavity with Ti:Sapphire crystal pumped by a Q-switched laser.
- The pulse travels around the cavity several (or more) times, until it is amplified to several mJ of energy.
- Amplified pulse is ejected from the cavity.
- The pulse is compressed (in time) back to several tens or hundreds of femtoseconds.
Stretching before and compressing after the amplification is necessary in order to avoid damage of optical components in the amplifier: if a pulse of 100 fs is amplified and focused in the laser crystal, its peak power becomes so high that the optical components in the amplifier break down.
Pulse stretching and compressing is done using diffraction gratings to spread the pulse spectrum in space and organizing the beam path in such a way that bluer frequency components travel a longer distance (in the stretcher) or shorter distance (in the compressor) than the red ones. An inset in fig. 7 schematically shows the electric field of a chirped pulse, where lower frequencies arrive at the observer earlier than the higher ones. Injection and ejection of pulses into the regenerative amplifier cavity is done using electronically controlled Pockels cells and reflective polarizers (their action is virtually identical to that of an electro-optical Q-switch, see above). More information on chirped pulse amplification is available in literature . After chirped pulse amplification, light parameters are typically as follows: pulse repetition rate – around 1 kHz., duration – around 50 fs, spectral width in the vicinity of 30 nm, pulse energy – 0.5 to 5 mJ.