Mode Locked Titanium Doped Sapphire Lasers

We have two Ti:Al₂O₃ lasers in our laboratory:

There has been a great deal of interest in the development of ultrashort pulse solid state lasers. Solid state lasers have several advantages over dye lasers including continuous operation for longer periods, and the capability for smaller size and simpler construction. A necessary requirement for ultrashort pulse operation of solid state lasers is a gain material with a large gain bandwidth capable of supporting the wide spectral width of subpicosecond pulses. Titanium doped sapphire is currently the leading solid state material for ultrashort pulse laser development. As a gain material for short pulse lasers Ti:Al₂O₃ has several positive features, including a large gain bandwidth of approximately 220 nm in the near IR. It also exhibits a broad absorption centered at 490 nm, optimal for optical pumping by an argon ion laser (see Wall, et. al., IEEE Journal of Quantum Electronics, vol. 24, p. 1016, 1988).

The passively mode locked ultrashort pulse Ti:Al₂O₃ laser on which most current designs are based was developed by Spence, Kean, and Sibbet in late 1990 (Spence, et. al. Optics Letters. vol. 16, p.42, 1991). This design was the first to achieve sub 100 fs operation. Previous designs had achieved pulses as short as 200 fs but relied on synchronous pumping, active mode locking, or nonlinear external cavities. Longer, 4 ps pulses were achieved by passively mode locking the laser with a saturable absorber jet similar to the CPM.

The mode locking process that occurs within the Ti:Al₂O₃ laser is due to self-focusing. Self-focusing is due to the same third order nonlinearity responsible for self phase modulation. In self focusing, however, the modulation of the index of refraction of the medium occurs as the intensity varies spatially across the beam profile rather than temporally across the pulse envelope. This spatial modulation of the index of refraction results in creation of a ``lens'' in the medium and leads to self-focusing. The process by which the self-focusing process results in ultrashort pulse operation of a laser is known as Kerr lens mode locking. The mode locking process occurs since self-focusing adds an intensity dependent optical element to the cavity by modulating the spatial dimensions of the intracavity beam. One way Kerr lens mode locking can be achieved is by designing the cavity so that a better overlap between the pump beam and intracavity beam is achieved at higher intensities. This provides more gain for higher intensities. An example of this is shown in the figure below. Another way to achieve Kerr lens mode locking is to put a slit or aperture in the cavity at a location where higher intensity mode locked beams are spatially smaller than the lower intensity CW beam. This requires very careful design of the laser cavity, however, since both theory and experiment have shown there are few places in the cavity where the loss is less for a higher power beam. Novel methods such as microdot mirrors have also been tried successfully.

Since the same third order nonlinearity responsible for self-focusing and mode locked operation is also responsible for self phase modulation, some form of dispersion compensation must be incorporated into the laser cavity to obtain transform limited pulses. As in the CPM cavity, prisms are used as the dispersive elements to provide an adjustable amount of negative GVD to compensate the positive SPM and compress the Ti:Al₂O₃ pulses. A schematic diagram of the Ti:Al₂O₃ laser is shown below.

Since mode locking is dependent upon a third order nonlinearity, a large initial intracavity intensity is required to initiate mode locking in the Ti:Al₂O₃ laser. This is usually accomplished by perturbing the cavity which results in a rapid shifting of the cavity modes and produces mode beating when several modes constructively interfere. Several methods have been used to provide a small perturbation to the cavity including vibration of optical elements and addition of an intracavity AO modulator.

Tuning in this laser is accomplished by means of an adjustable slit in front of the high reflector where the beam is spectrally and spatially dispersed. The slit is oriented vertically, and translated horizontally to adjust the wavelength. Adjustment of the slit width allows narrowing of the bandwidth and inhibits lasing at narrow linewidth spectral ``satellites''.