Date of Award


Degree Type


Degree Name

Doctor of Philosophy (PhD)




Professor B.K. Garside


Professor E.A. Ballik


It was the purpose of the work described in this thesis to develop a basic understanding of the passive mode-locking behaviour of TE CO₂ lasers employing the saturable absorbers SF₆ and germanium (Ge). An understanding of the dynamical processes, which occur in the gain and loss media on sub-ns to μs time scales, is necessary in order to predict the output characteristics of mode-locked pulse trains. In the case of SF₆ the complex dynamical behaviour was not known previous to our work. Therefore, a large portion of this thesis is devoted to investigating such processes. Transmission measurements, using ~200 ns TE CO₂ laser pulses, are used to obtain the first complete measurements of the saturation characteristics of SF₆. The experiments are performed for a range of CO₂ wavelengths which characterize the entire SF₆ absorption at the 10.4 μm band (P(12) - P(28)) and for a range of SF₆ gas pressures which are most frequently used in SF₆-CO₂ mode-locking systems. These experiments demonstrate that, for low J-value CO₂ lines (less than P(24)), the rise in transmission at increased pulse intensities is due to intensity saturation effects. The pressure dependence of the saturation curves indicates that the level relaxation rates scale with SF₆ pressure. However, for CO₂ lines greater than P(22) (longer wavelength), the transmission curves and the transmitted pulse shaping cannot be accounted for solely by intensity saturation effects. It is shown that a multi-vibrational and rotational level treatment for the absorption of CO₂ radiation by SF₆ can fully account for the discrepancies between theory and experiment observed at longer CO₂ wavelengths. The model treats all the SF₆ vibrational and rotational levels as belonging to a bath of levels characterized by a single virbational and a single rotational temperature. Energy absorbed for a laser pulse is rapidly distributed to the bath of levels, establishing new vibrational and rotational populations characterisitics of a higher bath temperature. This type of absorption process depends on the energy rather than intensity, of the laser pulse. Such a model, although expected to apply only at high SF₆ pressures (where fast equilibrium times exist), can in fact be used to predict double-resonance signals observed at low SF₆ pressures (1-10 Torr). Infrared double-resonance experiments with mode-locked pump pulse trains and with ~200 ns duration pump pulses demonstrate that the success of the model at these low pressure is due to very fast vibration-to-vibration and rotational energy transfer rates. The observed rates are much faster than those reported in the literature. It is shown that this heating model, combined with the previously mentioned intensity saturation processes, fully describes the transmission data for the wide range of CO₂ rotational lines and SF₆ pressures employed. This thesis also demonstrates that the mode-locking of a TE CO₂ laser using an SF₆ saturable absorber, can, for the first time, be understood from a knowledge of the above intensity saturation and vibrational bath heating effects. The mode-locking with SF₆ is reported over the widest range of CO₂ rotational lines and with the shortest pulse durations obtained to date. Pulse narrowing across the mode-locked pulse train (usually necessary to obtain very short pulses) is absent in the SF₆-CO₂ mode-locking system. It is indicated that processes such as multi-level saturation, vibrational bath heating and multiple-phonon absorptions may prevent such narrowing from occuring.

The mode-locking behaviour of p-type Ge is also investigated in this thesis. Computer simulations of the evolution of mode-locked pulse from noise are presented using a 2-vibrational level model, which includes rotational coupling, to describe the CO₂ amplifier. A steady-state inhomogeneously broadened saturation response is shown to be appropriate for the Ce absorber. A set of density equations is used to describe the growth of the pulse intensities in the amplifier. For the first time a theory is presented which quantitatively predicts the experimentally observed pulse narrowing across the mode-locked pulse trains, the pulse intensities and shape of the mode-locked train envelope. It is also shown that there is a progression from deterministic mode-locking at laser pressures of ~300 Torr to statistical mode-locking at pressures substantially in excess of this pressure. Furthermore it is demonstrate that there is an optimum value in how close to lasing threshold one can operate to achieve repeatable, clean, short-duration, high intensity mode-locked laser pulses. In addition, it is shown that too large a linear loss cavity and/or too low a ratio of the non-linear to linear loss is detrimental to the production of short-duration pulses. It is also indicated that a laser cavity which diverges the laser beam at the absorber must be used to obtain optimum mode-locking for high pressure lasers (e.g., 5 atm). The choice of the correct cavity is the single most important consideration necessary to achieve stable short-duration (<100 ps) mode-locked pulses at these pressures. A comparison is given between the mode-locking performances of SF₆ and Ge. The ability of Ge to produce multi-line, multi-band mode-locked pulses with sub-ns duration makes it a more viable saturable absorber than SF₆ for use in low intensity mode-locked oscillators.

Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Included in

Physics Commons