DELIMITACON DE LA IMPUTACION FISCAL HECHOS IMPUTADOS.

5.1 Introduction

The technique of ultrafast laser inscription (ULI) has been introduced as a method for creating crystalline channel waveguides [1, 2] and an overview of this technique was described in Chapter 1. The experimental set-up used to write the structures was described in Chapter 2, and some results of channel waveguides fabricated in a Quantum Dot doped glass were presented in Chapter 3. The technique has previously been employed to create lasing crystalline channel waveguides in LiF at 707 nm [3] and Nd:YAG [4-6] around 1 µm. In 2008 Raman gain was demonstrated in ULI channel waveguides of undoped KGdW [7], and in 2007 channel waveguides were fabricated in Yb:KYW but the resulting waveguides were not able to support lasing [8]. In this chapter the first examples of lasing in channel waveguides of Yb:KGdW and Yb:KYW are presented [9]. The samples were fabricated by Dr Robert Thomson at Heriot Watt University and the sample analysis and laser experiments were performed by myself at the University of St Andrews.

5.2 Sample Description and Writing Conditions

Fig. 5.1. Diagram of Yb:KGdW and Yb:KYW crystal illustrating the crystallographic axes, the positions of the waveguides and the writing directions used.

b c a L = 9 mm +ve -ve

ULI was used to transversally write waveguides along the c axis of crystalline Yb(5 at. %):KGdW and Yb(5 at. %):KYW, as shown in fig. 5.1. The samples had dimensions of 2(b)×10(a)×10(c) mm3 and 1(b)×10(a)×10(c) mm3 respectively. The inscription set-up was as described in section 2.2 using the Yb:fibre laser as the writing source. When using ULI in crystalline media the written track usually manifests itself as a lower index barrier, and so pairs of modified tracks were inscribed in order to enclose the waveguide in a well-defined region and to create symmetric waveguides. A schematic illustrating the principle of this technique is presented in fig. 5.2. This technique had previously been shown to improve confinement and waveguide performance [8]. Writing was performed to within less than 500 µm of the crystal facets, and the crystals were polished back to reveal the modified region. The final crystal length along the c axis was 9 mm after polishing.

(a)

(b)

(c)

Fig. 5.2. Schematic illustration of typical guiding regions found in ULI crystal waveguides, where written filaments are shown in grey, and guiding regions are shown in red. (a) and (b) show typical guiding areas in a region surrounding the written volume. The exact position of the guiding areas can vary with crystal, writing parameters and guiding polarisation. (c) shows the proposed writing scheme, utilising two parallel tracks to create a symmetric and more confined guiding region. This has previously been demonstrated to improve guiding results [8].

As explained in section 1.7 there are many writing parameters which are known to affect the performance of the waveguides, and thus substantial work is required to optimise the writing conditions. The waveguides reported here were fabricated with a constant writing wavelength of 1064 nm; repetition rate of 500 kHz; 1.3 ps pulse duration; 6 mms-1 writing velocity and 20× focusing objective. The varied parameters were the pulse energy, writing polarisation, scan separation, and writing direction. The incident pulse energy was varied from 296 nJ – 558 nJ in Yb:KGdW and 252 nJ –

578 nJ in Yb:KYW in steps of approximately 20 nJ. In the case of Yb:KGdW structures were inscribed 430 µm below the surface using linear polarisation, which was polarised along either the a or c axis, with scan separations ranging from 10 µm to 25 µm in 5 µm steps, and also using circular polarisation with scan separations varying from 20 µm to 35 µm. After inscribing the structures in Yb:KGdW it was noted that linear polarisation resulted in catastrophic crystal damage and so in Yb:KYW circular polarisation was used for all structures. Structures were inscribed 360 µm below the crystal’s surface.

The effect of writing direction on waveguide performance was also investigated, with pairs of tracks being written by translating the sample in the negative c axis direction as well as the positive direction for each pulse energy, scan separation and polarisation. At the time the experiments were performed the scan direction had recently been shown to affect the waveguide quality, but the reason for this was not known. This effect has now been attributed to wavefront tilt [10]. Unfortunately, as this was not the known cause at the time, the wavefront tilt during writing was not recorded. Therefore, although a difference in performance is evident between the two writing conditions, a quantitative measure of the cause of this is not. A total of 176 structures were written in the Yb:KYW crystal, and 168 structures in Yb:KGdW. The writing conditions used for each of these structures are presented in a table in Appendix A. In the following sections of this chapter the waveguiding behaviour is presented, together with results from lasing structures and selected non-lasing structures.

5.3 Experimental Procedures

5.3.1 Identifying Crystal Modification and Guiding Regions

The first task was to determine the positions of the structures and the energies needed to modify the crystal. This was simply done by imaging the crystal’s end facets with a CCD camera and 50× objective under incoherent white light illumination from a tungsten lightbulb.

Guiding regions were then identified using the 980 nm InGaAs laser diode described in section 2.3. Although this wavelength corresponds to an absorption peak of the crystals, the power available from the diode was sufficient to saturate the crystal absorption and so guiding structures were easily identifiable. For a given crystal and guiding polarisation the guided mode sizes for structures written under different conditions were similar. Therefore, by using this technique, it was also possible to identify the most promising guiding regions by observing the point of absorption saturation. For lower propagation loss the absorption became saturated at lower pump powers. Once saturation was reached the structures with lower propagation losses also gave a higher gradient of transmitted power to incident power. These two characteristics of the curve gave a clear qualitative indication of the waveguide’s quality. An example of this behaviour is shown in section 5.4.2.2.

5.3.2 Laser Cavity

Fig. 5.3. Photograph showing a waveguide in the Yb:KGdW crystal under pumping at 980 nm. The high reflector (HR) and output coupler (OC) have been drawn in to illustrate the laser cavity that was formed.

Attempts were made to obtain lasing in all of the identified waveguides using a compact laser cavity as described in section 2.3.3. A photo of the set-up is shown in fig. 5.3, where the end mirrors and beams have been drawn in for clarity. Various objectives were used to deliver the pump light to the channel waveguide in order to establish the optimum coupling. The best results were obtained when using a 30× microscope objective to collimate the pump light and a 10× objective to focus the light into the waveguide. This gave a spot size of approximately 18 µm e-2 diameter

Pump