Publication No: 1383Search all ORC publications    

Stable, high-power, single frequency generation at 532nm from a diode-bar-pumped Nd:YAG ring laser with intracavity LBO frequency doubler

Kevin I.Martin, W.Andrew Clarkson, and David C.Hanna


We obtained 2.5 W of single-frequency TEM00 output at 532 nm using a Brewster-angled LBO crystal for intracavity second-harmonic generation in a diode-bar-pumped Nd:YAG laser. By inserting a thin uncoated étalon, the 1061.4 nm laser transition can be selected, generating 1.6 W of output at 530.7 nm.

We have recently reported [1] the generation of 3W of cw single-frequency output at 532 nm using a potassium titanyl phosphate (KTP) crystal for intracavity second-harmonic generation (SHG) in a diode-bar-pumped Nd:YAG laser. To achieve that power level it was necessary to introduce a number of measures to prevent the laser from accessing oscillation schemes that escape from the nonlinear loss associated with efficient harmonic conversion. One such oscillation scheme observed under some conditions involved a misaligned path round the resonator that avoided efficient harmonic generation by misalignment from the phase-matching condition. Such behaviour can affect the power scaling of intracavity SHG. One way of reducing this problem would be to use a nonlinear crystal with a broader angular acceptance for harmonic generation so that avoidance of the phase-matching condition would carry too great a penalty in losses due to misalignment with respect to the optimum resonator round-trip optical path. LBO is a suitable candidate for such an approach, having an angular acceptance for 1.064 μm doubling [2] of 57 mrad cm1/2, compared with 15 mrad cm for KTP. There are other important differences in the parameters of LBO and KTP, so that significant differences are implied between the resonator designs appropriate for KTP and LBO. The much smaller nonlinearity of LBO is its main drawback, suggesting that it will have a more important role as powers are scaled up. Despite this drawback we show here that, using a single 20 W diode bar as the pump for a Nd:YAG laser, one can obtain for LBO, SHG efficiencies similar to those for KTP [1], i.e., 3.3 W of 532 nm light generated inside the LBO crystal (2.5 W actually extracted, the difference being lost mainly at the Brewster-angled face of the LBO). Before describing in detail the performance characteristics of this laser we first address the relative merits of KTP and LBO in this context.

In a comparison of LBO and KTP for cw high-power intracavity SHG, the other considerations, beside the larger angular acceptance of LBO and its smaller (~x4) nonlinearity, are of relevance. The damage threshold for LBO is much higher than for KTP, as it is free from the grey tracking effect seen in KTP. The temperature bandwidth for SHG is small, ~4.2 °C cm in LBO [2] compared with that of KTP [2] ~25 °C cm. LBO needs an elevated temperature (~150 °C) and temperature control to maintain optimum phase matching. KTP also needs temperature control, however, because of its type II phase matching it acts as a wave plate for the fundamental and must be adjusted to give full wave retardation [1]. For LBO, the type I phase matching allows polarization properties of the resonator that are generally simpler and it also allows the use of the Brewster angle orientation, thus enabling lower losses to be achieved. Another merit of LBO is the fact that it can be temperature tuned to give exact noncritical phase matching for a wide range of wavelengths, whereas in the case of KTP a special crystal cut [3] of an orientation specific to the 1.064 μm fundamental wavelength was necessary to achieve reduced double-refraction walk-off. This convenient feature of LBO is illustrated by our demonstration of efficient doubling of another Nd:YAG transition at 1061.4 - 530.7 nm.

The reduced nonlinearity of LBO implies a need for a higher fundamental intensity, which can be achieved with a smaller spot size in the nonlinear crystal. Also, reduced losses in the LBO resonator (Brewster angle crystal, fewer optical components as explained below) increased the circulating fundamental power. Despite this, the internal conversion efficiency [second-harmonic (SH) power and circulating fundamental power] for LBO was less, at ~1.7%, compared with the situation for KTP, ~3.4%. However, higher circulatory fundamental powers, ~190 W for LBO versus ~92 W for KTP, gave essentially the same level of generated SH power.

Fig. 1. Schematic diagram of the Nd:YAG ring laser resonator, with TGG Faraday rotator and LBO frequency-doubling crystal.

To obtain a smaller spot size in the LBO it would have been convenient to work with shorter radius of curvature concave mirrors adjacent to the crystal (see Fig. 1). However the oven housing of the LBO made this physically impossible, and it was necessary to use the same 125 mm radius of curvature concave mirrors as used with KTP, with the result that, for the smaller spot in the crystal (~38 μm in the Brewster plane by 26 μm for LBO compared with ~70 μm for KTP), the mode spot size was larger on the concave mirrors. To achieve mode matching to essentially the same spot size in the laser rod as used for KTP (~250 μm) implied a longer resonator path and necessitated some degree of compensation of the thermally induced lens in the laser rod (~150 mm focal length power). This compensation was produced by a convex mirror (~200 mm radius of curvature) with a dichroic coating so that pump light could enter through this mirror. To keep the induced thermal lens as central as possible to the collimated section of the resonator, the resonator was distorted from the usual symmetric arrangement (see Fig. 1). We applied the usual astigmatism compensation [4,  5] by an incidence angle at the concave mirrors to compensate for the astigmatism introduced by the 15 mm long Brewster-cut LBO crystal.

Other components in the resonator include the TGG Faraday rotator (providing 7.5° of rotation) and the half-wave plate to compensate the rotation. To maximize the loss difference between the two counterpropagating waves, for a cavity with a weakly polarizing element (e.g., a Brewster plate), the rotation compensation should be provided ideally by either a true rotator or a pair of half-wave plates that can be configured to achieve the same effect [6]. Under this scheme, the retardation plate effect of a birefringent element, such as the LBO crystal, would have a significant effect on the loss difference. We have used the walk-off enhanced polarizing properties of the Brewster-angled LBO crystal to eliminate this potential problem and to provide the necessary loss discrimination against the counterpropagating wave. The counterpropagating wave incident on the LBO crystal has a component of polarization (vertical) out of the plane (Fig. 1). This polarization component is refracted at a different angle from the component in the plane of the figure. The vertical component becomes separated from the horizontal component and is essentially completely removed from the counterpropagating wave, causing the LBO to act as a near-ideal polarizer. Our estimate indicates that this results in an additional loss of ~6.7% for the counterpropagating wave, independent of the retardation plate effect of the LBO. This loss difFerence exceeds the loss that is due to conversion efficiency, so unidirectional operation is maintained (and will be maintained to 6.7% conversion efficiency).

An estimate of total resonator losses was made from measurement of the losses of individual components, combined with an estimate of losses by reflection at the LBO faces due to depolarization resulting from thermally induced birefringence in the laser rod. Because the oven housing prevented access to these reflections, direct measurement of the loss was not possible and we assumed a depolarization loss of 0.6%, similar to that of the fundamental ring case,  [7] allowing for the stronger polarizing effect of the LBO crystal. The estimated linear losses of 1.8% were less than for the KTP case (2.5%), and the observed increase in circulating fundamental power, from ~92 W for KTP to ~190 W for LBO, was reasonably consistent with these loss figures when combined with the conversion losses.

With this resonator the highest SH power generated was 3.3 W, of which 2.5 W was useful output since 20% was lost at the Brewster-angled exit face of the LBO and 6% was lost through imperfect transmission of the output mirror. Single-frequency operation was confirmed by measurement with a Fabry-Perot interferometer (see Fig. 2) and was found to be maintained with excellent stability.

Fig. 2. Typical 7.6 GHz confocal scanning Fabry-Perot trace for the 1064 nm leakage, confirming single-frequency operation. (FSR:- free spectral range).

No evidence was seen of the tendency for self-misalignment, to avoid the SH loss, as had occurred for KTP. Figure 3(a) shows a graph of output against pump power. It can be seen from this figure that the output is maximum at 12 W of pump power. Above this, the laser output fell rapidly and could not be recovered by resonator adjustments, such as by an adjustment of the concave mirror separation. In fact, various values of the plane/convex mirror separation were tried, and 440 mm was found to give the best performance. As this roll-off effect was observed for a variety of different reflectivities of the plane mirrors, and hence different intracavity powers, it is not thought that any absorption loss (linear or nonlinear) in the LBO crystal is responsible for this roll off. It is probable that the reduction in out put power is due to pump-induced thermal distortion the laser rod. It was observed that the power r off was accompanied by a degradation in fundamental beam quality, thus compounding the roll off in SH efficiency.

Figure 3. Incident pump power versus (a) 532 nm output power and (b) the intracavity fundamental power.

At the 2.5 W output level (3.3 W generated), the harmonic beam had measured M2 values of Mx2 = 1.1, My2 = 1.05. Under these conditions the conversion efficiency from circulating fundamental intensity was ~1.7%, agreeing reasonably well with the calculated value (2.1%) based on a deff of 0.8 pm/V. Up to these power levels it did not prove necessary to include an étalon in the resonator prevent oscillation at 1061.4 nm. However, based on our estimates of resonator loss (linear and nonlinear), an étalon would be needed if any increase in internal conversion efficiency is achieved. In fact we have used an étalon (fused quartz, 100 μm thick to deliberately enforce oscillation at 1061.4 nm, that this could be frequency doubled. The phase matching temperature was increased by 2 °C, an an output of 1.6 W was obtained at 530.7 nm. Under the conditions of these measurements, the laser was operating at 2.0 W output at 532 nm when the étalon and LBO temperature were optimized for 1064 nm, indicating some excess losses in the laser. So the 530.7 nm performance, which is only slightly lower, suggests that good SH efficiency on a number of other Nd transitions should be achievable.

In conclusion, we have shown that LBO, despite its lower nonlinearity than KTP, can nevertheless achieve comparable intracavity SH conversion efficiency at the level of ~3 W of generated SH power The prospects for scalability are, however, better for LBO. The results reported here were obtained with a nonoptimum resonator, making use of a limited range of available mirrors. With optimization, further power improvements can be expected. How ever, it is also clear that a good understanding and control of pump-induced distortions in the laser rod will be important for scaling to higher powers [8].

This research was supported by the Engineering and Physical Sciences Research Council. K. I. Martin acknowledges the support of Lumonics, Ltd in the form of a Cooperative Award in Science and Engineering studentship.

1. K. I. Martin, W. A. Clarkson, and D. C. Hanna, "3W of single-frequency output at 532 nm by intracavity frequency doubling of a diode-bar-pumped Nd:YAG ring laser", Opt. Lett. 21, 875-877 (1996).
2. J. T. Lin, J. L. Montgomery, and K. Kato, "Temperature-tuned noncritically phase-matched frequency conversion in LiB305 crystal", Opt. Commun. 80, 159-165 (1990).
3. J. L. Nightingale and J. K. Johnson, "0.6 W stable, single-frequency, green laser", in Compact Blue-Green Lasers, Vol. 1 of 1994 OSA Technical Digest Series (Optical Society of America, Washington, D.C., 1994), pp. 10-13.
4. H. W. Kogelnik, A. Dienes, and C. V. Shank, "Astigmatically compensated cavities for cw dye lasers", IEEE J. Quantum Electron. QE-8, 373-379 (1972).
5. D. C. Hanna, "Astigmatic Gaussian beams produced by axially asymmetric laser cavities", IEEE J. Quantum Electron. QE-5, 483-488 (1969).
6. Chun-Ching Shih, "Direction-independent polarisation rotator", in Conference on Lasers and Electro-Optics, Vol. 15 of 1995 OSA Technical Digest Series (Optical Society of America, Washington, D.C., 1995), pp. 112-113.
7. W. A. Clarkson, K. I. Martin, and D. C. Hanna, High-power single-frequency operation and efficient intracavity frequency doubling of a Nd:YAG ring laser end-pumped by a 20 W diode bar", in Conference on Lasers and Electro-Optics, Vol. 15 of 1995 OSA Technical Digest Series (Optical Society of America, Washington, D.C., 1995), pp.19-20.
8. S. C. Tidwell, J. F. Seamans, M. S. Bowers, and A. K Cousins, "Scaling cw diode-end-pumped Nd:YAG lasers to high average powers", IEEE J. Quantum Electron. 28, 997-1009 (1992).

Applied Optics (1997) Vol.36(18) pp.4149-4152

doi: 10.1364/AO.36.004149

Southampton ePrint id: 77990




Copyright University of Southampton 2006