Open Access
Add to BibSonomyAbstract
We present the investigation of nonlinear mirror modelocking (NLM) of a bounce amplifier laser. This technique, a potential rival to SESAM modelocking, uses a nonlinear crystal and a dichroic mirror to passively modelock a Nd:GdVO4 slab bounce amplifier operating at 1063nm. At 11.3W, we present the highest power achieved using the NLM technique, using type-II phase-matched KTP, with a pulse duration of 57ps. Using type-I phase-matched BiBO, modelocking was achieved with a shorter pulse duration of 5.7ps at an average power of 7.1W.
©2010 Optical Society of America
1. Introduction
High average power, picosecond lasers are an invaluable tool in industrial, scientific and medical applications. Passive modelocking is a well established technique of obtaining such lasers and semiconductor saturable absorber mirror (SESAM) modelocking provides an efficient and simple scheme that can be easily incorporated into a laser system [1]. However, SESAM’s are not without their problems [2]; the inherent absorption process upon which they rely for operation can cause damage and pulse duration limitation and each SESAM must be tailored for a particular wavelength, adding complexity for interchanging wavelengths. Here, we present a potential alternative to SESAM modelocking.
Nonlinear mirror (NLM) modelocking, first proposed by Stankov [3,4], is an elegant way in which modelocking can be achieved and can address the issues that limit SESAM operation. Comprising a nonlinear crystal (NLC) and dichroic output coupler, the basis of its operation is that of second harmonic generation (SHG) and difference frequency generation (DFG). Since these are nonlinear processes and do not involve any absorptive processes, the NLM is not inherently susceptible to absorptive damage, save for any residual linear absorption of the fundamental or second harmonic wavelengths. Additionally, the SHG nonlinear process is instantaneous, so the pulse duration is not limited by any modelocking mechanism recovery time. It could instead be limited by the spectral bandwidth of phase-matching and dispersion effects in the NLC.
The basis of NLM modelocking involves the introduction of a χ(2) nonlinear medium into the laser cavity which is optimised for SHG. A dichroic mirror with high reflectivity (HR) for the second harmonic and partial reflectivity for the fundamental wavelength is used as the cavity’s output coupler. As the second harmonic is reflected back through the NLC, with the correct phasing it can be converted back to the fundamental via DFG. Stankov’s rigorous analysis shows that the nonlinear reflectivity, RNL, is dependent on the second harmonic conversion efficiency, η, as well as the dichroic mirror reflectivities for the fundamental and second harmonic, Rω and R2ω, given in Eqs. (1) and (2) [3]:
where:Since the second harmonic efficiency increases with fundamental intensity, the nonlinear reflectivity is thus intensity dependent. This means that the leading and trailing portions of a pulse will be attenuated more than the peak, by greater output coupling from the cavity, and the full width half maximum (FWHM) of the pulse will decrease with each pass through the NLM. Such pulse-shortening is the basis for the formation of modelocked pulses.
High energy picosecond generation has previously been achieved with Nd:YVO4 as the lasing medium, using the NLM as the modelocking mechanism [5] and recently with Nd:GdVO4 [6], but at relatively low output powers. In this paper, high power modelocking using the NLM technique and the bounce geometry laser [7,8] is investigated. The bounce laser has previously been modelocked using SESAM’s [9], giving some success; however, these SESAM’s frequently experienced gradual deterioration and sometimes catastrophic damage due to the high peak powers produced. By implementing NLM modelocking, we hope to extend the power range without experiencing these issues.
2. Experimental
The bounce geometry utilises a single total internal reflection at the pump face of the lasing slab, where maximum gain is accumulated. An experimental bounce laser cavity was constructed, as shown in Fig. 1 . The bounce amplifier was a 20 × 5 × 2 mm Nd:GdVO4 slab, contact cooled via the 20 × 5 mm faces. A 40W pump diode was used, lasing at 808nm, so the front 20 × 2mm face was anti-reflection (AR) coated for the pump wavelength. The diode polarisation was parallel to the c-axis of the laser crystal.
Fig. 1 Schematic of bounce geometry laser, including a nonlinear mirror composed of a nonlinear crystal and dichroic output coupler
The two 5 × 2mm emitting faces were AR coated for the lasing wavelength of 1063nm and angled at 5° to the pump normal to minimise parasitic oscillation in the high gain amplifier. A 12.7mm vertical cylindrical lens (VCLD), AR coated at the diode wavelength, was used to generate a line focus on the pump face and two further vertical cylindrical lenses (VCL) of focal length 50mm were implemented to efficiently match the vertical mode size of the cavity to that of the gain region. An internal bounce angle of ~7° with respect to the pump face was used. Such a system has been shown to yield high powers efficiently [8]. Figure 2 shows the power curve obtained with this system operating in the continuous-wave (CW) regime.
The power curve in Fig. 2 shows the output power vary with increasing pump power through two stability zones, separated by a zone of instability. The power increases almost linearly until approximately 20W, where the power decreases significantly and subsequently recovers after 30W. This variation of output power is due to the changing stability of the resonator with thermal lens power and the incidence of these two zones of stability is a general feature of an asymmetric resonator that includes a thermal lens [10]. Modelocked operation occurs in the second stability zone.
2.1. NLM modelocking using type-II phase-matched SHG
A pair of lenses with focal length f = 100mm were incorporated into the bounce laser cavity to form a focus in the NLC and subsequently recollimate the beam. A type-II phase-matched KTP crystal was positioned at the focus and a half-wave retardation plate (HWP) was placed in the cavity to ensure the correct polarisation of the fundamental beam entering the crystal. A dichroic output coupler of 70% reflectivity at 1063nm and highly reflecting (HR) at 532nm was employed, positioned at a distance from the NLC such that the correct phase difference between the fundamental and second harmonic was achieved for DFG, by utilising the air dispersion between the fundamental and second harmonic wavelengths. The HWP was rotated to an angle such that maximum second harmonic conversion was observed and minor adjustments were made to the phase-matching angle and NLC-output coupler distance until CW modelocking was observed.
Figure 3 shows a pulse train from the CW modelocked output. The pulse repetition rate was 137MHz. An autocorrelation was performed using a purpose built background-free autocorrelator and the results plotted. By applying a sech2 deconvolution factor to the autocorrelation pulse FWHM, the pulse duration of 57ps was obtained. The average output power of this system was measured to be 11.3W – the highest power recorded to date of a NLM modelocked system – for a pump power of 40W. The pulse energy was 82nJ.
In addition to the temporal aspect of modelocking, a spatial phenomenon was also observed: during modelocking the central portion of the beam appeared to dim. We believe that this is a spatial characteristic of the NLM technique, which is strongest at the centre of the beam – since the NLM is more reflecting at higher intensities, as discussed in Section 1. This is the first time, to the best of our knowledge, that such an observation has been reported. The beam quality profile shown in Fig. 4 had a measured M2 of 1.2 in the vertical and 1.5 in the horizontal.
The pulse duration in this system is limited by the different propagation times of the o and e polarisations of the fundamental and second harmonic which leads to a group velocity mismatch. This also has implications on the beam quality: when the fundamental passes back through the HWP, it does not return to linear polarisation. Since the Nd:GdVO4 itself is a birefringent material, lasing selectively along the c-axis, the elliptically polarised light will subsequently be degraded by the lasing medium.
2.2. NLM modelocking using type-I phase-matched SHG
To address the pulse duration and beam quality issues arising with type-II phase-matching, the KTP was replaced with type-I phase-matched BiBO. The HWP was removed and minor adjustments were made to accommodate the new crystal. Since BiBO has a small acceptance angle for phase-matching in the critical phase-matching direction, careful adjustments were required to achieve modelocking. Once again, CW modelocking was observed and an autocorrelation taken to determine the pulse duration, the result of which is shown in Fig. 5 . At an average output power of 7.1W, a pulse duration of 5.7ps was measured at repetition rate of 144MHz, corresponding to a pulse energy of 50nJ.
Fig. 5 Autocorrelation of modelocked pulses with FWHM of 8.8ps; applying a sech2 deconvolution factor, this gives a pulse duration of 5.7ps.
3. Conclusion
For the first time, the NLM modelocking technique has been used to modelock a bounce geometry laser. The aim was to develop a modelocked system that was durable and simple to implement, without experiencing the damage that SESAM’s often incur in high power systems. CW modelocking was attained using both type-I and type-II phase-matched nonlinear crystals.
Initially, type-II phase-matched KTP was used. The system was self-starting and remained modelocked over the course of several hours with a pulse duration of 57ps and an average power of 11.3W. This is the highest power ever recorded in a NLM modelocked system, exploiting the benefits of such a high-gain system as the bounce geometry laser. This system highlighted a pulse duration limitation in the use of a type-II NLC, due to a group velocity mismatch arising from the different fundamental and second harmonic polarisations. As such, the KTP was replaced with a type-I phase-matched BiBO.
Using BiBO, CW modelocking was achieved at an average power of 7.1W, with a pulse duration of 5.7ps. These results show that NLM modelocking could be a potential rival to the more commonly used SESAM modelocking; high power systems can be produced without the absorption damage incurred in SESAM systems.
Since it relies simply upon choice of NLC, it can be applied at any wavelength without any intricate design changes. It can be further power scaled without the absorption damage limitations by increasing pump power. Furthermore, a master-oscillator power-amplifier (MOPA) system can be implemented simply in bounce geometry systems [11].
References and links
1. U. Keller, K. J. Weingarten, F. X. Kartner, D. Kopf, B. Braun, I. D. Jung, R. Fluck, C. Honninger, N. Matuschek, and J. Aus der Au, “Semiconductor Saturable Absorber Mirrors (SESAM’s) for Femtosecond to Nanosecond Pulse Generation in Solid-State Lasers,” IEEE J. Sel. Top. Quantum Electron. 2(3), 435–453 (1996). [CrossRef]
2. P. Ji-ying, M. Jie-guang, W. Yong-gang, W. Bao-shan, T. Hui-ming, Q. Long-sheng, and M. Xiao-yu, “High average power and high conversion efficiency continuous-wave modelocked Nd:YVO4 laser with semiconductor absorber mirror,” Opt. Laser Technol. 39(6), 1135–1139 (2007). [CrossRef]
3. K. A. Stankov, “A Mirror with an Intensity-Dependent Reflection Coefficient,” Appl. Phys. B 45(3), 191–195 (1988). [CrossRef]
4. K. A. Stankov and J. Jethwa, “A new modelocking technique using a nonlinear mirror,” Opt. Commun. 66(1), 41–46 (1988). [CrossRef]
5. G. I. Petrov, V. V. Yakovlev, and N. I. Minkovski, “Broadband nonlinear optical conversion of a high-energy diode-pumped picosecond laser,” Opt. Commun. 229(1-6), 441–445 (2004). [CrossRef]
6. H. Iliev, D. Chuchumishev, I. Buchvarov, S. Kurimura, V. Petrov, and U. Griebner, “Passive modelocking of a Nd:GdVO4 laser by intracavity SHG in periodically-poled stoichiometric lithium tantalate”, OSA ASSP Conf. Proc. (2010)
7. M. J. Damzen, M. Trew, E. Rosas, and G. J. Crofts, “Continuous-wave Nd:YVO4 grazing-incidence laser with 22.5 W output power and 64% conversion efficiency,” Opt. Commun. 196(1-6), 237–241 (2001). [CrossRef]
8. A. Minassian, B. Thompson, and M. J. Damzen, “Ultrahigh-efficiency TEM00 diode-side pumped Nd:YVO4 laser,” Appl. Phys. B 76(4), 341–343 (2003). [CrossRef]
9. D. J. Farrell and M. J. Damzen, “High power scaling of a passively modelocked laser oscillator in a bounce geometry,” Opt. Express 15(8), 4781–4786 (2007). [CrossRef] [PubMed]
10. V. Magni, “Resonators for solid-state lasers with large-volume fundamental mode and high alignment stability,” Appl. Opt. 25(1), 107–117 (1986). [CrossRef] [PubMed]
11. A. Minassian, B. Thompson, G. Smith, and M. J. Damzen, “High-power scaling (>100 W) of a diode-pumped TEM00 Nd:GdVO4 laser system” IEEE J. Sel. Topics in Q. Elec. 11, 621–625 (2005). [CrossRef]
