We report a high power tunable femtosecond soliton-based source using a simple combination of fiber-amplified pulses at 1064nm and hollow-core photonic bandgap fiber. Compression of 5.5ps input pulses, strongly chirped by self phase modulation in the amplifier, results in stable 520fs-soliton formation with 77% conversion efficiency after only 8m propagation in the hollow-core fiber. The Raman self-frequency shift of the solitons was used to provide 33nm wavelength tuneability. The transform-limited output pulses were frequency doubled using a nonlinear crystal with high conversion efficiency of 60% to demonstrate a femtosecond green laser tunable from 534nm to 548nm with 180nJ pulse energy.
©2008 Optical Society of America
Compact and cost effective femtosecond fiber-based pulsed laser sources emitting in the near infrared and visible regions are attractive for a variety of applications including spectroscopy and microscopy. Since their first realization, Hollow-Core Photonic Bandgap Fibers (HC-PBGFs) have attracted significant interest due to their unusual optical properties . One of the most spectacular is their low nonlinear response, typically 3 orders of magnitude less that that of conventional fiber. In consequence, they have been used to demonstrate the propagation of high-power pulses . Another important property is the unusual group velocity dispersion curve. The group-velocity dispersion (GVD) of the guided mode is primarily determined by the dispersion of the waveguide, causing strong normal and anomalous dispersion at short- and long-wavelength edges of the bandgap respectively. Recently, linear compression in HC-PBGF of laser pulses which were strongly chirped by self phase modulation (SPM) has been shown to enable design of compact femtosecond sources, but with time-bandwidth product larger than 0.8 [3,4]. HC-PBGFs have also been used for generation and guidance of high-power optical solitons [5–7]. This effect arises at high energy when the greatly reduced Kerr nonlinearity balances the effect of anomalous dispersion, for a transform-limited input pulse. In this paper, we demonstrate both pulse compression and subsequent soliton propagation. We show that a HC-PBGF only 8m-long enables efficient and stable soliton generation from a strongly chirped input pulse by the combined effects of anomalous dispersion, Raman scattering and self-phase modulation. 5.5ps input pulses at wavelengths around 1064nm evolved into stable 520fs solitons with 300nJ maximum pulse energy. In a subsequent experiment, we used the self-frequency-shifting solitons to generate second harmonic (SH) radiation with high conversion efficiency (around 60%) and tuneability (from 534nm to 548nm) in a temperature-controlled LBO crystal. This simple single-pass combination of a chirped picosecond laser and HC-PBGF provides compact, efficient and tunable femtosecond sources at infra-red and green wavelengths.
2. Linear properties of HC-PBGF
An 8m-long piece of HC-PBGF was used in the following experiments. The transmission band was centered at 1064nm with a full width at half maximum (FWHM) of 150nm. At this wavelength, the attenuation reached a minimum of 50dB/km and the group-velocity dispersion (GVD) was around +93ps/nm/km, enabling soliton propagation. The measured values refer to one of the two polarization-modes which was used in subsequent experiments: values for the other mode are very similar (but not identical).
3. Pulse compression experiments: formation of solitons
3.1 Soliton source configuration
Our experiments were performed using the FemtoPower 1060-10 (Fianium Ltd) with average power of up to 10W, a central wavelength of 1064nm and a repetition rate of 20MHz. Inside the laser, 5.5ps transform-limited pulses from a seed oscillator undergo strong self-phase modulation (SPM) in a fiber amplifier to yield 5.5ps output pulses with broadened bandwidth and strong normal chirp. These pulses were coupled into the HC-PBGF using a ×10 objective lens. The efficiency obtained was greater than 70% (including transmission losses through the 8m of fiber). The pulse power was controlled using a polarizing waveplate and beamsplitter (PBS) combination, and a second waveplate was used to align the polarization axis of the beam with one of the polarization axes of the HC-PBGF. The fiber output was characterized using a spectrometer and an autocorrelator based on 2-photon absorption in a GaAs LED. Typical results for an input bandwidth of 12.5nm (6.4W amplifier power / 320nJ) are plotted in Fig. 2(a), showing strong pulse compression by a factor 10. The graph can be divided into two parts. At energies up to 80nJ, linear pulse compression appears due to the anomalous dispersion of the HC-PBGF. With the 5.5ps, 12nm bandwidth, normally chirped input pulses, linear propagation over 4.5m of HC-PBGF with an anomalous dispersion of 93ps/nm/km will compensate the chirp. Propagation over the remainder of the 8m of HC-PBGF gives anomalously chirped output pulses of 3ps duration. Above 80nJ output energy the output pulse duration again reduces, because now there is a contribution from the (weak) nonlinear response of the air core fiber, allowing the effects of SPM to be counteracted by the anomalous dispersion of the fiber. When the balance is reached (for 150nJ output energy / 220nJ input energy), pulse propagation is stable, with the pulse length remaining almost constant at an average value of 807fs (corresponding to 520fs assuming a sech2 pulse). Finally, 77% of the pulse energy (170nJ) is redistributed within the spectrum to yield the characteristic smooth spectrum and small time-bandwidth product of a soliton (measured product is 0.35). Spectral components not transferred into the soliton propagate as dispersive waves. Figure 2(b) shows the input and output spectra obtained after 8m-propagation at 320nJ laser pulse energy. In the attached movie [AVI, 311KB], intermediate output spectra have been recorded showing the soliton formation versus the laser pulse energy.
In order to better understand what occurs along the fiber length, we carried out cut back measurements to study the evolution of pulse length and spectrum. The results are shown in Figs. 3(a) and 3(b). For this study the laser pulse energy was fixed to 320nJ [as in Fig. 2(b)] and repeated autocorrelation and spectral measurements were made as the 8m length of fiber was cut back in 50cm increments. Figure 3(a) shows that after 4.5m-propagation, the initially chirped pulse was compensated by the anomalous dispersion of the hollow-core fiber. The pulse duration at this point was measured to be 550fs (autocorrelation width) and no noticeable pedestal was observed in the autocorrelation trace. The intensity was then high enough to enable soliton formation, which stabilized after 6.5m-propagation (after roughly one additional soliton length) to reach 520fs duration.
Moreover, the soliton component shifts to lower frequencies (4nm-shift giving an output central wavelength centered at 1073nm) due to the Raman self-scattering from the air and the overlap of the fundamental guided mode with the silica . In conclusion a compact and efficient soliton source has been made by the simple combination of a picosecond fiber laser and an 8m-long HC-PBGF.
3.2 Tuneability of the soliton source
In a second experiment, propagation at much higher energies up to 2µJ has also been studied. The experiments were done using a lower repetition rate of 1MHz rather than 20MHz. Similar results in terms of soliton energy threshold and pulse duration were found.
However, the output stability is greatly improved for a given energy as the thermal effects of the average power at the fibre end face are greatly reduced: at 1MHz, 320nJ pulses correspond to 320mW rather than the 6.4W at 20MHz. The results plotted in Fig. 4(a) show that the solitons formed have similar durations as the pulse energy is increased, but self-frequency shift to longer wavelengths (only fundamental soliton is observed). Figure 4(b) shows the output shift of the solitons in the spectral domain. For up to 1.6µJ input energy, the soliton is shifted to reach 1103nm (giving 33nm-tuneability) without significant change in characteristics despite the GVD increasing by a factor of four [see Fig. 1(b)]. Increasing the input energy has some consequences on the linear/nonlinear propagation boundary inside the fiber (leading to a nonlinear stage before 4.5m propagation) and could explain this result. Moreover, the fraction of the output energy redistributed into the soliton changes from 77% at 320nJ laser pulse energy down to 21% at 2µJ to achieve 300nJ soliton energy (graph not shown).
4. Tunable femtosecond green laser by frequency doubling
The results presented in the previous section demonstrate that the HC-PBGF acts as a stable pulse cleaner leading to a tunable femtosecond soliton source. We have frequency-doubled the output solitons to obtain tunable femtosecond laser at visible (green) wavelengths. Here due to the soliton characteristics (small time-bandwidth product compared to [8,9]) high conversion efficiency can be expected. The short temporal duration yields high power, and the narrow spectral bandwidth (3nm) allows the use of relatively long crystals whilst remaining within the phase-matching bandwidth.
We used the picosecond laser and the 8m-long piece of HC-PBGF as a pulse source, as illustrated in Fig. 5. A 5mm lithium triborate (LBO) crystal (type I non-critically phase-matched) with 2nm phase-matching bandwidth is added to generate the second harmonic. The conversion was optimized using a half-wave plate and by temperature-tuning the LBO crystal. At the output, dichroic mirrors, a spectrometer, and a power meter are used to select and analyze the green wavelength component. The spectral evolution and the SH radiation shift versus the laser input energy are plotted in Figs. 6(a) and 6(b) respectively. By exploiting the 33nm Raman self frequency shift of the fundamental soliton [see Fig. 4(b)], the second harmonic can be tuned from 534nm to 548nm while maintaining a clean spectral shape. The measured conversion efficiency (shown in Fig. 7) is around 60%, yielding a maximum 180nJ output pulse energy (or 180mW average power).
We believe that the 2nm-bandwidth of the LBO crystal is the limiting factor in conversion efficiency, and 3nm bandwidth phase-matching should allow us to achieve 70% conversion. Finally, the pulse duration of the green laser is still around 500fs due to the low dispersion of the 5mm LBO crystal.
In summary, a picosecond fiber-based source combined with pulse compression and soliton propagation in 8m of HC-PBGF provides a stable and compact sub-picosecond source of energetic pulses. The 5.5ps input pulses were compressed by a factor of over 10, giving 520fs-solitons (time bandwidth product 0.35) with 300nJ maximum energy and 33nm-tuneability. For further reduction of the pulse duration, tapered HC-PBGFs are under investigation . Frequency doubling of the solitons was then demonstrated, leading to tunable green femtosecond laser pulses. The second harmonic process was very efficient (60% measured and 70% expected with crystal bandwidth optimized) giving 180nJ output pulse energy and the Raman self-frequency shift was used to provide wavelength tuneability from 534nm to 548nm.
The authors would like to thank A.K. George and S. Renshaw for fiber fabrication. This work was funded by the Technology Strategy Board-led Technology Programme (TP/4/NGL/6/I/22227). W.J. Wadsworth is a Royal Society University Research Fellow.
References and links
1. R. F. Cregan, B. J. Mangan, J. C. Knight, T. A. Birks, P. St. J. Russell, P. J. Roberts, and D. A. Allan, “Single-mode photonic band gap guidance of light in air,” Science 285, 1537–1539 (1999). [CrossRef] [PubMed]
2. G. Humbert, J. C. Knight, G. Bouwmans, P. St. J. Russell, D. P. Williams, P. J. Roberts, and B. J. Mangan, “Hollow core photonic crystal fibers for beam delivery,” Opt. Express 12, 1477–1484 (2004). [CrossRef] [PubMed]
4. B. Ortaç, M. Plötner, T. Schreiber, J. Limpert, and A. Tünnermann, “Experimental and numerical study of pulse dynamics in positive net-cavity dispersion modelocked Yb-doped fiber lasers,” Opt. Express 15, 15595–15602 (2007). [CrossRef]
5. D. G. Ouzounov, F. R. Ahmad, D. Muller, N. Venkataraman, M. T. Gallagher, M. G. Thomas, J. Silcox, K. W. Koch, and A. L. Gaeta, “Generation of megawatt optical solitons in hollow-core photonic band-gap fibers,” Science 301, 1702–1704 (2003). [CrossRef] [PubMed]
6. F. Luan, J. C. Knight, P. St. J. Russell, S. Campbell, D. Xiao, D. T. Reid, B. J. Mangan, D. P. Williams, and P. J. Roberts, “Femtosecond soliton pulse delivery at 800nm wavelength in hollow-core photonic bandgap fibers,” Opt. Express 12, 835–840 (2004). [CrossRef] [PubMed]
7. F. Gérôme, K. Cook, A. K. George, W. J. Wadsworth, and J. C. Knight, “Delivery of sub-100fs pulses through 8m of hollow-core fiber using soliton compression,” Opt. Express 15, 7126–7131 (2007). [CrossRef] [PubMed]
8. G. McConnell and A. I. Ferguson, “Simultaneous stimulated Raman scattering and second harmonic generation in periodically poled lithium niobate,” Opt. Express 13, 2099–2104 (2005). [CrossRef] [PubMed]
9. C. J. S. de Matos, R. E. Kennedy, S. V. Popov, and J. R. Taylor, “20-kW peak power all-fiber 1.57-µm source based on compression in air-core photonic bandgap fiber, its frequency doubling, and broadband generation from 430nm to 1450nm,” Opt. Lett. 30, 436–438 (2005). [CrossRef] [PubMed]