Open Access
Add to BibSonomyAbstract
We investigated the characteristics and behavior of spectral compression in a quasi-dispersion-increasing comb-profile fiber (CPF). A periodical breathing behavior and sidelobe emission process in the CPF were observed in numerical analysis. Then, taking account of the numerical results, we developed an improved CPF in which the sidelobe suppression was dramatically improved to −24.2 dB while keeping a narrow spectral width of ~0.6 nm. As a seed pulse source, we developed a high-repetition-rate Er-doped ultrashort-pulse fiber laser with single-wall carbon nanotubes and used the improved CPF to realize a high-power, narrow-linewidth source with wide wavelength tunability in the 1.62–1.90 μm band.
© 2016 Optical Society of America
1. Introduction
Rapid, intense, wideband-wavelength-tunable, narrow-linewidth sources are highly desired for applications such as spectroscopy, nonlinear microscopy, and optical coherence tomography. So far, rapid wavelength-tunable laser sources have been developed in the field of swept-source optical coherence tomography (OCT) in particular [1–3]. In these light sources, the wavelength tuning bandwidth is limited by the bandwidth of the gain device, and wideband operation has been difficult to achieve. Since the axial resolution of OCT is determined by the bandwidth of the light source, the wideband sources have been desired to realize the ultrahigh resolution, swept source OCT. In addition, the maximum tuning speed has been ~a few hundred kHz, and only the continuous wavelength tuning operation has been possible in the current wavelength tunable source, and it has limited the range of applications.
By using ultrashort-pulse, anomalous-dispersion fibers, we can generate widely wavelength tunable soliton pulses through a combination of the soliton effect and intra-pulse stimulated Raman scattering [4–7]. As the fiber input power is increased, the wavelength of the generated soliton pulses is continuously shifted toward the longer wavelength side. Wide, continuous wavelength tuning operation can be achieved by intensity control [8], and fast tuning operation at a few tens of GHz can be achieved by using an electro-optic (EO) modulator. To the best of our knowledge, this is the fastest wavelength tunable light source. This also allows the wavelength of the generated pulse to be tuned shot by shot, and arbitrary wavelength tuning can be achieved by controlling the EO modulator. If spectral compression of the wavelength tunable ultrashort pulse can be achieved, it will be possible to realize a narrow-linewidth, rapid, wideband, and freely wavelength tunable pulse source. This light source would enable us to demonstrate ultrafast, and ultrahigh resolution swept source OCT, and functional spectroscopic applications.
Spectral compression of ultrashort pulses was first observed in a single-mode fiber (SMF) [9]. A prechirped ultrashort pulse was coupled into SMF, and spectral compression was induced by self-phase modulation (SPM) [9,10]. Spectral compression was also observed in gain fibers [11,12] and photonic crystal fibers (PCF) [13–16]. Large spectral compression ratios of 16 and 21 were achieved in a gain fiber [11] and PCF [13], respectively. However, sidelobe components generally arise in the compressed pulse spectra, and it is difficult to obtain a high-quality narrow spectrum.
In 2010, we demonstrated wideband spectral compression by adiabatic soliton spectral compression in a quasi-dispersion-increasing fiber (DIF) [17]. Using adiabatic soliton spectral compression in a DIF, a high-quality, large spectral compression can be obtained by exploiting the soliton effect [17–19]. We realized a quasi-DIF by use of a comb-profile fiber (CPF) technique, which was originally developed for temporal pulse compression and high-repetition-rate pulse train generation [20,21]. A conventional single-mode fiber (SMF) and a dispersion-shifted fiber (DSF) were spliced alternately, and the magnitude of the average dispersion was designed to increase along the fiber. Using the CPF technique, we can control the dispersion profile by changing the length of spliced conventional fibers. Spectral compression of a wavelength-tunable soliton pulse has been demonstrated in a wide wavelength range [17]. The observed maximum spectral compression factor was up to 25.6 using a 200 fs soliton pulse. Real-time absorption spectroscopy has been demonstrated by rapid wavelength tuning using an EO modulator [22]. Recently, the largest spectral compression ratio of 28.6 was reported using 1 km of DIF and a 112 fs ultrashort pulse [23].
The spectral compression of optical pulses has also been investigated in the field of all-optical signal processing, especially all-optical analog to digital conversion [24,25]. Spectral compression of a wavelength-tunable soliton pulse in a CPF was demonstrated using a combination of highly nonlinear fiber and single-mode fiber [26,27]. Numerical analysis of the spectral compression of a Gaussian-shaped pulse in a CPF consisting of highly nonlinear fiber was reported in the field of optical signal processing [28].
In the fields of OCT and accurate spectroscopic measurement, pedestal components degrade the signal-to-noise ratio. Therefore, improvement of the spectral compression is necessary to widen the range of applications of this kind of light source. A few studies on analysis of spectral compression have been reported [19,28–30]; however, the characteristics of adiabatic spectral compression of soliton pulses in CPF have not been investigated, and the detailed performance has not been clarified yet.
In this work, we investigated the characteristics of adiabatic soliton spectral compression in quasi-dispersion-increasing CPF. The characteristics and behavior of spectral compression in a conventional DIF and the CPF that we developed were investigated numerically, especially in terms of sidelobe suppression. By taking the numerical results into account, we developed an improved CPF, and we succeeded in dramatically improving the suppression of spectral pedestals. Then, we developed a high-repetition-rate, Er-doped ultrashort pulse fiber laser using single-wall carbon nanotubes and experimentally realized a high-power, high-quality, wideband-wavelength-tunable, narrow-linewidth source with the improved CPF. The characteristics of the wavelength dependence were examined in terms of wide wavelength-tuning operation.
2. Numerical analysis of adiabatic soliton spectral compression in dispersion-increasing fibers
In this section, we discuss the numerical analysis of adiabatic soliton spectral compression in dispersion-increasing fibers. As the input pulse, we assumed a transform-limited sech2-shaped pulse with a temporal width of 200 fs full-width at half maximum (FWHM). First, we examined input pulse center wavelengths of 1620 nm. Then the wavelength dependence of the spectral compression in the improved CPF was discussed in the following sections.
The pulse propagation along the fibers was investigated using the strict nonlinear Schrödinger equations [17,31],
where A = A(z,T) represents the complex electric field envelope, z is distance, and T = t-β1z. The symbols β1, β2, and β3 represent the magnitudes of first-, second-, and third-order dispersions. The symbol α corresponds to the optical loss. The right-hand side represents the nonlinear effects. The individual terms correspond to SPM, self-steepening, and Raman scattering. The symbols γ, ω0, and TR represent the nonlinear coefficient, center angular frequency, and parameter corresponding to Raman response time, respectively. The split-step Fourier method was used for the numerical simulation.Since the spectral compression occurs through soliton effect, which is the combination of self-phase modulation and chromatic dispersion, the performance depends on the peak power of the input pulse [17,23]. In this work, the peak power was optimized to achieve the highest quality spectral compression with both the narrower spectral width and the smaller pedestal components under each condition.
For the fiber parameters, the dispersion properties and mode-field diameter (MFD) of the fibers were experimentally measured as a function of wavelength using the wavelength tunable soliton pulses and wideband supercontinuum [32–34].
2.1.1 Dispersion-increasing fiber (DIF) at λ = 1620 nm
First, we examined the spectral compression in a conventional DIF at a wavelength of 1620 nm. The initial spectral width was 12.3 nm FWHM for a 200 fs soliton pulse. Figure 1 shows the characteristics of spectral compression in the DIF. In this work, we used the following function to model the dispersion increasing behavior:
where β2(x) is the magnitude of second-order dispersion at position x, L is the total length of the DIF, β2L is the magnitude of second-order dispersion at position x = L, and Δβ2 is the magnitude of variation of β2. At the wavelength of 1620 nm, β2L was −30 ps2/km, and Δβ2 was 22 ps2/km. This means that the magnitude of dispersion gradually decreased from −8 to −30 ps2/km. The third order dispersion β3 = 0.16 ps3/km, and the MFD was assumed to be constant at 11.0 μm.
Fig. 1 Characteristics of spectral compression in DIF at λ = 1620 nm; propagation characteristics for (a) spectral width and β2, (b) temporal width and soliton order N. (c),(d) input and output pulse spectra on (c) linear (Visualization 1) and (d) log scales, (e),(f) instantaneous wavelength and temporal shapes of output pulse on (e) linear and (f) log scales.
Figure 1(a) shows the variation of the spectral width and magnitude of second-order dispersion β2 as a function of the fiber length. As the fiber length was increased, the magnitude of β2 decreased continuously and smoothly, and it saturated around −30 ps2/km at a length of 500 m. The spectral width narrowed continuously along the fiber propagation direction, and reached 0.78 nm at the fiber output.
Figures 1(c) and 1(d) show the optical spectra at both the fiber input and output. The variation of the pulse spectrum as a function of propagation length is also shown as a movie in Visualization 1. A compressed spectrum with pedestal components was generated at the fiber output. On the log scale, a skirt-like pedestal component was observed. The pedestal level was −15.0 dB from the peak. For the temporal shape, a small but wide pedestal component was observed around the main pulse. From the instantaneous wavelength, we can see that the pedestal component in the temporal shape corresponds to those on the spectral one. Since the fiber has anomalous dispersion properties, the shorter wavelength components are at the leading part, and the longer ones are at the trailing part. Almost linear down-chirping was confirmed for the pedestal components. For the main pulse, slight nonlinear chirping by the SPM was observed.
Figure 1(b) shows the variation of soliton order N and temporal width of the propagating pulse in the DIF. The temporal width increased from 0.2 ps up to 2.7 ps around 350 m, and then started decreasing slowly. The soliton order started from 0.65 and increased rapidly up to 1.36, and then it gradually decreased and approached 1.0, which corresponds to the fundamental soliton.
The spectral compression characteristics in the DIF depend on the parameters, such as the manner in which the dispersion increases, and they can be optimized even further. Actually, better performance was achieved in [19,23].
2.1.2 Original quasi-dispersion-increasing comb-profile fiber (CPF) at λ = 1620 nm
Next, we discuss spectral compression in a quasi-dispersion-increasing comb-profile fiber. This fiber was originally developed by us in our previous work [17]. Here, we used a conventional single-mode fiber (SMF) and a dispersion-shifted fiber (DSF). The MFDs were 11.0 μm in SMF and 8.5 μm in DSF at wavelength of 1620 nm. The third order dispersion β3 = 0.16 ps3/km in both fibers. These two fibers were spliced alternately by changing the ratio of their lengths, and the average value of the dispersion was increased.
Figure 2(a) shows the dispersion map and variation of the spectral width for the initially developed CPF. The CPF was designed so that the magnitude of average dispersion fitted the dispersion curve of the DIF. Considering the effect of fusion splicing loss, the fiber length was set to be longer than that of the DIF to achieve adequate spectral compression. As shown in Fig. 2(a), the spectral width was compressed continuously from 12.3 nm down to 0.78 nm.
Fig. 2 Characteristics of spectral compression in original CPF at λ = 1620 nm; propagation characteristics for (a) spectral width and β2, and (b) temporal width and soliton order. (c),(d) Input and output pulse spectra on (c) linear (Visualization 2) and (d) log scales. (e),(f) Instantaneous wavelength and temporal shapes and of output pulse on (e) linear and (f) log scales.
Figures 2(c) and 2(d) show the spectral shape of the output pulse. The variation of the pulse spectrum is also shown as a movie in Visualization 2. A spectrally compressed pulse with low sidelobe components was generated. On the log scale, we can see the pedestal components at both sides of the main pulse. The sidelobe level was −17.3 dB from the peak. For OCT and spectroscopy applications, these pedestal components cause background noise and degrade the signal-to-noise ratio of the measurement.
Figure 2(b) shows the variation of pulse duration and soliton order as a function of the fiber length. As shown in the figures, the pulse width was gradually increased with periodical breathing along the pulse propagation direction. The soliton order was also periodically changed and finally reached 1, which corresponds to the fundamental soliton.
Figure 3 shows the characteristics of spectral compression at short propagation length. The variation of the pulse spectrum and temporal pulse shape as a function of propagation length are also shown as movies in Visualization 3 and Visualization 4. From Fig. 3 and animations of thetemporal and spectral variations, we can confirm that the anomalous dispersion effect is dominant in the SMF, and the temporal width is broadened. In the DSF, on the other hand, SPM is dominant, and pulse compression occurs. As a result, periodical breathing of the temporal width of the pulse was observed at the comb-profile part.
Fig. 3 Characteristics of initial process of spectral compression in original CPF at λ = 1620 nm: (a) spectral (Visualization 3) and (b) temporal shape and instantaneous wavelength (Visualization 4) of propagating pulse at 25 m length; and (c)(d) propagation characteristics for pulse duration and soliton order.
In Fig. 3 and Visualization 3, the pedestal components were generated at short propagation length. In Visualization 4, we can see the generation process of the sidelobe components. It is considered that a large soliton order N induces the generation of pedestal components by dispersive wave emission during the soliton formation [31,35]. We noticed that the pedestal components were generated in the initial process of spectral compression. It is considered that improvement of the CPF configuration, especially at short propagation length, is important for obtaining a clean, ideal compressed pulse.
2.1.3 Improved quasi-dispersion-increasing comb-profile fiber (CPF) at λ = 1620 nm
By taking account of the results in the previous section, we improved the design of the CPF. Figure 4(a) shows the dispersion map and variation of spectral width for the improved CPF. We used the same dispersion curve of DIF given by Eq. (2) as the base, and we examined a lot of configurations of CPF in terms of the length, interval, and number of the pairs. Especially we focus on the initial part, and the averaged dispersion was changed precisely at lengths shorter than 30 m. The total number of splicing points was decreased from 19 to 16. The splice loss was 0.13 dB/point, and the corresponding reduction of splice loss was estimated to be ~0.39 dB. So far, this is the best profile in which we can achieve the highest quality spectral compression with both the narrower spectrum width and lower sidelobe components.
Fig. 4 Characteristics of spectral compression in improved CPF at λ = 1620 nm; propagation characteristics for (a) spectral width and b2, (b) temporal width and soliton order. (c),(d) Input and output pulse spectra on (c) linear (Visualization 5) and (d) log scales. (e),(f) Instantaneous wavelength and temporal shape of output pulse on (e) linear and (f) log scales.
Figures 4(c) and 4(d) show the optical spectra at the CPF output. The variation of the pulse spectrum is also shown as a movie in Visualization 5. We can see that the pedestal components were dramatically suppressed with the improved CPF. At the output of the CPF, the pedestal level was reduced below −24.2 dB from the spectral peak, which was lower than that of the original one by 6.9 dB. The spectral width of the output pulse was 0.61 nm, which is narrower than that of the original one by 0.17 nm. The corresponding spectral compression ratio was 20.2.
Figures 4(e) and 4(f) show the temporal waveforms of the output pulses. We can see that the pedestal components were negligibly small even on the log scale. For the instantaneous wavelength, the main part shows the nonlinear chirping property due to the effect of SPM. On the other hand, the sidelobe components had almost linear down-chirping properties due to the anomalous dispersion in the CPF.
Figure 4(b) shows temporal width and soliton order N as a function of propagation length. As the propagation length was increased, the temporal width was broadened after smallbreathing at the CPF part. It took a maximum of 4.4 ps at a length of around 500 m, and then it decreased slightly. The soliton order N reached 1.0 after breathing.
Figure 5 shows characteristics of spectral compression in the CPF at short propagation length. The pulse spectrum and temporal pulse shape as a function of propagation length are also shown as movies in Visualization 6 and Visualization 7.We can see that the soliton order was close to 1 at the initial part of the fiber, and N was lower than that in the original CPF and also increased gradually along the propagation length. The temporal pedestal component was also suppressed well. It is considered that since the soliton order N was close to 1 at the initial part, and increased slowly along the propagation length, the generation of pedestal components was suppressed well. As a result, the pedestal component was suppressed well using the improved CPF.
Fig. 5 Characteristics of initial process of spectral compression in improved CPF at λ = 1620 nm: (a) spectral (Visualization 6) and (b) temporal shape and instantaneous wavelength (Visualization 7) of propagating pulse at 25 m length; and (c)(d) propagation characteristics for pulse duration and soliton order.
2.2 Wideband operation of improved comb-profiled fiber (CPF)
2.2.1 Spectral compression properties at λ = 1770 nm
In order to examine the wideband operation properties, here we discuss the spectral compression performance at a 1770 nm wavelength region in the improved CPF.
Figure 6(a) shows the dispersion map and variation of the spectral width. The initial spectral width was 16.4 nm FWHM for a 200 fs soliton pulse. The peak power was optimized to achieve the best spectral compression performance. At 1770 nm, β2 = −48 ps2/km and MFD = 11.8 μm in SMF, and β2 = −25 ps2/km and MFD = 9.8 μm in DSF.
Fig. 6 Characteristics of spectral compression in improved CPF at λ = 1770 nm: propagation characteristics for (a) spectral width and β2, (b) temporal width and soliton order N; (c) output pulse spectra; and (d) temporal pulse shape and instantaneous wavelength of output pulse.
Smooth variation was observed for the spectral width and temporal width. The spectral width was narrowed continuously along the propagation length. The temporal width was broadened smoothly, took a maximum at a length of around 600 m, and then slightlydecreased. The minimum spectral width was 0.60 nm, and the corresponding spectral compression ratio was as large as 27.3. As for the soliton order N, it showed step-like varying behavior, and then increased along the propagation direction, and reached 1.0. As for the spectral shape, although a skirt-like pedestal was observed, the pedestal level was −14.5 dB from the peak.
For comparison, we examined the spectral compression properties in the DIF and original CPF at a wavelength of 1770 nm. As the results, the minimum spectral width was 0.71 nm and the pedestal level was −10.0 dB for the DIF. For the original CPF, the minimum spectral width was 0.81 nm, and the pedestal levels was −13.5 dB.
From these results, the highest results were obtained in the improved CPF, compared with those in the DIF and the original CPF. Although there were sidelobe components, a narrow linewidth of 0.6 nm was achieved, which is almost the same value as that at 1620 nm. Therefore, we confirmed that the improved CPF is effective both at the wavelength of 1620 nm and the wavelength of 1770 nm, and wideband operation can be expected.
2.2.2 Wavelength dependence of spectral compression in improved CPF
Next, we investigated the wavelength dependence of spectral compression in improved CPF numerically. The 200 fs sech2 shaped pulses at several different wavelengths were propagated along the improved CPF, and the properties of spectral compression were examined. Figure 7shows the experimentally measured dispersion properties and MFD of the used fibers as a function of wavelength. The wavelength dependence of the parameters shown in Fig. 7 was considered in the numerical analysis.
Fig. 7 Experimentally measured fiber parameters in SMF and DSF as a function of wavelength, (a) second-order dispersions β2, and (b) mode field diameter (MFD).
Figure 8 shows the numerical results of variation of compressed spectral width and sidelobe level at the output of improved CPF. The peak power was optimized to achieve the highest quality spectral compression. From Fig. 8, the compressed spectral width was narrower than 0.6 nm, and the sidelobe level was smaller than −13.5 dB for the whole wavelength range. The high spectral compression ratio of 20-30 was obtained.
Fig. 8 Numerical results of compressed spectral width and side lobe level of output pulses from improved CPF as a function of wavelength.
From Fig. 8, the sidelobe level was degraded as the wavelength was increased. The ratio of magnitude of dispersion is important factor for spectral and pulse compression [23]. From Fig. 7, we can see that the ratio of the dispersion was decreased as the wavelength was increased. It is considered that this decrement of dispersion ratio affects the sidelobe level of the compressed pulse.
3. Development of high-power wavelength-tunable narrow linewidth source using improved CPF
Next, we developed a high-power, wavelength-tunable, narrow-linewidth source using the improved CPF, and we examined the characteristics of spectral compression experimentally. The experimental setup is shown in Fig. 9. As the seed pulse source, we developed a passively mode-locked ultrashort-pulse Er-doped fiber laser using single-wall carbon nanotubes (SWNTs) [36]. The SWNTs work as a transmission-type saturable absorber. In this work, a thin polyimide film in which the SWNTs were dispersed was inserted between the fiber connectors inside the fiber laser cavity, and was used as a saturable absorber for passive mode-locking [36].
Fig. 9 Experimental setup of wavelength-tunable narrow-linewidth source using CPF and SWNT fiber laser. PBC: polarization beam combinor, LPF: long pass filter, VA: variable attenuator.
Since a single ultrashort pulse generates a single spectral compressed pulse, we can increase the average power of the output pulses by using a high-repetition-rate pulse source. In this work, a high-repetition-rate fiber laser operating at ~100 MHz was developed, and we used this as the seed pulse source. A high-concentration Er-doped fiber with anomalous dispersion properties was used as the gain fiber, and it was pumped by a 980 nm high-power LD. The cavity consisted of non-polarization-maintaining conventional single-mode fiber devices. A polarization controller was used to optimize the polarization conditions in the cavity.
Figure 10 shows the characteristics of output pulses from the developed SWNT fiber laser. The net dispersion of the fiber laser cavity was negative, and stable soliton mode-locking operation was achieved. As shown in Fig. 10(a), a sech2-like pulse was generated stably. The spectral width was 9.0 nm. The temporal width of the autocorrelation trace was 440 fs FWHM, which corresponds to a 266 fs ultrashort pulse, assuming a sech2 pulse. A pulse train with a repetition rate of 95 MHz was stably achieved. Clear RF spectra were observed, and stable mode-locking was confirmed. The average power of the laser output was 24 mW.
Fig. 10 Characteristics of output pulses: (a) optical spectra, (b) autocorrelation trace, (c) pulse train observed with fast photodiode and digital oscilloscope, and (d) rf spectrum of pulse train.
The output pulses from the fiber laser were introduced into the Er-doped fiber amplifier. Four high-power LDs with an oscillation wavelength of 1480 nm were used. A 2 m length of non-polarization-maintaining EDF with normal dispersion properties was used as the gain fiber. The average power of the amplified pulse train was 403 mW.
The output pulse was introduced into 100 m of anomalous-dispersion polarization-maintaining single-mode fiber (PMF), and was Raman shifted, generating a wavelength-tunable soliton pulse. The output pulse from the PMF was passed through a long pass filter, and only the generated soliton pulses were picked off. Figure 11(a) shows the optical spectra of the generated soliton pulse. Clean sech2-shaped pulses were generated. The center wavelength could be shifted from 1620 nm to 1920 nm by changing the pump power in the EDF. The temporal width of the soliton pulse was 200 fs FWHM [17]. Figure 11(b) shows the variation of the average output power as a function of the center wavelength of the generated soliton pulse. The properties obtained with the previous system based on a 50 MHz pulse source are also shown for comparison. Thanks to the high repetition rate of the fiber laser, the average power was about twice as large as the previous ones using a 50 MHz fiber laser system.
Fig. 11 Characteristics of generated wavelength-tunable soliton pulses: (a) optical spectra, and (b) wavelength dependence of average output power when passively mode-locked ultrashort-pulse fiber lasers with repetition rates of 50 and 95 MHz were used.
Next, the generated soliton pulses were coupled into the newly developed CPF. The improved CPF was developed based on the numerically designed dispersion map, as mentioned in the previous section. The splice loss was 0.13 dB at each point, and the total splice loss was ~2.1 dB.
Figure 12 shows the experimentally obtained optical spectra of the output pulses from the improved CPF. Spectrally compressed pulses were achieved at each wavelength. Optical spectra similar to the numerical ones were obtained. A small amount of red-shift was observed, especially at a wavelength of 1620 nm. It is considered that this red-shift was caused by stimulated Raman scattering in the CPF [31].
Fig. 12 Observed optical spectra at output of improved CPF for wavelengths of (a) 1620, (b) 1660, and (c) 1770 nm.
Figure 13 shows the characteristics of output pulses from the developed system using high-repetition-rate fiber lasers and the improved CPF. The properties obtained in the previous system using the 50 MHz pulse source and the original CPF are also shown for comparison [17]. The input power into CPF was adjusted by a variable attenuator. Spectrally compressed narrow pulse spectra with low sidelobe components were successfully generated over a wide wavelength range of 1620–1900 nm. The compressed spectral width was 0.60–0.67 nm over the whole operation bandwidth. Thanks to the high-repetition-rate pulse source and improved CPF, the output power was 4–9 mW, which was more than twice as large as the previously achieved output powers. The sidelobe suppression ratio was −14 to −24.5 dB, which was improved by as much as 1–9 dB from the previous level. A spectral compression ratio of 20–28 was achieved. From these results, we confirmed that the improved CPF is effective for both spectral compression and sidelobe suppression over a wide wavelength range. The experimental results were almost in agreement with the numerical ones. Using the improved CPF and the high-repetition-rate fiber laser, we succeeded in demonstrating a high-power, widely wavelength tunable, narrow-linewidth source.
Fig. 13 Characteristics of generated spectral compressed pulse with original and improved CPFs; (a) optical spectra of spectral compressed pulse in improved CPF, (b)-(d) wavelength dependence of (a) compressed spectral width, (b) output power, and (c) sidelobe level.
4. Conclusion
In this work, we investigated the characteristics of adiabatic spectral compression in quasi-dispersion-increasing comb-profile fibers (CPF) to demonstrate a high-power, narrow-linewidth, wavelength tunable pulse source. The behavior of spectral compression in conventional dispersion-increasing fibers (DIFs) and CPFs was examined numerically, especially in terms of sidelobe suppression. The periodical breathing behavior and the process of generating sidelobe components in the CPFs were analyzed numerically. By taking account of the numerical results, we designed an improved CPF in which the dispersion was varied more precisely in the short propagation length region. A dramatic improvement in the sidelobe suppression ratio to −24.2 dB was achieved while maintaining a narrow linewidth of ~0.6 nm, and high spectral compression was realized.
Then, we developed a high-power wavelength-tunable narrow-linewidth source using an ultrashort-pulse fiber laser and an improved CPF. A high-repetition-rate ultrashort-pulse Er-doped fiber laser using single-wall carbon nanotubes (SWNTs) was developed as a seed pulse source. An improved CPF was developed experimentally based on the numerical analysis. High spectral compression was obtained using the improved CPF and the developed fiber laser system. The compressed spectra with a narrow linewidth of ~0.6 nm was obtained over a wide wavelength range of 1.62–1.90 μm. The sidelobe suppression ratio was −14 to −24 dB. The average power was more than twice as large as those achieved previously. The experimental results were in good agreement with the numerical ones. This developed light source will be useful for optical measurement, especially spectroscopic applications.
References and Links
1. S. H. Yun, C. Boudoux, G. J. Tearney, and B. E. Bouma, “High-speed wavelength-swept semiconductor laser with a polygon-scanner-based wavelength filter,” Opt. Lett. 28(20), 1981–1983 (2003). [CrossRef] [PubMed]
2. R. Huber, M. Wojtkowski, and J. G. Fujimoto, “Fourier Domain Mode Locking (FDML): A new laser operating regime and applications for optical coherence tomography,” Opt. Express 14(8), 3225–3237 (2006). [CrossRef] [PubMed]
3. I. Grulkowski, J. J. Liu, B. Potsaid, V. Jayaraman, C. D. Lu, J. Jiang, A. E. Cable, J. S. Duker, and J. G. Fujimoto, “Retinal, anterior segment and full eye imaging using ultrahigh speed swept source OCT with vertical-cavity surface emitting lasers,” Biomed. Opt. Express 3(11), 2733–2751 (2012). [CrossRef] [PubMed]
4. P. Beaud, W. Hodel, B. Zysset, and H. P. Weber, “Ultrashort pulse propagation, pulse breakup, and fundamental soliton formation in a single-mode optical fiber,” IEEE J. Quantum Electron. 23(11), 1938–1946 (1987). [CrossRef]
5. N. Nishizawa and T. Goto, “Compact system of wavelength tunable ultrashort soliton pulse generation system,” IEEE Photonics Technol. Lett. 11, 325 (1999). [CrossRef]
6. J. H. Lee, J. van Howe, X. Liu, and C. Xu, “Soliton self-frequency shift: Experimental demonstrations and applications,” IEEE J. Sel. Top. Quantum Electron. 14(3), 713–723 (2008). [CrossRef] [PubMed]
7. N. Nishizawa, “Highly functional all-optical control using ultrafast nonlinear effects in optical fibers,” IEEE J. Quantum Electron. 45(11), 1446–1455 (2009). [CrossRef]
8. K. Sumimura, T. Ohta, and N. Nishizawa, “Quasi-super-continuum generation using ultrahigh-speed wavelength-tunable soliton pulses,” Opt. Lett. 33(24), 2892–2894 (2008). [CrossRef] [PubMed]
9. M. Oberthaler and R. A. Hopfel, “Spectral narrowing of ultrashort laser pulses by self-phase modulation in optical fibers,” Appl. Phys. Lett. 63(8), 1017 (1993). [CrossRef]
10. B. R. Washburn, J. A. Buck, and S. E. Ralph, “Transform-limited spectral compression due to self-phase modulation in fibers,” Opt. Lett. 25(7), 445–447 (2000). [CrossRef] [PubMed]
11. J. Limpert, T. Gabler, A. Liem, H. Zellmer, and A. Tunnermann, “SPM-induced spectral compression of picosecond pulses in a single-mode Yb-doped fiber amplifier,” Appl. Phys. B 74(2), 191–195 (2002). [CrossRef]
12. J. Limpert, N. Deguil-Robin, I. Manek-Hönninger, F. Salin, T. Schreiber, A. Liem, E. Röser, H. Zellmer, A. Tünnermann, A. Courjaud, C. Hönninger, and E. Mottay, “High-power picosecond fiber amplifier based on nonlinear spectral compression,” Opt. Lett. 30(7), 714–716 (2005). [CrossRef] [PubMed]
13. E. R. Andresen, J. Thøgersen, and S. R. Keiding, “Spectral compression of femtosecond pulses in photonic crystal fibers,” Opt. Lett. 30(15), 2025–2027 (2005). [CrossRef] [PubMed]
14. D. A. Sidorov-Biryukov, A. Fernandez, L. Zhu, A. Pugzlys, E. E. Serebryannikov, A. Baltuska, and A. M. Zheltikov, “Spectral narrowing of chirp-free light pulses in anomalously dispersive, highly nonlinear photonic-crystal fibers,” Opt. Express 16(4), 2502–2507 (2008). [CrossRef] [PubMed]
15. A. B. Fedotov, A. A. Voronin, I. V. Fedotov, A. A. Ivanov, and A. M. Zheltikov, “Spectral compression of frequency-shifting solitons in a photonic-crystal fiber,” Opt. Lett. 34(5), 662–664 (2009). [CrossRef] [PubMed]
16. E. R. Andresen, J. M. Dudley, D. Oron, C. Finot, and H. Rigneault, “Transform-limited spectral compression by self-phase modulation of amplitude-shaped pulses with negative chirp,” Opt. Lett. 36(5), 707–709 (2011). [CrossRef] [PubMed]
17. N. Nishizawa, K. Takahashi, Y. Ozeki, and K. Itoh, “Wideband spectral compression of wavelength-tunable ultrashort soliton pulse using comb-profile fiber,” Opt. Express 18(11), 11700–11706 (2010). [CrossRef] [PubMed]
18. R. Liang, X. Zhou, Z. Zhang, Z. Qin, H. Li, and Y. Liu, “Numerical investigation on spectral compression of femtosecond soliton in a dispersion-increasing fiber,” Opt. Fiber Technol. 15(5-6), 438–441 (2009). [CrossRef]
19. H. P. Chuang and C. B. Huang, “Wavelength-tunable spectral compression in a dispersion-increasing fiber,” Opt. Lett. 36(15), 2848–2850 (2011). [CrossRef] [PubMed]
20. S. V. Chernikov, R. Kashyap, and J. R. Taylor, “Comblike dispersion-profiled fiber for soliton pulse train generation,” Opt. Lett. 19(8), 539–541 (1994). [CrossRef] [PubMed]
21. K. Igarashi, J. Hiroishi, T. Yagi, and S. Namiki, “Comb-like profiled fiber for efficient generation of high quality 160 GHz sub-picosecond soliton train,” Electron. Lett. 41(12), 688 (2005). [CrossRef]
22. N. Nishizawa and K. Takahashi, “Time-domain near-infrared spectroscopy using a wavelength-tunable narrow-linewidth source by spectral compression of ultrashort soliton pulses,” Opt. Lett. 36(19), 3780–3782 (2011). [CrossRef] [PubMed]
23. W. T. Chao, Y. Y. Lin, J. L. Peng, and C. B. Huang, “Adiabatic pulse propagation in a dispersion-increasing fiber for spectral compression exceeding the fiber dispersion ratio limitation,” Opt. Lett. 39(4), 853–856 (2014). [CrossRef] [PubMed]
24. C. Xu and X. Liu, “Photonic analog-to-digital converter using soliton self-frequency shift and interleaving spectral filters,” Opt. Lett. 28(12), 986–988 (2003). [CrossRef] [PubMed]
25. T. Nishitani, T. Konishi, and K. Itoh, “Resolution improvement of all-optical analog-to-digital conversion employing self-frequency shift and self-phase-modulation-induced spectral compression,” IEEE J. Sel. Top. Quantum Electron. 14(3), 724–732 (2008). [CrossRef]
26. T. Konishi, K. Takahashi, H. Matsui, T. Satoh, and K. Itoh, “Five-bit parallel operation of optical quantization and coding for photonic analog-to-digital conversion,” Opt. Express 19(17), 16106–16114 (2011). [CrossRef] [PubMed]
27. K. Takahashi, H. Matsui, T. Nagashima, and T. Konishi, “Resolution upgrade toward 6-bit optical quantization using power-to-wavelength conversion for photonic analog-to-digital conversion,” Opt. Lett. 38(22), 4864–4867 (2013). [CrossRef] [PubMed]
28. Y. Chen, Z. Zhang, X. Zhou, X. Chen, and Y. Liu, “Spectral compression in a comb-like distributed fiber and its application in 7-bit all-optical quantization,” Opt. Eng. 53(12), 126106 (2014). [CrossRef]
29. S. N. Li, H. P. Li, J. K. Liao, X. G. Tang, R. G. Lu, and Y. Liu, “Numerical investigation on frequency-shifting-induced spectral compression of femtosecond solitons in highly nonlinear fiber,” Optik (Stuttg.) 124(16), 2281–2284 (2013). [CrossRef]
30. C. Finot and S. Boscolo, “Design rules for nonlinear spectral compression in optical fibers,” J. Opt. Soc. Am. B 33(4), 760 (2016). [CrossRef]
31. G. P. Agrawal, Nonlinear Fiber Optics, 4th ed. (Academic, 2007), Ch. 5.
32. F. Koch, S. V. Chernikov, and J. R. Taylor, “Dispersion measurement in optical fibers over the entire spectral range from 1.1 μm to 1.7 μm,” Opt. Commun. 175(1-3), 209–213 (2000). [CrossRef]
33. N. Nishizawa, A. Muto, and T. Goto, “Measurement of chromatic dispersion of optical fibers using wavelength-tunable soliton pulses,” Jpn. J. Appl. Phys. 39(1), 4990–4992 (2000). [CrossRef]
34. N. Kuwaki and M. Ohashi, “Waveguide dispersion measurement technique for single-mode fibers using wavelength dependence of mode field radius,” J. Lightwave Technol. 7(6), 990–996 (1989). [CrossRef]
35. J. P. Gordon, “Dispersive perturbations of solitons of the nonlinear Schrödinger equation,” J. Opt. Soc. Am. B 9(1), 91 (1992). [CrossRef]
36. N. Nishizawa, Y. Seno, K. Sumimura, Y. Sakakibara, E. Itoga, H. Kataura, and K. Itoh, “All-polarization-maintaining Er-doped ultrashort-pulse fiber laser using carbon nanotube saturable absorber,” Opt. Express 16(13), 9429–9435 (2008). [CrossRef] [PubMed]
