Enhancement of a nonlinear optical interaction through waveguides or resonators disclose unconventional interplay among multiple lights. Microresonator-based optical frequency comb (OFC) generation via third order nonlinearity is a typical example of such enhancements. Recently, quadratic-nonlinearity-based OFC with an external cavity configuration has been found and its on-chip implementation is highly demanded. Here we for the first time demonstrate such an on-chip OFC with a quadratic nonlinear waveguide resonator. Furthermore, we controlled the comb spectra separation by adjusting frequency difference of two pump light. This on-chip quadratic device will be useful for not only metrologies but also integrated quantum information technologies.
© 2018 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
Recent progress in study of third order optical nonlinearity using monolithic photonic devices enables a frequency combs [1–3] based on Kerr nonlinearity [4–8]. The Kerr nonlinear systems use whispering-gallery-mode resonators which provide frequency combs with a high-repetition rate from several GHz to THz and a wide-ranged sidebands over several hundred THz . Moreover, strong confinement of light supported by resonators with a high quality (Q) factor ranging from 106 to 1010 has drastically reduced a pump power consumption. Such enhancements can be applied to a frequency comb generation in the cascaded quadratic nonlinear optical system [10–12] studied both in theory and experiments with an external cavity [13–18].
Here, we employ a resonator integrating a PPLN waveguide with a high reflective coating as a quadratic nonlinear waveguide resonator (QNWR) (which we call this device by PPLN-QNWR hereafter). Figure 1(a) depicts a conceptual illustration for frequency comb generation in a PPLN-QNWR which confines 1.5-μm light by the high reflective coating. The quadratic nonlinearity has several features different from those found in Kerr nonlinearity. First the quadratic nonlinear materials have a larger nonlinearity than the Kerr nonlinearity and require a lower pump power. This reduces the necessary intensity inside the cavity and allows the larger system size, which is suitable for generating denser comb spectra. In addition, the quadratic nonlinear materials allow to arrange the comb spectra by controlling the phase matching condition with periodically-poling techniques [19,20]. Such features of quadratic nonlinearity are useful for not only conventional photonics but also quantum information processing  such as frequency-engineered photon sources [22,23] and a large-scale frequency-domain quantum computing [24–26].
2. Experimental design
2.1. Basic concept
We explain a frequency comb formation via quadratic nonlinear optical effects with cavity structure around a fundamental input pump light [15,18]. The generation process has hierarchical two stages similar to Kerr-comb formation through four wave mixing processes . We illustrate a visualization of the hierarchical frequency comb formation in Fig. 1(b). At first, second harmonic generation (SHG) of the fundamental pump light at ω is generated. Then, optical parametric oscillation (OPO) of the SH light at 2ω produces a signal and an idler light at ω ± Δ. Through cascaded processes of SHG and OPO of the lights, higher-order sidebands with separation Δ are generated, resulting in a primary comb around the initial pump light. By increasing the initial pump power, secondary sidebands with a separation corresponding to one FSR of the resonator are generated in the vicinity of the primary comb. This secondary combs are caused by OPO of SHG from each line among the primary comb. Note that the SH lights need to exceed a threthold power of OPO. Finally, when the comb clusters are merged, they form a widely spanning comb spectra.
In the above explanation, the separation Δ is determined by the effect of the group velocity dispersion only with SHG and OPO process for the comb generation. However, other quadratic nonlinear processes such as sum frequency generation (SFG) and difference frequency generation (DFG) complicate the mechanism of the comb growth as theoretically discussed in Refs. [16,18]. Of particular importance for the first step is SFG of the fundamental pump light with the signal and idler light. This process leads the nonlinear loss, and plays a fundamental role in determining Δ. From the second or later steps of the cascaded SHG-OPO processes, both SFG and DFG of not only the pump light but also the sidebands are included, and the effects modulate the value of Δ.
2.2. Experimental setup
Our experimental setup for the frequency comb generation in the PPLN-QNWR is shown in Fig. 1(c). A fundamental pump light for the frequency combs is prepared by a distributed feedback laser (DFBL) at a center wavelength of 1540 nm with a linewidth of 15 kHz which is amplified by an erbium-doped fiber amplifier (EDFA). The pump power is adjusted by a half wave plate (HWP) and a polarizing beamsplitter (PBS). The pump light is set to vertical (V) polarization by another HWP, and then it is focused on the PPLN-QNWR.
The PPLN consists of Zinc-doped lithium niobate as a core and lithium tantalite as a clad. The waveguide is a ridged type fabricated through a dicing process. A waveguide cross-section is 8 μm × 8 μm and its length is 20 mm. The periodically-poling period is about 18 μm for the type-0 quasi-phase-matched SHG of 1560 nm at 50 °C. By a dielectric multilayer, a high reflective coating for 1.5-μm light and an anti reflective coating for its SH light are achieved at both ends of the PPLN waveguide. The reflectance is designed to ∼ 99 % from 1535 nm to 1770 nm, and less than 1 % from 745 nm to 790 nm.
After the PPLN-QNWR, a dichroic mirror (DM) divides the 1.5-μm light and the SH light. After coupling into single mode optical fibers, they are connected to proper equipment such as I, II, III and IV in the figure which we will explain in section 3.1, 3.2 and 3.3, to evaluate the nonlinear optical signals.
3. Experimental result
3.1. Characterization of our PPLN-QNWR
The PPLN-QNWR in our experiment has been characterized as follows. Instead of using the DFBL in Fig. 1(c), we used an external cavity diode laser (ECDL) at a center wavelength of 1560 nm with a linewidth of 300 kHz with scanning a frequency at the range of 10 GHz. Figure 1(d) shows the experimental results of the transmission spectra from the PPLN-QNWR measured by a digital sampling oscilloscope (DSO) connected to a photodetector (PD1) for 1.5-μm light (see Fig. 1(c)–I). From the results, we estimated the quality (Q) factor, full width at the half maximum (FWHM) and free spectral range (FSR) to be 5.5 × 106, 35 MHz and 3.5 GHz, respectively. The observed FSR is in good agreement with length 20 mm of the PPLN. The overall Q factor is determined by the external and the internal photon losses. From the experimental results, we estimated the effective reflectance ∼ 97 % of the high reflective coating when we assume there is no internal photon loss in the PPLN. On the other hand, if the reflectance of the coated mirrors is assumed to be 99 % as we designed, the internal photon loss over a round-trip in the PPLN is estimated to be 1 %. The pump power dependence of quadratic nonlinearities in our device was measured by connecting the output of PDs for light around 1.5 μm and 780 nm into DSO.
The signals of the sidebands without the pump light were obtained by filtering out the pump light using a bandpass filter (BPF) before PD1 for 1.5μm light (see Fig. 1(c)–II). The bandwidth is 15 nm, and we set the center wavelength to the longer side of the pump wavelength such that the pump light is sufficiently suppressed. By simultaneously monitoring the signal of the sidebands around the fundamental light and that around the SH light, we extracted the peak intensity of their signals in DSO while scanning the laser frequency as increasing the pump power which is measured in front of the PPLN (see Fig. 1(e)). We see that the SHG signals gradually increased, but there exists a transition point where the power dependence of the SHG signal clearly changed at 70 mW. In this point, the sidebands around the fundamental light appeared via the OPO process. We notice that the threthold pump power of 70 mW is less than the previously reported values [8,18]. In addition, considering the scan range of the pump light over 1 FSR, the genuine threthold power is expected to be much lower than the observed value. From the ratio of FSR to FWHM, the duty cycle of the pump light coupling into the PPLN-QNWR is about 0.01, and thereby the expected threthold power coupling to the cavity is roughly estimated to be <1 mW. We guess such a very low threthold property is caused by the best phase-matching condition of the PPLN about 1560 nm and its SH light at 780 nm, and a well mode matching between the cavity mode and the propagation mode through the PPLN-QNWR due to the monolithic structure.
3.2. Frequency comb generation
We then performed the experiments for frequency comb generation. The laser frequency around 1540 nm was stabilized to near the cavity resonance by using an error signal from PD1 connected to a laser servo system. For several pump powers from 0 mW to 280 mW, we measured frequency comb spectra around 1.5 μm (see Fig. 1(c)–III). The result is shown in Fig. 2(a). The comb spectra appeared at the pump power of > 90 mW, and the span of them gradually became wider as increasing the pump power. For 280-mW pump power, we observed a frequency comb spectra over the spectral spanning of 60 nm and several sharp primary combs in the 1.5-μm band (see Fig. 2(b)). At the same time, we monitored the spectra around the SH light at 770 nm. Similar to the spectra in the 1.5-μm band, there also exists comb-like multiple spectra in spite of no resonant structure for SH lights (we could not resolve each dense spectra limited by the resolution 0.02 nm of our optical spectrum analyzer (OSA)). This is due to the cascaded SHG and OPO processes. We note the envelope of the spectra has peaks around 780 nm in addition to ones around 770 nm. It might reflect difference of the coupling strength of the SH and fundamental light because the best phase-matching condition of our PPLN is achieved for 780 nm and 1560 nm.
For 280-mW pump power, we measured the radio frequency (rf) beat notes of sidebands around 1.5 μm extracted from the BPF with its bandwidth of 15 nm. In order to improve the signal to noise ratio, we amplified the filtered signal using an additional EDFA (depicted as EDFA2 in Fig. 1(c)–IV). Figure 3(a) shows the rf spectra observed in the resolution bandwidth (RBW) of 10 Hz and Figs. 3(b) and (c) show the optical spectra at the 1.5-μm band when the laser frequency was tuned on the cavity resonance as much as possible. We observed a very sharp rf spectrum with the bandwidth of 113 Hz at the center frequency of 3.5 GHz which corresponds to the FSR of the cavity. On the other hand, when we set the laser frequency slightly detuned from the peak of the resonance, we obtained a broaden rf spectrum with the same center frequency and the bandwidth of ∼796 kHz as shown in Fig. 3(d). The optical spectra are shown in Figs. 3(e) and (f). These results of the beat signals are explained as follows. When the frequency of the pump light is adjusted to a peak of the cavity resonance, the signal and idler light by the cascaded SHG-OPO processes also appear at peaks of the resonances. As a result, when the several combs are merged, all of the spectra met at inside a single resonance have the same frequency. This leads to the sharp beat note of the frequency combs and the strong intracavity field to provide clear secondary frequency comb spectra by avoiding the mismatch of secondary combs as observed in the enlarged spectra around the pump spectrum in Fig. 3(c). On the other hand, when the frequency of the pump light is detuned from a peak of the cavity resonance and the pump power is enough to grow the frequency combs, there may exist multiple lines inside a single resonance of the cavity. In this situation, there is the mismatch of the secondary combs due to the weak intracavity field, and the optical spectra were formed with a narrower spanning without any sharp primary combs as shown in Fig. 3(f). We notice that if we measure the beat note with a wider range, the linewidth of the beat signal might be broader due to the cavity dispersion even when the pump is perfectly resonant [16,18].
3.3. Control of the frequency comb separation
Finally, we performed the demonstration of the controll of the comb separation by using the pump light at 1540 nm from the DFBL and the second pump light from the ECDL. The two pump lights were amplified by using the EDFA1 and the EDFA2. They were combined by a BS, and then input to the PPLN-QNWR with their power of ∼ 100 mW. We denote the frequency of the lasers by with their frequency difference of . When the two pump lights are injected to the material, SH light of the them are produced at . Each SH light interacts with their strong subharmonic pump light, and then sideband peaks at by DFG are produced. The repeat of the cascaded processes generates the comb structure around the pump with the separation of . The experimental results are shown in Fig. 4. For example, in the middle of the upper figures, we observed comb spectra with separation ∼ 2.5 nm. This value is in good agreement with the wavelength difference between the two pump lights; the first pump light (pump 1) and the second pump light (pump 2) are at 1540 nm and 1542.5 nm, respectively. About the other wavelength settings of the second pump light, we see the clear comb peaks with the separation corresponding to the wavelength difference δp between the two pump lights. The separation of the peaks corresponds to ∼ 100 times larger than the FSR of the cavity. This means that we can widely controll the comb separation from 1 FSR to 100 FSR by proper adjusting the frequency difference between the two pump lights.
In conclusion, we, for the first time, have performed the experimental demonstration of frequency comb generation by a PPLN-QNWR. We characterized the SHG and OPO processes in the PPLN-QNWR, and generated the frequency comb via the cascaded SHG-OPO processes. The observed threthold pump power was smaller than the previous reported values in the quadratic and Kerr nonlinearity-based comb generators [8,18]. By setting the pump frequency to on-resonance, we observed a wide spectral spanning up to 60 nm with a narrow rf beat spectrum at 3.5 GHz corresponding to 1 FSR due to the well overlap of the merged secondary combs. When we use two pump lights, we can controll the frequency comb with a separation corresponding to the frequency difference up to 100 FSR of the input pump light. The spectral span range demonstrated here is comparable with that based on conventional mode-locked lasers, but is smaller than the typical values based on the microresonators . The spectral range was limited by the wavelength range of the high reflective coating and phase-matching condition, but we could overcome the problems by commercial dielectric multilayer techniques and chirped quasi-phase matching techniques [11,27]. The small mode spacing of 3.5 GHz complements the missing region existing in the comb generators based on conventional systems and microresonators . We believe that the demonstrated monolithic simple system will be useful for a new comb generator device based on quadratic nonlinear optical interaction. In addition, together with recent technologies of frequency manipulation of photons by the PPLN waveguide [28–30], the on-chip quadratic device will be applied in the field of quantum information processing such as frequency-multiplexed photon sources and a large-scale quantum computing by using multiple frequency modes [24,25].
CREST, JST JPMJCR1671; MEXT/JSPS KAKENHI Grant Number JP15H03704, JP16H02214, JP16H01054, JP15KK0164 and JP16K17772; Research Foundation for Opto-Science and Technology, the Asahi Glass Foundation, and Murata Science Foundation.
We thank Kenichiro Matsuki for helpful discussion.
References and links
1. R. Holzwarth, T. Udem, T. W. Hänsch, J. C. Knight, W. J. Wadsworth, and P. S. J. Russell, “Optical frequency synthesizer for precision spectroscopy,” Phys. Rev. Lett. 85, 2264–2267 (2000). [CrossRef] [PubMed]
2. D. J. Jones, S. A. Diddams, J. K. Ranka, A. Stentz, R. S. Windeler, J. L. Hall, and S. T. Cundiff, “Carrier-envelope phase control of femtosecond mode-locked lasers and direct optical frequency synthesis,” Science. 288, 635–639 (2000). [CrossRef] [PubMed]
4. P. Del’Haye, A. Schliesser, O. Arcizet, T. Wilken, R. Holzwarth, and T. Kippenberg, “Optical frequency comb generation from a monolithic microresonator,” Nature. 450, 1214–1217 (2007). [CrossRef]
5. A. A. Savchenkov, A. B. Matsko, V. S. Ilchenko, I. Solomatine, D. Seidel, and L. Maleki, “Tunable optical frequency comb with a crystalline whispering gallery mode resonator,” Phys. Rev. Lett. 101, 093902 (2008). [CrossRef] [PubMed]
6. Y. K. Chembo, D. V. Strekalov, and N. Yu, “Spectrum and dynamics of optical frequency combs generated with monolithic whispering gallery mode resonators,” Phys. Rev. Lett. 104, 103902 (2010). [CrossRef] [PubMed]
8. T. Herr, K. Hartinger, J. Riemensberger, C. Wang, E. Gavartin, R. Holzwarth, M. Gorodetsky, and T. Kippenberg, “Universal formation dynamics and noise of kerr-frequency combs in microresonators,” Nat. Photonics 6, 480–487 (2012). [CrossRef]
10. G. I. Stegeman, “cascading: nonlinear phase shifts,” Quantum Semiclassical Opt. J. Eur. Opt. Soc. Part B 9, 139 (1997). [CrossRef]
11. C. Langrock, M. Fejer, I. Hartl, and M. E. Fermann, “Generation of octave-spanning spectra inside reverse-proton-exchanged periodically poled lithium niobate waveguides,” Opt. letters 32, 2478–2480 (2007). [CrossRef]
12. C. Phillips, C. Langrock, J. Pelc, M. Fejer, J. Jiang, M. E. Fermann, and I. Hartl, “Supercontinuum generation in quasi-phase-matched linbo 3 waveguide pumped by a tm-doped fiber laser system,” Opt. Lett. 36, 3912–3914 (2011). [CrossRef] [PubMed]
13. V. Ulvila, C. Phillips, L. Halonen, and M. Vainio, “Frequency comb generation by a continuous-wave-pumped optical parametric oscillator based on cascading quadratic nonlinearities,” Opt. Lett. 38, 4281–4284 (2013). [CrossRef] [PubMed]
14. V. Ulvila, C. Phillips, L. Halonen, and M. Vainio, “High-power mid-infrared frequency comb from a continuous-wave-pumped bulk optical parametric oscillator,” Opt. express 22, 10535–10543 (2014). [CrossRef] [PubMed]
15. I. Ricciardi, S. Mosca, M. Parisi, P. Maddaloni, L. Santamaria, P. De Natale, and M. De Rosa, “Frequency comb generation in quadratic nonlinear media,” Phys. Rev. A 91, 063839 (2015). [CrossRef]
16. F. Leo, T. Hansson, I. Ricciardi, M. De Rosa, S. Coen, S. Wabnitz, and M. Erkintalo, “Walk-off-induced modulation instability, temporal pattern formation, and frequency comb generation in cavity-enhanced second-harmonic generation,” Phys. Rev. Lett. 116, 033901 (2016). [CrossRef] [PubMed]
17. F. Leo, T. Hansson, I. Ricciardi, M. De Rosa, S. Coen, S. Wabnitz, and M. Erkintalo, “Frequency-comb formation in doubly resonant second-harmonic generation,” Phys. Rev. A 93, 043831 (2016). [CrossRef]
18. S. Mosca, I. Ricciardi, M. Parisi, P. Maddaloni, L. Santamaria, P. De Natale, and M. De Rosa, “Direct generation of optical frequency combs in χ (2) nonlinear cavities,” Nanophotonics 5, 316–331 (2016). [CrossRef]
19. M. Houe and P. Townsend, “An introduction to methods of periodic poling for second-harmonic generation,” J. Phys. D: Appl. Phys. 28, 1747 (1995). [CrossRef]
20. V. Pruneri, G. Bonfrate, P. Kazansky, C. Simonneau, P. Vidakovic, and J. Levenson, “Efficient frequency doubling of 1.5 μm femtosecond laser pulses in quasi-phase-matched optical fibers,” Appl. Phys. Lett. 72, 1007–1009 (1998). [CrossRef]
21. O. Alibart, V. D’Auria, M. De Micheli, F. Doutre, F. Kaiser, L. Labonté, T. Lunghi, É. Picholle, and S. Tanzilli, “Quantum photonics at telecom wavelengths based on lithium niobate waveguides,” J. Opt. 18, 104001 (2016). [CrossRef]
22. E. Pomarico, B. Sanguinetti, C. I. Osorio, H. Herrmann, and R. T. Thew, “Engineering integrated pure narrow-band photon sources,” New J. Phys. 14, 033008 (2012). [CrossRef]
23. K.-H. Luo, H. Herrmann, S. Krapick, B. Brecht, R. Ricken, V. Quiring, H. Suche, W. Sohler, and C. Silberhorn, “Direct generation of genuine single-longitudinal-mode narrowband photon pairs,” New J. Phys. 17, 073039 (2015). [CrossRef]
25. Z. Xie, T. Zhong, S. Shrestha, X. Xu, J. Liang, Y.-X. Gong, J. C. Bienfang, A. Restelli, J. H. Shapiro, F. N. Wong, and C. W. Wong, “Harnessing high-dimensional hyperentanglement through a biphoton frequency comb,” Nat. Photonics 9, 536 (2015). [CrossRef]
26. T. Kobayashi, R. Ikuta, S. Yasui, S. Miki, T. Yamashita, H. Terai, T. Yamamoto, M. Koashi, and N. Imoto, “Frequency-domain hong-ou-mandel interference,” Nat. Photonics 10, 441–444 (2016). [CrossRef]
27. A. Tanaka, R. Okamoto, H. H. Lim, S. Subashchandran, M. Okano, L. Zhang, L. Kang, J. Chen, P. Wu, T. Hirohata, S. Kurimura, and S. Takeuchi, “Noncollinear parametric fluorescence by chirped quasi-phase matching for monocycle temporal entanglement,” Opt. Express 20, 25228–25238 (2012). [CrossRef] [PubMed]
28. R. Ikuta, Y. Kusaka, T. Kitano, H. Kato, T. Yamamoto, M. Koashi, and N. Imoto, “Wide-band quantum interface for visible-to-telecommunication wavelength conversion,” Nat. Commun. 2, 1544 (2011). [CrossRef] [PubMed]
29. P. Manurkar, N. Jain, M. Silver, Y.-P. Huang, C. Langrock, M. M. Fejer, P. Kumar, and G. S. Kanter, “Multidimensional mode-separable frequency conversion for high-speed quantum communication,” Optica. 3, 1300–1307 (2016). [CrossRef]
30. T. Kroh, A. Ahlrichs, B. Sprenger, and O. Benson, “Heralded wave packet manipulation and storage of a frequency-converted pair photon at telecom wavelength,” Quantum Sci. Technol. 2, 034007 (2017). [CrossRef]