Open Access
Add to BibSonomy
Abstract
Microresonators facilitate enhanced light-matter interactions within a limited space, showing great promise for nonlinear optics. Here, we demonstrate a high-quality (Q) factor Fabry-Perot microresonator (FPR) for second harmonic generation (SHG) on an x-cut thin film lithium niobate (TFLN) platform. The FPR exhibits Q factors of Qpump = 1.09 × 105 and QSH = 1.15 × 104 at the 1560 nm pump wavelength and 780 nm second harmonic wavelength, respectively. Under low pump power, a normalized SHG efficiency of 158.5 ± 18.5%/W is attained. We experimentally verify that increased temperatures mitigate photorefractive effects that degrade SHG performance. This work highlights the immense capabilities of one-dimensional planar optical waveguide resonators for efficient on-chip nonlinear wavelength conversion.
© 2024 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Thanks to the mass production of high-quality thin film lithium niobate (TFLN) wafers [1–3], lithium niobate materials have become one of the most promising platforms in the field of integrated optics. TFLN has been widely studied in classical optics and quantum optics due to its wide transparent window from ultraviolet to mid-infrared, high electro-optical coefficient and high second-order nonlinear coefficient. Efficient second-order nonlinear transformations such as second harmonic generation (SHG) [4–8], optical parametric oscillation (OPO) [9] and spontaneous parametric down conversion (SPDC) [10] have been rapidly developed on TFLN platforms.
Micro-cavities [11,12] with high quality factors are capable of confining light in a limited space, which can both enhance the interaction time between light and material and the optical power density to improve nonlinear conversion efficiency. SHG enhancement through microcavities [13,14] on TFLN has been widely reported. However, the anisotropy of lithium niobate materials [15] results in different crystal axis directions having different refractive indices. To maintain consistent phase matching conditions in the chip plane, most existing microring resonators are fabricated on Z-cut TFLN platforms, where wave vectors in different in-plane directions have the same refractive index. For x-cut thin film lithium niobate platforms, when the light wave vector is bent 90 degrees along the in-plane Y-axis onto the Z-axis, the effective refractive index of the waveguide changes, causing decreased phase matching and nonlinear conversion efficiency. Notably, Fabry-Perot microresonator (FPR) [16,17] can limit light to a one-dimensional direction and reduce the footprint of the resonator. Utilizing FPRs in SHG helps mitigate challenges such as phase mismatch [18], mode hybridization [19], and phase delay [20] within the resonant cavity. Jiang et al. have realized nonlinear frequency conversion in a one-dimensional photonic crystal nanocavity [21]; however, nonlinear phase matching conditions and signal light resonance were not concurrently considered, resulting in unsatisfactory conversion efficiency.
The waveguide structure confines the optical field to a very small area; in particular, the optical field density in a microcavity increases exponentially, significantly enhancing photorefractive effects [22–24]. In SHG studies, the infrared light field typically undergoes conversion into higher-energy visible or near-infrared light, which scatters on both sides of the waveguide, generating photogenerated charge carriers that induce refractive index changes in the crystal. This degrades phase matching conditions and modulates the light field phase, affecting resonator resonance conditions. Thus, understanding photorefractive impact in the SHG process is crucial.
Here, we propose an FPR based on a dual-wavelength distributed Bragg reflector (DBR) to enhance SHG on x-cut TFLN. It exhibits high Q factors of ${Q_{pump}} = 1.09 \times {10^5}$ and ${Q_{SH}} = 1.15 \times {10^4}$ in the pump and second harmonic bands, respectively. In particular, the second harmonic (SH) is deliberately overcoupled to the resonator, thereby enabling maximum SHG efficiency through straightforward temperature regulation. At 36°C, a maximum normalized conversion efficiency of 158.5 ± 18.5%/W is attained. Furthermore, we experimentally prove that increasing temperature can reduce the photorefractive effect on SHG. This research makes nonlinear optics and quantum optics more compatible with excellent optical devices on the x-cut TFLN platform.
2. Design and simulation of the FPR
Figure 1 shows the structure diagram of the FPR, which consists of two dual-wavelength reflectors (DWRs), a uniform waveguide with a length of 1 mm and two directional couplers. The input light field can be confined between the two DWRs to enhance nonlinear interaction. The waveguide between the two DWRs undergoes carefully engineered dispersion to satisfy the mode phase matching condition, i.e., $2{k_{pump}} = {k_{SHG}}$. To take advantage of the maximum nonlinear coefficient d33, the TE0 mode of the pump and the TE2 mode of the SH are selected for nonlinear interaction. The upper directional coupler (DC) couples 1560 nm pump light into the FPR, where the SH is generated at doubled frequency through second-order nonlinear interaction with the lithium niobate crystal. An asymmetric directional coupler (ADC) downloads the SH TE2 mode into a narrow waveguide, converts it to the fundamental TE0 mode, and efficiently collects it into a 780 nm single-mode fiber. The ADC employs a shortcuts to adiabaticity (STA) design approach [18,25,26], offering significant fabrication tolerance and high coupling efficiency.
Fig. 1. (a) Structure diagram of the Fabry-Perot resonator (FPR). The FPR consists of two dual-wavelength reflectors (DWRs), a 1 mm uniform waveguide, and two directional couplers. The inset shows a magnified view of the DWR structure comprising an input waveguide, a cone grating waveguide with embedded linear taper, and a Bragg grating. (b) The variation of refractive index of pump light and second harmonic waveguide modes with waveguide width.
Figure 1 illustrates the DWR structure, formed by periodic contour modulation on the waveguide sidewall. The DWR comprises three parts: input waveguide, cone grating waveguide with embedded linear taper, and Bragg grating. The 5µm taper grating enables smooth mode transition between the input waveguide and Bragg grating, avoiding additional scattering loss from refractive index jumps. The DWR satisfies the Bragg condition at both pump and SH wavelengths $nef{f_a} \cdot \Lambda = m\frac{{{\lambda _a}}}{2}, \;nef{f_b} \cdot \Lambda = n\frac{{{\lambda _b}}}{2},$ where $nef{f_{a(b)}}$ is the equivalent refractive index of the periodic Bragg grating at wavelength λ, obtained via finite element simulation. Here, m and n are positive integer grating orders and Λ is the grating period. Figure 1(b) depicts the simulation of the effective refractive index of the optical waveguide within the resonator. The two curves intersect at w = 0.733µm, indicating that the phase-matching condition $nef{f_{pump,T{E_0}}} = nef{f_{SH,T{E_2}}}$ is satisfied. Through numerical solution, a dual-wavelength reflector is realized with input waveguide width w = 0.733µm, periodic groove depth d = 0.3µm, and grating period Λ=0.475µm. The structure is verified via finite-difference time-domain (FDTD) simulation. The simulated reflection spectra in Figs. 2(a) and 2(c) show high reflectance for the 1560 nm and 780 nm bands; central wavelength reflectance is −0.06 dB and −0.12 dB, respectively. Figures 2(b) and 2(d) show the transmission field diagrams at 1560 nm and 780 nm central wavelengths, demonstrating near-complete reflection of input energy. The white arrow denotes the direction of the incident field, while the black line delineates the exterior outline of the DWR. Two such reflectors placed in mirror symmetry form an FPR to confine the light field.
Fig. 2. FDTD simulation results of the dual-wavelength reflector (DWR). (a) Reflectance spectrum of the DWR in the infrared band. (b) Electric field distribution showing transmission of the DWR in the infrared spectral range. (c) Reflectance spectrum of the DWR in the near-infrared band. (d) Electric field distribution showing transmission of the DWR in the near-infrared spectral range.
3. Device fabrication and experimental test results
The FPRs are fabricated on x-cut TFLN with a 300 nm top lithium niobate layer and 4.7µm SiO2 substrate, using a dry etching process. First, a 100 nm Cr layer was deposited on the lithium niobate substrate via electron beam evaporation (EBE). This Cr layer serves as a hard mask for lithium niobate etching. Subsequently, the Cr layer was coated with ARP-6200.013, an electron beam lithography (EBL) photoresist that can be easily removed post-exposure by immersing in developer solution. The graphics are then transferred concurrently to the Cr and LN layers in a two-step etching process. The lithium niobate was completely etched via inductively coupled plasma (ICP) etching based on Ar. A 3µm SiO2 cladding layer was deposited by plasma enhanced chemical vapor deposition (PECVD). Figures 3(a)–3(d) show optical and electron microscope images of the fabricated FPR, directional coupler (DC), asymmetric directional coupler (ADC), and dual-wavelength reflector (DWR), respectively.
Fig. 3. (a) Optical micrograph of the fabricated Fabry-Perot resonator (FPR) on thin film lithium niobate (TFLN). Scanning electron micrographs of (b) the directional coupler (DC), (c) asymmetric directional coupler (ADC), and (d) dual-wavelength reflector (DWR) structures.
The experimental setup is diagrammed in Fig. 4(a). A C-band tunable laser and near-infrared source tested device linear transmission and resonance spectra. TE-polarized 1560 nm and 780 nm light at desired orientations was generated using polarizers and a polarization controller. A 1560 nm narrow linewidth laser characterized nonlinear performance. Lens fibers coupled light in/out through the chip endface. A wavelength division multiplexer (WDM) separated residual pump light from the SH signal. Pump power was monitored via a 1560 nm photodetector, while a near-infrared optical spectrum analyzer (OSA) measured generated SH signals.
Fig. 4. (a) Diagram of the experimental characterization setup. Measurements involve tuning laser sources, polarization control, wavelength division multiplexing, and optical detection with a photodetector and optical spectrum analyzer. (b) Measured resonance spectra of Fabry-Perot resonator (FPR) devices with varying dual-wavelength reflector (DWR) grating periods. The resonance wavelength red-shifts for increasing periods.
To account for fabrication variability and experimental contingencies, five FPR groups with varying DBR periods were integrated on-chip. Figure 4(b) shows measured transmission spectra, revealing only FPRs exhibited a resonance peak within the DWR reflection band. The ∼20 nm experimental resonant bandwidth agrees well with simulated DWR reflection spectra in Fig. 2(a). Increasing the DWR lattice constant red-shifts the resonant wavelength. With a 480 nm grating period, the resonance peak centers at 1560 nm, prompting nonlinear characterization.
The designed microresonator enables SHG from a 1560 nm pump. The 480 nm DWR period and engineered waveguide dispersion fulfil phase matching requirements. Figures 5(a) and 5(b) show the resonance spectra at 1560 nm and 780 nm, verifying resonant enhancement at both pump and SH wavelengths with high quality factors of 109,000 and 11,500, respectively. The 780 nm second harmonic is extracted via an asymmetric directional coupler and detected by coupling to a lensed fiber. Measured fiber-chip coupling efficiencies are 5.2 dB/facet at 1560 nm and 9 dB/facet at 780 nm. Figure 5(c) shows chip temperature-dependent SHG efficiency with fixed input pump power and wavelength aligned to the FPR resonance. Peak conversion occurs at 36°C. Figure 5(d) then depicts SH power versus pump power at this optimized 36°C temperature. At low powers, the SH and the square of pump power show a linear relationship with slope 1.19, agreeing with the theoretical quadratic pump-SH dependence. A normalized conversion efficiency of 158.5%/W is obtained.
Fig. 5. (a) Resonance spectrum of the Fabry-Perot resonator (FPR) at the 1560 nm pump wavelength. (b) Resonance spectrum of the FPR at the 780 nm second harmonic wavelength. (c) Generated second harmonic optical power versus microchip temperature. Peak efficiency is achieved at 36°C. (d) Quadratic relationship between second harmonic power and 1560 nm pump power at optimized 36°C device temperature. Inset: Top-down charge coupled device (CCD) image of red second harmonic light generated in the FPR between two dual-wavelength reflectors.
Theoretically, the SHG efficiency in microresonators can be calculated by the following formula.
where g represents the nonlinear coupling strength, which depends on the mode overlap integral of the pump and SH as well as the effective mode area. Parameter $\hbar$ is the Dirac constant, ${\omega _1}$ represents the angular frequency of the pump. Q1(2) and Qe1(e2) denote the total quality factor and external quality factor of the pump and SH, respectively. They are related to the intrinsic quality factor Qi1(i2) by $1/Q = 1/{Q_e} + 1/{Q_i}$. In Fig. 5(a, b), the experimentally obtained values correspond to the external Q, and the intrinsic Q can be calculated from the resonance spectrum [27]. By extracting the simulated results of the pump and SH mode field intensities in Fig. 1(b) and the Q factors, the theoretical SHG efficiency of the F-P cavity is calculated as 395%/W/cm2 according to Eq. (1). This value closely matches the normalized SHG efficiency measured experimentally. The discrepancy between theoretical and experimental results mainly arises from inaccuracies in the waveguide width during fabrication, leading to imperfect phase-matching in the resonator.Furthermore, the cavity loss of the resonator can be approximated from the Q factor obtained experimentally [27]. The total in-cavity losses at the pump and SH wavelengths are 0.14 dB and 2.42 dB, respectively. Neglecting the transmission loss of the direct waveguide in the F-P cavity, the reflectivity of a single mirror is calculated as −0.07 dB and −1.21 dB at the pump and SH wavelengths, respectively. The large deviation observed in the near-infrared band may result from substantial loss of the reflector for higher-order SH modes or a mismatch between the reflector's central wavelength and the SH wavelength.
The intense red illumination of the FP microresonator between the two DWRs is visible via a top-down charge coupled device (CCD) camera when both pump and SH resonate within the FPR. Greater visible scattering is observed from the DWR end faces, consistent with simulations showing stricter DWR accuracy requirements for SH that satisfies the second-order Bragg reflection condition. Scattering losses likely arise due to the dry etching process, which creates angled rather than vertical waveguide sidewalls, allowing leakage into the cladding and substrate. By optimizing lithium niobate etching, intra-cavity SH scattering losses can be further reduced to improve conversion efficiency.
The pronounced confinement within the nano-waveguide necessitates intense pump light to facilitate nonlinear processes, while concurrently amplifying detrimental photorefractive effects in microresonators. Within the limited cavity space, the light field interacts with defects in the lithium niobate crystal, generating charge carriers that induce refractive index perturbations:
where n0 is the intrinsic lithium niobate refractive index, reff the effective electro-optic coefficient, and Es the optical field-induced space charge. Such refractive index disturbances adversely impact phase matching and resonance conditions for nonlinear wavelength conversion. However, the photorefractive effect diminishes significantly at elevated temperatures as charge carrier mobility increases around the waveguide.We examined the influence of photorefraction on SHG by monitoring second harmonic output fluctuations. Figure 6 depicts the SH power stability over 200 s with the chip temperature set to 30°C, 60°C, and 90°C. Since the primary optical power within the cavity originates from the input pump light, which remains consistent during temperature-dependent power fluctuation tests, and resonates within the FP cavity. Furthermore, the pump power coupled into the resonator is deliberately increased sufficiently to induce a significant photorefractive effect within the cavity. At 30°C, SH intensity is clearly modulated by refractive index changes. As temperature rises, fluctuations gradually subside as expected. Although the average SH optical power exhibits some fluctuations at different temperatures, a comparison of the test results at 30°C and 60°C demonstrates that small variations in SHG power are not the primary source of noise. Our results demonstrate increasing device temperature helps mitigate photorefraction, enabling robust nonlinear performance.
Fig. 6. Fluctuations in second harmonic optical power over 200 seconds measured by an optical power meter at microchip temperatures of (a) 30°C, (b) 60°C, and (c) 90°C. Increasing device temperature reduces intensity fluctuations, mitigating adverse photorefractive effects in the nonlinear optical process.
4. Conclusion
In summary, we have designed and demonstrated a high-performance Fabry-Perot microresonator for on-chip second harmonic generation on an x-cut thin film lithium niobate platform. Based on dual-wavelength Bragg reflectors, the microresonator achieves high quality factors in both the fundamental pump and second harmonic wavelength bands. The ∼20 nm resonant bandwidth enables flexible spectral tuning through the Bragg grating period. Under optimized conditions, record normalized conversion efficiency reaching 158.5 ± 18.5%/W is attained. Further, we experimentally validate that elevated temperatures help mitigate adverse photorefractive effects that deteriorate nonlinear interactions. This work enhances the compatibility of integrated nonlinear and quantum photonics with existing x-cut lithium niobate electro-optic and periodic poling capabilities.
Funding
National Natural Science Foundation of China (61975062, U21A20511); Wuhan National Laboratory for Optoelectronics.
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. G. Poberaj, H. Hu, W. Sohler, et al., “Lithium niobate on insulator (LNOI) for micro-photonic devices,” Laser Photonics Rev. 6(4), 488–503 (2012). [CrossRef]
2. D. Zhu, L. Shao, M. Yu, et al., “Integrated photonics on thin-film lithium niobate,” Adv. Opt. Photonics 13(2), 242–352 (2021). [CrossRef]
3. Y. Qi and Y. Li, “Integrated lithium niobate photonics,” Nanophotonics 9(6), 1287–1320 (2020). [CrossRef]
4. C. Wang, X. Xiong, N. Andrade, et al., “Second harmonic generation in nano-structured thin-film lithium niobate waveguide,” Opt. Express 25(6), 6963–6973 (2017). [CrossRef]
5. R. Luo, Y. He, H. Liang, et al., “Highly tunable efficient second-harmonic generation in a lithium niobate nanophotonic waveguide,” Optica 5(8), 1006–1011 (2018). [CrossRef]
6. Y. Li, Z. Huang, W. Qiu, et al., “Recent progress of second harmonic generation based on thin film lithium niobate,” Chin. Opt. Lett. 19(6), 060012 (2021). [CrossRef]
7. J. J. Lu, J. B. Surya, X. Liu, et al., “Periodically poled thin-film lithium niobate microring resonators with a second-harmonic generation efficiency of 250,000%/W,” Optica 6(12), 1455–1460 (2019). [CrossRef]
8. J. Lin, N. Yao, Z. Hao, et al., “Broadband quasi-phase-matched harmonic generation in an on-chip monocrystalline lithium niobate microdisk resonator,” Phys. Rev. Lett. 122(17), 173903 (2019). [CrossRef]
9. J. J. Lu, A. A. Sayem, Z. Gong, et al., “Ultralow-threshold thin-film lithium niobate optical parametric oscillator,” Optica 8(4), 539–544 (2021). [CrossRef]
10. J. J. Lu, M. Li, C. L. Zou, et al., “Toward 1% single-photon anharmonicity with periodically poled lithium niobate microring resonators,” Optica 7(12), 1654–1659 (2020). [CrossRef]
11. P. F. Liu, H. Wen, L. H. Ren, et al., “χ(2) nonlinear photonics in integrated microresonators,” Front. Optoelectron. 16(1), 18 (2023). [CrossRef]
12. W. Bogaerts, P. D. Heyn, and T. V. Vaerenbergh, “Silicon microring resonators,” Laser Photonics Rev. 6(1), 47–73 (2012). [CrossRef]
13. J. Y. Cheng, Y. M. Sua, H. Fan, et al., “Modal phase matched lithium niobate nanocircuits for integrated nonlinear photonics,” OSA Continuum 1(1), 229–242 (2018). [CrossRef]
14. R. Luo, Y. He, H. Liang, et al., “Optical parametric generation in a lithium niobate microring with modal phase matching,” Phys. Rev. Appl. 11(3), 034026 (2019). [CrossRef]
15. D. E. Zelmon, D. L. Small, and D. Jundt, “Infrared corrected Sellmeier coefficients for congruently grown lithium niobate and 5 mol. % magnesium oxide-doped lithium niobate,” J. Opt. Soc. Am. B 14(12), 3319–3322 (1997). [CrossRef]
16. H. N. Xu, Y. Qin, G. L. Hu, et al., “Million-Q integrated Fabry-Perot cavity using ultralow-loss multimode retroreflectors,” Photonics Res. 10(11), 2549–2559 (2022). [CrossRef]
17. C. L. Sun, C. Y. Zhong, M. L. Wei, et al., “Free-spectral-range-free filters with ultrawide tunability across the S + C + L band,” Photonics Res. 9(6), 1013–1018 (2021). [CrossRef]
18. J. Cheng, D. S. Gao, J. J. Dong, et al., “Ultra-efficient second harmonic generation via mode phase matching in integrated lithium niobate racetrack resonators,” Opt. Express 31(22), 36736–36744 (2023). [CrossRef]
19. D. X. Dai and M. Zhang, “Mode hybridization and conversion in silicon-on-insulator nanowires with angled sidewalls,” Opt. Express 23(25), 32452–32464 (2015). [CrossRef]
20. T. Yuan, J. Wu, Y. Liu, et al., “Chip-scale spontaneous quasi-phase matched second harmonic generation in a micro-racetrack resonator,” Sci. China Phys. Mech. Astron. 66(8), 284211 (2023). [CrossRef]
21. H. Jiang, H. Liang, R. Luo, et al., “Nonlinear frequency conversion in one dimensional lithium niobate photonic crystal nanocavities,” Appl. Phys. Lett. 113(2), 021104 (2018). [CrossRef]
22. A. Ashkin, G.D. Boyd, J. M. Dziedzic, et al., “Optically-induced refractive index inhomogeneities in LiNbO3 and LiTaO3,” Appl. Phys. Lett. 9(1), 72–74 (1966). [CrossRef]
23. H. W. Jiang, R. Luo, H. X. Liang, et al., “Fast response of photorefraction in lithium niobate microresonators,” Opt. Lett. 42(17), 3267–3270 (2017). [CrossRef]
24. Y. T. Xu, A. A. Sayem, C. L. Zou, et al., “Photorefraction-induced Bragg scattering in cryogenic lithium niobate ring resonators,” Opt. Lett. 46(2), 432–435 (2021). [CrossRef]
25. D. Guo and T. Chu, “Silicon mode (de)multiplexers with parameters optimized using shortcuts to adiabaticity,” Opt. Express 25(8), 9160–9170 (2017). [CrossRef]
26. D. Guo and T. Chu, “Compact broadband silicon 3 dB coupler based on shortcuts to adiabatici,” Opt. Lett. 43(19), 4795–4798 (2018). [CrossRef]
27. L. W. Luo, G. S. Wiederhecker, J. Cardwnas, et al., “High quality factor etchless silicon photonic ring resonators,” Opt. Express 19(7), 6284–6289 (2011). [CrossRef]
