Thin-film periodically poled lithium niobate (TF-PPLN) optical waveguides, given its strong second-order nonlinearity and light confinement, has been applied to many important fields such as optical communications and quantum information processing. The effective mode area for thin-film optical waveguide cross-section is ~ 1 , which is 1/10 of their bulk counterparts. Therefore, the pump power that is required to generate the same second harmonic (SH) light intensity for the thin-film optical waveguide can be reduced by 100 times. In recent years, TF-PPLN devices with ultra-high normalized conversion efficiency have been applied to many different scenarios, including classical optical parametric amplifier and all-optical ultrafast optical switches, as well as quantum entangled photon sources, quantum squeezed light sources, and up-conversion single photon detectors.
Despite the rapid progress, fabrication of TF-PPLN devices today still relies on waveguide patterning and crystal poling processes performed on individually centimeter-sized chips based on electron-beam lithography and manual poling. This limitation greatly hinders the development of large-scale nonlinear photonic integrated circuits that could demand multiple TF-PPLN devices on the same chip. Additionally, although the normalized conversion efficiencies of TF-PPLN devices are much higher than their bulk counterparts, the absolute nonlinear efficiencies have been lagging behind. This is because the thin-film optical waveguides are more sensitive to the geometric parameter variation along the optical waveguide, such as the top width, etch depth, and film thickness, which will lead to the variation of local quasi-phase-matching (QPM) wavelength and distorted QPM spectrum from the ideal sinc-like transfer function.
To address these issues, researchers from City University of Hong Kong have recently developed fabrication methods to produce TF-PPLN devices on a 4-inch wafer scale and proposed to use segmented micro-heaters along the TF-PPLN waveguides to locally fine tune the QPM wavelength and restore the distorted QPM spectrum, thus enhancing the absolute nonlinear conversion efficiency of TF-PPLN devices. Relevant research results were recently published in Photonics Research, Volume 12, Issue 8, 2024. [Xiaoting Li, Haochuan Li, Zhenzheng Wang, Zhaoxi Chen, Fei Ma, Ke Zhang, Wenzhao Sun, and Cheng Wang. Advancing large-scale thin-film PPLN nonlinear photonics with segmented tunable micro-heaters[J]. Photonics Research, 2024, 12(8): 1703]
Figure 1 illustrates the conceptual schematic of the segmented micro-heater design for wafer-level TF-PPLN devices. Before the application of segmented micro-heaters, the QPM spectrum of the TF-PPLN device is often severely broadened and distorted [Fig. 1(c)] from the theoretical spectrum, which is caused by the inhomogeneous film thickness variation or other changes in the waveguide geometry (such as the etching depth or the top width). By applying different thermal tuning powers [Fig. 1(a)] to the segmented micro-heaters, researchers could precisely adjust the local QPM wavelength in each section and re-align them together back to the target QPM wavelength, as shown in Fig. 1(c).
Figure 2 demonstrates the experimentally measured improvement of nonlinear conversion efficiency of TF-PPLN devices with segmented micro-heaters. Fig. 2 (a) illustrates the original QPM spectrum of a 6-mm-long TF-PPLN device, which exhibits three main peaks at 1545.1 nm, 1548.8 nm, and 1554.9 nm, respectively. The highest peak conversion efficiency clearly falls short of the theoretical optimum (∼64% compared with ideal QPM peak, blue dashed curve) due to the inhomogeneous broadening of the QPM spectrum. Researchers then applies DC powers to the four segmented micro-heaters integrated alongside the TF-PPLN device. Through iteratively fine-tuning the DC powers on each micro-heater [Fig. 4(e)], a single-main-peak QPM spectrum could be achieved as depicted in Fig. 4(b). The measured peak second-harmonic (SH) conversion efficiency after thermal tuning is 3802% W−1cm−2, which is increased by 32% compared with the initial value (2878% W−1cm−2), which corresponds to 84% the theoretical conversion efficiency (4500% W−1cm−2). The remaining minor discrepancy from an ideal conversion efficiency is mainly attributed to the small sub-peak at 1560.9 nm, which could not be merged into the main SHG peak in this particular set of devices, possibly due to a larger thickness variation than expected at a certain location of the chip. Moreover, it is also feasible to further re-shape the QPM spectra by applying another set of tuning currents, as exhibited in Fig. 4(c).
Fig. 1 Schematic illustration of the wafer-scale PPLN optical waveguides featuring segmented micro-heaters. Insets: (a), (b) recovered QPM spectrum (b) after thermal tuning with segmented heating power (a). (c) Broadened QPM spectrum due to thickness variation before thermal tuning.
Fig. 2 (a)-(c) Measured SHG intensities as functions of pump wavelengths for a 6 mm TF-PPLN waveguide before applying tuning currents (a), after optimization of the heater powers (b), and with an arbitrary set of tuning parameters (c). (d)-(f) DC powers applied to each segmented micro-heater for the scenarios in (a)-(c), respectively.
In the near future, the research team will further optimize the segmented micro-heater design by increasing the number of micro-heaters to achieve better control of the QPM spectrum. They also plan to develop control algorithms improving the QPM spectrum alignment efficiency.
The post-fabrication QPM tuning enabled by segmented micro-heaters, together with the wafer-scale fabrication method, could be of great importance in promoting large-scale integrated nonlinear photonic circuits that require consistency in the QPM wavelength and nonlinear conversion efficiency across many devices. Such applications include phase-sensitive amplifiers in optical fiber communications that could potentially offer lower noise figures than current erbium-doped fiber amplifiers, as well as quantum squeezed and entangled light sources.
