High-power handling and bias stability of thin-film Lithium Tantalate microring and coupling resonators

TL;DR

This work targets the PR limitations of LN-based photonics by evaluating thin-film lithium tantalate (TFLT) devices. Through oxide cladding and thermal annealing, TFLT microring resonators achieve watt-scale power handling with TO effects dominating residual performance, and PR effects become negligible. The study also demonstrates the first coupling modulator on the TFLT platform, achieving a low $V_\pi$ of approximately $3~\mathrm{V}$ in a $2~\mathrm{mm}$ electrode, with stable bias and phase control suitable for high-speed operation. Together, these results position TFLT as a competitive platform for practical high-power and stable EO photonics in both classical and quantum contexts, enabling compact, energy-efficient devices at telecommunication wavelengths.

Abstract

In this paper, we demonstrate the ultra-high-power handling capability and DC bias stability of optical microring and electro-optic (EO) coupling resonators on the thin-film lithium tantalate (TFLT) platform. We show that, with annealing, oxide-cladded TFLT resonators can handle several watts (4W) of circulating power with minimal frequency shift and no observable photo-refractive effect. Furthermore, we demonstrate a compact 2mm coupling modulator achieving a low Vpi of 3V with stable bias and phase control in the telecom C-band.

Figures (3)

• Figure 1: Device architecture and high-power characterization of the TFLN micro-ring resonator: (a) Optical microscope image of the fabricated thin-film lithium niobate (TFLN) micro-ring resonator ($50\text{ }\mu\text{m}$ scale bar). The inset shows a cross-sectional schematic of the coupling section, highlighting the etched rib waveguide structure on a silicon substrate. (b) Schematic of the high-power measurement setup, incorporating an Erbium-Doped Fiber Amplifier (EDFA) and a tunable filter to characterize the resonator's nonlinear response. (c–d) Power-dependent transmission spectra before (c) and after (d) thermal annealing. The spectra compare forward (blue-to-red) and backward (red-to-blue) wavelength scans at various speeds ($1\text{ nm/s}$ and $2\text{ nm/s}$). Post-annealing results show significantly reduced resonance distortion and thermal bistability. (e–f) Extracted maximum resonance frequency shift as a function of intracavity power before (e) and after (f) annealing. The annealed device demonstrates superior power handling, with resonance shifts remaining within the Full-Width at Half-Maximum (FWHM) for intracavity powers up to approximately $1\text{ W}$.
• Figure 2: Characterization of the coupling modulator: (a) Schematic of the experimental setup for characterizing the device, including a laser source, polarization controller (PC), and power meter (PM). The arbitrary waveform generator (AWG) provides voltage to the Signal (S) and Ground (G) electrodes of the coupling modulator, which has an active length of $2\text{ mm}$. (b) Relative transmission spectra at bias voltages of $0\text{ V}$, $1\text{ V}$, and $3\text{ V}$. (c) Top: Applied DC bias voltage stepped from $0\text{ V}$ to $3\text{ V}$ in $1\text{ V}$ increments every 10 minutes. Bottom: Time-resolved transmission heatmap showing the spectral response over a 40-minute duration.
• Figure 3: Phase stability and repeatability: (a) Schematic of the experimental setup used for phase stability measurements, featuring a laser source, polarization controller (PC), power meter (PM), and an arbitrary waveform generator (AWG) to drive the electrodes, with the output captured by a photodetector (PD). (b) Top: Applied phase bias voltage applied in discrete steps. Bottom: Corresponding transmission spectra over time, showing the resonance shift as a function of the stepped voltage. (c) Stability analysis showing the change in resonance position over 25 minutes. The resonance shift remains below $0.1 \text{ pm}$, well within the half-maximum (FWHM) boundaries.