🔬 physics

Integrated photon-pair sources on periodically poled thin-film lithium tantalate

This paper demonstrates the first realization of high-efficiency, high-purity, and high-fidelity integrated photon-pair sources on periodically poled thin-film lithium tantalate (TFLT) in both traveling-wave and resonant configurations, establishing the platform as a competitive contender for scalable quantum light generation and photonic quantum information processing.

Original authors: Di Zhu, Sakthi Sanjeev Mohanraj, Xiaodong Shi, Ran Yang, Lin Zhou

Published 2026-06-29

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Di Zhu, Sakthi Sanjeev Mohanraj, Xiaodong Shi, Ran Yang, Lin Zhou

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are trying to build a super-secure communication network, but instead of sending letters or emails, you are sending "ghostly" pairs of light particles called photons. To make this work, you need a machine that can reliably split one beam of light into two perfectly synchronized twins. This paper introduces a new, highly efficient machine built on a tiny chip made of a special crystal called Lithium Tantalate.

Here is a breakdown of what the researchers did, using simple analogies:

1. The Problem: Finding the Right Crystal

For years, scientists have been trying to build these "photon twins" on computer chips. They usually use a material called Lithium Niobate. It's good, but it has some flaws: it gets easily damaged by bright light (like a camera sensor getting blown out by the sun) and it gets "foggy" (photorefractive) when you shine light on it, messing up the signal.

The researchers decided to try a cousin of that material: Lithium Tantalate.

The Analogy: Think of Lithium Niobate as a delicate glass vase. It works, but if you hit it too hard or shine a bright light on it, it might crack or get cloudy. Lithium Tantalate is like a tough, high-quality ceramic. It can handle much brighter lights without breaking or getting foggy, and it's still very good at doing the magic trick of splitting light.

2. The Magic Trick: "Periodic Poling"

To make the light split correctly, the crystal needs to be organized in a very specific pattern, like a zipper or a barcode. This process is called periodic poling.

The Analogy: Imagine the crystal is a long hallway. To get the light to split into twins, you need to place a series of mirrors on the floor in a repeating pattern. The researchers used high-voltage electricity to "stamp" this pattern into the crystal. They proved this pattern was perfect by shining a laser through it and seeing if it produced the expected "echo" (a phenomenon called Second-Harmonic Generation). It worked perfectly.

3. The Two Machines They Built

The team built two different types of "factories" on this chip to create the photon twins:

Machine A: The Straight Highway (Waveguide)

How it works: They made a straight, tiny road for the light to travel on. They pumped a laser into one end, and out the other end came pairs of photons.
The Result: This machine was incredibly efficient. It produced a massive number of photon pairs very quickly.
The Quality: The twins were so perfectly matched that they were almost indistinguishable from each other. The researchers tested this by checking if the twins were "entangled" (linked by quantum physics). They found a 98.9% match, which is like flipping two coins and having them land on the same side almost every single time, even when they are miles apart.

Machine B: The Race Track (Resonator)

How it works: Instead of a straight road, they built a racetrack. The light zooms around in circles, bouncing back and forth many times before exiting. This traps the light and makes the "factory" much brighter.
The Result: This machine didn't just make one pair of twins; it made a whole comb of them.
The Analogy: Imagine a piano. The straight road played one note. The race track played a whole scale of notes at once. They created a "Quantum Frequency Comb" that spans a huge range of colors (from the C-band to the L-band in telecommunications). This is like having a factory that can produce thousands of different types of photon twins simultaneously, ready to be used for sending many messages at once (wavelength multiplexing).

4. Why This Matters (According to the Paper)

The paper claims this is a major step forward because:

It's the first time this specific material (Lithium Tantalate) has been used to make photon pairs this way.
It's very bright: It produces more photon pairs per second than many other current technologies.
It's very clean: The "noise" (unwanted extra photons) is extremely low.
It's versatile: Because it can make a "comb" of colors, it is perfectly suited for wavelength-multiplexed quantum communications.
- Simple explanation: This means you can send many different quantum messages down the same fiber optic cable at the same time, just like how a radio station can broadcast many different channels on different frequencies.

Summary

The researchers took a tough, high-performance crystal (Lithium Tantalate), stamped a perfect pattern into it, and built two tiny chip-based machines. These machines successfully split light into high-quality, entangled photon pairs. One machine is great for speed and simplicity, while the other is a powerhouse that can generate a whole spectrum of these pairs at once, making it a strong contender for building the future of secure, high-speed quantum networks.

Technical Summary: Integrated Photon-Pair Sources on Periodically Poled Thin-Film Lithium Tantalate

Problem Statement

While chip-integrated photon-pair sources based on spontaneous parametric down-conversion (SPDC) are critical for scalable quantum technologies, the specific potential of thin-film lithium tantalate (TFLT) for this application remained unexplored. Although TFLT offers compelling advantages over its counterpart, thin-film lithium niobate (TFLN)—including a weaker photorefractive response, a higher optical-damage threshold, and lower birefringence—previous demonstrations on TFLT were limited to spontaneous four-wave mixing (SFWM). The native $\chi^{(2)}$ nonlinearity of TFLT, which enables efficient SPDC with low noise and easy pump filtering, had not yet been utilized for integrated photon-pair generation.

Methodology

The authors addressed this gap by combining high-quality periodic electrical poling with low-loss nanophotonic waveguides on x-cut TFLT platforms. The study utilized two distinct device architectures:

Traveling-Wave Source: A 3-mm-long periodically poled lithium tantalate (PPLT) waveguide was fabricated on a 310-nm-thick TFLT platform. The device featured a 2 µm top width, 200 nm etch depth, and a poling period of 3.24 µm designed for quasi-phase matching (QPM) between 1560 nm and 780 nm modes.
Resonant Source: PPLT racetrack resonators were fabricated on a 575-nm-thick TFLT platform (2 µm width, 300 nm etch depth) with an 800 µm poled section and a 4.32 µm poling period. These air-clad resonators were designed to enhance SPDC via cavity effects.

Fabrication and Characterization:
The devices were fabricated using electron-beam lithography (EBL), inductively coupled plasma (ICP) dry etching, and plasma-enhanced chemical vapor deposition (PECVD) for cladding. Periodic poling was achieved using high-voltage pulses after depositing a hafnium dioxide buffer layer. The quality of the poling was verified via second-harmonic generation (SHG) confocal microscopy.

Experimental Setup:

Classical Characterization: SHG and sum-frequency generation (SFG) were used to map phase-matching functions and determine conversion efficiencies.
Quantum Characterization: The devices were pumped with continuous-wave (CW) lasers at their respective phase-matching wavelengths. Photon pairs were detected using superconducting nanowire single-photon detectors (SNSPDs).
Metrics: The study measured pair generation rates (PGR), coincidence-to-accidental ratios (CAR), heralded second-order correlation ( $g^{(2)}_H(0)$ ), and Franson interference visibility to assess entanglement fidelity.

Key Contributions and Results

1. Traveling-Wave PPLT Waveguide

Efficiency: The device achieved a normalized SHG conversion efficiency of $841 \pm 15 \text{ %W}^{-1}\text{cm}^{-2}$ .
Photon-Pair Generation: The on-chip pair generation efficiency reached 2.1 GHz mW $^{-1}$ .
Noise Performance: A maximum coincidence-to-accidental ratio (CAR) of $3.8 \times 10^5$ was observed at low pump powers.
Single-Photon Purity: Heralded second-order correlation measurements yielded a $g^{(2)}_H(0)$ of $0.0018 \pm 0.0002$ , confirming deep single-photon regime operation.
Entanglement: Franson-type two-photon interferometry demonstrated a raw visibility of $98.9 \pm 0.5\%$ , verifying high-fidelity time-energy entanglement.
Bandwidth: The source generated broadband photon pairs spanning over 100 nm in the telecom band with strong frequency correlation.

2. Resonant PPLT Racetrack Microcavity

Cavity Enhancement: The racetrack resonator exhibited loaded quality factors of $6.22 \times 10^5$ (fundamental harmonic) and $5.31 \times 10^5$ (second harmonic).
Quantum Frequency Comb (QFC): The device generated a broadband QFC spanning the telecom C- and L-bands (1525 nm to 1610 nm) with 50 GHz spacing.
Brightness: The spectral brightness was estimated at 11 GHz mW $^{-1}$ GHz $^{-1}$ (or 1.3 THz mW $^{-1}$ nm $^{-1}$ ), calculated by normalizing the pair generation efficiency to the resonance linewidth.
Performance: For selected frequency modes, the pair generation efficiency was $(3.3 \pm 0.4)$ GHz mW $^{-1}$ , with a CAR of up to $2.8 \times 10^3$ .
Non-Classicality: The resonator source maintained non-classical operation with a measured $g^{(2)}_H(0)$ of $0.06 \pm 0.01$ .

Significance and Claims

The authors position these results as establishing TFLT as a competitive platform for integrated quantum photonics. The paper claims that the demonstrated performance is competitive with the state of the art across various $\chi^{(2)}$ integrated platforms (including TFLN, InGaP, and AlGaAs).

Specifically, the work highlights:

First Demonstration: This is the first demonstration of SPDC-based photon-pair generation on TFLT.
Material Advantages: The results leverage TFLT's inherent material properties, such as high optical-damage thresholds and weak photorefractive effects, which support stable operation at elevated optical powers and facilitate broadband nonlinear interactions.
Application Potential: The high spectral brightness and broad coverage of the resonator source make the platform naturally suited for dense wavelength division multiplexing (DWDM) and wavelength-multiplexed quantum communications.
Scalability: The ability to co-integrate nonlinear sources, modulators, and multiplexers on a single TFLT chip is presented as a pathway toward large-scale quantum photonic systems.

The authors also note fabrication challenges, specifically the partial reversion of periodically poled domains during etching and cladding processes due to the relatively low Curie temperature of lithium tantalate, suggesting a need for careful thermal management in future process optimization.