⚛️ quantum physics

Epitaxial single T centres in silicon-on-insulator

This paper demonstrates the successful integration of single T centres into silicon-on-insulator wafers using high-purity molecular-beam epitaxy, achieving narrow optical linewidths of 30 MHz and establishing a robust platform for coherent spin-photon interfaces in silicon quantum photonics.

Original authors: Christian H. Christiansen, Kasper H. Nielsen, Alisha Nanwani, Sebastiano Guaraldo, E. Laurits Piehorsch, Arnulf J. Snedker-Nielsen, Magnus L. Madsen, David R. Gongora, Emanuele Brusaschi, Rodrigo A. T

Published 2026-07-08

📖 4 min read🧠 Deep dive

CC BY 4.0

Original authors: Christian H. Christiansen, Kasper H. Nielsen, Alisha Nanwani, Sebastiano Guaraldo, E. Laurits Piehorsch, Arnulf J. Snedker-Nielsen, Magnus L. Madsen, David R. Gongora, Emanuele Brusaschi, Rodrigo A. Thomas, Ian Farrer, D. Hieu Nguyen, Georgios Kountouris, Beñat M. d. A. Jokisch, Mohammad Khalifa, Claudia Piccinini, Mathias Ø. Augustesen, Hugo Laurell, Kokeb B. Benti, Maria S. Gonzalez, Amedeo Carbone, Elvedin Memisevic, Sangeeth Kallatt, Mark K. Svendsen, Marianne E. Bathen, Lasse Vines, Peter Granum, Peter Krogstrup, Stefano Paesani

Original paper licensed under CC BY 4.0 (http://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-fast, ultra-secure communication network using light. To do this, you need tiny "transmitters" inside a computer chip that can send out single particles of light (photons) in a perfectly synchronized rhythm. In this paper, scientists are working with a specific type of transmitter called a T centre, which is a tiny defect inside a silicon chip that acts like a quantum light bulb.

Here is the story of how they improved these light bulbs, explained simply:

The Problem: A Messy Construction Site

Think of a silicon chip as a pristine, empty city block. To create a T centre (the light bulb), scientists usually use a method called ion implantation.

The Analogy: Imagine trying to plant a single, perfect flower in a garden by firing seeds at it with a high-powered cannon.
The Result: While you do get your flower, the cannon blast also smashes up the soil, creates craters, and scatters debris everywhere. In the chip, this "cannon blast" creates a lot of messy damage and unwanted defects nearby.
The Consequence: These messy surroundings act like static on a radio. They cause the light from the T centre to wobble, change color slightly, and lose its perfect rhythm. This makes the signal "fuzzy" and hard to use for high-tech computing.

The Solution: Precision Gardening

The researchers wanted to grow these T centres without the cannon blast. They used a technique called Molecular Beam Epitaxy (MBE).

The Analogy: Instead of firing seeds with a cannon, they decided to build the garden layer by layer, like stacking LEGO bricks or baking a perfect cake.
The Process:
1. They took a standard silicon chip and shaved off the top layer to make it thin.
2. They placed it in a super-clean, high-vacuum oven.
3. They slowly grew a new layer of silicon, but at a specific moment, they added a tiny, precise sprinkle of carbon atoms (the "seed" for the T centre).
4. They covered it with more silicon to seal it in.
The Result: Because they built it carefully from the bottom up, there was no "cannon blast" damage. The T centres were born in a perfectly clean, quiet neighborhood.

The Big Discovery: Silence is Golden

When they tested these new, carefully grown T centres, the difference was huge.

The Old Way (Implanted): The light was "noisy." It was like trying to hear a whisper in a crowded, noisy room. The light's color would jump around wildly (a problem called "spectral wandering").
The New Way (Epitaxial): The light was incredibly pure and steady. It was like hearing that same whisper in a soundproof library.
The Numbers: They measured how "fuzzy" the light was. The new method reduced the fuzziness (broadening) by ten times compared to the old method. They achieved a level of clarity so sharp that the light's color stayed steady within a range of just 30 MHz.

Putting It to Work

The scientists didn't just grow these centres; they built a tiny highway (a nanophotonic waveguide) right next to them.

The Analogy: Imagine the T centre is a singer, and the waveguide is a microphone cable.
The Test: They showed that they could hook up a single T centre to this "microphone," turn on a laser to make it sing, and catch the light perfectly.
The Proof: They proved it was a single light source (not a group of them) by showing that the light came out one photon at a time, like a steady drip from a faucet, rather than a gush of water.

Why This Matters (According to the Paper)

The paper claims this is a major step forward because it proves you can create these perfect quantum light sources directly inside the silicon chips used for modern technology, without damaging the chip. By growing them cleanly, they removed the "noise" that was previously stopping these chips from working as well as they could. This creates a solid, reliable foundation for building future quantum computers and networks that rely on silicon.

Technical Summary: Epitaxial Single T Centres in Silicon-on-Insulator

Problem Statement
Silicon quantum emitters, particularly the T centre, represent a promising platform for solid-state spin-photon interfaces due to their excellent spin properties, operation at telecom wavelengths (O-band), and compatibility with industry-standard silicon-on-insulator (SOI) manufacturing. However, the realization of coherent photon emission is currently hindered by stringent material quality requirements. Standard formation methods, specifically carbon and hydrogen ion implantation, introduce significant material damage and a variety of unwanted defects (e.g., vacancies, interstitial clusters). These defects create a noisy charge environment that leads to spectral wandering and homogeneous broadening, resulting in optical linewidths of hundreds of MHz to GHz at 4 K. This performance is orders of magnitude worse than that observed in bulk silicon, limiting the utility of T centres for scalable quantum networking and computing. Furthermore, ion implantation offers poor depth control, typically spreading the defect distribution over a $\sim$ 100 nm region, which complicates integration with nanophotonic structures.

Methodology
To address these limitations, the authors developed a growth-based approach using high-purity molecular-beam epitaxy (MBE) to incorporate single T centres directly into an SOI material stack. The process involves:

Substrate Preparation: Starting with a standard float-zone (FZ) SOI wafer with a 220 nm top silicon layer, the layer is thinned via back-etching to 70 nm.
Epitaxial Growth: The wafer is transferred to a dedicated ultra-high-purity MBE system. A silicon buffer layer is grown at 650°C to ensure crystal quality.
$\delta$ -Doping: The substrate temperature is lowered to 220–350°C to facilitate kinetically limited carbon incorporation. A 9 nm layer of carbon-doped silicon (Si:C) is deposited, creating a $\delta$ -doped region with a high concentration of carbon atoms.
Capping: A 105 nm silicon capping layer is grown at approximately 400°C to complete the stack while preventing carbon diffusion out of the $\delta$ -layer.
Post-Processing: The samples undergo thermal annealing at 510°C to facilitate the recombination of carbon and hydrogen atoms into T centres and to thermally cure the material.
Device Fabrication: The wafers are processed using electron-beam lithography to create nanobeam waveguides with grating couplers for input and output coupling.
Characterization: Devices are operated in a dilution refrigerator at 100 mK. Optical characterization includes resonant photoluminescence excitation (PLE), second-order autocorrelation measurements ( $g^{(2)}$ ), and hole-burning spectroscopy to determine homogeneous linewidths.

Key Contributions and Results

Demonstration of Epitaxial T Centres: The authors successfully demonstrated the formation of single T centres within an epitaxially grown SOI stack. Photoluminescence measurements confirmed emission at the expected 1325.8 nm wavelength, alongside other centres such as G and I centres.
Single-Photon Operation: Integrated waveguide devices successfully coupled to individual epitaxial T centres. Resonant excitation yielded a second-order autocorrelation value of $g^{(2)}(0) = 0.26(3)$ , confirming operation at the single-photon level.
Suppression of Homogeneous Broadening: The most significant result is the observation of a substantial reduction in homogeneous broadening. Using hole-burning measurements, the authors extracted instantaneous homogeneous optical linewidths as narrow as 30 MHz for epitaxial T centres grown in natural silicon.
Comparison with Implantation: When compared to reference samples created via standard ion implantation on the same SOI base substrate and evaluated under identical conditions, the epitaxial method showed a ten-fold reduction in broadening. While the inhomogeneous broadening ( $\sim$ 30 GHz) remained similar between the two methods (suggesting a common origin in oxide-layer stress), the homogeneous linewidths were drastically improved in the epitaxial samples.
Spectral Diffusion: Measurements of spectral wandering indicated a decay time of 34(5) $\mu$ s, which is comparable to or slightly longer than values reported for implanted devices in similar nanophotonic cavities.

Significance
The paper establishes epitaxial T centres as a robust foundation for coherent spin-photon interfaces in silicon quantum photonics. By utilizing high-purity MBE and $\delta$ -doping, the authors circumvent the disorder and damage inherent to ion implantation, achieving optical linewidths that approach the fundamental limits required for coherent quantum operations. This work demonstrates that it is possible to grow high-quality, single T centres within a material stack compatible with nanophotonic device manufacturing, offering a scalable pathway toward integrating quantum emitters with silicon-on-insulator technology.