Electrically-triggered spin-photon devices in silicon

This paper demonstrates the first electrically triggered single-photon emission from a silicon T centre integrated with nanophotonic devices, achieving high-fidelity spin initialization and establishing a scalable pathway for electrically controlled quantum networking and computing.

The Gist

Imagine you are trying to build a super-fast, super-secure internet for the future, one that uses the laws of quantum physics instead of just electricity. To do this, you need tiny, reliable "workers" that can hold information (like a bit of data) and send it out as light (photons) to talk to other workers.

For a long time, scientists have been looking for the perfect worker to live inside silicon—the same material used in your computer chips. They found a promising candidate called the T centre. Think of the T centre as a tiny, glowing speck of dust trapped inside the silicon crystal. It has a special "spin" (like a tiny spinning top) that can store information, and it glows with light that is perfect for traveling through the fiber-optic cables used in our current internet.

However, there was a big problem: until now, getting these T centres to work required shining a very precise, expensive laser on them from the outside. It was like trying to start a car by pushing it from the outside every time you wanted to go somewhere. You couldn't just flip a switch inside the car.

The Breakthrough: Flipping the Switch
In this paper, the researchers built a new kind of "car" for these T centres. They created a tiny electronic device (a diode) right next to the T centre. Instead of using an outside laser to wake the T centre up, they simply sent an electric current through the device.

• The Analogy: Imagine a row of streetlights. Before, you had to walk down the street with a giant flashlight to turn each one on. Now, the researchers have installed a switch right at the base of each light. You can flip a switch, and zap, the light turns on instantly.

What They Discovered

1. Electric Light from Silicon: They successfully made the T centre glow just by applying electricity. This is the first time anyone has made a single T centre emit a single photon (a single particle of light) using only electricity. It's like turning a silicon chip into a tiny, electric lightbulb that speaks the language of quantum physics.
2. The "Herald" Trick: Here is the clever part. When the T centre glows, the color of the light it emits depends on which way its "spin" is pointing (up or down).
   • The researchers used a special filter (like a pair of sunglasses that only lets through a specific color) to watch the light.
   • If they saw a flash of light through the filter, they knew instantly that the T centre's spin was set to a specific direction.
   • This is called "heralding." It's like a waiter ringing a bell to tell the kitchen, "Table 4 is ready!" In this case, the "bell" (the flash of light) tells the computer, "The memory bit is now set to '1'."

Why This Matters
The researchers showed that they could set the spin state of the T centre with very high accuracy (about 92% success rate) just by flipping an electrical switch and watching for a specific color of light.

• Scalability: Because this method uses electricity, you don't need a giant, complex laser setup for every single T centre. You could potentially have thousands of them on a single chip, all controlled by electrical wires, just like the transistors in your phone today.
• Speed: Electrical switches are much faster and easier to control than moving lasers around.

The Bottom Line
This paper proves that we can take a quantum "worker" (the T centre) living inside a silicon chip and control it using simple electricity, just like we control the lights in our house. They demonstrated that these workers can be turned on, set to a specific state, and ready to send information, all without needing external lasers. This is a major step toward building a quantum computer that can be mass-produced using the same factories that make our current computer chips.

Technical Summary

1. Problem Statement

Scalable quantum networking and computing require high-speed control methods for dense networks of quantum devices. While solid-state platforms like silicon color centers offer promising scalability and integration with existing semiconductor infrastructure, controlling them typically relies on complex optical switching networks. This approach faces significant overhead when scaling to thousands of devices. Furthermore, while silicon T centers possess excellent spin-photon interface (SPI) properties (long-lived spins, telecom O-band emission), previous demonstrations have largely relied on optical excitation. There is a critical need for electrically triggered single-photon sources and spin initialization methods that allow for parallel, on-chip control without the complexity of massive optical switching arrays.

2. Methodology

The authors developed integrated optoelectronic devices combining nanophotonic waveguides/cavities with lateral p-i-n diodes on a Silicon-on-Insulator (SOI) platform.

• Device Fabrication:

  • Platform: Commercial (100) SOI chips with a 220 nm device layer.
  • Defect Creation: The silicon T centers were generated via ion implantation: blanket Carbon ($^{12}$C) implantation followed by masked Hydrogen ($H^+$) implantation to define active regions and minimize dopant passivation.
  • Doping: High-density p-type (Boron) and n-type (Phosphorus) implants ($>10^{20} \text{ cm}^{-3}$) were used to form ohmic contacts and prevent carrier freeze-out at cryogenic temperatures (1.5 K).
  • Device Types: Two configurations were fabricated:
    1. Tapered Waveguide: Coupled to grating couplers (GCs) for both the Zero-Phonon Line (ZPL) and Phonon Sideband (PSB).
    2. Photonic Crystal Cavity: A zero-length 1D cavity coupled to a single ZPL GC, designed to enhance emission via the Purcell effect.

• Experimental Setup:

  • Devices were operated in a closed-cycle cryostat at 1.5 K.
  • Electrical Control: Forward bias was applied via p-i-n diodes. For single-photon experiments, pulsed electrical excitation (8 V, 150 ns pulses at 160 kHz) was used to minimize heating and cavity resonance shifts.
  • Optical Control: Resonant optical excitation was used for Photoluminescence Excitation (PLE) spectroscopy and spin readout.
  • Detection: Emission was collected via fiber arrays and detected using Superconducting Nanowire Single-Photon Detectors (SNSPDs) and spectrometers. Spectral filtering (free-space bandpass + Fiber Fabry-Pérot Interferometer) was employed to isolate single T center emission from background defects.

3. Key Contributions

1. First Electrically Injected Single-Photon Source in Silicon: The paper reports the first demonstration of electrically triggered single-photon emission from a silicon color center (the T center).
2. Electrically Triggered Spin Initialization: The authors demonstrate a new method for initializing the electron spin state of a T center by "heralding" on spectrally resolved electroluminescence (EL), eliminating the need for complex optical switching networks for initialization.
3. Integrated Optoelectronic Platform: The successful integration of p-i-n diodes with photonic cavities and waveguides in silicon, enabling simultaneous electrical and optical access to quantum defects.

4. Key Results

• Electroluminescence (EL) from T Center Ensembles:

  • Tapered waveguide devices successfully generated EL from T center ensembles in the telecommunications O-band (approx. 1326 nm).
  • Peak count rates reached 198 kcps for the ZPL and 111 kcps for the PSB at 250 $\mu$A current.
  • The devices functioned as on-chip silicon LEDs.

• Single-Photon Emission:

  • A single T center coupled to a photonic crystal cavity ($Q \approx 2960$) was electrically excited.
  • The second-order autocorrelation function under pulsed electrical excitation yielded $g^{(2)}(0) = 0.05(2)$ after background subtraction, confirming single-photon emission.
  • The lifetime under electrical excitation was measured at 570 ns, showing a Purcell enhancement factor of 1.65 compared to the bulk lifetime.

• Spin Initialization via EL Heralding:

  • By applying a magnetic field (350 mT), the spin-dependent optical transitions (A, B, C, D) were resolved.
  • The authors used electrical pulses to "repump" the spin state and spectral filtering to herald the final spin state.
  • Fidelity: The State Preparation and Measurement (SPAM) fidelity for initializing the ground state electron spin ($|\downarrow_E\rangle$ or $|\uparrow_E\rangle$) was 92(8)%.
  • The fidelity is currently limited by background luminescence from other defects and the time-bin size, but the method proves the concept of electrical spin initialization.

5. Significance and Impact

• Scalability: This work addresses a major bottleneck in scaling solid-state quantum computers. Traditional optical initialization requires individual optical switches for every qubit, which is unscalable. The proposed electrical spin initialization allows for the parallel control of thousands of T centers using on-chip filters and detectors, compatible with cryo-CMOS circuitry.
• Telecom Integration: The T center operates in the low-loss O-band, making it ideal for integration with existing silicon photonics and long-distance quantum networks.
• Versatility: The demonstrated ability to electrically inject, control, and read out spin states in silicon color centers establishes the T center as a versatile platform for both quantum memories and distributed quantum computing.
• Future Applications: The technology paves the way for electrically driven quantum networks, remote entanglement distribution between T centers, and the development of scalable, fault-tolerant quantum processors based on silicon defects.

In summary, this paper represents a foundational step toward practical, scalable quantum hardware by demonstrating that silicon T centers can be fully controlled via electrical interfaces, offering a pathway to high-density, parallel quantum processing.