🔬 mesoscale physics

Control of threshold voltages in Si/SiGe quantum devices via optical illumination

This paper demonstrates that systematic threshold voltage shifts in Si/SiGe quantum devices can be achieved and precisely controlled via near-infrared optical illumination under an applied gate voltage, providing a reproducible method to set stable operating conditions and explaining the mechanism behind the successful recovery of qubit devices from charge injection events.

Original authors: M. A. Wolfe, Brighton X. Coe, Justin S. Edwards, Tyler J. Kovach, Thomas McJunkin, Benjamin Harpt, D. E. Savage, M. G. Lagally, R. McDermott, Mark Friesen, Shimon Kolkowitz, M. A. Eriksson

Published 2026-02-05

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: M. A. Wolfe, Brighton X. Coe, Justin S. Edwards, Tyler J. Kovach, Thomas McJunkin, Benjamin Harpt, D. E. Savage, M. G. Lagally, R. McDermott, Mark Friesen, Shimon Kolkowitz, M. A. Eriksson

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 a quantum computer as a tiny, ultra-precise orchestra. Each instrument in this orchestra is a "quantum dot," a microscopic trap that holds a single electron to act as a bit of information (a qubit). For the orchestra to play in tune, every instrument must be perfectly calibrated. In these silicon-based devices, this calibration is controlled by a "threshold voltage"—think of it as the specific amount of pressure you need to apply to a gate to let the electron in.

The problem is that these gates are notoriously finicky. Because of tiny imperfections and trapped electrical charges at the microscopic interfaces (like dust on a lens), the pressure needed to open the gate can vary wildly from one device to another, or even change after the device gets cold. This makes it hard to get the orchestra to start playing.

Scientists often use a trick called "optical illumination" (shining a light on the device while it's freezing cold) to fix this. It's like hitting the "reset" button on a glitchy video game. However, nobody really understood how the light fixed the problem or if they could use it to tune the instrument to a specific note, rather than just resetting it to a default.

This paper is about discovering how to use that light not just as a reset button, but as a precise tuning knob.

The Experiment: Shining a Light with a Push

The researchers built a special silicon device and cooled it down to near absolute zero. They then shined a near-infrared laser on it while applying different amounts of electrical "push" (voltage) to the gate.

Here is what they found, explained through simple analogies:

1. The "Magic Match" (Small Pushes)
When they shined the light while applying a small electrical push, something magical happened. The "threshold voltage" (the pressure needed to open the gate) shifted to match the push they were applying almost perfectly.

The Analogy: Imagine a crowded hallway where people (electrons) are stuck in a jam. If you shine a light, it wakes them up and lets them move. If you gently push the crowd from one side, the light lets them rearrange themselves to fill that space perfectly. When you stop pushing and turn off the light, the crowd stays in that new formation. The researchers found that by choosing how hard they pushed, they could "freeze" the device into a specific, stable state. If they pushed with 0.5 volts, the device would now require exactly 0.5 volts to turn on.

2. The "Full Parking Lot" (Medium Pushes)
As they increased the push, they hit a limit. The threshold voltage stopped moving and stayed flat.

The Analogy: Think of the interface between the silicon and the glass (oxide) as a parking lot with a fixed number of spots. The light helps cars (electrons) find empty spots. Once every spot is filled, no matter how hard you push or how bright the light is, you can't fit any more cars in. The system has reached "saturation." The researchers calculated that this parking lot holds a specific number of charges, and once it's full, the tuning stops.

3. The "High-Speed Tunnel" (Large Pushes)
When they pushed even harder (above 1.5 volts), the threshold voltage started shifting again, but this time it wasn't because of the light filling spots.

The Analogy: The electrical push became so strong that it created a "tunnel" through the barrier (a process called Fowler-Nordheim tunneling). It's like the cars in the parking lot suddenly gaining enough speed to drive through the wall instead of just parking in the lot. This allowed extra charge to get trapped in places the light couldn't reach before, shifting the threshold voltage in a new way.

4. The "Two-Photon Dance" (Negative Pushes)
When they pushed in the opposite direction (negative voltage), the behavior changed again. The amount of tuning depended on the square of the light's brightness.

The Analogy: This suggests a "two-photon process." Imagine trying to open a heavy door. A single photon (a particle of light) isn't strong enough to knock it open. But if two photons hit the door at the exact same time, they combine their energy to knock it open. The researchers found that in this negative voltage regime, the light needed to work in pairs to free up trapped charges.

Why This Matters

The paper concludes that this method gives scientists a powerful new tool. Instead of just blindly hoping a quantum device works after it cools down, they can now use a laser and a specific voltage to "dial in" the exact operating point they need.

It explains why the old "reset" trick works: the light wakes up trapped charges, allowing them to rearrange and screen out electrical noise. But now, by adding a voltage "push" while shining the light, they can control exactly how those charges rearrange. This turns a chaotic, unpredictable device into a precisely tuned instrument, ready to join the quantum orchestra.

Technical Summary: Control of Threshold Voltages in Si/SiGe Quantum Devices via Optical Illumination

Problem Statement
Gate-defined semiconductor quantum-dot devices offer a platform for solid-state qubits but suffer from significant variability in operating points, even between nominally identical devices on the same chip. A primary source of this instability is the presence of charge traps at the gate-oxide interface, which lead to electrostatic disorder, charge noise, and unpredictable threshold voltages ( $V_T$ ) upon cooling. While optical illumination at cryogenic temperatures is a common experimental technique to "reset" devices by rearranging trapped charge, the underlying mechanisms, limitations, and potential for systematic control of this process remain poorly understood. Specifically, the relationship between the applied gate bias during illumination and the resulting stable threshold voltage has not been systematically characterized.

Methodology
The authors investigated a dopant-free Si/SiGe field-effect transistor (H-FET) and a Si/SiGe quadruple quantum-dot device. The experimental setup involved cooling the devices to cryogenic temperatures (1.2 K to 3 K) and utilizing a near-infrared (780 nm) laser diode for illumination.

The core methodology involved a systematic "biased illumination" protocol:

Initialization: Devices were reset by illuminating at zero gate bias ( $V_B = 0$ V) to establish a consistent starting condition.
Biased Illumination: The device was illuminated for a fixed duration (30 s) while a specific DC gate bias ( $V_B$ ) was applied.
Characterization: Following illumination, the source-drain current ( $I_{SD}$ ) was measured as a function of gate voltage to extract the new threshold voltage ( $V_T$ ), defined as the voltage required to achieve $I_{SD} = 1$ nA.
Parameter Variation: The experiments varied the magnitude and sign of the applied bias ( $V_B$ ), the laser current (light intensity), and the duration of the bias application to distinguish between different charge trapping mechanisms.

Key Results
The study reveals distinct regimes of threshold voltage tuning based on the applied gate bias during illumination:

Linear Tuning at Low Bias: For small gate biases, the resulting threshold voltage shifts linearly with the applied bias ( $V_T \approx V_B$ ) with a slope of $0.94 \pm 0.01$ . The authors attribute this to photo-generated electron-hole pairs that migrate to the oxide-semiconductor interface to screen the electric field created by the gate bias and work function differences. These carriers become trapped upon turning off the light, effectively "freezing" the interface charge density to match the applied bias.
Saturation at Moderate Positive Bias: As the positive bias increases, $V_T$ reaches a plateau (observed around $V_B \approx 1.1$ V). This saturation corresponds to the filling of all available interface states below the SiGe conduction band edge. The authors estimate the maximum trapped charge density ( $\sigma_{max}$ ) and the average interface state density ( $\bar{D}_{it} \approx 3.6 \times 10^{12} \text{ eV}^{-1}\text{cm}^{-2}$ ), consistent with low-temperature thermal growth interfaces.
Fowler-Nordheim Tunneling at High Positive Bias: Beyond the saturation plateau ( $V_B > 1.5$ V), $V_T$ shifts upward again. This behavior is attributed to Fowler-Nordheim tunneling, where electrons in the accumulated quantum well tunnel through the triangular SiGe barrier into the oxide or localized states in the silicon cap layer. This process allows for the trapping of charge density exceeding the limit of the interface states. The tunneling current density follows the expected $J \propto E^2 e^{-B/E}$ relationship.
Quadratic Dependence at Negative Bias: Under large negative bias, the threshold voltage shift depends quadratically on the laser intensity. This suggests a competing two-photon process that liberates trapped charge while photo-generated carriers attempt to screen the field. Additionally, at very large negative biases, Fowler-Nordheim dynamics re-emerge.

Key Contributions and Model
The paper presents a simple, intuitive model explaining these phenomena based on the control of trapped charge density at the oxide-semiconductor interface. The model posits that:

Illumination generates mobile carriers that screen the internal electric field.
The sign and magnitude of the applied bias determine which carriers (electrons or holes) accumulate and become trapped at the interface.
The tunability is limited by the density of available interface states, with Fowler-Nordheim tunneling providing a mechanism to access higher charge densities at large biases.

Significance
The authors claim that these results provide a systematic method for the in-situ tuning of threshold voltages in Si/SiGe quantum devices over a wide voltage range. By demonstrating that the threshold voltage can be set to match a specific gate bias during illumination, the work offers a precise tool for calibrating quantum-dot qubit devices. Furthermore, the study elucidates the physical mechanisms behind the widely used practice of cryogenic illumination, explaining its success in resetting devices after undesired charging events and clarifying the roles of interface state saturation and tunneling in device stability. The findings suggest that illumination can be utilized not just for resetting, but potentially for setting non-zero operating points on individual gates within a device.

The Experiment: Shining a Light with a Push

Why This Matters

More like this