⚛️ quantum physics

Spin Kerr-cat qubits

This paper introduces the spin Kerr-cat qubit encoding, which utilizes clock transitions in quadrupolar nuclei like antimony-123 to suppress dephasing and achieve estimated coherence times of 100 seconds and 99% gate fidelity for two-qubit operations.

Original authors: Z. M. McIntyre, Daniel Loss

Published 2026-04-22

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Z. M. McIntyre, Daniel Loss

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 send a secret message across a noisy room. If you whisper, the wind and chatter might change your words before they reach the listener. In the world of quantum computing, this "chatter" is called noise, and it destroys the delicate information stored in qubits (the basic units of quantum data) almost instantly.

This paper introduces a clever new way to protect that information, using a concept called the "Spin Kerr-cat" qubit. Here is the story of how it works, explained without the heavy math.

1. The Problem: The Shaky Table

Think of a standard quantum bit (qubit) as a spinning top on a table. If the table shakes (due to magnetic noise or electrical fluctuations), the top wobbles and eventually falls over. This is decoherence.

In silicon chips, scientists use the nuclei of atoms (like Antimony-123) as these spinning tops. They are very stable, but they still wobble. The main enemy here is dephasing: the noise makes the top spin slightly faster or slower than intended, scrambling the timing of the message.

2. The Solution: The "Clock" Trick

The authors propose a special encoding called the Spin Kerr-cat. To understand it, imagine a clock face.

Normal Qubits: If you try to tell time by looking at the second hand, a tiny breeze might push the hand forward or backward. Your time reading is wrong.
The "Clock Transition": The authors found a special spot on the clock face (a specific energy level) where the hand is momentarily paused. At this exact point, if a breeze pushes the clock, the hand doesn't move forward or backward; it just wiggles in place.

In physics terms, they tuned the system to a "clock transition." At this specific setting, the qubit's frequency becomes immune to the first order of noise. It's like finding a "sweet spot" where the noise cancels itself out.

3. The "Cat" in the Room

Why call it a "Cat"? This is a nod to Schrödinger's Cat, the famous thought experiment where a cat is both dead and alive at the same time.

The Analogy: Imagine a cat that is simultaneously sleeping on the left side of the bed and the right side of the bed. This is a "superposition."
The Spin Cat: In this new qubit, the information is stored in a state that is a superposition of two distinct nuclear spin directions. It's like the atom is spinning "left" and "right" simultaneously.
The Protection: Because the information is spread out between these two states (like the cat being in two places), a small nudge of noise can't easily flip the whole thing. It's much harder to knock over a cat that is stretching across the whole bed than one sitting in a corner.

4. The Magic Ingredients: Strain and Spin

To make this work, the scientists use a specific type of atom (Antimony) trapped in silicon.

The Squashed Ball: These atoms have a "squashed" shape (quadrupole moment). When you squeeze the silicon crystal (using strain), it creates an electric field gradient.
The Result: This interaction creates the "double-well" potential (the two sides of the bed) and the "clock transition" (the pause point) automatically. You don't need complex external lasers to create this; the physics of the atom itself does the heavy lifting.

5. The Results: From Seconds to Hours

The paper does some impressive number-crunching:

Current State: Without this trick, these nuclear spins might hold information for about 50 milliseconds (0.05 seconds).
With the Trick: By using this "Spin Kerr-cat" encoding, they estimate the information could last for 100 seconds.
Why it matters: That is a 2,000-fold improvement. It turns a fleeting whisper into a shout that can be heard clearly across the room.

6. How to Talk to the Cat

You might ask, "If the qubit is so protected, how do we read it or change it?"
The authors propose using a hopping electron as a messenger.

Imagine the nuclear spin (the cat) is in a cage.
An electron (a tiny, fast messenger) can hop into the cage, talk to the cat, and hop out.
Because the electron interacts strongly with the nucleus, it can read the cat's state or perform a "gate" (a logic operation) with another cat nearby.
They estimate they can perform these operations with 99% accuracy, which is good enough to start building real quantum computers.

The Big Picture

This paper is like inventing a noise-canceling headphone for quantum computers. Instead of trying to build a perfectly silent room (which is impossible), they built a special ear (the qubit encoding) that naturally ignores the background noise.

By leveraging the unique physics of heavy atoms in silicon, they found a way to make quantum memory last hundreds of times longer than before. This brings us one giant step closer to building quantum computers that can actually solve problems without falling apart from the slightest vibration.

1. Problem Statement

Quantum computing faces a central challenge: protecting quantum information from decoherence. While nuclear spins in silicon (specifically donors like $^{31}$ P and $^{123}$ Sb) offer exceptionally long coherence times compared to electron spins, they are still susceptible to dephasing caused by environmental noise (magnetic and electric).

The Limitation: Standard nuclear spin qubits typically suffer from dephasing times ( $T_2^*$ ) in the millisecond range, limited by $1/f$ noise and charge fluctuations.
The Goal: To extend the lifetime of quantum information at the hardware level by designing a qubit encoding that is inherently robust against noise, specifically targeting the dominant error source (dephasing) without relying solely on active quantum error correction (QEC) or dynamical decoupling.

2. Methodology and Theoretical Framework

The authors propose a new encoding scheme called the Spin Kerr-cat qubit, which leverages the physics of high-spin nuclei ( $I \ge 1$ ) interacting with electric-field gradients (EFG).

Physical System: The proposal focuses on high-spin donors in silicon, specifically Antimony-123 ( $^{123}$ Sb, $I=7/2$ ) and Bismuth-209 ( $^{209}$ Bi, $I=9/2$ ).
Hamiltonian and Symmetry:
- The nuclear spin Hamiltonian includes the Zeeman term and a quadrupole interaction term ( $H_q$ ) arising from the interaction between the nuclear quadrupole moment and the EFG.
- By aligning the external magnetic field with the principal axis of the EFG tensor, the Hamiltonian acquires a $Z_2$ symmetry (generated by $\hat{\Pi} = e^{i\pi \hat{I}_z}$ ).
- This symmetry ensures that the eigenstates of the system have definite parity (even or odd) with respect to $\hat{I}_z$ .
Clock Transition: The qubit is defined by the two lowest energy eigenstates of this Hamiltonian. The system is tuned to a specific magnetic field strength ( $B_0$ ) where the energy splitting ( $\Delta$ ) between these two states reaches a local maximum. At this point, the first derivative of the splitting with respect to magnetic field fluctuations vanishes ( $\partial \Delta / \partial B = 0$ ). This is known as a clock transition, rendering the qubit first-order insensitive to noise.
Spin Cat States: The authors demonstrate that these two lowest eigenstates are well-approximated by spin cat states (superpositions of spin coherent states with opposite azimuthal phases).
- Analogy: This is the spin analog of the circuit QED "Kerr-cat" qubit. In the limit $I \to \infty$ , the quadrupole Hamiltonian maps onto a squeezed Kerr-nonlinear oscillator via the Holstein-Primakoff transformation.
- Distinction: Unlike circuit QED Kerr-cats which suppress bit-flips (photon loss), the Spin Kerr-cat encoding is designed to suppress phase-flips (dephasing), which is the dominant error channel for spin qubits.

3. Key Contributions

Novel Encoding: Introduction of the Spin Kerr-cat encoding, utilizing the quadrupolar interaction of high-spin nuclei to create a hardware-protected qubit.
First-Order Noise Suppression: Demonstration that operating at a clock transition in the nuclear-spin spectrum leads to a vanishing first-order sensitivity to magnetic and electric noise, significantly extending dephasing times.
Analytical Modeling of Decoherence:
- Dephasing: Calculated the dephasing time ( $T_{2,ct}^*$ ) under a realistic $1/f$ noise model. The authors show that because the first derivative of the splitting is zero, the decay follows a slower power law ( $t^{-1/2}$ ) rather than a Gaussian decay, leading to massive enhancements in coherence time.
- Relaxation: Analyzed the relaxation channel caused by charge noise breaking the $Z_2$ symmetry. They derived that bit-flip errors due to charge fluctuators (TLFs) are exponentially suppressed with distance, estimating bit-flip times $>1000$ s for TLFs at 10 nm.
Control Protocols: Proposed a complete suite for qubit operation:
- Initialization: Using ESR pulses and parity measurements to prepare the spin in the cat subspace.
- Single-Qubit Gates: Using NMR-style RF fields or virtual Z-gates.
- Two-Qubit Gates: A scheme mediated by a mobile electron spin that shuttles between donors, utilizing the hyperfine interaction to generate an entangling gate (equivalent to a Controlled-Z).
- Readout: Mapping the nuclear spin parity onto an electron spin for measurement.

4. Results and Estimates

Using measured parameters for $^{123}$ Sb donors in silicon:

Coherence Time: The authors estimate that with a quadrupole coupling $Q \approx 100$ kHz and a standard high-field dephasing time $T_2^* \approx 10-50$ ms, the Spin Kerr-cat encoding could achieve a dephasing time of $T_{2,ct}^* \approx 100$ seconds. This represents an enhancement of several orders of magnitude.
Gate Fidelity:
- They estimate that a two-qubit gate mediated by hopping electrons could achieve 99% fidelity.
- This requires a modest enhancement (factor of $\approx 3-5$ ) of the quadrupole splitting compared to current experimental values to ensure the secular approximation holds and the spin-cat approximation is accurate ( $F > 99\%$ ).
Robustness: The encoding is robust against charge noise as long as charge fluctuators are located further than $\sim 10$ nm from the donor. Phonon-induced relaxation is found to be negligible compared to charge noise.

5. Significance

Hardware-Level Protection: This work offers a path to long-lived quantum memories by embedding error suppression directly into the qubit's physical encoding, reducing the overhead required for active quantum error correction.
Scalability: The use of nuclear spins in silicon, combined with the ability to shuttle electrons for gate operations, aligns well with scalable silicon quantum computing architectures.
New Physics: It bridges the gap between continuous-variable bosonic codes (Kerr-cats) and discrete-variable spin systems, showing that high-spin nuclei can naturally realize "cat-like" states with built-in protection against the dominant noise source in solid-state spin qubits.
Feasibility: The proposal relies on parameters (quadrupole couplings, magnetic fields) that are within reach of current or near-future experimental capabilities in silicon quantum devices, making it a highly actionable roadmap for next-generation quantum processors.