⚛️ quantum physics

Optimal Two-Qubit Gates for Group-IV Color-Centers in Diamond

This paper demonstrates a scalable strategy for implementing robust two-qubit gates with fidelities exceeding 99.9% between the electron and nuclear spins of Germanium-vacancy centers in diamond using quantum optimal control, thereby addressing a critical requirement for distributed quantum computing and quantum repeaters.

Original authors: Jurek Frey, Katharina Senkalla, Philipp J. Vetter, Fedor Jelezko, Frank K. Wilhelm, Matthias M. Müller

Published 2026-04-14

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Jurek Frey, Katharina Senkalla, Philipp J. Vetter, Fedor Jelezko, Frank K. Wilhelm, Matthias M. Müller

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

The Big Picture: Building a Quantum Internet

Imagine you want to build a "Quantum Internet." To do this, you need tiny, super-powerful computers (nodes) scattered across the globe that can talk to each other instantly using "spooky" quantum connections called entanglement.

The researchers in this paper are working on one specific type of hardware for these nodes: Diamonds with special defects (called Group-IV color centers). Think of these diamonds not as jewelry, but as tiny, ultra-stable quantum factories. Inside these factories, they use two main workers:

The Electron: A fast, energetic worker who can talk to the outside world (via light) but gets tired and confused easily (decoherence).
The Nucleus: A slow, calm, and incredibly patient worker (a Carbon-13 atom) who can hold onto information for a long time (memory).

The Problem: The "Too-Tight" Dance

To make these computers work, the Electron and the Nucleus need to perform a specific dance together. They need to swap information or flip each other's states (like a CNOT or SWAP gate).

In many diamond setups, these two are too close. They are so tightly linked that they constantly bump into each other.

The Analogy: Imagine trying to teach a toddler (the Electron) and a grandparent (the Nucleus) to dance a complex waltz. The toddler is energetic but clumsy and gets distracted by noise. The grandparent is steady but moves slowly. Because they are holding hands so tightly, when the toddler stumbles, they drag the grandparent down with them. This "stumbling" is noise and decoherence, which ruins the dance (the calculation).

Previously, scientists tried to control this dance using simple, standard moves. But because the toddler was so jittery and the connection so strong, the dance often ended in a mess, with low accuracy (fidelity).

The Solution: The "Choreographer" (Quantum Optimal Control)

The authors of this paper didn't try to separate the dancers or wait for the toddler to calm down. Instead, they hired a super-choreographer (Quantum Optimal Control, specifically an algorithm called dCRAB).

This choreographer doesn't just say "do a spin." Instead, they design a custom, split-second dance routine that accounts for every stumble, every noise, and every bump.

The Metaphor: Instead of telling the toddler to "spin left," the choreographer says, "At 0.001 seconds, lean slightly right to counter the wind, then pivot hard left, then pause exactly when the grandparent is ready." It's a highly complex, mathematically perfect sequence of microwave pulses that guides the system through the noise.

What They Achieved

Using this "choreographer," they managed to teach the Electron and Nucleus to perform three critical dances:

The CNOT: A move where one dancer flips the other based on their own state.
The SWAP: A move where they completely exchange places (the Electron's info goes to the Nucleus, and vice versa).
The Hadamard: A move that puts the dancer into a superposition (a state of being both "up" and "down" at once).

The Result?
They achieved 99.9% accuracy.

In everyday terms: If you asked this system to perform this dance 1,000 times, it would only mess up once or twice. In the world of quantum computing, where errors are usually rampant, this is a massive victory. It means the system is now reliable enough to be used in real-world quantum networks.

The "Secret Sauce": Tuning the Music

The researchers also discovered a clever trick. They realized they could change the "music" (the magnetic field strength) to change how fast the dancers move.

The Analogy: By turning up the volume of the music (increasing the magnetic field), they could make the dancers move faster. While moving faster usually makes it harder to stay in sync, their choreography was so good that they could speed up the dance significantly without losing accuracy. This means they can process information much faster than before.

The "Magic Trick": Redefining the Goal

For the "SWAP" dance (exchanging places), they tried a new approach. Instead of demanding the dancers end up in exactly the same position as a textbook SWAP, they asked: "Can you end up in a position that acts like a SWAP, even if you did a few extra spins along the way?"

The Result: By allowing this flexibility (using something called Cartan decomposition), they found a path that was even more robust against noise. It's like realizing that to get to the destination, you don't have to walk in a straight line; a slightly curved path that avoids the puddles is actually faster and safer.

Why This Matters

This paper is a blueprint for building the future of the Quantum Internet.

Reliability: They proved that even with "noisy" and "clumsy" components, we can build reliable quantum nodes.
Scalability: The method they used (the choreographer) can be applied to other types of diamond defects, not just this one.
Speed: They showed how to make these operations faster, which is crucial for keeping quantum information alive before it fades away.

In short: They took a messy, noisy quantum system and used advanced math to choreograph a perfect, high-speed dance, proving that diamond-based quantum computers are ready to join the global network.

1. Problem Statement

Group-IV color centers in diamond (specifically the Germanium-vacancy or GeV center) are promising candidates for distributed quantum computing and quantum repeaters due to their long-lived nuclear spin memory and strong optical transitions. However, realizing high-fidelity two-qubit gates between the electron spin and a strongly coupled nuclear spin (e.g., $^{13}$ C) presents significant challenges:

Strong Hyperfine Coupling: While strong coupling enables fast gate operations, it introduces complex dynamics that can degrade gate fidelity if not managed carefully.
Decoherence: The electron spin is susceptible to dephasing noise, which limits the coherence time and gate performance.
Control Limitations: Traditional perturbative approaches or dynamical decoupling sequences often face time limitations or hardware constraints that prevent achieving the high fidelities (>99.9%) required for fault-tolerant quantum networks.
Lack of Intrinsic Nuclear Spins: Unlike Nitrogen-Vacancy (NV) centers, Group-IV centers lack intrinsic host nuclear spins, making the control of nearby random $^{13}$ C spins more challenging due to symmetry constraints in hyperfine interactions.

The core problem is to design robust, fast, and high-fidelity control pulses for two-qubit gates (CNOT and SWAP) in the presence of realistic dephasing noise, overcoming the limitations imposed by strong hyperfine coupling.

2. Methodology

The authors employ Quantum Optimal Control (QOC) to synthesize control pulses that maximize gate fidelity.

System Model:
- Hamiltonian: The system consists of an electron spin ( $S=1/2$ ) and a nuclear spin ( $I=1/2$ ). The Hamiltonian includes Zeeman splittings ( $\omega_e, \omega_I$ ), longitudinal ( $A_{zz}$ ) and transverse ( $A_{zx}$ ) hyperfine couplings, and a time-dependent microwave control field acting on the electron.
- Noise Model: Dephasing is modeled using an Ornstein–Uhlenbeck (OU) process to simulate non-Markovian noise with finite correlation time, parameterized by experimental values ( $T_2^*$ and $T_2$ ).
Optimization Algorithm:
- The study utilizes the QuOCS suite with the dCRAB (differential Chopped Random Basis) algorithm.
- Control pulses are parameterized as a truncated Fourier series with a Gaussian envelope to ensure smooth switching.
Optimization Strategies:
1. Standard Average Fidelity (FoM $_{av}$ ): Minimizes the average distance between the noisy evolution and a specific target unitary (e.g., exact CNOT or SWAP) over 5,000 noise realizations.
2. Invariant-Based Optimization (FoM $_{cb}$ ): Recognizing that gates differing only by local single-qubit rotations are equivalent for entanglement generation, the authors optimize for the non-local content of the gate.
  - They use the Cartan (KAK) decomposition to isolate the non-local component ( $A$ ) of the gate.
  - The cost function (FoM $_{cb}$ ) combines the non-local fidelity (matching Weyl chamber coordinates) with the noise robustness of the ideal (noiseless) evolution. This allows the optimizer to find pulses that are robust against dephasing even if they do not produce the exact target unitary, but rather a locally equivalent one.

3. Key Contributions

High-Fidelity Gate Synthesis: The paper demonstrates the numerical realization of robust two-qubit gates (CeNOT $_n$ , CnNOT $_e$ , and SWAP) with fidelities exceeding 99.9% under realistic noise conditions.
Invariant-Based Control: A significant methodological contribution is the application of invariant-based optimization (using Cartan decomposition) to two-qubit gates in noisy solid-state systems. This approach relaxes local constraints, allowing the optimizer to find solutions that are significantly more robust against dephasing than those targeting a specific unitary.
Parameter Space Exploration: The study investigates the impact of the nuclear Larmor frequency ( $\omega_I$ ) on gate performance, identifying regimes where higher magnetic fields enable faster gates without sacrificing fidelity.
Scalable Framework: The proposed control strategy is adaptable to other Group-IV color centers (e.g., Silicon-vacancy, Tin-vacancy) as they share the same electron spin-1/2 structure.

4. Key Results

Gate Fidelities:
- CeNOT $_n$ (Electron-controlled NOT): $99.91\%$ fidelity at $T = 4.450$ $\mu$ s.
- CnNOT $_e$ (Nuclear-controlled NOT): $99.94\%$ fidelity at $T = 5.2$ $\mu$ s.
- SWAP Gate: $99.91\%$ fidelity at $T = 8.2$ $\mu$ s. A faster alternative ($4.825$ $\mu$ s) achieved $99.89\%$ .
Impact of Invariant Optimization:
- When optimizing the SWAP gate using the invariant-based approach (FoM $_{cb}$ ), the gate error was reduced by almost one order of magnitude compared to standard average-fidelity optimization.
- For a gate duration of $2.950$ $\mu$ s, the non-local infidelity ( $1-F_{nl}$ ) reached $2.3 \times 10^{-6}$ , significantly outperforming standard methods.
Speed vs. Robustness:
- Increasing the nuclear Larmor frequency ( $\omega_I$ ) significantly reduces the required gate duration (speeding up operations) without degrading the maximum achievable fidelity, though it increases the variance in performance at very high frequencies.
Comparison to Direct Driving: The optimized pulses are approximately one order of magnitude faster than direct RF-driven approaches or those relying on perturbative schemes.

5. Significance

This work provides a critical pathway toward scalable quantum networks using Group-IV color centers.

Fault Tolerance: Achieving fidelities >99.9% is a prerequisite for fault-tolerant quantum error correction and long-distance entanglement distribution.
Experimental Feasibility: The control pulses and parameter regimes identified are compatible with current experimental hardware (microwave amplitudes up to 15 MHz), making immediate experimental verification possible.
Quantum Memory: The ability to perform fast, robust CNOT gates is essential for mapping nuclear spin states to electron spins for single-shot readout, minimizing back-action on the nuclear memory.
Generalizability: The invariant-based optimization strategy offers a new paradigm for designing robust gates in other solid-state qubit architectures where local noise and strong couplings are prevalent.

In conclusion, the paper establishes that Quantum Optimal Control, particularly when leveraging local symmetries and invariant-based cost functions, can overcome the inherent limitations of strong hyperfine coupling and decoherence in Group-IV color centers, enabling them to function as high-performance nodes in a future quantum internet.