🔬 atomic physics

Quantum science with arrays of metastable helium-3 atoms

This paper presents a comprehensive architectural blueprint for using light, metastable helium-3 atoms in programmable optical tweezer arrays to overcome inertia limitations, enabling significantly faster atomic transport and hopping, novel qubit manipulation in trap potentials, and enhanced resource efficiency for fermionic quantum simulation and computation.

Original authors: Zheyuan Li, Rupsa De, Rishi Sivakumar, William Huie, Hao-Tian Wei, Justin D. Piel, Chris H. Greene, Kaden R. A. Hazzard, Zoe Z. Yan, Jacob P. Covey

Published 2026-07-13

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Zheyuan Li, Rupsa De, Rishi Sivakumar, William Huie, Hao-Tian Wei, Justin D. Piel, Chris H. Greene, Kaden R. A. Hazzard, Zoe Z. Yan, Jacob P. Covey

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 not as a giant, humming supercomputer in a cold room, but as a tiny, high-speed train system made of invisible tracks and single atoms. For years, scientists have been building these tracks using "optical tweezers"—lasers that act like pincers to grab and move individual atoms. But there's a catch: the atoms they've been using, like Lithium-6, are a bit like heavy freight cars. They move, but they're sluggish.

This paper proposes a radical upgrade: swapping those heavy freight cars for the lightest, fastest "race cars" possible in the atomic world—metastable Helium-3 atoms ( $^3$ He $^*$ ). Because these atoms are so incredibly light, the entire quantum system speeds up dramatically, opening doors to new types of physics simulations that were previously too slow or too clunky to build.

The Speed Boost: From Freight to Formula 1

The main finding here is all about speed. In the quantum world, the lighter the object, the faster it can move and the more quickly it can change its state. The authors calculate that by using Helium-3, the "hopping" of an atom from one laser trap to another can be $\gtrsim 3\times$ faster than what has been demonstrated with Lithium-6.

Think of it like this: if your current quantum computer is a bicycle, this new proposal turns it into a jet ski. The paper suggests that this speed isn't just a nice-to-have; it's essential. Quantum computers have a "time limit" before their information gets scrambled (decoherence). If your operations are too slow, the computer forgets what it was doing before it finishes the math. By using the lightest trappable atom, the authors suggest we can perform these operations so quickly that we stay ahead of the clock.

The "Magic" Magnetic Field and the Qubit

To make this work, the atoms need to be stable. The paper identifies a specific "magic" magnetic field of 803.5 G (Gauss). At this exact setting, the atoms become immune to tiny wobbles in the magnetic field, making them perfect for storing information (qubits).

The authors also argue against the idea that light atoms are too messy to control. While lighter atoms usually scatter light more easily (which can be bad), the specific structure of Helium-3's energy levels is actually a "sweet spot." It has a large gap between its energy states, which helps keep the information clean. In simulations, the authors show that this setup could achieve gate fidelities (accuracy) of 0.9995, which is comparable to the best heavier atoms like Sodium, but much faster.

New Tricks with the "Bouncy" Trap

Here is where it gets really playful. Usually, scientists use the "spin" of an atom (like a tiny magnet pointing up or down) to store data. But because Helium-3 is so light and the laser traps are so tight, the atoms bounce around inside the trap with a frequency of about 400 kHz.

The paper suggests we can use this bouncing motion itself as a qubit! Imagine a ball in a bowl. Usually, you just care if the ball is in the bowl or not. But because the bowl is so small and the ball is so light, the ball can only vibrate in specific, distinct steps. The authors propose encoding information in these vibration steps (like the ball being at the bottom vs. one step up). Because the "steps" are so far apart in energy (about 30% anharmonicity, compared to the 5–10% found in superconducting circuits), we can easily tell them apart without accidentally hitting the wrong step. This is a "new toolbox" that only works because the atom is so light.

The "Anti-Tweezer" and the Tunneling

To move these atoms, the team proposes using "blue-detuned" light. Instead of shining a bright light to hold the atom in the center (like a red-detuned trap), they suggest using a "dark hole" in a bright field. The atom sits in the darkness, avoiding the light that might knock it out of its state.

In this setup, the atoms can "tunnel" (quantum jump) between traps incredibly fast. The authors simulate that with a trap depth of 10 MHz and a separation of 1.2 µm, the tunneling rate could reach $\sim$ 1 kHz. This is a massive improvement over previous experiments, which struggled to get rates above 300 Hz. This speed is crucial for simulating "fermionic" systems (like electrons in a metal), where particles need to swap places constantly.

What This Can (and Can't) Do

The paper outlines three main ways this fast, light system could be used:

Fermionic Quantum Simulation: Instead of translating complex electron rules into simple "on/off" switches (which is slow and error-prone), this system uses the atoms' natural "fermionic" nature. The atoms naturally follow the rules of electrons, so we can simulate materials like high-temperature superconductors or complex molecules directly. The authors suggest this could help solve mysteries like why some materials conduct electricity without resistance at high temperatures.
Molecular Energy Calculations: By using a method called VQE (Variational Quantum Eigensolver), the system could calculate the energy of molecules. The authors estimate that for 1D chains, this fermionic approach could be 2–3 times more efficient in terms of the number of gates needed compared to standard qubit methods.
Precision Measurements: The paper suggests that these atom arrays could be used as ultra-precise clocks. Because the atoms are trapped in a grid, we can avoid the "collisions" that mess up measurements in gas clouds. This could help solve the "proton charge radius puzzle" by comparing the clock ticks of Helium-3 and Helium-4.

The Bottom Line

This paper doesn't claim to have built the machine yet; it is a comprehensive "architectural blueprint." It suggests that by switching to Helium-3, we can build a quantum computer that is faster, more efficient, and capable of simulating complex physics that current heavy-atom systems struggle with. The authors are confident in the physics (the math checks out), but the actual construction of such a system remains a future challenge. They argue that the lightness of Helium-3 isn't just a minor tweak; it's the key to unlocking a new generation of quantum science.

Technical Summary: Quantum Science with Arrays of Metastable Helium-3 Atoms

Problem Statement
Optical tweezer arrays of neutral atoms have emerged as a leading platform for quantum science, enabling programmable connectivity, bosonic error correction, and native fermionic encoding. However, the utility of atomic motion for these applications is fundamentally limited by the inertia of the atoms. The timescales for coherent transport, inter-tweezer tunneling, and motional state manipulation scale inversely with atomic mass. Consequently, heavier species (e.g., $^{87}$ Rb, $^{133}$ Cs) suffer from slower operation speeds, limiting the resource efficiency of quantum algorithms and the fidelity of fermionic simulations due to finite coherence times. While lighter species like $^{6}$ Li offer faster dynamics, they present significant challenges in laser cooling, state preparation, and hyperfine structure resolution.

Methodology and Architectural Blueprint
The authors propose a comprehensive architectural blueprint utilizing metastable Helium-3 ( $^{3}\text{He}^*$ ), the lightest trappable atomic species, in programmable optical tweezer arrays. The paper provides a detailed analysis of the atomic structure, optical trapping potentials, and interaction mechanisms required to realize this platform.

Atomic Structure and Level Scheme: The analysis focuses on the unique level structure of $^{3}\text{He}^*$ , which features a "metastable ground state" ( $1s2s\ ^3S_1$ ) 20 eV above the true ground state with a lifetime of $\approx 7800$ s. This state possesses a nuclear spin of $I=1/2$ , sharing features with alkaline-earth-like atoms (favorable level structure for Raman coupling) and heavy alkalis (strong S-P transitions at 1083 nm).
Optical Trapping and Cooling: The authors calculate the optical polarizability for key states to identify suitable trapping wavelengths. They propose using red-detuned tweezers at 1150 nm or blue-detuned "anti-tweezers" at 1013 nm. The small mass of $^{3}\text{He}^*$ results in high trap frequencies ( $\omega \sim m^{-1/2}$ ), enabling efficient Raman Sideband Cooling (RSC) to the motional ground state. The paper details a bichromatic imaging scheme (1083 nm cooling, 389 nm probe) to achieve high-fidelity fluorescence readout while minimizing photoionization losses.
Qubit Encoding and Gates:
- Hyperfine Qubits: A "magic" magnetic field of 803.5 G is identified where the differential Zeeman shift between the $|F=3/2, m_F=-1/2\rangle$ and $|F=1/2, m_F=-1/2\rangle$ states vanishes, creating a field-insensitive qubit.
- Rydberg Interactions: A two-photon excitation scheme (389 nm and 785 nm) is proposed to access Rydberg states. The authors analyze the $C_6$ coefficients, noting that while $^{3}\text{He}^*$ requires higher principal quantum numbers ( $n \sim 70$ ) than alkalis to achieve similar interaction strengths due to smaller quantum defects, the light mass offers advantages in motional sensitivity and blockade regimes.
- Motional Qubits: The high trap frequency and significant anharmonicity of shallow tweezers allow for the encoding of qubits directly in the motional states ( $|n=0\rangle$ and $|n=1\rangle$ ). These can be manipulated via trap position modulation.
Coherent Transport: The paper models coherent inter-tweezer tunneling in a double-well potential. Due to the mass ratio between $^{3}\text{He}^*$ and $^{6}$ Li, the tunneling rate scales as $t \sim m^{-1}$ , predicting significantly faster hopping.

Key Results

Enhanced Dynamics: The authors demonstrate that inter-tweezer hopping of $^{3}\text{He}^*$ atoms can be $\gtrsim 3\times$ faster than previous demonstrations with $^{6}$ Li. The trap frequency is $\approx 6\times$ higher than for $^{87}$ Rb, enabling faster coherent transport and motional state manipulation.
Raman Coupling Fidelity: Calculations of the Raman coupling figure of merit ( $|\beta|$ ) show that at the magic field (800 G), $^{3}\text{He}^*$ achieves a $|\beta|$ ratio $\sim 100\times$ larger than $^{6}$ Li, resulting in a single-qubit gate fidelity of $F \approx 0.9995$ , comparable to $^{23}$ Na.
Tunneling Rates: Numerical simulations of a 3D double-well potential predict absolute tunneling rates on the $\sim 1$ kHz scale for $^{3}\text{He}^*$ at 1.2 $\mu$ m separation, compared to $\sim 300$ Hz for $^{6}$ Li.
Measurement-Based Cooling: A protocol is proposed using a tunnel-coupled auxiliary tweezer to detect and remove thermal errors, enabling the preparation of defect-free, zero-temperature arrays.
Anharmonicity: The paper shows that at shallow trap depths ( $\sim 75$ kHz), the anharmonicity of the $^{3}\text{He}^*$ trap reaches $\approx 30\%$ , allowing for high-fidelity addressing of motional qubits without coupling to higher levels.

Significance and Claims
The paper claims that the $^{3}\text{He}^*$ platform offers a unique combination of minimal mass and optimal level structure, addressing the inertia limitations of current neutral atom quantum science. By leveraging the light mass, the platform enables:

Resource Efficiency: Faster gate times and transport reduce the wall-clock time for deep, error-corrected algorithms.
Native Fermionic Simulation: The ability to perform direct fermionic hopping operations (via tunneling) without the overhead of mapping fermionic operators to qubit operators (e.g., Jordan-Wigner transformations).
New Encoding Schemes: The high trap frequency enables a new toolbox for encoding qubits in motional states, facilitating bosonic error correction and dense encoding of gauge bosons.
Advanced Simulations: The architecture supports the simulation of complex fermionic systems, including geometrically frustrated lattices for high- $T_c$ superconductivity, multi-orbital many-body physics (e.g., Kondo lattices, moiré materials), and 3D strongly correlated systems with single-site resolution.
Precision Metrology: The platform offers a pathway to "atomic array optical clocks" using metastable helium, potentially improving isotope shift measurements to probe the proton charge radius puzzle with reduced collisional and Doppler shifts compared to bulk gas experiments.

The authors conclude that this architecture provides a scalable, resource-efficient route for quantum simulation and computation, particularly for problems involving itinerant fermions coupled to bosonic baths, such as lattice gauge theories and quantum chemistry beyond the Born-Oppenheimer approximation.