Trapped-ion qubits

Trapped-ion qubits store quantum information in the internal electronic states of individual atomic ions confined in vacuum by oscillating and static electromagnetic fields. Manipulated with laser or microwave pulses, they achieve the highest single- and two-qubit gate fidelities of any modality and offer all-to-all connectivity within a trap, at the cost of comparatively slow gate speeds.

#How they work

Ions such as ytterbium, calcium, or beryllium are ionized and held in a linear radio-frequency (Paul) trap, forming a crystal spaced by their mutual Coulomb repulsion. Each ion is a Qubit encoded in two long-lived electronic levels, often hyperfine ground states that are insensitive to magnetic noise. Single-qubit gates are driven by resonant pulses. Two-qubit gates use the shared motional modes of the ion crystal as a bus: a laser couples each ion's internal state to the collective vibration, entangling ions that are not physically adjacent. This scheme was proposed in 1995 and remains the foundation of the field (Cirac and Zoller 1995).

#Strengths

Because all ions of a species are identical and naturally isolated from their environment, trapped ions reach exceptional performance. Two-qubit gate fidelities above 99.9 percent and single-qubit fidelities above 99.99 percent have been demonstrated (Ballance et al. 2016). Any ion in a trap can be entangled with any other through the shared motion, giving full connectivity without the routing overhead that sparse solid-state chips face. Coherence times are long, often seconds, far exceeding superconducting circuits (Bruzewicz et al. 2019).

#Limitations

The main tradeoff is speed. Gates rely on driving motional modes and typically run in microseconds to tens of microseconds, slower than superconducting gates by one to three orders of magnitude. Scaling a single trap is hard: as more ions share one crystal, the motional spectrum becomes crowded and gates degrade. The leading answer is the quantum charge-coupled device architecture, in which ions are physically shuttled between separate trap zones for storage and gate operations, which adds control complexity. Laser systems and vacuum requirements also make the apparatus intricate.

#Who uses them

• Quantinuum builds the H-series trapped-ion systems using a QCCD architecture and reports high-fidelity gates and all-to-all connectivity (Quantinuum).
• IonQ develops trapped-ion processors and cloud access, using ytterbium ions and characterizing performance with algorithmic benchmarks rather than raw qubit count alone.

Both distinguish between what has been demonstrated in published experiments and what appears on forward-looking roadmaps.

#Readout and coherence

State readout is done by state-dependent fluorescence: a laser illuminates the ions, and a qubit in one state scatters many photons and appears bright, while the other state stays dark. A camera or photomultiplier then reads each ion with very high accuracy. Combined with hyperfine encodings that are insensitive to stray magnetic fields, this gives trapped ions their long coherence and low measurement error. The same properties that make ions excellent qubits, their isolation and identical structure, are also what make them slow: the weak coupling that protects them from noise also means gates must work through the shared motion rather than a strong direct interaction.

#Relevance to error correction and cryptography

High native fidelity means fewer physical qubits are needed per Logical qubit once Quantum error correction is applied, and trapped-ion groups have demonstrated small logical qubits and fault-tolerant primitives. Still, the systems remain firmly in the NISQ era, and running Shor's algorithm at cryptographic scale would require far larger machines than exist as of early 2026. Trapped ions therefore inform, but do not change, the uncertain timeline to Q-Day.