Trapped Rydberg Ions: Control and Interactions

• Trapped Rydberg ions are atomic ions with a coherently excited Rydberg electron, merging precise trapping with strong, tunable dipolar interactions for advanced quantum operations.
• Microwave dressing and state-dependent control mitigate challenges from trap-induced Stark shifts, micromotion, and phonon excitations, ensuring robust system performance.
• Experimental studies demonstrate fast, high-fidelity entangling gates and scalable multiqubit architectures, underpinning applications in quantum simulation, metrology, and thermodynamic studies.

Searching arXiv for papers on trapped Rydberg ions to ground the article in the current literature. Trapped Rydberg ions are trapped atomic ions whose valence electron is coherently excited to a high-lying Rydberg level while the ion remains confined in an electromagnetic trap. The subject combines the precise control of internal and motional states familiar from trapped-ion platforms with the large electric polarizability and strong, long-range electric dipolar interactions characteristic of Rydberg physics. In the literature, this combination has been developed as a route to fast entangling gates, tunable spin models, vibronic quantum simulation, microscopic thermal machines, and hybrid cavity-based metrology, with central technical themes including state-dependent trapping, microwave dressing, micromotion control, and dipole–dipole engineering (Mokhberi et al., 2020).

1. Platform, level structure, and defining interactions

In a linear Paul trap, a singly charged ion experiences a time-dependent quadrupole potential and, in the secular approximation, an effective harmonic confinement. A representative form used in the literature is

$$\Phi(R,t)=\gamma'(X^2-Y^2)\cos(\Omega_{\rm RF} t)-\gamma[(1+\epsilon)X^2+(1-\epsilon)Y^2-2Z^2],$$

with an external Hamiltonian

$$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$

When the ion is excited to a Rydberg state $|r\rangle$ with electric polarizability $\alpha_r$, the trapping frequencies and Stark shifts become state dependent. The polarizability scales as $\alpha_r\propto n^7$, and for ${\rm Sr}^+$ the value $$\alpha_{50S}\approx1.0\times10^{-30}\,{\rm C}^2{\rm m}^2/{\rm J}\approx97\,{\rm MHz}/({\rm V}/{\rm cm})^2$$ was quoted explicitly. This giant polarizability underlies both the opportunities and the complications of the platform (Mokhberi et al., 2020).

Rydberg excitation has been implemented in both single-step and two-step schemes. For ${\rm Ca}^+$, a single-step $3D_{3/2}\to nF$ excitation at $\approx122\,{\rm nm}$ was discussed, whereas two-step excitation in $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$0 or $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$1 uses an intermediate $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$2 state and, in the large-detuning limit, yields an effective two-photon Rabi frequency

$$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$3

Typical ultraviolet laser powers of tens of mW give $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$4 up to several MHz; in an $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$5 experiment at $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$6, the effective two-photon Rabi frequency was reported as $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$7 (Mokhberi et al., 2020, Thomm et al., 28 May 2026).

The inter-ion interaction central to most proposals is the resonant dipole–dipole coupling

$$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$8

with $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$9 and therefore $|r\rangle$0. For two dressed $|r\rangle$1 ions at $|r\rangle$2 and $|r\rangle$3, $|r\rangle$4 was reported in the sub-microsecond gate experiment (Zhang et al., 2019). A recurring theme across the field is that these interactions can be made strong while the ions remain trapped, rather than by releasing the crystal from confinement (Zhang et al., 2019).

2. Trap-induced physics: quadrupole coupling, polarizability, and micromotion

The first direct observation of a trapped Rydberg ion in a Paul trap used a single $|r\rangle$5 ion and established two fundamental trap effects. First, the quadrupolar trap field couples to Rydberg $|r\rangle$6-states but not to Rydberg $|r\rangle$7-states at first order. In the notation of that work,

$|r\rangle$8

and the electron–trap coupling generates Floquet sidebands for $|r\rangle$9-states because the radio-frequency term cannot be handled by a simple rotating-wave approximation when its coupling is comparable to the trap drive. Experimentally, $\alpha_r$0 spectra exhibited prominent rf sidebands spaced by $\alpha_r$1, while $\alpha_r$2 spectra showed only Zeeman splitting (Higgins et al., 2016).

Second, the large Rydberg polarizability modifies the secular confinement experienced by the ion’s center-of-mass motion. In the same experiment, excitation to $\alpha_r$3 produced asymmetric broadening from the motional-state-dependent shift of the resonance. After Doppler cooling, the measured linewidth was $\alpha_r$4, whereas radial sideband cooling to $\alpha_r$5 and $\alpha_r$6 reduced it to $\alpha_r$7, limited by the lasers. The extracted radial shift per phonon was $\alpha_r$8 (Higgins et al., 2016).

A complementary theoretical treatment analyzed the trap-induced quadrupole interaction in an effective representation. For a four-level model with laser-driven couplings and an rf-induced $\alpha_r$9 quadrupole term, the transformed Hamiltonian replaces the explicit time dependence by rescaled laser intensities and additional Stark shifts: $\alpha_r\propto n^7$0 This led to the conclusion that detrimental quadrupole coupling can be compensated by increasing the physical Rabi frequencies and pre-shifting the detunings (Simeonov et al., 2019).

Micromotion and field-axis misalignment remain a separate constraint. A 2024 Floquet analysis of a single trapped Rydberg ion showed that if the dc and rf quadrupole axes are misaligned by an angle $\alpha_r\propto n^7$1, a term $\alpha_r\propto n^7$2 produces extrinsic micromotion and can broaden the excitation spectrum into a forest of lines. The paper identified a regime in which isolated Rydberg lines persist: $\alpha_r\propto n^7$3, $\alpha_r\propto n^7$4, $\alpha_r\propto n^7$5, and $\alpha_r\propto n^7$6, giving a single well-separated peak with full-width at half-maximum below $\alpha_r\propto n^7$7 (Martins et al., 2024).

These results establish that the platform is not intrinsically “motion free.” Rather, motion can be rendered perturbative or spectroscopically benign only after careful control of polarizability, trap quadrupole effects, and micromotion.

3. Coherent control and microwave-dressed Rydberg-state engineering

Coherent manipulation of trapped Rydberg ions began with a single-$\alpha_r\propto n^7$8 experiment that realized coherent Rydberg excitation and a single-qubit Rydberg gate. The three-level ladder $\alpha_r\propto n^7$9, ${\rm Sr}^+$0, ${\rm Sr}^+$1 was driven by ${\rm Sr}^+$2 and ${\rm Sr}^+$3 fields with ${\rm Sr}^+$4 up to ${\rm Sr}^+$5 and ${\rm Sr}^+$6. On resonance, the dark state

${\rm Sr}^+$7

was used in STIRAP. With ${\rm Sr}^+$8, ${\rm Sr}^+$9 was quoted. The measured $$\alpha_{50S}\approx1.0\times10^{-30}\,{\rm C}^2{\rm m}^2/{\rm J}\approx97\,{\rm MHz}/({\rm V}/{\rm cm})^2$$0 lifetime was $$\alpha_{50S}\approx1.0\times10^{-30}\,{\rm C}^2{\rm m}^2/{\rm J}\approx97\,{\rm MHz}/({\rm V}/{\rm cm})^2$$1, and the geometric-phase single-qubit gate reached $$\alpha_{50S}\approx1.0\times10^{-30}\,{\rm C}^2{\rm m}^2/{\rm J}\approx97\,{\rm MHz}/({\rm V}/{\rm cm})^2$$2 (Higgins et al., 2017).

Microwave dressing then became the standard method to reconcile two otherwise competing goals: near-zero polarizability and strong dipolar interaction. In the theoretical gate proposal based on trapped $$\alpha_{50S}\approx1.0\times10^{-30}\,{\rm C}^2{\rm m}^2/{\rm J}\approx97\,{\rm MHz}/({\rm V}/{\rm cm})^2$$3, a strong microwave field couples $$\alpha_{50S}\approx1.0\times10^{-30}\,{\rm C}^2{\rm m}^2/{\rm J}\approx97\,{\rm MHz}/({\rm V}/{\rm cm})^2$$4 and $$\alpha_{50S}\approx1.0\times10^{-30}\,{\rm C}^2{\rm m}^2/{\rm J}\approx97\,{\rm MHz}/({\rm V}/{\rm cm})^2$$5 via

$$\alpha_{50S}\approx1.0\times10^{-30}\,{\rm C}^2{\rm m}^2/{\rm J}\approx97\,{\rm MHz}/({\rm V}/{\rm cm})^2$$6

producing dressed states

$$\alpha_{50S}\approx1.0\times10^{-30}\,{\rm C}^2{\rm m}^2/{\rm J}\approx97\,{\rm MHz}/({\rm V}/{\rm cm})^2$$7

By choosing $$\alpha_{50S}\approx1.0\times10^{-30}\,{\rm C}^2{\rm m}^2/{\rm J}\approx97\,{\rm MHz}/({\rm V}/{\rm cm})^2$$8 so that $$\alpha_{50S}\approx1.0\times10^{-30}\,{\rm C}^2{\rm m}^2/{\rm J}\approx97\,{\rm MHz}/({\rm V}/{\rm cm})^2$$9, one can tune the dressed-state polarizability to zero, so that the Rydberg trapping potential matches that of ${\rm Ca}^+$0 and the Franck–Condon factors become trivial, ${\rm Ca}^+$1. At the same time the dressed state acquires a permanent dipole moment and an interaction ${\rm Ca}^+$2 (Li et al., 2013).

This principle has now been demonstrated experimentally. In 2026, coherent transfer between different trapped-ion Rydberg states was reported in a single ${\rm Ca}^+$3 ion. Between Rydberg ${\rm Ca}^+$4 and ${\rm Ca}^+$5 states, the experiment achieved a population transfer efficiency of ${\rm Ca}^+$6 in a single microwave ${\rm Ca}^+$7-pulse. For ${\rm Ca}^+$8, the fitted microwave Rabi frequency was ${\rm Ca}^+$9, giving $3D_{3/2}\to nF$0. The same work also demonstrated adiabatic transfer between a zero-polarizability dressed state and a maximally interacting dressed state. For $3D_{3/2}\to nF$1, the zero-polarizability point occurred at $3D_{3/2}\to nF$2 and $3D_{3/2}\to nF$3; a $3D_{3/2}\to nF$4 linear sweep $3D_{3/2}\to nF$5 achieved adiabaticity $3D_{3/2}\to nF$6 in simulation (Thomm et al., 28 May 2026).

The broader significance is operational rather than merely spectroscopic. Zero-polarizability states are robust to trapping-field-induced dephasing but have $3D_{3/2}\to nF$7 and therefore $3D_{3/2}\to nF$8, whereas maximally interacting states at $3D_{3/2}\to nF$9 maximize $\approx122\,{\rm nm}$0 and $\approx122\,{\rm nm}$1 but are more sensitive to stray fields and trap-induced level shifts (Thomm et al., 28 May 2026). Microwave dressing provides a way to move between these two regimes within a single sequence.

4. Entangling gates and quantum-computing architectures

The first explicit Rydberg-ion two-qubit gate theory proposed a controlled adiabatic phase gate between two trapped $\approx122\,{\rm nm}$2 ions, with qubit states $\approx122\,{\rm nm}$3 and $\approx122\,{\rm nm}$4, and a laser driving $\approx122\,{\rm nm}$5. For $\approx122\,{\rm nm}$6, $\approx122\,{\rm nm}$7, $\approx122\,{\rm nm}$8, $\approx122\,{\rm nm}$9, one finds $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$00, so at $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$01, $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$02. Using pulse shapes $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$03 and $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$04, an example choice $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$05, $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$06, $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$07 yields $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$08. The same analysis emphasized that phonon excitations enter only at $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$09 once $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$10, although the $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$11 lifetime at $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$12, $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$13, still gives a loss probability of order $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$14 for the example pulse (Li et al., 2013).

The first experimental entangling gate then demonstrated a $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$15 two-ion gate in trapped $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$16, using two-photon excitation to approximately $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$17 and a $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$18 microwave coupling to $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$19. The dressed interaction reached $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$20 at $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$21, and the Bell-state fidelity was $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$22. The paper identified error sources and projected a total error below $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$23 for experimentally achievable parameters, while estimating a residual motional contribution of $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$24 even in a crystal of $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$25 ions (Zhang et al., 2019).

More recent gate studies focus on pulse optimization under realistic lifetimes. A 2024 analysis derived an effective four-level Hamiltonian for microwave-dressed $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$26 ions and compared three controlled-phase protocols. In a conservative regime $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$27, Protocol A gave $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$28 and Protocol B gave $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$29. In an optimistic regime $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$30, Protocol B reached $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$31 without decay, and with a finite Rydberg lifetime $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$32 a $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$33 gate still achieved $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$34 (Wilkinson et al., 2024).

A different architecture uses the large polarizability of a Rydberg state to shift collective vibrational mode frequencies and then applies a shaped electric waveform to the trap electrodes. In a $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$35 proposal, excitation to $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$36 gives a transverse shift $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$37 for $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$38, and optimized continuous waveforms yield $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$39 for two ions and $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$40 in a six-ion crystal, with all-to-all pair selectivity in a linear chain (Bao et al., 2024).

The logic of extending two-qubit control to native multiqubit gates has also been made explicit. A 2025 proposal introduced a native microwave-dressed CCZ gate for three $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$41 ions. For cryogenic parameters $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$42 and $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$43, a global single-pulse protocol with $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$44, $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$45, $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$46, and $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$47 achieved $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$48. The same work embedded native CCZ and CZ operations into a measurement-free Bacon–Shor error-correction cycle and obtained logical-error scaling $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$49, consistent with distance-3 fault tolerance (Bolsmann et al., 18 Dec 2025).

5. Many-body, vibronic, and thermodynamic regimes

Beyond gates, trapped Rydberg ions support many-body interactions that differ qualitatively from the pairwise blockade picture. In a quasi one-dimensional $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$50 chain, coupling between Rydberg pair interactions and collective phonons generates effective two-, three-, and four-body terms after a Lang–Firsov transformation: $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$51 For $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$52, the linear–zigzag transition occurs at $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$53. Near this soft mode, the effective couplings are enhanced; the paper quoted $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$54, $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$55, and $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$56 as $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$57 is tuned from $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$58 to $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$59. The resulting three-body anti-blockade provides a spectroscopic signature of the structural transition (Gambetta et al., 2020).

A different branch of the literature uses trapped Rydberg ions to emulate non-adiabatic molecular physics. In the single-$$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$60-excitation manifold $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$61, the reduced Hamiltonian

$$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$62

produces adiabatic Born–Oppenheimer surfaces

$$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$63

These surfaces cross when $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$64 and $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$65, giving a conical intersection whose position can be tuned by microwave dressing and a static bias field. In the symmetric case $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$66, the geometric phase produces destructive interference and inhibits nuclear motion, whereas for $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$67 the wavepacket transfers slowly between wells on a timescale $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$68 (Gambetta et al., 2020).

The effect of finite Rydberg lifetime on this vibronic dynamics has also been analyzed. For two $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$69 ions with $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$70, $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$71, $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$72, $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$73, and $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$74, the master equation predicts several underdamped oscillations in $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$75, $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$76, and the phonon populations before relaxation to the steady state. With $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$77, the paper estimated $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$78–$$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$79 full oscillation periods before decay dominates (Chaudhary et al., 2024).

Trapped Rydberg ions have also been used as a model thermal device. In a one-dimensional harmonic trap, two laser-driven Rydberg ions with a state-dependent interaction

$$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$80

can drive their relative vibrational mode as a flywheel. Under a periodically modulated two-stroke protocol with period $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$81, the stored work is identified as

$$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$82

For $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$83, $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$84, $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$85, and $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$86, simulations reached $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$87–20 phonons in $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$88 (Martins et al., 2023).

6. Geometries, scaling, and open problems

The original motivation for trapped Rydberg ions was to realize quantum operations independently of the crowded normal-mode spectrum of large crystals (Li et al., 2013). That objective remains visible in more recent architectures, but the literature equally shows that geometry matters. In linear Paul traps, the principal open challenges include precise control of stray electric and magnetic fields, suppression of micromotion sidebands, coherent global microwave dressing across extended chains, black-body and photo-ionization losses of high-$$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$89 states, and engineered geometries with inter-ion spacing $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$90 (Li et al., 2013).

One route around rf-specific issues is the Penning trap. A 2026 proposal for planar Rydberg-ion crystals in a Penning trap emphasized that static confinement avoids rf micromotion and supports two-dimensional arrays with direct dipolar couplings. In the co-rotating frame the center-of-mass Hamiltonian is a three-dimensional harmonic oscillator, while the internal Rydberg structure contains linear Zeeman, diamagnetic, and quadrupole Stark terms. For $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$91, $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$92, and $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$93, the dipolar coupling was estimated as $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$94, with spin-phonon corrections of order $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$95 (Martins et al., 4 Jan 2026).

Hybridization with cavity QED provides another direction. A proposal based on a single trapped ion in circular Rydberg states inside a high-$$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$96 microwave cavity used a beam-splitter interaction and a cross-Kerr term to create hybrid high-N00N states between the ion’s motion and a cavity mode. Representative near-term parameters were $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$97, $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$98, $$H_{\rm trap}= \frac{P^2}{2M}+\frac12 M(\omega_X^2 X^2+\omega_Y^2 Y^2+\omega_Z^2 Z^2).$$99, $|r\rangle$00, and $|r\rangle$01, with an estimated fidelity $|r\rangle$02 for $|r\rangle$03 (Mohseni et al., 2018).

A common misconception is that “phonon-independent” Rydberg-ion operation removes all trap engineering requirements. The published record indicates the opposite. What has been demonstrated is that microwave dressing can make Franck–Condon factors trivial, that residual gate errors from phonons can be suppressed to $|r\rangle$04 or to $|r\rangle$05 in large crystals, and that isolated Rydberg lines can survive in realistic Paul traps. None of these results eliminates the need for sideband cooling, micromotion compensation, field alignment, or control of finite Rydberg lifetimes [(Li et al., 2013); (Zhang et al., 2019); (Martins et al., 2024)].

Taken together, the literature defines trapped Rydberg ions as a platform in which the usual trapped-ion advantages—state preparation, readout, long-lived low-lying qubits, and controlled motional degrees of freedom—are combined with tunable polarizability and strong, switchable dipolar interactions. This suggests a unifying perspective: the field is no longer centered only on “fast gates,” but on a broader capability to engineer electronic, motional, and vibronic Hamiltonians within a single ion-trap architecture (Mokhberi et al., 2020, Thomm et al., 28 May 2026).