Increased cyclicity of atomic transitions via coherent interference of decay paths

AI summary

Overview

Research area: Quantum optics and quantum-defect physics — specifically optical cycling transitions, coherent interference of spontaneous decay, and single-shot spin-state readout for atomic and solid-state qubits.

Technical level: Advanced. The paper works with master equations, cross-damping (coherent) terms between decay paths, transition dipole moments (TDMs), dressed states, and Rabi frequencies; it assumes familiarity with open quantum systems.

Scope (one sentence): The paper proposes and numerically models two ways to create artificial cycling transitions — one using a single strong drive and one using an auxiliary dressing field — so that spontaneous emission from a chosen superposition state into an unwanted ground state is cancelled by destructive interference, and it quantifies the resulting photon-count increase for generic two-excited-state models and for the NiV⁻ center in diamond.

What This Paper Is About

Optical readout of a quantum state is only as good as the number of photons you can collect, and that number collapses if the excited state has a decay path into the "wrong" ground state — the system goes dark after a handful of photons. A truly cycling transition, where the electron repeatedly returns to the state it started in, avoids this, but many promising emitters lack one. The paper shows that when two decay paths have suitably related dipole moments, a carefully applied drive can make them interfere destructively on the unwanted channel, converting an ordinary transition into a cycling one and boosting the emitted photon count by orders of magnitude.

Key Contributions

1. A general criterion for drive-induced cyclicity. The authors identify the specific relationship between transition dipole moments (Eq. 2 in the paper) under which a single strong, suitably polarized drive converts a non-cycling two-excited-state system into one that cycles between one ground state and an excited superposition, with decay into the other ground state forbidden.

2. A second, more flexible scheme using an auxiliary field. An auxiliary laser coupling the excited manifold to an auxiliary state |a⟩ creates dressed excited states whose superposition coefficients are tunable by the auxiliary field's amplitude and detuning, so cyclicity no longer depends on the TDMs having one particular form — only on the two ground-state TDMs being parallel or antiparallel.

3. Quantified photon-count gains, including tolerance analysis. Simulations show increases of multiple orders of magnitude in average emitted photons over traditional weak-drive readout, plus explicit studies of how the gain degrades when TDM amplitudes (Fig. 2b) or polarizations (Fig. 2c) deviate from the ideal form.

4. A concrete material proposal: NiV⁻ in diamond. The negatively charged nickel vacancy center is analyzed as a prime candidate, including a scheme that removes the usual tradeoff between efficient optical spin initialization (misaligned magnetic field) and high-fidelity readout (aligned field), along with an assessment of sensitivity to drive polarization error.

Main Findings

• Traditional readout saturates quickly. For a system with branching ratio b = 0.9 driven weakly on a single transition (Ω = 0.1 ω_ess), the population is optically pumped into the dark ground state on a timescale of about 10/γ₀, and only about 10 photons are emitted on average before the system goes dark.
• A strong drive suppresses the dark channel. With a drive coupling the starting ground state to both excited states at Ω_S = 20 ω_ess, population cycles continuously between the ground state and the superposition |e_S⟩ with negligible transfer to the unwanted ground state.
• Orders-of-magnitude gain for ideal dipole moments. For TDMs of the required form, the strong-drive technique increases the average emitted photon number by multiple orders of magnitude over traditional readout, and systems with higher original branching ratio b reach higher photon numbers still.
• Deviations from the ideal TDM form cap the gain. When TDM amplitudes or polarizations depart from Eq. 2, decay from |e_S⟩ into the unwanted ground state no longer cancels, and the photon number saturates as drive power increases rather than continuing to grow. The tolerance study uses |μ_e1g1| = 3|μ_e2g1| = 3|μ_e1g2| (b = 0.9) and γ₀ = 0.1 ω_ess/2π.
• NiV⁻ gains more than two orders of magnitude. Simulations using the optical and spin Hamiltonian of prior NiV⁻ work, an optical decay time of 10 ns, an orbital decay time of 30 ps, and B = 500 G (excited-state splitting about 1.5 GHz) show more than two orders of magnitude more emitted photons than traditional readout, across various magnetic field angles. Larger misalignment degrades the bare-state cyclicity but can be compensated with a stronger drive.
• Cavity-based alternatives are constrained for this defect. The narrowest diamond optical cavities are still on the order of a few GHz in linewidth, too broad to selectively enhance spin-conserving transitions that are separated by at least a few GHz.
• Strong-drive readout is polarization-sensitive. With ideal polarization, a system starting in the unwanted ground state stays dark. With imperfect polarization it is driven and contributes spurious photons. The paper simulates the ratio n_ph,g1/n_ph,g2 for t = 1e3/γ₀ at Ω_S = 100 GHz (assuming a uniform Purcell enhancement of 100, raising γ₀ to 1/100 ps) and finds the ratio insensitive to magnetic field orientation (5°, 15°, 45° give nearly equivalent results); the resulting measurement fidelity also depends on collection efficiency and the experimental noise floor, which the paper does not assign a value.
• Power requirement for the strong-drive route. Based on prior NiV⁻ experimental work, the authors estimate roughly 10 mW is needed to reach Ω_S = 50 GHz.
• The auxiliary-field route works at much lower power. For the TDM set equivalent to the b = 0.5 case of Fig. 2(a), with auxiliary Rabi frequencies |Ω_e1,a| = |Ω_e2,a| = ω_ess and excitation Rabi frequencies 0.02 ω_ess, the simulation reaches n_ph,g1 = 3040 photons over t = 1e4/γ₀ at zero detunings — a photon count that would otherwise require Ω_S > 50 ω_ess under a single strong drive.
• The auxiliary scheme also rescues "bad" dipole moments. For a TDM set (Eq. 9) that saturates at a few tens of photons under a single strong drive, an auxiliary detuning of Δ_aux = −0.31 ω_ess plus an excitation detuning of Δ_exc = −0.10 ω_ess produces 2810 photons over t = 1e4/γ₀, with |Ω_e1,g1| = 3|Ω_e2,g1| = 0.02 ω_ess.
• A rate-versus-total tradeoff exists in the auxiliary scheme. Raising excitation power increases the emission rate but also off-resonantly excites other dressed states that decay to the unwanted ground state, lowering the total photon count; increasing the auxiliary field amplitude improves both metrics.
• Interference requires simultaneous coupling. The appendix shows that decay paths can only interfere if the states are coupled by the same fields to the same states; a field that directly couples the two excited states (Fig. 5b) introduces time dependence into the collapse operators that prevents observable interference.

Methodology in Plain English

The authors treat the emitter as a small quantum system with two ground states and two excited states and write down a master equation that includes not just the ordinary decay rates but also cross terms between different decay paths, whose size depends on the angle between the two dipole moments (Γ_jk = √(Γ_jΓ_k) cos θ_jk). When those cross terms are right, the decay amplitudes from a superposition of the two excited states cancel on one ground state — a "dark" superposition that does not spontaneously decay, distinct from the usual coherent-population-trapping dark state, which cannot be excited.

They then arrange for the laser to drive the system into exactly that superposition. In the first scheme, a single strong field polarized along one dipole axis drives both excited states at once; when the Rabi rate greatly exceeds the excited-state splitting ω_ess, the drive effectively defines a new eigenstate and the unwanted decay channel closes. In the second scheme, a separate auxiliary laser mixes a third state into the excited manifold, so the superposition coefficients become knobs set by laser power and detuning rather than fixed by the material's dipole moments.

To quantify performance, they numerically integrate the master equation and count emitted photons by summing the decay contributions to the ground-state populations (Appendix B), rather than simply looking at the excited-state population. They sweep drive strength, dipole amplitude ratios, dipole polarization angles, magnetic field angle, and both laser detunings, and they assume a uniform Purcell enhancement of 100 in some NiV⁻ simulations purely to shorten computation time, noting this does not change the relevant steady-state photon numbers in the strong-drive regime.

Why This Matters

Cycling transitions underpin high-fidelity state measurement in trapped ions and atoms and in solid-state quantum defects, and they are also what makes laser cooling and trapping efficient. The standard fix for a non-cycling transition — putting the emitter in a narrow cavity — demands a cavity linewidth below the spectral separation of the wanted and unwanted decay paths, which the authors argue may be technologically infeasible for some atomic systems (and which they show is currently out of reach for NiV⁻ in diamond). Interference-based cyclicity needs no such cavity and is controlled entirely by laser fields.

The NiV⁻ result carries a specific practical payoff: it breaks the usual tradeoff in which a well-aligned magnetic field gives good readout but poor optical initialization, while a misaligned field gives good initialization but poor readout. With the strong-drive scheme, one can initialize efficiently at a misaligned field and then switch on the drive to read out.

Potential applications (as motivated by the paper's cited context):

• Single-shot spin readout for solid-state quantum-defect qubits in quantum computing and quantum networking.
• High-fidelity state detection for trapped-ion and neutral-atom platforms.
• More efficient laser cooling and trapping of atoms and ions.
• Improved readout and initialization protocols for optically addressable defects in diamond photonics.

Industry relevance: the work targets quantum-hardware builders working with diamond photonics, integrated photonic cavities, and high-power on-chip optical drive delivery; it also bears on materials engineering efforts to improve crystal quality and reduce spurious optical absorption, and on molecular-design programs aimed at better cycling transitions.

Future Directions

• Extending interference to three or more decay paths. Because three non-parallel vectors can sum to zero, multi-path destructive interference could engineer cyclicity in systems whose dipole moments are not parallel, widening the range of atomic and solid-state emitters that qualify.
• Generalizing beyond two excited states. All analysis here is restricted to systems with only two decaying excited states; the authors state the principles should generalize to more complicated excited-state manifolds.
• Experimental demonstration and practical readout engineering. The paper is a proposal with simulations only; it suggests separating the emitted photons from the strong drive by detecting on the phonon sideband or on a Mollow sideband displaced by ±Ω_S from the drive frequency, and it notes the eventual fidelity will depend on collection efficiency and the noise floor of the specific setup.
• Identifying good auxiliary states. Appendix D sets conditions on |a⟩: decay and decoherence among |e₁⟩, |e₂⟩ and |a⟩ must be negligible (plausible when |a⟩ lies close to the excited states relative to the ground-state separation), and a decay channel from |a⟩ to the unwanted ground state would suppress cyclicity. Finding real systems that meet these conditions remains open.
• Exploiting favorable level structures. In systems where the ground-state splitting greatly exceeds the excited-state splitting (ω_gss ≫ Ω_S ≫ ω_ess), the residual polarization-error problem could be strongly suppressed because spurious driving of the second ground state would be far off-resonant.

Target Audience

Quantum-optics and quantum-photonics researchers, particularly those working on spin-photon interfaces and optically addressed solid-state defects; AMO physicists interested in coherent population trapping, interference of decay, and lasing without inversion; diamond-defect specialists evaluating NiV⁻ and related centers; and quantum-hardware engineers designing readout, initialization, and high-power optical delivery for defect-based qubits. Readers without graduate-level quantum optics will find the master-equation treatment and dressed-state arguments demanding, though the underlying idea — cancel an unwanted decay path by interference — is conceptually accessible.

Authors’ abstract

Optical readout is a fundamental tool in atomic state measurement, yet the fidelity of optical readout is frequently limited by imperfect photon collection. This can be mitigated when readout occurs on a cycling transition which continuously fluoresces under resonant excitation, thus increasing signal and enabling single-shot readout. We present a method to extend the cyclicity of atomic transitions via coherent destructive interference between spurious decay paths. We describe the characteristics of atomic systems in which this method can be implemented and model several examples in which the number of emitted photons is increased by multiple orders of magnitude.