---
title: Qubit–Magnon Dipole Coupling
url: https://www.emergentmind.com/topics/qubit-magnon-dipole-coupling
type: topic
---

# Qubit–Magnon Dipole Coupling

Searching arXiv for recent and foundational papers on qubit–magnon dipole coupling to ground the article in the current literature.
arxiv_search.query({"search_query":"all:\"qubit magnon\" OR all:\"quantum magnonics\" OR all:\"magnon qubit coupling\"", "start": 0, "max_results": 10, "sort_by": "submittedDate", "sort_order": "descending"})
Qubit–magnon dipole coupling is the interaction between a two-level qubit and a magnonic collective spin excitation, most commonly the uniform Kittel mode of a yttrium-iron-garnet (YIG) sphere or the uniform mode of a magnetic particle. In superconducting quantum magnonics, this coupling appears in several distinct but related forms: a cavity-assisted interaction in which a transmon couples electrically to a microwave cavity while the magnon couples magnetically to the same cavity mode; a direct near-field magnetic-dipole or inductive interaction between a qubit loop and a ferromagnet; and a direct dispersive $S_z\sigma_z$-type interaction used to probe noneigenmode magnons. The effective low-energy descriptions are typically a Jaynes–Cummings exchange term, $H_{\rm eff}\propto \sigma^+ m+\sigma^- m^\dagger$, or a dispersive term, $H_{\rm disp}\propto \sigma_z m^\dagger m$, and these have been used to demonstrate strong coupling, magnon-vacuum-induced Rabi splitting, magnon-number resolution, dissipation-based sensing, squeezing, blockade, and mediated quantum gates [1508.05290], [1410.3781], [1610.00839], [2005.09250], [2306.05065].

## 1. Physical basis and device geometries

The canonical cavity–magnon–qubit architecture consists of a single-mode microwave cavity, a transmon-type superconducting qubit, and a magnon mode in a YIG sphere. In this arrangement the qubit couples to the cavity through an electric-dipole interaction, while the magnon couples to the same cavity through a magnetic-dipole interaction. For the Kittel mode, the magnon corresponds to a uniform precession of all spins in the ferromagnet, so its collective dipole moment is macroscopically enhanced; in the 2014 experiment the YIG sphere had net spin $N\approx1.4\times10^{18}$, yielding a collective magnetic dipole moment $N\mu_B$ [1410.3781].

In the rotating-wave approximation, this architecture is described by a three-mode Hamiltonian with cavity operator $a$, magnon operator $m$ or $c$, and qubit Pauli operators or transmon ladder operators. The electric-dipole term has the form $g_{qc}(a^\dagger b+a b^\dagger)$ or $g_q(a\sigma^+ + a^\dagger\sigma^-)$, whereas the magnetic-dipole term has the form $g_{mc}(a^\dagger c+a c^\dagger)$ or $g_m(a m^\dagger+a^\dagger m)$. The physical distinction is central: the superconducting qubit is tied to the cavity electric field, while the magnon is tied to the cavity magnetic field, so the effective qubit–magnon interaction is a mixed electric–magnetic process mediated by the cavity vacuum field [1508.05290], [2304.10760].

A different geometry places a superconducting transmon qubit in the near field of a millimetre-size YIG sphere so that the qubit loop and the collective Kittel-mode magnon share a direct magnetic-dipole interaction. In that case the device Hamiltonian contains a transverse exchange-like term $J(m\sigma_+ + m^\dagger\sigma_-)$ and a longitudinal term $g_{\rm lp}\sigma_z(m+m^\dagger)$, both derived from the flux dependence of the SQUID Josephson energy and from the flux pull generated by the YIG macrospin fluctuation [2303.13823].

A third formulation arises in quantum magnonics models that begin from direct magnetic Zeeman coupling. In the resonance-fluorescence treatment of a qubit and a magnon mode, the coupling strength is extracted from $H_Z=-\hat\mu_q\cdot\hat B$, giving $g=(\gamma_e\mu_B B_{\rm zpf})/\hbar$ with $B_{\rm zpf}=\sqrt{\mu_0\hbar\omega_m/(2V_{\rm eff})}$. In typical circuit-magnon experiments, this yields $g/2\pi\sim10$–$50\,{\rm MHz}$ [2501.01254].

## 2. Effective Hamiltonians and the emergence of qubit–magnon exchange

For cavity-assisted systems, the central theoretical step is adiabatic elimination of the far-detuned cavity mode. In the regime $\Delta_m=\omega_0-\omega_m\simeq \Delta_q=\omega_0-\omega_q\equiv\Delta\gg g_q,g_m$, a second-order Schrieffer–Wolff transformation produces an effective Hamiltonian
$$
H_{\rm eff}=\omega_m m^\dagger m+\frac{\omega_q}{2}\sigma_z+\frac{g_q^2}{\Delta}\sigma^+\sigma^-+\frac{g_m^2}{\Delta}m^\dagger m+g_{\rm eff}(\sigma^+m+\sigma^-m^\dagger),
$$
with
$$
g_{\rm eff}= \frac{g_q g_m}{\Delta}.
$$
The terms proportional to $g_q^2/\Delta$ and $g_m^2/\Delta$ are dispersive frequency shifts, while the exchange term is the operative qubit–magnon coupling [2304.10760].

In the more general asymmetric-detuning form, the same elimination yields
$$
H_{\rm eff}\simeq g_{qm}\bigl(\sigma^+ c+\sigma^- c^\dagger\bigr), \qquad
g_{qm}=g_{qc}g_{mc}\Bigl(\frac{1}{\Delta_q}+\frac{1}{\Delta_m}\Bigr)\Big/2.
$$
When $\Delta_q\simeq\Delta_m\equiv\Delta$, this reduces to $g_{qm}=g_{qc}g_{mc}/\Delta$ [1410.3781], [1903.12498], [2605.06297].

The physical mechanism is explicitly second order. In the cavity-mediated scheme of Guo et al., the qubit and the magnon do not interact directly; rather, the qubit flips and creates or annihilates a virtual cavity photon, and that virtual photon is converted into or from a magnon. This produces the Jaynes–Cummings–type interaction $H_{\rm eff}\supset g_{\rm eff}(\sigma^+m+\sigma^-m^\dagger)$ [2304.10760].

In the dispersive limit of the resulting two-mode theory, the exchange interaction is converted into a cross-Kerr-type coupling. For a two-level qubit one obtains
$$
H_{\rm disp}/\hbar=\chi_{q-m}\sigma_z c^\dagger c,
$$
or equivalently a dressed Hamiltonian in which the qubit frequency shifts by $\chi$ per magnon, with leading-order $\chi=g^2/\Delta$. For a weakly anharmonic transmon, the correction $\chi=-g^2\alpha/[\Delta(\Delta+\alpha)]$ is also standard in this context [1610.00839], [2005.09250].

Not all dispersive couplings are second order. In the anisotropic-ferromagnet treatment of Romling et al., the interaction Hamiltonian contains a direct first-order term $\chi a^\dagger a\,\sigma_z$, arising from the $S_z\sigma_z$ part of an interfacial spin–spin exchange. There the dispersive coupling is present already at first order rather than being generated perturbatively from a Jaynes–Cummings exchange [2306.05065].

## 3. Coupling regimes and experimental benchmarks

The development of the field can be traced through a sequence of increasingly controlled regimes, from cavity-assisted strong coupling to strong dispersive readout and direct near-field implementations.

| Work | Architecture | Reported coupling or signature |
|---|---|---|
| [1410.3781] | cavity-assisted qubit–magnon hybrid | $g_{qm,s}/2\pi=10.0\,{\rm MHz}$; tunable $g_{qm,p}/2\pi$ up to $3.4\,{\rm MHz}$ |
| [1508.05290] | cavity-mediated YIG–transmon system | $g_{qm}/2\pi\approx11.4\,{\rm MHz}$ measured; $2g_{qm}/2\pi\approx22.8\,{\rm MHz}$ |
| [1903.12498] | driven cavity-mediated hybrid | $g_{q-m}/2\pi\approx20.1\,{\rm MHz}$; vacuum Rabi splitting $\approx40\,{\rm MHz}$ |
| [1610.00839] | strong-dispersive quantum magnonics | $g_{q-m}/2\pi=7.79\,{\rm MHz}$; $\chi_{q-m}/2\pi=1.5\pm0.1\,{\rm MHz}$ |
| [2005.09250] | dispersive sensing regime | $g/2\pi=7.07\,{\rm MHz}$; sensitivity $1.42(3)\times10^{-3}\,{\rm magnons}/\sqrt{\rm Hz}$ |

The 2014 demonstration of coherent coupling between a ferromagnetic magnon and a superconducting qubit used a re-entrant rectangular copper microwave cavity with TE$_{102}$ mode frequency $\omega_c/2\pi=8.488\,{\rm GHz}$, a transmon qubit at $\omega_q/2\pi\approx8.136\,{\rm GHz}$ with $g_{qc}/2\pi=121\,{\rm MHz}$, and a YIG sphere realizing the Kittel mode with $g_{mc}/2\pi=21.0\,{\rm MHz}$. Using $\Delta/2\pi\approx215\,{\rm MHz}$, the effective-coupling estimate was $g_{qm}/2\pi\approx11.8\,{\rm MHz}$, while the observed static coupling was $g_{qm,s}/2\pi=10.0\,{\rm MHz}$. The measured linewidths were $\kappa_{102}/2\pi=2.5\,{\rm MHz}$, $\gamma_m/2\pi=1.4\,{\rm MHz}$, and $\gamma_q/2\pi=1.2\,{\rm MHz}$, so $g_{qm,s}\gg\kappa_{102},\gamma_m,\gamma_q$ and the system was in the strong-coupling regime [1410.3781].

In the 2015 “single-magnon limit” experiment, the TE$_{101}$ cavity–magnon coupling was $g_{mc}/2\pi=47\,{\rm MHz}$, the TE$_{102}$ qubit–cavity coupling was $g_{qc}/2\pi=117\,{\rm MHz}$, the effective detuning was $\Delta/2\pi\approx183\,{\rm MHz}$, and the resulting qubit–magnon coupling was measured as $g_{qm}/2\pi\approx11.4\,{\rm MHz}$, with a predicted value $13.4\,{\rm MHz}$. At the qubit–magnon degeneracy point the observed splitting was $2g_{qm}/2\pi\approx22.8\,{\rm MHz}$. The same work reported magnon–cavity cooperativity $C_{mc}\approx3\times10^3$ and qubit–magnon cooperativity $C_{qm}\approx1.4\times10^2$ [1508.05290].

The 2019 driven qubit–magnon experiment reported a larger cavity–qubit coupling $g_{q-c}/2\pi\approx239\,{\rm MHz}$ and cavity–magnon coupling $g_{m-c}/2\pi\approx43.0\,{\rm MHz}$, with detuning $\Delta\approx0.5\,{\rm GHz}$ and effective qubit–magnon coupling $g_{q-m}/2\pi\approx20.1\,{\rm MHz}$. In the absence of drive, the vacuum Rabi splitting was $2g_{q-m}\approx40\,{\rm MHz}$ [1903.12498].

Representative far-detuned parameters for the two-tone squeezing proposal are cavity frequency $\omega_0/2\pi\simeq7.5\,{\rm GHz}$, qubit and magnon frequencies $\omega_q/2\pi\approx\omega_m/2\pi\simeq7.2\,{\rm GHz}$, couplings $g_q/2\pi\simeq36\,{\rm MHz}$ and $g_m/2\pi\simeq36.6\,{\rm MHz}$, and detuning $\Delta/2\pi\simeq300\,{\rm MHz}$, which give $g_{\rm eff}/2\pi\simeq4.4\,{\rm MHz}$. Dissipation rates were quoted as magnon linewidth $\kappa/2\pi\simeq1\,{\rm MHz}$ and qubit decay $\gamma/2\pi\simeq20\,{\rm kHz}$ [2304.10760].

## 4. Spectroscopy, dispersive readout, and number resolution

The most direct signature of coherent qubit–magnon exchange is an avoided crossing or vacuum-Rabi splitting at resonance. In both the 2014 and 2015 cavity-mediated experiments, qubit spectroscopy revealed the hybridization of a single qubit excitation with a single magnon, thereby establishing strong coupling at the single-excitation level [1410.3781], [1508.05290].

Moving away from resonance into the large-detuning regime yields the strong-dispersive regime of quantum magnonics. In the 2016 number-state-resolution experiment, the lowest-order model was
$$
H=H_q+H_m+H_{\rm int}, \qquad H_{\rm int}=\hbar g_{q-m}(\sigma_+ c+\sigma_- c^\dagger),
$$
which in the limit $|\Delta|=|\omega_q-\omega_m|\gg g_{q-m}$ reduces to
$$
H_{\rm disp}=\hbar \chi_{q-m}\sigma_z c^\dagger c.
$$
The reported values were $g_{q-m}/2\pi=7.79\,{\rm MHz}$, $\chi_{q-m}/2\pi=1.5\pm0.1\,{\rm MHz}$, $| \Delta |/2\pi=89\,{\rm MHz}$, $\gamma_m/2\pi=1.3\,{\rm MHz}$, and narrowed qubit linewidth $\gamma_q/2\pi\approx0.78\,{\rm MHz}$. Because $2|\chi_{q-m}|> \max[\gamma_q,\gamma_m]$, the qubit spectral line split into well-resolved peaks spaced by $2\chi$ per magnon. The same experiment reported that resolving $0.026\pm0.012$ magnons corresponds to detecting the flip of one spin among $\sim5\times10^{19}$ [1610.00839].

The 2020 dissipation-based sensing experiment used the same dispersive principle but focused on qubit dephasing induced by magnon-number fluctuations. There the effective qubit–magnon coupling was $g/2\pi=7.07\,{\rm MHz}$ and the dispersive shift per magnon was $\chi/2\pi=-1.74\,{\rm MHz}$, extracted from $2\chi/2\pi=-3.48\,{\rm MHz}$. With qubit linewidth $\gamma_q/2\pi=0.36\,{\rm MHz}$ and magnon linewidth $\gamma_m/2\pi=1.6\,{\rm MHz}$, Ramsey interferometry yielded a steady-state magnon-population sensitivity $\mathcal S=1.42(3)\times10^{-3}\,{\rm magnons}/\sqrt{\rm Hz}$ [2005.09250].

A complementary spectroscopic regime is resonance fluorescence. For a weakly driven qubit–magnon Jaynes–Cummings system,
$$
H=\frac{1}{2}\Delta_q\sigma_z+\Delta_m m^\dagger m+g(\sigma_+m+\sigma_-m^\dagger)+\Omega(\sigma_++\sigma_-),
$$
the steady-state qubit and magnon spectra show two sidebands at $\omega=\omega_L\pm g$ and a central $\delta$-function “Heitler peak” at $\omega=\omega_L$ with vanishing linewidth in the weak-driving limit. The sideband splitting is $\Delta\omega=2g$, and the effect persists at both strong and weak coupling [2501.01254].

The dispersive paradigm has also been generalized to noneigenmode magnons in anisotropic ferromagnets. There the qubit frequency depends on the bare-magnon number even though the true eigenmode is a squeezed-magnon Bogoliubov mode. Qubit excitation spectroscopy can then resolve the even bare-magnon Fock components of the squeezed-magnon ground state through transitions $\lvert g;0\rangle\rightarrow\lvert e;2n\rangle$ [2306.05065].

## 5. Drive engineering, tunability, and nonclassical-state generation

Time-dependent control of qubit–magnon dipole coupling has become a central method for generating nonclassical magnon states. In the 2014 experiment, when the qubit and magnon were detuned by $274\,{\rm MHz}$ so that the static coupling was negligible, a microwave drive at
$$
\omega_d=\frac{\omega_q+\omega_m}{2}
$$
activated a third-order nonlinearity of the transmon. On two-photon resonance this produced a tunable interaction with experimentally observed $g_{qm,p}/2\pi$ up to $3.4\,{\rm MHz}$ and slope $g_{qm,p}/\sqrt{P_d}\approx4.0\,{\rm MHz/pW}$ [1410.3781].

Guo et al. proposed a two-tone-driving protocol in a cavity–magnon–qubit system to generate magnon squeezing. After cavity elimination, the qubit–magnon interaction is driven so as to engineer an effective parametric Hamiltonian $H_P\propto\chi(m^2+m^{\dagger 2})$ with $\chi=G^2/(4\Omega_2)$ and $G=g_{\rm eff}$. The required conditions are $|\delta_2|=2\Omega_1\gg\Omega_2\gg G/2$, together with residual-detuning tuning to cancel fast phases. Numerically, moderate squeezing with variance $\lesssim0.3\times$ vacuum was found for $\Omega_1\approx50\,G$ and $\Omega_2\approx3\,G$, with $G/2\pi\sim4\,{\rm MHz}$ and drive strengths in the hundreds of MHz range. The proposal states that the generated squeezed states involve more than $10^{18}$ spins and are therefore macroscopic quantum states [2304.10760].

Driven qubit–magnon coupling can also produce double dressing. In the 2019 experiment, a coherent drive resonant with the qubit led first to classical dressing of the qubit and then to hybridization of the dressed states with the magnon. The resulting four eigenmodes formed two “particle-like” and two “hole-like” branches whose splittings depended on the drive power through $\Omega_d\propto\sqrt{P_d}$, creating a particle–hole symmetry and enabling quantum simulation of composite fermion-boson quasi-particles [1903.12498].

For directly coupled transmon–YIG devices, driving can instead be used to generate strong antibunching. Under the optimized condition $\Delta_m=\Delta_q=+J$ and with probe and drive amplitudes satisfying $\Omega_q=3\,\Omega_m$, the equal-time second-order correlation function is analytically minimized. The reported optimum reaches $g^{(2)}(0)\sim10^{-7}$, and the robustness of the blockade was analyzed in the presence of thermal noise and moderate longitudinal coupling [2303.13823].

The same two-tone-control logic has been extended to multimagnetic-mode settings. In a system with one transmon, two YIG spheres, and one cavity mode, adiabatic elimination gives qubit–magnon couplings
$$
g_{qk}= \frac{g_{cq}g_{ck}}{2}\Bigl(\frac{1}{\Delta_q}+\frac{1}{\Delta_k}\Bigr),
$$
and a cavity-mediated magnon–magnon beam-splitter coupling $g_{12}$. Under two-tone driving of the qubit and appropriate detuning choices, a fourth-order elimination yields an effective two-mode-squeezing term
$$
g_{\rm eff}(m_1^\dagger m_2^\dagger + m_1m_2)\sigma_z,
$$
with $g_{\rm eff}=r_g g_{q1}g_{q2}$. For the quoted parameters, $g_{q1}/2\pi\approx7.54\,{\rm MHz}$, $g_{q2}/2\pi\approx7.08\,{\rm MHz}$, $g_{12}/2\pi\approx4.55\,{\rm MHz}$, and $g_{\rm eff}/2\pi\approx0.5$–$1\,{\rm MHz}$ [2605.06297].

## 6. Conceptual distinctions, applications, and extensions

A recurring conceptual distinction is that “qubit–magnon dipole coupling” does not refer to a single microscopic mechanism. In cavity-assisted YIG–transmon systems, the operative interaction is second order and mediated by virtual cavity photons, combining electric-dipole qubit coupling and magnetic-dipole magnon coupling. In direct transmon–YIG geometries, the dominant couplings are transverse and longitudinal terms generated by the flux response of a SQUID to the YIG stray field. In anisotropic-ferromagnet treatments, a direct first-order dispersive term can exist independently of Jaynes–Cummings exchange. This suggests that the same low-energy algebra can arise from physically different microscopic couplers [2304.10760], [2303.13823], [2306.05065].

The applications already established or proposed in the cited literature are correspondingly diverse. The earliest cavity-mediated work identified quantum control and measurement of magnon excitations, including generation of single-magnon Fock states, Bell states, and arbitrary superpositions using standard circuit-QED pulse sequences, with the ferromagnetic mode functioning as a quantum memory at the single-excitation level [1410.3781]. Subsequent work emphasized non-classical magnon states, strong-dispersive number resolution, and a coherent interface between superconducting qubits and optical photons [1508.05290], [1610.00839].

Sensing and spectroscopy form a second major branch. Dissipation-based interferometry turns magnon-induced dephasing into a metrological resource at the $10^{-3}\,{\rm magnons}/\sqrt{\rm Hz}$ scale [2005.09250], while resonance-fluorescence theory shows that elastic qubit-to-magnon scattering can produce a magnonic analog of the Heitler effect and indicates a path toward coherent magnon sources [2501.01254]. Direct dispersive access to noneigenmode magnons further enables spectroscopy of equilibrium squeezing and deterministic generation of squeezed even Fock states [2306.05065].

A third branch uses magnons as mediators rather than endpoints. In NV-center/YIG-film hybrids, the Hamiltonian
$$
H_{\rm int}=\sum_{\mathbf k} g_{\mathbf k}(\hat\sigma^+ a_{\mathbf k}+\hat\sigma^- a_{\mathbf k}^\dagger)
$$
describes dipolar NV–magnon coupling, and the magnon-induced self-energy determines an effective NV–NV interaction
$$
g_{\rm eff}(r)=\sum_{\mathbf k}|g_{\mathbf k}|^2\frac{\cos(\mathbf k\cdot\mathbf r)}{\omega_{\rm NV}-\omega_{\mathbf k}}.
$$
The experimentally extracted value at $r=0.5\,\mu{\rm m}$ was $|g_{\rm eff}(0.5\,\mu{\rm m})|\approx2\pi\times1.2\,{\rm Hz}$, compared with direct dipole–dipole coupling $2\pi\times0.4\,{\rm Hz}$ [2308.11710].

For superconducting-qubit networks, a related idea is to use a magnetic particle as a magnonic bus. In the transmon–magnet hybrid analyzed by Dols et al., inductive qubit–magnon couplings $J_i$ and $g_i$ lead, after a Schrieffer–Wolff transformation, to three effective qubit–qubit interactions: transverse $(XX+YY)$, longitudinal $(ZZ)$, and mixed $(ZX)$. Under realistic parameters the proposed gate performances were average fidelity $\gtrsim99\%$ for iSWAP and CZ, and $\gtrsim88\%$ for iCNOT [2406.14967].

Taken together, these results define qubit–magnon dipole coupling as a family of hybrid interactions that link superconducting or spin qubits to collective magnetic excitations across resonant, dispersive, and parametrically driven regimes. The common technical theme is the conversion of a macroscopically large magnetic mode into a controllable quantum degree of freedom without sacrificing single-quantum addressability.

Source: https://www.emergentmind.com/topics/qubit-magnon-dipole-coupling