Spontaneous Magnon-Pair Creation

• Spontaneous magnon-pair creation is the intrinsic process where correlated two-magnon excitations emerge from a polarized vacuum without external pumping.
• It encompasses distinct mechanisms including equilibrium instabilities leading to spin-nematic order, nonequilibrium vacuum emissions, and quasiparticle decay into two-magnon continua.
• Observable signatures through optical, microwave, and Raman spectroscopy clarify these phenomena and distinguish true vacuum instabilities from externally driven pair generation.

Spontaneous magnon-pair creation denotes a set of many-body processes in which correlated two-magnon excitations, or magnon-antimagnon excitations, appear without direct single-magnon injection. In the literature, the phrase is used in several technically distinct senses. In frustrated high-field magnets it can mean the equilibrium, field-driven instability of a fully polarized state toward a condensate of bound two-magnon states. In driven or inverted magnetic environments it can mean vacuum emission of a magnon-antimagnon pair through anomalous bosonic mode mixing between positive- and negative-energy branches. In the decay literature it is closely related to spontaneous one-magnon $\to$ two-magnon conversion at $T=0$. By contrast, parametric pumping, Suhl processes, two-photon resonance, and Raman scattering create or probe magnon pairs through externally driven channels and therefore occupy a different conceptual category, even when they access the same pair sector [(Zhitomirsky et al., 2010); (Kleinherbers et al., 2024); (Adorno et al., 2023)].

1. Terminological scope and conceptual distinctions

In the high-field spin-nematic setting, “spontaneous magnon-pair creation” refers to a ground-state instability of the fully polarized vacuum: as the field is lowered, a bound two-magnon mode reaches zero energy before the one-magnon branch, so the vacuum becomes unstable to equilibrium creation of magnon pairs which then Bose condense. The crucial point is that no external pumping is involved; the process is field-driven and equilibrium in nature (Zhitomirsky et al., 2010).

A different usage appears in nonequilibrium magnetic interfaces and driven synthetic antiferromagnets. There, one subsystem is maintained in an inverted or negative-energy configuration, and anomalous interface terms of the form $\psi_L^\dagger\psi_R^\dagger+\psi_L\psi_R$ mix annihilation and creation operators. The incoming vacuum of quasiparticles is then not inert: outgoing magnon pairs are emitted spontaneously, in close analogy with two-mode squeezing, bosonic Klein scattering, or Hawking-like pair production. This entangling vacuum emission is explicitly absent in thermal equilibrium (Kleinherbers et al., 2024, Bassant et al., 2024).

A third, closely related but not identical, meaning occurs in the spontaneous-decay literature. There the central process is the cubic-channel conversion of one magnon into two magnons, schematically $a^\dagger a^\dagger a+\mathrm{h.c.}$. Viewed from the final-state side, this is spontaneous creation of a magnon pair by an excited quasiparticle, but it is not vacuum pair creation from the ground state (Zhitomirsky et al., 2012).

Driven optical and microwave processes must therefore be separated from strict spontaneous pair creation. In room-temperature YIG, microwave parametric pumping injects magnons in pairs because photons at $\omega_{\mathrm p}$ split into magnons at $\omega_{\mathrm p}/2$ with opposite wavevectors, but the injection itself is externally driven; the spontaneous element in that work is the later emergence of condensate coherence from the overpopulated magnon gas (Noack et al., 2021). Similarly, two-photon excitation of a bound magnon pair near a spin-nematic phase and two-magnon Raman processes are driven spectroscopies of the pair sector rather than vacuum instabilities (Sato et al., 2020, Fedianin et al., 2024).

2. Equilibrium field-driven pair creation and spin-nematic order

The canonical equilibrium realization is the frustrated spin-$\tfrac12$ Heisenberg magnet in a magnetic field,

$$\hat{\mathcal H} = \frac12 \sum_{i,\mathbf r} J(\mathbf r)\,\mathbf S_i\cdot \mathbf S_j - H\sum_i S_i^z.$$

For sufficiently strong field the ground state is the fully polarized ferromagnetic state $|0\rangle=|\!\uparrow\uparrow\uparrow\cdots\rangle$, which serves as the magnon vacuum. In ordinary antiferromagnets the first instability on lowering $H$ is a one-magnon Bose condensation at

$T=0$0

In the frustrated ferro/antiferromagnetic case, however, ferromagnetic bonds generate an attraction between nearby spin flips, so a two-magnon bound state can split off below the continuum. If its binding energy is $T=0$1, the pair instability occurs at

$T=0$2

and therefore preempts one-magnon condensation whenever $T=0$3 (Zhitomirsky et al., 2010).

For the quasi-1D model motivated by $T=0$4, with $T=0$5, $T=0$6, and $T=0$7, bound states exist for $T=0$8, the minimum occurs at $T=0$9, and the numerical binding energy is $\psi_L^\dagger\psi_R^\dagger+\psi_L\psi_R$0. The resulting saturation fields are $\psi_L^\dagger\psi_R^\dagger+\psi_L\psi_R$1 and $\psi_L^\dagger\psi_R^\dagger+\psi_L\psi_R$2, so the first soft mode is a magnon pair rather than a single magnon. In this precise sense, spontaneous magnon-pair creation means field-induced, equilibrium onset of a condensate of bound two-magnon states from the fully polarized vacuum (Zhitomirsky et al., 2010).

Below $\psi_L^\dagger\psi_R^\dagger+\psi_L\psi_R$3, the condensate is represented by the coherent pair state

$\psi_L^\dagger\psi_R^\dagger+\psi_L\psi_R$4

which is described as the bosonic equivalent of the BCS pairing wavefunction. The anomalous average

$\psi_L^\dagger\psi_R^\dagger+\psi_L\psi_R$5

shows that the condensed object is the pair field, not the single-magnon field. The order parameter is quadrupolar: $\psi_L^\dagger\psi_R^\dagger+\psi_L\psi_R$6 Consequently $\psi_L^\dagger\psi_R^\dagger+\psi_L\psi_R$7 although $\psi_L^\dagger\psi_R^\dagger+\psi_L\psi_R$8, which is the defining signature of spin-nematic order (Zhitomirsky et al., 2010).

The zero-temperature nematic phase retains a gapped one-magnon branch. For the $\psi_L^\dagger\psi_R^\dagger+\psi_L\psi_R$9 parameters, the one-magnon gap at $a^\dagger a^\dagger a+\mathrm{h.c.}$0 is

$a^\dagger a^\dagger a+\mathrm{h.c.}$1

At the same time the pair condensate supports an additional gapless collective mode associated with phase fluctuations of the nematic order parameter. Static dipolar correlations remain short-ranged, the transverse structure factor is diffuse rather than Bragg-like, and the paper predicts a high-field long-range ordered spin-nematic phase for $a^\dagger a^\dagger a+\mathrm{h.c.}$2 in the window $a^\dagger a^\dagger a+\mathrm{h.c.}$3, with $a^\dagger a^\dagger a+\mathrm{h.c.}$4 from the mean-field comparison to the lower-field spin-cone state (Zhitomirsky et al., 2010).

3. Nonequilibrium vacuum pair creation, anomalous scattering, and Schwinger analogues

A nonequilibrium route appears in coupled ferromagnets where one magnet is in its ground state and the other is maintained in an inverted state by spin torque pumping. In that setting the left side supports positive-energy magnons with $a^\dagger a^\dagger a+\mathrm{h.c.}$5, while the right side supports negative-energy excitations with $a^\dagger a^\dagger a+\mathrm{h.c.}$6. The interface Hamiltonian density

$a^\dagger a^\dagger a+\mathrm{h.c.}$7

contains explicit anomalous pair terms. The term $a^\dagger a^\dagger a+\mathrm{h.c.}$8 creates simultaneously one positive-energy magnon and one negative-energy magnon, and the process is resonant in the window $a^\dagger a^\dagger a+\mathrm{h.c.}$9. In the incoming vacuum $\omega_{\mathrm p}$0, the paper finds a nonzero vacuum-state quantum spin current

$\omega_{\mathrm p}$1

which is the direct signature of spontaneous magnon-antimagnon pair creation (Kleinherbers et al., 2024).

The same anomalous structure yields nonlocal dissipation for two spatially separated color centers. The reduced dynamics contains jump operators

$\omega_{\mathrm p}$2

and at the symmetric point the steady state is

$\omega_{\mathrm p}$3

Using $\omega_{\mathrm p}$4, the concurrence approaches $\omega_{\mathrm p}$5 when $\omega_{\mathrm p}$6. The paper states explicitly that this entanglement is absent in thermal equilibrium (Kleinherbers et al., 2024).

A closely related interface mechanism is realized in a driven synthetic antiferromagnet. There one ferromagnetic layer is aligned with its field, the other is dynamically stabilized opposite to its field by spin-orbit torque, and the RKKY interface generates anomalous terms

$\omega_{\mathrm p}$7

The positive-energy left branch $\omega_{\mathrm p}$8 overlaps a negative-energy right branch $\omega_{\mathrm p}$9 in the window $\omega_{\mathrm p}/2$0. In that regime the scattering matrix is pseudo-unitary and mixes annihilation and creation operators. The outgoing vacuum occupations are

$\omega_{\mathrm p}/2$1

and the zero-temperature current is

$\omega_{\mathrm p}/2$2

The paper further reports nonclassical left-right current correlations and entanglement detected through violation of a Cauchy-Schwarz inequality (Bassant et al., 2024).

A field-theoretic extreme of the same general idea is the Schwinger mechanism for antiferromagnetic magnons in a spatially inhomogeneous magnetic field. The low-energy theory maps the magnon field to a charged complex scalar,

$\omega_{\mathrm p}/2$3

A critical magnetic step satisfies $\omega_{\mathrm p}/2$4. In the Klein zone, in- and out-vacua are inequivalent, and the differential mean number of created magnon-antimagnon pairs is

$\omega_{\mathrm p}/2$5

For smooth-gradient steps the pair flux has the universal form

$\omega_{\mathrm p}/2$6

The paper also identifies a distinctly bosonic “statistically assisted Schwinger effect,” in which preexisting bosons enhance the pair-production rate rather than suppressing it (Adorno et al., 2023).

This suggests a unifying nonequilibrium picture: spontaneous pair creation arises when anomalous bosonic couplings connect positive- and negative-energy channels, so that the vacuum defined by incoming modes is no longer a stationary no-particle state.

4. Driven pair generation and optical access to the pair sector

External driving accesses the same pair sector through explicitly nonequilibrium channels. In room-temperature YIG, parametric pumping in a longitudinal microwave field creates magnons because “external microwave photons of frequency $\omega_{\mathrm p}/2$7 and wavenumber $\omega_{\mathrm p}/2$8 split into two magnons with the frequency $\omega_{\mathrm p}/2$9 and wavevectors $\tfrac12$0.” In the experiment $\tfrac12$1, so the initially amplified magnons are at $\tfrac12$2, above the condensate minimum $\tfrac12$3. The injected population forms a parametrically overpopulated magnon gas that redistributes by four-magnon $\tfrac12$4 scattering, then evolves toward a coherent Bose-Einstein condensate. The late-time line narrows to a Lorentzian with $\tfrac12$5, and the paper identifies the spontaneous element as the later appearance of coherence, not the initial pair injection itself (Noack et al., 2021).

A more explicit pair instability is the second-order Suhl process for Dirac magnons in a honeycomb ferromagnet. A uniform circularly polarized transverse pump cannot create a single Dirac magnon because the Dirac points are at finite momentum, but it can create opposite-momentum pairs through absorption of two pump quanta. At the resonance $\tfrac12$6, the anomalous pairing amplitude is

$\tfrac12$7

and the pair-mode spectrum is

$\tfrac12$8

Near the Dirac points, at $\tfrac12$9,

$$\hat{\mathcal H} = \frac12 \sum_{i,\mathbf r} J(\mathbf r)\,\mathbf S_i\cdot \mathbf S_j - H\sum_i S_i^z.$$0

so there is an instability for

$$\hat{\mathcal H} = \frac12 \sum_{i,\mathbf r} J(\mathbf r)\,\mathbf S_i\cdot \mathbf S_j - H\sum_i S_i^z.$$1

The paper introduces a bosonic analog of the Cooper ladder, which enhances the pairing vertex and leads to a Dirac magnon paired state with reduced or zero magnetization, with estimates suggesting relevance for CrBr$$\hat{\mathcal H} = \frac12 \sum_{i,\mathbf r} J(\mathbf r)\,\mathbf S_i\cdot \mathbf S_j - H\sum_i S_i^z.$$2 or CrCl$$\hat{\mathcal H} = \frac12 \sum_{i,\mathbf r} J(\mathbf r)\,\mathbf S_i\cdot \mathbf S_j - H\sum_i S_i^z.$$3 below but near the Curie temperature (Zyuzin, 2020).

Near a spin-nematic phase, bound magnon pairs can also be driven directly by nonlinear resonance. In a frustrated four-spin model, the single-magnon resonance is

$$\hat{\mathcal H} = \frac12 \sum_{i,\mathbf r} J(\mathbf r)\,\mathbf S_i\cdot \mathbf S_j - H\sum_i S_i^z.$$4

whereas the magnon-pair resonance is

$$\hat{\mathcal H} = \frac12 \sum_{i,\mathbf r} J(\mathbf r)\,\mathbf S_i\cdot \mathbf S_j - H\sum_i S_i^z.$$5

Because the pair carries angular momentum $$\hat{\mathcal H} = \frac12 \sum_{i,\mathbf r} J(\mathbf r)\,\mathbf S_i\cdot \mathbf S_j - H\sum_i S_i^z.$$6, it is excited by simultaneous absorption of two photons. Solving a Lindblad master equation, the paper concludes that intense THz laser with ac magnetic field of $$\hat{\mathcal H} = \frac12 \sum_{i,\mathbf r} J(\mathbf r)\,\mathbf S_i\cdot \mathbf S_j - H\sum_i S_i^z.$$7 is enough to observe magnon-pair resonance, and uses the pair peak as a signature of nearby spin-nematic order (Sato et al., 2020).

Optical spectroscopy of two-magnon modes provides another route. In spontaneous Raman scattering from a cubic antiferromagnet, the relevant excitation is a pair of magnons with opposite wavevectors and total momentum approximately zero. The Raman signal is

$$\hat{\mathcal H} = \frac12 \sum_{i,\mathbf r} J(\mathbf r)\,\mathbf S_i\cdot \mathbf S_j - H\sum_i S_i^z.$$8

so RS probes the imaginary part of the pair Green function only. In impulsive stimulated Raman scattering, by contrast, the detected ellipticity depends on the absolute value of the full Green function, and therefore on both real and imaginary parts. The paper concludes that RS probes a continuum of incoherent pair populations, whereas ISRS probes coherent pair amplitudes and phases; as a consequence, the ISRS spectrum is shifted, broadened, and extends above the maximum two-magnon mode frequency (Fedianin et al., 2024).

5. Spontaneous decay into two-magnon continua and related non-Hermitian structures

The most extensively studied adjacent phenomenon is spontaneous magnon decay at $$\hat{\mathcal H} = \frac12 \sum_{i,\mathbf r} J(\mathbf r)\,\mathbf S_i\cdot \mathbf S_j - H\sum_i S_i^z.$$9. Its microscopic origin is the cubic interaction

$|0\rangle=|\!\uparrow\uparrow\uparrow\cdots\rangle$0

which appears whenever symmetry allows mixing between one- and two-magnon sectors. In collinear antiferromagnets cubic terms are forbidden by the residual SO(2) symmetry about the Néel vector. They become allowed in noncollinear magnets, in field-canted antiferromagnets, and in certain triplon systems. The kinematic condition is overlap with the two-magnon continuum,

$|0\rangle=|\!\uparrow\uparrow\uparrow\cdots\rangle$1

and the on-shell decay rate is

$|0\rangle=|\!\uparrow\uparrow\uparrow\cdots\rangle$2

The papers in this line treat spontaneous creation of the two-magnon final state by an already excited magnon, rather than vacuum pair creation (Zhitomirsky et al., 2012).

In a honeycomb ferromagnet with in-plane second-neighbor Dzyaloshinskii-Moriya exchange, the same $|0\rangle=|\!\uparrow\uparrow\uparrow\cdots\rangle$3 channel can be absorbed into an effective non-Hermitian Hamiltonian for the one-magnon sector. The one-loop self-energy is

$|0\rangle=|\!\uparrow\uparrow\uparrow\cdots\rangle$4

so entry into the two-magnon continuum gives an imaginary part and finite lifetime. Near a Dirac point, this anti-Hermitian structure splits the touching into topologically protected exceptional points or lines, and produces a characteristic anisotropy of spectral broadening in the spectral function and dynamical structure factor (McClarty et al., 2019).

Hybrid magnetoelastic excitations furnish an experimentally sharp realization. In $|0\rangle=|\!\uparrow\uparrow\uparrow\cdots\rangle$5, noncollinear $|0\rangle=|\!\uparrow\uparrow\uparrow\cdots\rangle$6 order allows cubic magnon anharmonicities, and exchange striction generates linear magnon-phonon hybridization. The resulting high-energy magnetoelastic mode near $|0\rangle=|\!\uparrow\uparrow\uparrow\cdots\rangle$7 crosses the line of singularities of the two-magnon continuum even where the pure magnon branch does not. Anharmonic $|0\rangle=|\!\uparrow\uparrow\uparrow\cdots\rangle$8 calculations reproduce strong intrinsic linewidth broadening of the top mode near the B and D points, which the paper identifies as spontaneous decay of the hybrid magnetoelastic excitation into two magnons (Oh et al., 2016).

A related but distinct extension occurs in two-dimensional altermagnets, where the leading spontaneous zero-temperature channel is not $|0\rangle=|\!\uparrow\uparrow\uparrow\cdots\rangle$9 but $H$0. There the decay rate obeys

$H$1

with strong directional and chirality selectivity. This does not describe magnon-pair creation directly, but it clarifies that anomalous magnon-number-nonconserving interactions and dispersion convexity can also favor higher-order spontaneous generation channels (Cichutek et al., 27 Feb 2025).

6. Order parameters, observables, and conceptual boundaries

Across these settings, the most important distinction is between pair creation as a microscopic channel and pair order as a macroscopic phase. In the spin-nematic problem, the order parameter is $H$2, not $H$3, so the ordered state exhibits quadrupolar long-range order without dipolar transverse order. The predicted signatures are a sharp change in the slope of the magnetization curve, absence of transverse dipolar Bragg peaks, diffuse transverse structure factor, a gapped one-magnon branch, and a longitudinal low-energy response from the collective mode of the nematic order parameter (Zhitomirsky et al., 2010).

In nonequilibrium pair-emission problems, the primary observables are instead currents, outgoing occupations, and nonlocal correlations. For the coupled-ferromagnet interface, the vacuum-state spin current is nonzero even though the incoming quasiparticle modes are empty, and the pair correlations induce nonlocal Lindblad jump operators that drive remote color centers into an entangled steady state. For the driven synthetic antiferromagnet, the spontaneous contribution survives as a temperature-independent current and as left-right correlations in the outgoing magnon channels (Kleinherbers et al., 2024, Bassant et al., 2024).

Optical probes access the pair sector differently. Parametric pumping and Suhl processes reveal instabilities of the pair channel under strong drive, two-photon spectroscopy resolves bound magnon pairs near a spin-nematic instability, and Raman scattering measures the two-magnon continuum either incoherently through $H$4 or coherently through $H$5. These experiments do not by themselves establish vacuum instability, but they provide frequency scales, selection rules, and line shapes of the pair sector that can diagnose or constrain spontaneous mechanisms (Noack et al., 2021, Sato et al., 2020, Fedianin et al., 2024).

A common misconception is therefore to treat every two-magnon process as the same phenomenon. The literature distinguishes at least four physically different cases: equilibrium softening of a bound pair mode in a saturated magnet, vacuum pair emission by anomalous positive-/negative-energy mode mixing, spontaneous decay of an excited quasiparticle into a two-magnon continuum, and externally driven pair excitation by optical or microwave fields. They share recurring mathematical structures—anomalous averages $H$6, anomalous quadratic terms, cubic vertices $H$7, or pseudo-unitary Bogoliubov scattering—but the physical interpretation depends on whether the relevant instability is thermodynamic, dynamical, spectroscopic, or decay-like.

This dependence on context is especially clear in bosonic Bogoliubov-de Gennes treatments of magnon condensates. In the ferroelectric instability of a magnon Bose-Einstein condensate, the fluctuation Hamiltonian contains explicit off-diagonal anomalous terms

$H$8

with $H$9 representing pair-creation and pair-annihilation channels relative to the condensate. Yet the paper frames the main transition as a spontaneous ferroelectric instability of the magnon superfluid, not as standalone spontaneous magnon-pair creation. This suggests that anomalous pair structure is often a necessary component of the theory, but not always the primary physical phenomenon (Yamamoto et al., 26 Dec 2025).

Spontaneous magnon-pair creation is therefore not a single mechanism but a technically stratified concept. In one limit it is the equilibrium emergence of a bound-pair condensate and spin-nematic order from the fully polarized vacuum. In another it is vacuum decay into magnon-antimagnon pairs in a nonequilibrium or inhomogeneous background, with associated currents and entanglement. In a third it is the pair-emission side of spontaneous quasiparticle decay. Around these core meanings lies a broad spectroscopic and driven literature that accesses the same pair sector from outside equilibrium and has made its dynamics experimentally visible.