Motion-Induced Magnon Transport

• Motion-induced magnon transport is a phenomenon where magnons and magnon condensates actively carry spin angular momentum, orbital moments, and heat across materials through mechanisms like phase gradients, Doppler shifts, and Berry-curvature effects.
• Experimental studies using yttrium-iron-garnet (YIG) waveguides and Pt detectors reveal time-resolved propagation with clear delays indicative of finite group velocities and contributions from secondary scattering.
• Theoretical models integrate wave-packet kinematics, electromagnetic effects of moving dipoles, and emergent gauge fields to explain transport behavior, informing advances in magnon spintronics and thermal management applications.

Searching arXiv for the cited papers to ground the article in current records. Motion-induced magnon transport denotes transport phenomena in which the motion of magnons, magnon condensates, or magnonic wave packets carries spin angular momentum, orbital moment, heat, or related observables across space, between interfaces, or between relatively moving subsystems. In the literature, the concept spans time-resolved propagation of spin-wave packets, phase-gradient-driven condensate flow, drift-controlled nonlocal propagation, Berry-curvature-induced transverse motion, and Doppler-driven transport or instability under relative motion (Chumak et al., 2011, Bozhko et al., 2018, Matsumoto et al., 2011, Oue, 28 Jun 2026). It is also closely tied to the broader spin-transport program in magnetic insulators, where magnons provide an angular-momentum channel decoupled from itinerant electronic charge flow inside the magnetic medium (Althammer, 2021).

1. Fundamental picture and transport variables

A magnon is a quantum of a spin wave and carries spin angular momentum. In ferromagnets discussed in the literature, the magnon magnetic moment is taken as

${\mathbf{\mu}_{\rm{m}}= - g\mu_{\rm{B}}{\mathbf{e}_z},$

so a propagating magnon is simultaneously a carrier of spin and a moving magnetic dipole (Nakata et al., 2016). This dual character underlies two complementary viewpoints. The first treats transport through the wave-packet kinematics of spin waves, with group velocity

$v_g=\frac{\partial \omega}{\partial k},$

and with transport controlled by propagation delay, damping, scattering, and interface conversion (Chumak et al., 2011). The second emphasizes the electromagnetism of moving magnetic dipoles: magnons acquire an Aharonov–Casher phase in electric fields and, conversely, steady magnon currents generate static electric fields (Nakata et al., 2016).

In all-electrical nonlocal magnon transport, charge-to-spin conversion in a normal metal creates an interfacial spin accumulation, inelastic spin-flip processes inject or absorb magnons in the magnetically ordered insulator, and the resulting nonequilibrium magnon spin chemical potential diffuses to a remote detector where the inverse process generates a voltage (Althammer, 2021). This diffusive description is appropriate once fast magnon-magnon and magnon-phonon processes have established local quasi-equilibrium while spin relaxation remains slow. A recurring theme is that transport is governed not only by the existence of magnons, but by the mechanism that gives them directed motion or a directional bias: an antenna-launched wave packet, a phase gradient, a Berry-curvature anomalous velocity, a drift term, or a Doppler shift from relative motion.

The moving-magnon picture is not universal. In one-dimensional antiferromagnetic spin-$1/2$ chains, the review literature stresses that the Néel order is destroyed by quantum fluctuations and the magnon quasiparticle picture breaks down; the low-energy transport is instead described by a Tomonaga–Luttinger liquid (Nakata et al., 2016). This qualifies any blanket identification of spin transport with magnon transport.

2. Time-resolved propagation and direct electrical conversion

The clearest direct demonstration of motion-induced magnon transport is the time-resolved experiment in a yttrium-iron-garnet waveguide with a spatially separated Pt detector (Chumak et al., 2011). A short spin-wave packet is launched by a $50~\mu$m-wide Cu microstrip antenna into a $2.1~\mu$m-thick YIG film, and a $10$ nm Pt strip placed $3$ mm away converts the arriving magnonic spin current into a voltage through spin pumping and the inverse spin Hall effect. The key observation is a delayed electrical pulse: both the inductively detected AC signal and the DC inverse-spin-Hall signal appear only after approximately $200$ ns, consistent with finite spin-wave group velocity over the $3$ mm propagation distance. For that path length and delay, the implied group velocity is approximately

$v_g \approx 1.5 \times 10^4~\text{m/s},$

which the paper identifies as fully consistent with magnetostatic spin-wave propagation in YIG (Chumak et al., 2011).

The transport geometry is deliberately nonlocal. The source and detector are separated by millimeters, so a signal tied to the arrival of the packet cannot be explained by immediate direct electromagnetic pickup. The experiment also exploits the symmetry of the inverse spin Hall effect,

$v_g=\frac{\partial \omega}{\partial k},$0

so reversing the YIG magnetization reverses the DC voltage sign. Together, spatial separation, propagation delay, and magnetization-dependent sign reversal provide the central evidence that the electrical pulse is generated by traveling magnons rather than by circuit artifact or local spurious response (Chumak et al., 2011).

The detected voltage does not represent only the initially launched primary wave. The temporal profile of the DC signal is broader and slightly delayed relative to the AC envelope, and the paper attributes this to secondary dipolar-exchange spin waves generated by elastic two-magnon scattering. Their contribution is modeled through

$v_g=\frac{\partial \omega}{\partial k},$1

which leads, under $v_g=\frac{\partial \omega}{\partial k},$2, to

$v_g=\frac{\partial \omega}{\partial k},$3

With $v_g=\frac{\partial \omega}{\partial k},$4 and $v_g=\frac{\partial \omega}{\partial k},$5, the inferred total magnon density under Pt exceeds the primary-wave density by factors of about $v_g=\frac{\partial \omega}{\partial k},$6 for $v_g=\frac{\partial \omega}{\partial k},$7 ns and $v_g=\frac{\partial \omega}{\partial k},$8 for $v_g=\frac{\partial \omega}{\partial k},$9 ns (Chumak et al., 2011). This does not negate the transport interpretation; it refines it by showing that the detected electrical signal reflects the total magnon population arriving under the detector region.

The same work also addresses wavelength and power dependence. After correcting inductive detection efficiency by

$1/2$0

the inferred spin-pumping efficiency is essentially independent of $1/2$1 within experimental uncertainty. At low power, the inverse-spin-Hall voltage is nearly proportional to transmitted spin-wave intensity; at intermediate power, nonlinear multi-magnon scattering suppresses the primary packet; at still higher power, the DC signal rises again because secondary magnons contribute strongly to spin pumping (Chumak et al., 2011). A common misconception is therefore that a propagating-wave ISHE signal must track inductive transmission one-to-one; in this geometry it does not, because spin pumping is sensitive to local dynamic magnetization at the interface rather than to antenna efficiency alone.

3. Coherent, collective, and microwave-driven regimes

A distinct regime of motion-induced transport arises in a room-temperature magnon Bose–Einstein condensate in YIG (Bozhko et al., 2018). In that system, local laser heating creates a thermally induced magnetic inhomogeneity, or “frequency well,” inside a parametrically pumped condensate. The local frequency reduction is estimated as

$1/2$2

at the highest heating power, and it produces a condensate phase gradient that drives a magnon supercurrent out of the heated region. The expelled condensed magnons form two density humps that propagate through the condensate as Bogoliubov waves, interpreted as a second-sound-like mode. The transport extends over hundreds of micrometers, with maximum observed propagation length $1/2$3, limited by the sample edge, and one pulse reflects from the sample boundary and propagates back (Bozhko et al., 2018).

The effective dynamics are described by a one-dimensional Gross–Pitaevskii equation,

$1/2$4

with repulsive interaction $1/2$5. Small perturbations on the condensate satisfy the Bogoliubov dispersion

$1/2$6

which reduces in the long-wavelength limit to

$1/2$7

Experimentally, the pulse velocity is approximately $1/2$8, far above the estimated diffusive velocity scale of about $1/2$9, and this disparity is a principal argument against interpreting the propagating humps as ordinary diffusion (Bozhko et al., 2018). The paper is explicit, however, that the observed motion is not perfectly dissipationless. The best characterization is weakly damped or low-loss collective transport rather than strict lossless superflow on experimental timescales.

Microwave pumping provides a second coherent route to induced transport (Nakata et al., 2015). In a single ferromagnetic insulator under a rotating ac magnetic field,

$50~\mu$0

the pumped coherent magnon amplitude is resonantly enhanced at ferromagnetic resonance,

$50~\mu$1

The pumped magnon density is

$50~\mu$2

In a junction between a quasi-equilibrium magnon condensate and pumped magnons, phase locking can convert the applied ac drive into a dc magnon current; in a junction between pumped and noncondensed magnons, the current is essentially dc and is strongly enhanced by direct and indirect ferromagnetic resonance (Nakata et al., 2015). The same paper extends the mechanism to a single FI with an Aharonov–Casher phase,

$50~\mu$3

yielding a persistent current

$50~\mu$4

This establishes a drive-induced, phase-controlled transport regime in which microwave pumping sets the nonequilibrium population while the electric field supplies the geometric phase bias (Nakata et al., 2015).

4. Coupling between magnon flow and magnetic textures

Motion-induced magnon transport is strongly coupled to domain-wall dynamics. In a one-dimensional ferromagnetic wire with a transverse wall, linear spin-wave theory yields a reflectionless Pöschl–Teller scattering problem,

$50~\mu$5

and the wall responds only at second order in spin-wave amplitude (Risinggård et al., 2017). The resulting collective-coordinate equations show that magnonic spin-transfer torque produces wall translation, while the wall rotation rate is exactly zero in the dissipationless limit and negligibly small for realistic low-damping materials. In that model,

$50~\mu$6

and

$50~\mu$7

for $50~\mu$8, so the dominant effect of a transmitted magnon current is translation without rotation-induced Walker breakdown (Risinggård et al., 2017). The same paper also shows that the apparent frequency dependence of wall velocity can arise not only from intrinsic magnon-wall coupling but from the wavevector selectivity of the microwave source, through factors such as $50~\mu$9 for a square-box field profile.

Bulk Dzyaloshinskii–Moriya interaction introduces a second mechanism (Wang et al., 2014). In a chiral ferromagnetic nanowire, a magnon traversing a DMI-twisted wall switches between the nonreciprocal domain dispersions

$2.1~\mu$0

so its wavevector changes by $2.1~\mu$1. The associated momentum transfer $2.1~\mu$2 does not primarily translate the wall; it rotates the wall plane, giving

$2.1~\mu$3

or equivalently an effective longitudinal field

$2.1~\mu$4

With an easy-plane anisotropy $2.1~\mu$5, this internal rotation is hindered and converted into efficient translational motion. In the regime studied, the DMI-induced linear-momentum-transfer contribution can exceed the conventional angular-momentum-driven velocity by about an order of magnitude (Wang et al., 2014). The literature therefore distinguishes two texture-coupled transport channels: angular-momentum transfer from transmitted magnons and DMI-enabled linear-momentum transfer from branch-switching across a nonreciprocal wall.

These mechanisms have been realized experimentally in a magnetic-insulator channel based on $2.1~\mu$6 nm Bi-doped YIG with perpendicular magnetic anisotropy (Fan et al., 2022). In that system, a static domain wall placed between two antennas attenuates transmitted spin-wave signal by a factor of $2.1~\mu$7, corresponding to a power reduction of $2.1~\mu$8. Conversely, coherent magnons drive wall motion toward the source antenna, i.e. opposite to the direction of magnon propagation, as expected for angular-momentum transfer by transmitted magnons. The wall can be moved over $2.1~\mu$9 at zero applied field by a $10$0 ns RF pulse at $10$1 GHz and $10$2 dBm, and the estimated upper-bound energy cost is roughly $10$3 pJ (Fan et al., 2022). The same work emphasizes that this is a reciprocal coupling result: the wall is both a transport-active scatterer and an object reconfigurable by magnon current.

5. Geometric, transverse, and orbital motion

A separate branch of motion-induced magnon transport originates in band geometry. In the semiclassical theory of magnon wave packets, the equations of motion are

$10$4

so a confining potential near a boundary drives an anomalous velocity parallel to the edge (Matsumoto et al., 2011). This edge current produces the thermal Hall effect, while the wave packet also undergoes self-rotation that contributes orbital corrections to transport coefficients. A major conceptual point of this theory is that Kubo-only treatments miss the orbital-motion contribution $10$5; the full thermal Hall coefficient requires both $10$6 and $10$7 terms (Matsumoto et al., 2011).

Berry-curvature-induced transport need not imply a nonzero Chern number. In a modified Lieb lattice with two magnon bands, easy-axis anisotropy opens the gap

$10$8

at $10$9, while out-of-plane DMI on diagonal $3$0-$3$1 bonds generates finite local Berry curvature

$3$2

The Chern numbers nevertheless vanish for all parameters considered, yet the system retains intrinsic Hall-like transport coefficients $3$3, $3$4, and $3$5 weighted by bosonic occupation (Oliveira et al., 2022). This directly addresses a common misconception: Hall-like magnon transport can be generated by local Berry curvature even in a globally topologically trivial magnon insulator.

In topological honeycomb magnonic systems, a thermal gradient can induce a transverse in-plane magnon current through Berry-curvature anomalous velocity,

$3$6

with anomalous magnon Nernst coefficient

$3$7

The sign of this response is controlled by the Berry-curvature distribution and magnon populations rather than by Chern number alone, and the literature predicts temperature-induced sign reversal when weak low-energy curvature near $3$8 competes with stronger opposite-sign curvature near $3$9 and $200$0 (Wang et al., 2017).

Static noncollinear textures can also generate spin-Hall-like transport. In a synthetic antiferromagnetic skyrmion lattice, opposite skyrmion textures in the two layers give opposite Hall tendencies and yield counterpropagating, layer-polarized in-gap magnon edge modes. A crucial qualification is that these modes are not a genuine bosonic $200$1 phase: the fully coupled bilayer lacks the pseudo-time-reversal symmetry required for that classification, so the observed edge states are spin-Hall-like rather than $200$2-protected helical modes (Zheng et al., 27 May 2026).

Orbital transport extends the same logic to magnon orbital moment and the electric dipole moment generated by magnon motion. In an hexagonal altermagnet, the magnon orbital moment operator

$200$3

and the effective electric dipole operator

$200$4

support Seebeck- and Nernst-type transport with both Drude-like and intrinsic generalized-Berry-curvature contributions (Sarkar et al., 8 Mar 2026). The proposed electrical probe converts the resulting EDM current into edge polarization and a measurable voltage, with an estimated signal of approximately $200$5 (Sarkar et al., 8 Mar 2026). This constitutes a direct electrical detection route for magnon orbital transport.

6. Relative motion, drift, gauge driving, and nonlocal platforms

Relative motion between interacting magnonic subsystems generates a qualitatively different transport mechanism (Oue, 28 Jun 2026). For two interacting ferromagnets with relative velocity $200$6, a magnon mode in the moving subsystem is Doppler shifted,

$200$7

producing an emergent frequency scale

$200$8

In the perturbative regime, the exchange channel drives magnon transport even without temperature or chemical-potential bias; for identical symmetric ferromagnets, the low-velocity current scales as $200$9 (Oue, 28 Jun 2026). When the nonconserving channel becomes resonant, relative motion destabilizes the magnonic vacuum and creates magnon pairs above a threshold velocity

$3$0

more precisely

$3$1

for $3$2 (Oue, 28 Jun 2026). The same paper explicitly connects this regime to quantum friction, Cherenkov emission, and Zeldovich superradiance.

A drift-based nonequilibrium bias can also be induced in a single magnetic insulator. In nonlocal magnon transport, the magnon chemical potential obeys

$3$3

so a finite drift velocity renormalizes the effective propagation length to

$3$4

The essential requirement is broken spatial symmetry in the magnonic description. Representative mechanisms are magnetic-field gradient, asymmetric dispersion, and temperature gradient, with the literature identifying the temperature gradient as particularly effective (de-la-Peña et al., 2021).

Gauge fields furnish another non-thermal drive. In a Dirac-magnon field theory on a honeycomb ferromagnet, emergent gauge fields $3$5 generate magnon density accumulation and spin current at $3$6 without conventional thermal or chemical-potential gradients (Fernández et al., 15 Dec 2025). In the dc limit, the transverse spin conductivity is

$3$7

while in the optical regime the conductivity exhibits a resonance when $3$8, reflecting interband transitions across the topological gap (Fernández et al., 15 Dec 2025). This is motion-induced transport in the broad sense that time-dependent strain, textures, or rotations can generate the emergent gauge fields.

Direct electrical access to moving-magnon transport remains a persistent theme. The review literature proposes measuring the static electric field generated by steady magnon currents, with

$3$9

and estimates voltage drops in the nV range for condensed magnon currents and in the mV range for noncondensed magnon currents (Nakata et al., 2016). In all-electrical nonlocal geometries on magnetic insulators, spin Hall injection, diffusion of magnon spin chemical potential, and inverse spin Hall detection provide the established experimental workhorse (Althammer, 2021). A bridging metallic-ferromagnet architecture shows that angular momentum can also be transferred over micron distances between electrically insulated ferromagnetic strips on a diamagnetic substrate, with phononic and dipolar interactions discussed as likely inter-strip channels and magnons implicated through local electron-magnon conversion in each ferromagnet (Schlitz et al., 2023).

Across these platforms, a recurrent interpretive issue is the distinction between literal motion of a texture and transport induced by a static or effective field. The literature is explicit that not every directional magnon current is generated by moving magnetic matter. Some cases are driven by traveling magnons themselves, some by condensate phase gradients, some by static Berry curvature of textured or crystalline backgrounds, some by emergent gauge fields, and some by relative motion between subsystems. Taken together, these works define motion-induced magnon transport as a family of nonequilibrium transport phenomena in which magnon motion is not merely a passive consequence of excitation, but the active mediator of spin, orbital, heat, or polarization flow (Chumak et al., 2011, Bozhko et al., 2018, Matsumoto et al., 2011, Oue, 28 Jun 2026).