Graviton Bremsstrahlung: Mechanisms & Signatures

• Graviton bremsstrahlung is the gravitational analogue of electromagnetic bremsstrahlung, where accelerating energy flows emit gravitons via universal coupling to the stress tensor.
• It plays a key role in processes like reheating, heavy-particle decays, and scattering events, with rates often suppressed by the Planck scale yet enhanced under extreme conditions.
• The phenomenon bridges universal soft-graviton theorems with hard process dynamics, offering insights into high-frequency gravitational wave backgrounds and potential indirect detection through decoherence.

Graviton bremsstrahlung is the emission of a graviton as an additional radiated quantum during an otherwise ordinary decay, scattering, or deflection process. In the modern literature, it is treated as the gravitational analogue of electromagnetic bremsstrahlung: whenever an accelerating or nontrivially rearranged energy-momentum flow is present, the universal coupling of gravity to the stress tensor permits the radiation of a graviton. This mechanism appears in perturbative reheating, heavy-particle decays, hyperbolic encounters, post-Minkowskian two-body scattering, plasma and nucleon collisions, and open-system analyses of decoherence. Across these settings, the subject combines soft-graviton universality, model-dependent hard dynamics, cosmological redshift, and detector-dependent observability into a single framework (Cline et al., 15 May 2026).

1. Universal coupling and basic processes

In linearized treatments, the metric is expanded around flat spacetime as

$g_{\mu\nu}=\eta_{\mu\nu}+\kappa h_{\mu\nu},$

with conventions such as $\kappa=\sqrt{32\pi G}=2/M_P$ or $\kappa=2/M_{\mathrm{Pl}}$ depending on normalization (Cline et al., 15 May 2026). In this regime, graviton bremsstrahlung is generated by the universal interaction of $h_{\mu\nu}$ with the energy-momentum tensor. Reheating analyses write this as

$$\sqrt{-g}\,\mathcal{L}\supset \frac{1}{M_P}h_{\mu\nu}\sum_k T_k^{\mu\nu},$$

while decay calculations often use

$$\sqrt{-g}\,\mathcal{L}_{\rm int}^{(g)}\supset -\frac{2}{M_P}\,h_{\mu\nu}T^{\mu\nu}$$

(Xu, 2024). The common point is that graviton emission is not an extra model-building option once the hard process is present; it is the gravitational dressing of that process.

The elementary channels studied in the literature are correspondingly diverse. In reheating, the basic radiative processes are three-body decays such as

$$\phi \to \varphi\,\varphi\,h,\qquad \phi \to \bar\psi\,\psi\,h,$$

and related annihilation or scattering channels (Barman et al., 2024). In heavy-particle cosmology, one finds decays such as

$S \to \Phi^\dagger \Phi + h_{\mu\nu},$

or scalar-leptoquark channels like

$$\Delta_a \to \ell_{iR}\,u_{ajR}\,h,\qquad \Delta_a \to \nu_{iL}\,d_{ajL}\,h$$

(Inui et al., 2024). In classical collision problems, gravitational bremsstrahlung is the radiation emitted when bodies scatter in a $1/r$ or Yukawa potential, and in post-Minkowskian treatments it is the radiated momentum carried by emitted gravitons during hyperbolic two-body scattering (Steane, 2023).

Because the coupling is Planck-suppressed, the effect is typically small unless enhanced by very large parent masses, high center-of-mass energy, or cosmological accumulation. This is explicit in decay calculations where differential widths scale as $\kappa=\sqrt{32\pi G}=2/M_P$0 for superheavy scalar leptoquarks, and in reheating calculations where the total GW signal inherits factors of $\kappa=\sqrt{32\pi G}=2/M_P$1 or $\kappa=\sqrt{32\pi G}=2/M_P$2 (Wang et al., 27 Jun 2026).

2. Soft-graviton structure and infrared universality

A central organizing principle is Weinberg’s soft-graviton theorem. For inflaton decay, the theorem is written as

$\kappa=\sqrt{32\pi G}=2/M_P$3

with

$\kappa=\sqrt{32\pi G}=2/M_P$4

This factorization implies that the soft limit depends only on external momenta and the gravitational coupling, not on the microscopic operator responsible for the hard decay (Cline et al., 15 May 2026).

Applied to $\kappa=\sqrt{32\pi G}=2/M_P$5, the squared amplitude acquires the universal singular form

$\kappa=\sqrt{32\pi G}=2/M_P$6

The $\kappa=\sqrt{32\pi G}=2/M_P$7 pole is the soft-graviton singularity. In the Boltzmann description of reheating, it yields a collision term with $\kappa=\sqrt{32\pi G}=2/M_P$8 behavior,

$\kappa=\sqrt{32\pi G}=2/M_P$9

and after cosmological redshift it becomes a present-day infrared branch

$\kappa=2/M_{\mathrm{Pl}}$0

The low-frequency linear rise is therefore a theorem-controlled result rather than an operator-specific feature (Cline et al., 15 May 2026).

This infrared universality does not imply that the soft sector always dominates the total emitted energy. In ultra-planckian classical collisions in $\kappa=2/M_{\mathrm{Pl}}$1, the low-frequency part of the radiation is present, but the paper on multidimensional collisions finds that its contribution to the total emitted energy is negligible for all $\kappa=2/M_{\mathrm{Pl}}$2, with the dominant classical emission scale instead at

$\kappa=2/M_{\mathrm{Pl}}$3

(Gal'tsov et al., 2012). In post-Minkowskian EFT, the zero-frequency limit is nevertheless physically meaningful, because it controls the soft energy spectrum and radiated angular momentum. There the emitted four-momentum starts at $\kappa=2/M_{\mathrm{Pl}}$4, and the $\kappa=2/M_{\mathrm{Pl}}$5 behavior is extracted directly from the classical emission amplitude (Mougiakakos et al., 2021).

A further implication is that infrared universality and hard-process sensitivity coexist. Reheating papers repeatedly separate these roles: the soft pole structure is universal, while the normalization and endpoint behavior depend on the hard decay rate, phase-space averages, and the microscopic channel (Cline et al., 15 May 2026).

3. Early-universe production and reheating backgrounds

The most extensive recent development concerns perturbative reheating, where graviton bremsstrahlung sources a stochastic high-frequency GW background. In the irreducible-floor analysis, inflaton decay during reheating inevitably emits soft gravitons, and after cosmological evolution this produces an SGWB with an infrared branch fixed by Weinberg’s theorem and a maximum signal that can reach

$\kappa=2/M_{\mathrm{Pl}}$6

at frequencies above the GHz scale. The same work shows that for inflaton $\kappa=2/M_{\mathrm{Pl}}$7-body decays the soft normalization is reduced by

$\kappa=2/M_{\mathrm{Pl}}$8

relative to the two-body case, so the maximal spectrum for $\kappa=2/M_{\mathrm{Pl}}$9-body decays is suppressed by $h_{\mu\nu}$0 (Cline et al., 15 May 2026).

This reheating program began with explicit three-body decay calculations. A detailed treatment of $h_{\mu\nu}$1 for scalar, fermion, and vector final states recalculated the differential decay rates and emphasized that earlier literature had obtained different results. The subsequent GW spectrum was evolved through full reheating rather than instantaneous reheating, with the signal typically peaking in the GHz to THz ballpark and contributing to $h_{\mu\nu}$2 at a level that is generally tiny unless the inflaton mass is close to $h_{\mu\nu}$3 (Barman et al., 2023). Closely related studies extended the same mechanism to monomial inflaton potentials $h_{\mu\nu}$4, distinguishing decay reheating from annihilation reheating. For $h_{\mu\nu}$5, inflaton annihilation can generate a larger GW amplitude than decay, whereas for $h_{\mu\nu}$6 annihilation typically produces a smaller amplitude but shifts the spectrum to higher frequencies because the graviton can carry more energy at emission (Bernal et al., 2023).

The reheating literature also broadened the source taxonomy. One study showed that the same reheating couplings responsible for $h_{\mu\nu}$7 bremsstrahlung also induce thermal $h_{\mu\nu}$8 inflaton–daughter scattering,

$h_{\mu\nu}$9

and that this channel can dominate over bremsstrahlung if the reheating temperature is larger than the inflaton mass (Xu, 2024). Another study of polynomial potentials distinguished a particle picture from a coherently oscillating classical-field picture: in the former, only gravitons with energies below $$\sqrt{-g}\,\mathcal{L}\supset \frac{1}{M_P}h_{\mu\nu}\sum_k T_k^{\mu\nu},$$0 are produced, whereas in the latter, for $$\sqrt{-g}\,\mathcal{L}\supset \frac{1}{M_P}h_{\mu\nu}\sum_k T_k^{\mu\nu},$$1, the harmonic decomposition of the oscillation allows arbitrarily high-energy gravitons and produces multiple peaks in the spectrum (Jiang et al., 2024).

Several model-specific applications embed the same mechanism in beyond-Standard-Model settings. Non-thermal leptogenesis from inflaton decay into right-handed neutrinos yields a high-frequency background with a low-frequency linear rise $$\sqrt{-g}\,\mathcal{L}\supset \frac{1}{M_P}h_{\mu\nu}\sum_k T_k^{\mu\nu},$$2, a peak near $$\sqrt{-g}\,\mathcal{L}\supset \frac{1}{M_P}h_{\mu\nu}\sum_k T_k^{\mu\nu},$$3, and a lower bound $$\sqrt{-g}\,\mathcal{L}\supset \frac{1}{M_P}h_{\mu\nu}\sum_k T_k^{\mu\nu},$$4 tied to seesaw perturbativity and reheating constraints (Ghoshal et al., 2022). In a kination-era scenario with gravitationally produced superheavy particles, bremsstrahlung from three-body decay is often buried under the inflationary background enhanced by kination, but for parent masses near the Planck scale it can surpass the inflationary component (Inui et al., 2024). In the KSVZ axion model, decay of the heavy PQ scalar or the vector-like heavy quark generates ultrahigh-frequency GW spectra whose high-frequency distinguishability improves if those particles induce an early matter-dominated era (Wang et al., 28 Jan 2026). A separate study of superheavy scalar leptoquarks in minimal renormalizable $$\sqrt{-g}\,\mathcal{L}\supset \frac{1}{M_P}h_{\mu\nu}\sum_k T_k^{\mu\nu},$$5 finds spectra peaking around $$\sqrt{-g}\,\mathcal{L}\supset \frac{1}{M_P}h_{\mu\nu}\sum_k T_k^{\mu\nu},$$6, with heavier leptoquarks giving stronger and harder signals because the differential widths scale as $$\sqrt{-g}\,\mathcal{L}\supset \frac{1}{M_P}h_{\mu\nu}\sum_k T_k^{\mu\nu},$$7 (Wang et al., 27 Jun 2026).

4. Scattering, collisions, and post-Minkowskian radiation

Outside reheating, graviton bremsstrahlung has a long classical and semiclassical history in scattering theory. In arbitrary $$\sqrt{-g}\,\mathcal{L}\supset \frac{1}{M_P}h_{\mu\nu}\sum_k T_k^{\mu\nu},$$8 dimensions, ultra-planckian small-angle collisions of massive point particles were analyzed with a multidimensional post-linear formalism. There, the effective source splits into accelerated-particle terms and a nonlocal gravitational stress tensor quadratic in the first-order field, and the paper shows a destructive interference in the high-frequency beamed regime, schematically

$$\sqrt{-g}\,\mathcal{L}\supset \frac{1}{M_P}h_{\mu\nu}\sum_k T_k^{\mu\nu},$$9

The characteristic classical emission frequency is $$\sqrt{-g}\,\mathcal{L}_{\rm int}^{(g)}\supset -\frac{2}{M_P}\,h_{\mu\nu}T^{\mu\nu}$$0, while the soft part contributes negligibly to the total radiated energy (Gal'tsov et al., 2012).

Modern amplitude-based formulations recast the same problem in inclusive quantum language. In the KMOC formalism, the radiated four-momentum $$\sqrt{-g}\,\mathcal{L}_{\rm int}^{(g)}\supset -\frac{2}{M_P}\,h_{\mu\nu}T^{\mu\nu}$$1 is the total momentum carried away by emitted gravitons, extracted from cuts of quantum scattering amplitudes. Using generalized unitarity and reverse unitarity, the $$\sqrt{-g}\,\mathcal{L}_{\rm int}^{(g)}\supset -\frac{2}{M_P}\,h_{\mu\nu}T^{\mu\nu}$$2 radiated momentum for two spinless black holes was obtained to all orders in the relative velocity,

$$\sqrt{-g}\,\mathcal{L}_{\rm int}^{(g)}\supset -\frac{2}{M_P}\,h_{\mu\nu}T^{\mu\nu}$$3

with analytic continuation to the bound-state regime giving the corresponding $$\sqrt{-g}\,\mathcal{L}_{\rm int}^{(g)}\supset -\frac{2}{M_P}\,h_{\mu\nu}T^{\mu\nu}$$4 energy loss in elliptic orbits (Herrmann et al., 2021). A complementary post-Minkowskian EFT calculation derived the conserved stress-energy tensor linearly coupled to radiation, the classical emission amplitude at leading and next-to-leading order in $$\sqrt{-g}\,\mathcal{L}_{\rm int}^{(g)}\supset -\frac{2}{M_P}\,h_{\mu\nu}T^{\mu\nu}$$5, and the soft limit of the emitted spectrum. Its small-velocity expansion of the radiated energy function,

$$\sqrt{-g}\,\mathcal{L}_{\rm int}^{(g)}\supset -\frac{2}{M_P}\,h_{\mu\nu}T^{\mu\nu}$$6

agrees with amplitude-based results (Mougiakakos et al., 2021).

The subject also includes nonrelativistic collision media. For particles scattering in a $$\sqrt{-g}\,\mathcal{L}_{\rm int}^{(g)}\supset -\frac{2}{M_P}\,h_{\mu\nu}T^{\mu\nu}$$7 potential, classical and Born-approximation results were combined into an approximate gravitational “Gaunt factor” for the total emitted energy. In that framework, close hyperbolic encounters in a gravitating cluster are identified as the dominant source of very high frequency gravitational noise in the solar system, and the total gravitational wave power of the Sun is estimated as $$\sqrt{-g}\,\mathcal{L}_{\rm int}^{(g)}\supset -\frac{2}{M_P}\,h_{\mu\nu}T^{\mu\nu}$$8 MW (Steane, 2023). A parallel Yukawa-based analysis, including the tensor part of the internucleon interaction, studies screened Coulomb collisions in the Sun and nucleon collisions in neutron stars; there the solar luminosity estimate becomes $$\sqrt{-g}\,\mathcal{L}_{\rm int}^{(g)}\supset -\frac{2}{M_P}\,h_{\mu\nu}T^{\mu\nu}$$9 MW and a representative neutron-star power is $$\phi \to \varphi\,\varphi\,h,\qquad \phi \to \bar\psi\,\psi\,h,$$0 (Steane, 2024). These numerical differences are tied to different interaction models and approximations rather than to a single universal solar prediction.

At trans-Planckian energies, a different structure emerges. The analysis of small-angle gravitational scattering beyond the Planck scale argues that spin-2 helicity-transformation phases unify soft and Regge emission and produce a spectrum with characteristic graviton energies

$$\phi \to \varphi\,\varphi\,h,\qquad \phi \to \bar\psi\,\psi\,h,$$1

This “Hawking-like” scale is accompanied by a suppressed fragmentation region and a reduced rapidity plateau (Ciafaloni et al., 2015).

5. Quantum decoherence, collapse models, and indirect graviton evidence

A distinct branch of the literature treats graviton bremsstrahlung as an environment that carries which-path information. In a quantum Boltzmann equation analysis of a fermion prepared in a spatial superposition, the off-diagonal density-matrix element obeys a pure-dephasing law,

$$\phi \to \varphi\,\varphi\,h,\qquad \phi \to \bar\psi\,\psi\,h,$$2

with

$$\phi \to \varphi\,\varphi\,h,\qquad \phi \to \bar\psi\,\psi\,h,$$3

The rate vanishes as $$\phi \to \varphi\,\varphi\,h,\qquad \phi \to \bar\psi\,\psi\,h,$$4, increases with separation, and in the many-body coherent regime scales as

$$\phi \to \varphi\,\varphi\,h,\qquad \phi \to \bar\psi\,\psi\,h,$$5

The paper interprets this as a microscopic route from graviton emission to decoherence and a bridge to dissipative CSL-like behavior (Zarei, 13 May 2026).

A more speculative but conceptually influential proposal considers a relativistic Planck-mass interferometer. There the decoherence factor $$\phi \to \varphi\,\varphi\,h,\qquad \phi \to \bar\psi\,\psi\,h,$$6 is the overlap of radiation states on two paths, and in the regime $$\phi \to \varphi\,\varphi\,h,\qquad \phi \to \bar\psi\,\psi\,h,$$7 the visibility loss becomes essentially the probability of no graviton emission. The key scaling is

$$\phi \to \varphi\,\varphi\,h,\qquad \phi \to \bar\psi\,\psi\,h,$$8

so the Planck mass is the natural threshold where gravitational bremsstrahlung becomes strong enough to drive substantial decoherence (Riedel, 2013). This proposal does not rely on absorbing gravitons in a conventional detector; instead, decoherence is taken as the observable imprint of individual graviton emission.

Related ideas also appear in photon propagation through external gravitational fields. A paper on “bremsstrahlung of light” models a photon, technically a massless scalar field, scattering off a weak Newtonian potential and coupling to TT gravitational fluctuations through a Markovian master equation. The result is spontaneous emission of soft gravitons during light bending, with an associated redshift of the outgoing photon and a putative sub-Hz stochastic background in systems such as X-ray binaries (Wang et al., 2021). This suggests that the conceptual domain of graviton bremsstrahlung extends beyond massive sources to radiative corrections in gravitational lensing-like environments.

6. Spectral signatures, detectors, and recurring interpretive issues

The most robust observational theme is the high-frequency placement of the signal. Reheating and heavy-decay scenarios repeatedly produce present-day spectra in the MHz-to-GHz, GHz-to-THz, or higher bands, with representative endpoints or peaks near $$\phi \to \varphi\,\varphi\,h,\qquad \phi \to \bar\psi\,\psi\,h,$$9, $S \to \Phi^\dagger \Phi + h_{\mu\nu},$0, or $S \to \Phi^\dagger \Phi + h_{\mu\nu},$1 depending on the microscopic scale and reheating history (Cline et al., 15 May 2026). This frequency placement excludes conventional low-frequency observatories as primary probes in most cases. Several papers instead single out resonant-cavity high-frequency GW detectors based on the inverse Gertsenshtein effect as the most relevant experimental avenue, particularly for leptoquark-induced or reheating-induced signals (Wang et al., 27 Jun 2026).

Indirect probes are therefore central. In reheating, the integrated graviton energy density is often translated into $S \to \Phi^\dagger \Phi + h_{\mu\nu},$2, with CMB and BBN bounds constraining the overall abundance (Barman et al., 2023). In non-thermal leptogenesis, an inflaton mass near $S \to \Phi^\dagger \Phi + h_{\mu\nu},$3 can yield $S \to \Phi^\dagger \Phi + h_{\mu\nu},$4 as large as $S \to \Phi^\dagger \Phi + h_{\mu\nu},$5, which is described as within projected reach of CMB-S4 and CMB-HD (Ghoshal et al., 2022). In kination-era models, however, the inflationary tensor background itself rises at high frequency, $S \to \Phi^\dagger \Phi + h_{\mu\nu},$6, so bremsstrahlung is often hidden unless the parent mass is extremely large (Inui et al., 2024).

Several recurring interpretive issues appear across the literature. One misconception is that graviton bremsstrahlung is intrinsically model dependent. The reheating soft-theorem analyses show instead that the infrared branch is fixed by universal soft physics, while only the normalization and hard turnover remain operator dependent (Cline et al., 15 May 2026). A second misconception is that a soft singularity necessarily dominates total energy release. Classical collision studies show the opposite can occur: soft gravitons are present, but the total radiated energy is controlled by nonsoft regions of phase space (Gal'tsov et al., 2012). A third misconception is that gravitons could only be evidenced by direct absorption in a conventional detector. Decoherence-based proposals argue that path-visibility loss can, in principle, serve as evidence for individual graviton emission, although the required mass, isolation, and relativistic control are far beyond current capability (Riedel, 2013).

Taken together, the literature portrays graviton bremsstrahlung not as a single niche process but as a recurring radiative phenomenon that connects soft-theorem universality, reheating microphysics, post-Minkowskian scattering, collision-induced stochastic backgrounds, and open-system decoherence. A plausible implication is that future progress will depend less on a single canonical observable than on correlating spectral shape, frequency scale, and cosmological history across multiple experimental windows.