Kinetic Screening: Dynamic Control of Interactions

• Kinetic screening is a dynamic process where effective force suppression arises from non-linear, derivative, or transport mechanisms rather than static responses, observed in fields ranging from gravity to materials.
• It plays a critical role in modified gravity models by reducing scalar-mediated forces through non-linear kinetic terms, exemplified by mechanisms like k-mouflage and variable screening radii.
• In condensed matter and plasma physics, kinetic screening emerges from time-dependent charge reorganization or kinetic collision modifications, influencing domain nucleation and transport properties.

Searching arXiv for papers on “kinetic screening” across fields to ground the overview. Kinetic screening denotes a class of phenomena in which screening is controlled by dynamical, derivative, transport, or kinetic processes rather than by a purely static dielectric response or a simple density-dependent potential. The term is used in several technically distinct ways across contemporary research. In scalar–tensor gravity it usually refers to suppression of fifth forces by non-linear kinetic self-interactions or derivative couplings; in ferroelectrics it can denote time-dependent charge screening that reshapes domain-nucleation kinetics at grain boundaries; in correlated-electron systems it can describe screening generated by kinetic-energy–bound charge complexes; in cosmology it can refer to screening of observables through kinetic mixing or through reconstructed optical-depth fields; and in plasma or QCD kinetic theory it can mean medium-regulated interactions implemented directly in the kinetic collision operator (Tian et al., 15 Dec 2025, Kapetanović et al., 2024, Brax et al., 14 Mar 2026, Pîrvu et al., 2023, Boguslavski et al., 2024).

1. Scalar–tensor gravity and derivative self-interactions

In scalar–tensor theories, kinetic screening is the mechanism whereby non-linear dependence on the scalar kinetic invariant suppresses scalar-mediated forces in regions of large field gradients or large effective source density. A representative action is written in Einstein frame as

$$S = \int d^4x\,\sqrt{-g}\,\bigg[\frac{P^2}{2}R + K(X)\bigg] + S_m\!\left[A(\varphi) g_{\mu\nu},\Psi_m\right], \qquad X \equiv g^{\mu\nu}\partial_\mu\varphi\,\partial_\nu\varphi,$$

with an EFT expansion

$$K(X) = -\frac{X}{2} + \frac{\beta}{4\Lambda^4}X^2 - \frac{\gamma}{8\Lambda^8}X^3 + O(X^4).$$

The field equation takes the form

$$\nabla_\mu\!\big[K'(X)\,\nabla^\mu\varphi\big] = \Sigma,$$

so that the effective kinetic coefficient $K'(X)$ controls how the scalar responds to matter sources. For large $|X|$, higher powers in $K(X)$ dominate and the scalar-mediated force is reduced. In the literature summarized here, this is also called $k$-mouflage (Cayuso et al., 1 May 2026, Bezares et al., 2021, Cayuso et al., 2024).

For static spherical sources, these theories admit a screening radius $r_*$. In one cubic $K$-essence truncation,

$r_* = \Big(\tfrac{3}{4}\Big)^{1/8}\frac{1}{\Lambda}\sqrt{\frac{\alpha M}{4\pi P},$

so that for $$K(X) = -\frac{X}{2} + \frac{\beta}{4\Lambda^4}X^2 - \frac{\gamma}{8\Lambda^8}X^3 + O(X^4).$$0 the scalar behaves like a linear Fierz–Jordan–Brans–Dicke (FJBD) field, while for $$K(X) = -\frac{X}{2} + \frac{\beta}{4\Lambda^4}X^2 - \frac{\gamma}{8\Lambda^8}X^3 + O(X^4).$$1 non-linear kinetic terms dominate and the scalar force is suppressed (Cayuso et al., 1 May 2026). Related Jordan-frame formulations use a non-linear function $$K(X) = -\frac{X}{2} + \frac{\beta}{4\Lambda^4}X^2 - \frac{\gamma}{8\Lambda^8}X^3 + O(X^4).$$2 with field equation

$$K(X) = -\frac{X}{2} + \frac{\beta}{4\Lambda^4}X^2 - \frac{\gamma}{8\Lambda^8}X^3 + O(X^4).$$3

where $$K(X) = -\frac{X}{2} + \frac{\beta}{4\Lambda^4}X^2 - \frac{\gamma}{8\Lambda^8}X^3 + O(X^4).$$4 in the screened region, suppressing the effective source term $$K(X) = -\frac{X}{2} + \frac{\beta}{4\Lambda^4}X^2 - \frac{\gamma}{8\Lambda^8}X^3 + O(X^4).$$5 (Shibata et al., 2022).

A distinct but closely related use of “kinetic/derivative screening” appears in multi-field scalar–tensor theories with curved target-space metric. In an axio–dilaton model with

$$K(X) = -\frac{X}{2} + \frac{\beta}{4\Lambda^4}X^2 - \frac{\gamma}{8\Lambda^8}X^3 + O(X^4).$$6

the dilaton equation contains the two-derivative source

$$K(X) = -\frac{X}{2} + \frac{\beta}{4\Lambda^4}X^2 - \frac{\gamma}{8\Lambda^8}X^3 + O(X^4).$$7

For $$K(X) = -\frac{X}{2} + \frac{\beta}{4\Lambda^4}X^2 - \frac{\gamma}{8\Lambda^8}X^3 + O(X^4).$$8 relative to the matter coupling, the axion backreaction can cancel the dilaton gradient and suppress the exterior scalar charge $$K(X) = -\frac{X}{2} + \frac{\beta}{4\Lambda^4}X^2 - \frac{\gamma}{8\Lambda^8}X^3 + O(X^4).$$9. The paper terms the resulting energy-minimization mechanism the “BBQ mechanism,” and emphasizes that screening is achieved by cancelling the dilaton’s gradient rather than localising it (Brax et al., 14 Mar 2026).

This body of work also makes clear that kinetic screening is not synonymous with thin-shell screening. In the multi-field setting, axion gradients can make the scalar profile more localized without suppressing the exterior charge, while in the unpinned regime the BBQ mechanism can strongly suppress $$\nabla_\mu\!\big[K'(X)\,\nabla^\mu\varphi\big] = \Sigma,$$0 without forming a thin shell (Brax et al., 14 Mar 2026). A plausible implication is that “screening” and “localisation” are dynamically distinct notions once multi-field or derivative couplings are present.

2. Compact stars, scalar radiation, and robustness questions

For compact objects, kinetic screening is tested not only in static configurations but also in radiative and strong-field dynamics. Numerical and semi-analytic studies of oscillating neutron stars, neutron star–black hole binaries, and binary neutron stars show that suppression depends on the relation between the screening scale and the emitted wavelength.

For non-spherical oscillations of neutron stars, the monopole scalar mode is reported to be always suppressed, regardless of the size of the screening radius $$\nabla_\mu\!\big[K'(X)\,\nabla^\mu\varphi\big] = \Sigma,$$1, whereas quadrupole suppression occurs only when $$\nabla_\mu\!\big[K'(X)\,\nabla^\mu\varphi\big] = \Sigma,$$2 (Shibata et al., 2022). In that regime, both monopole and quadrupole amplitudes scale approximately as

$$\nabla_\mu\!\big[K'(X)\,\nabla^\mu\varphi\big] = \Sigma,$$3

When $$\nabla_\mu\!\big[K'(X)\,\nabla^\mu\varphi\big] = \Sigma,$$4, quadrupole scalar waves remain comparable to the unscreened case (Shibata et al., 2022).

For neutron star–black hole binaries in the decoupling limit, the dipole mode is analytically and numerically found to be strongly screened. Matching near- and far-zone solutions yields

$$\nabla_\mu\!\big[K'(X)\,\nabla^\mu\varphi\big] = \Sigma,$$5

and the numerically extracted dipole amplitude scales approximately as $$\nabla_\mu\!\big[K'(X)\,\nabla^\mu\varphi\big] = \Sigma,$$6, close to the analytic $$\nabla_\mu\!\big[K'(X)\,\nabla^\mu\varphi\big] = \Sigma,$$7 expectation. By contrast, the quadrupole amplitude flattens once $$\nabla_\mu\!\big[K'(X)\,\nabla^\mu\varphi\big] = \Sigma,$$8 and is suppressed only by a factor $$\nabla_\mu\!\big[K'(X)\,\nabla^\mu\varphi\big] = \Sigma,$$9 relative to FJBD in the simulated regime (Cayuso et al., 2024).

For equal-mass binary neutron stars, kinetic screening acts non-monotonically on the dominant $K'(X)$0 scalar radiation. When $K'(X)$1, the amplitude is suppressed relative to FJBD and scales as

$K'(X)$2

whereas for $K'(X)$3 it is amplified above FJBD, with approximate scaling

$K'(X)$4

For unequal masses, a scalar dipole re-emerges and grows linearly with mass asymmetry, while quadrupolar screening remains close to the equal-mass case down to mass ratios $K'(X)$5 (Cayuso et al., 1 May 2026).

Fully relativistic simulations of binary neutron star mergers in a related screened scalar–tensor theory reported “no evidence of kinetic screening” for the dominant scalar quadrupole: screening suppresses the dipole but not the dominant quadrupole scalar mode, which can be as large as or larger than in FJBD for the MeV-range strong-coupling scales that were simulated (Bezares et al., 2021). This is not a contradiction of the compact-binary literature as a whole; rather, it indicates model dependence and wavelength dependence. The combined evidence suggests that radiative screening is less universal than static screening, and that the ratio $K'(X)$6 is often more informative than the existence of a screening radius alone.

A separate robustness problem concerns matter couplings. In neutron star solutions with derivative self-interactions in both the action and the matter coupling, kinetic screening is found to prevail over Damour–Esposito-Farèse scalarization, while kinetic couplings to matter can further suppress scalar gradients inside the star. Fine tuning the kinetic matter coupling against the derivative self-interactions can partially cancel the latter and weaken screening inside matter sources, providing a new mechanism for partial unscreening in the stellar interior (Lara et al., 2022). This suggests that screened exterior fields do not by themselves determine the strength of interior deviations from general relativity.

3. Ferroelectrics: kinetic screening as grain-boundary charge rearrangement

In polycrystalline ferroelectric thin films, “kinetic screening” is used in a different but precise sense: mobile charges slowly rearrange to screen bound charges at grain boundaries, and this time-dependent screening modifies the kinetics of domain nucleation at those boundaries. The mechanism is described as a “bulk imprint mechanism” because it is governed by grain boundaries inside the ferroelectric rather than solely by interface charges at the dead layer or electrode (Tian et al., 15 Dec 2025).

The model distinguishes polarization-bound charge

$K'(X)$7

from free charge density

$K'(X)$8

coupled through Poisson’s equation

$K'(X)$9

Polarization dynamics follow a time-dependent Ginzburg–Landau equation,

$|X|$0

while donor and acceptor trap populations obey thermally activated rate equations. For ionized donors,

$|X|$1

with an analogous equation for acceptors (Tian et al., 15 Dec 2025).

The physical picture is state dependent. Before imprint, bound charge at grain boundaries produces intense local fields and promotes nucleation there. During imprint, electrons or holes are trapped primarily on grain boundaries, driven by the attraction of bound charges. After imprint, for the same poling state, trapped charges screen the bound charge, weaken the local field, increase the nucleation barrier, and raise the effective coercive field. For the opposite poling state, the existing trapped charges enhance the local field after polarization reversal and assist nucleation and growth. The macroscopic coercive-field shift follows

$|X|$2

with thermal acceleration reflected in increasing $|X|$3 and decreasing $|X|$4 as temperature increases (Tian et al., 15 Dec 2025).

The paper contrasts this bulk mechanism with interface charge screening (ICS). ICS models reproduce logarithmic time dependence and thermal acceleration but tend to give symmetric branch shifts unless additional assumptions are added, whereas grain-boundary charge screening naturally yields asymmetric branch shifting because it directly reshapes the local nucleation landscape at specific sites (Tian et al., 15 Dec 2025). A plausible implication is that in fine-grained hafnia-based ferroelectrics, screening should be treated as a microstructurally localized kinetic process rather than only as a uniform internal-bias field.

4. Correlated electrons and impurity transport

In correlated-electron systems, “kinetic screening” can refer to screening generated by low-energy charge excitations whose binding is itself kinetic-energy driven. In the doped Hubbard model,

$|X|$5

the paper identifies nearest-neighbor doublon–holon binding as a kinetic effect: there is no explicit attractive interaction between doublon and holon on different sites, but their adjacent configuration lowers the energy by enabling virtual hopping processes of order $|X|$6 (Kapetanović et al., 2024).

Upon doping, the low-energy excitation manifold expands to include holon–doublon–holon trions. To characterize how nonlocal Coulomb repulsion $|X|$7 is absorbed into an effective local Hubbard interaction $|X|$8, the paper uses the variational relation

$|X|$9

with screening factor

$K(X)$0

At half filling and strong coupling, $K(X)$1 is associated with a nearest-neighbor exciton of energy $K(X)$2; upon doping, trion-like complexes lower the relevant excitation energy toward $K(X)$3, and $K(X)$4 over a broad range of fillings (Kapetanović et al., 2024). In this usage, kinetic screening means that hopping-generated bound states create low-energy charge fluctuations that more effectively screen the local repulsion.

A distinct transport setting appears in dilute magnetic alloys with strong spin–orbit coupling and Kondo screening. There the term denotes the incorporation of Kondo-screened resonant impurities into a quantum kinetic theory. The collision kernel is written non-perturbatively in terms of the impurity T-matrix and includes a contribution proportional to the momentum derivative of the impurity scattering S-matrix. This yields corrections to spin diffusion and spin-charge conversion and captures the side-jump process without relying on the Born approximation, which fails for resonant scattering (Huang et al., 2020). The authors find a large zero-temperature spin Hall conductivity and a transverse spin diffusion mechanism that modifies the standard Fick’s diffusion law. This suggests that, in impurity problems, “screening” may refer less to suppression of a force than to many-body dressing of the scatterer that reorganizes transport coefficients.

5. Cosmology, dark sectors, and observable screening

In cosmology, the phrase appears in at least three separate senses.

First, in the dark-photon model with kinetic mixing

$K(X)$5

the resonant conversion of CMB photons into dark photons in an inhomogeneous plasma produces a direction- and frequency-dependent optical depth. Resonance occurs when

$K(X)$6

and in the small-$K(X)$7 limit the conversion probability is

$K(X)$8

Because converted dark photons are invisible to the detector, this acts as an attenuation of the CMB, termed “patchy dark screening” (Pîrvu et al., 2023). Here “kinetic screening” is literal kinetic mixing between gauge fields.

Second, the reionization literature uses “patchy screening” for anisotropic Thomson optical depth $K(X)$9, reconstructed from CMB polarization and cross-correlated with a kSZ-squared field. The paper describes this joint use of the kinetic Sunyaev–Zel’dovich effect and patchy screening as a probe of the reionization-era ionized gas distribution, with sensitivity to the electron-density bispectrum and especially to the first half of reionization (Kramer et al., 13 Jan 2025). This is not a screening mechanism in modified gravity; it is an observable screening of primary anisotropies by free electrons.

Third, in Kinetic Field Theory applied to modified gravity, screening mechanisms are encoded phenomenologically through a scale- and time-dependent effective coupling

$k$0

The paper adopts the interpolation

$k$1

with $k$2 the screening scale and parameters controlling amplitude and transition sharpness (Oestreicher et al., 2024). Kinetic screening models such as k-mouflage are then represented by specific choices of $k$3, and their impact on the non-linear matter power spectrum is computed within a mean-field KFT approximation. This suggests that, in large-scale-structure applications, the microscopic derivative mechanism can often be traded for an effective screened coupling.

6. Plasma and kinetic-theory usages

In QCD kinetic theory, screening regulates soft gluon exchange in collision integrals. The vacuum $k$4-channel singularity is replaced by a medium-screened propagator. Standard simulations often use a Debye-like screening prescription, while the cited work implements the isotropic hard-thermal-loop (HTL) matrix element in the elastic gluon–gluon collision term (Boguslavski et al., 2024). The Debye mass is

$k$5

and the isotropic HTL propagator introduces explicit $k$6 dependence and both longitudinal and transverse sectors. For isotropic systems, the evolution is nearly unchanged, but in a Bjorken expanding plasma the choice of screening prescription changes several moments of the distribution and decreases the maximum pressure anisotropy by up to 50% (Boguslavski et al., 2024). In this context, kinetic screening means medium-regulated soft scattering implemented directly inside the kinetic collision operator.

A different plasma usage appears in stellarator neoclassics, where the “kinetic contribution of the electrons” to the ambipolar radial electric field can bring the plasma close to impurity temperature screening. The ambipolar field is determined from transport coefficients in

$k$7

and, including electrons,

$k$8

At very low collisionality, electron transport coefficients become large enough that $k$9 remains negative but small, so the ion-temperature-gradient-driven outward impurity flux can dominate or nearly cancel the inward $r_*$0 pinch (Velasco et al., 2017). The same paper shows that when $r_*$1 is small, tangential magnetic drift must be retained in order to compute the tangential electric field $r_*$2, which can be larger than previously expected and can strongly affect impurity transport (Velasco et al., 2017).

These two plasma usages are technically unrelated, but they share a structural feature: screening is not imposed as a static constitutive relation; it emerges from kinetic transport equations and their collision or drift terms.

7. Conceptual unification and recurrent misconceptions

Across these literatures, kinetic screening is not a single mechanism but a family resemblance. The repeated motifs are a dynamical screening variable, a self-consistent feedback loop, and a suppression or reshaping of effective interactions that depends on derivatives, trapping kinetics, or transport rather than only on static charge rearrangement. In gravity, the relevant object is typically $r_*$3, a screening radius $r_*$4, or a multi-field gradient contribution such as $r_*$5 (Cayuso et al., 1 May 2026, Brax et al., 14 Mar 2026). In ferroelectrics it is the slow trap population $r_*$6 and the induced screening of grain-boundary depolarization fields (Tian et al., 15 Dec 2025). In correlated systems it is the kinetic-energy stabilization of bound complexes that enhance low-energy charge susceptibility (Kapetanović et al., 2024). In cosmology it may literally be kinetic mixing or reconstructed optical-depth modulation (Pîrvu et al., 2023, Kramer et al., 13 Jan 2025).

One common misconception is to treat “screening” as automatically implying suppression of all observable effects. The compact-binary literature shows that kinetic screening can suppress dipole radiation while leaving quadrupole radiation only mildly suppressed or even enhanced, depending on the ratio of screening scale to wavelength (Cayuso et al., 2024, Cayuso et al., 1 May 2026, Bezares et al., 2021). Another misconception is to equate localization with screening: the multi-field dilaton analysis shows that a thinner shell need not imply a smaller exterior charge, and conversely a strongly suppressed charge can arise without thin-shell localization (Brax et al., 14 Mar 2026). In materials, an analogous simplification would be to identify imprint solely with interface screening, whereas the grain-boundary model attributes it to localized, time-dependent screening that alters nucleation kinetics rather than merely shifting an average internal field (Tian et al., 15 Dec 2025).

A plausible implication is that the most useful definition of kinetic screening is operational rather than ontological: it refers to cases in which the screened response cannot be specified without solving a kinetic, derivative, or transport problem. Under that broad definition, the term legitimately spans modified gravity, condensed matter, plasma physics, ferroelectrics, and cosmological radiative transfer, while retaining a common emphasis on dynamical self-consistency.