Strange Spins: New Calculation Maps Hidden Spin-Momentum Correlations Inside the Kaon

Deep inside every kaon, its quarks are not simply drifting about in a featureless fog. They carry spin, they carry momentum, and — according to a new theoretical study — the two are subtly entangled with one another in ways that could soon be measured at the world’s most powerful particle accelerators. A team of theorists led by Dan-Dan Cheng, Minghui Ding, Daniele Binosi and Craig D. Roberts has published, in The European Physical Journal C, the first complete symmetry-preserving calculation of the kaon’s Boer–Mulders function, a quantity that describes how the transverse spin of a valence quark inside an unpolarised meson correlates with that quark’s motion perpendicular to the meson’s direction of travel. The work extends an earlier analysis of the pion and delivers a four-part portrait of the kaon’s valence structure that is, in the authors’ words, complete at the hadron scale.

To understand why this matters, it helps to recall what a transverse momentum dependent parton distribution, or TMD, actually is. Whereas an ordinary parton distribution function tells you the probability of finding a quark carrying a given fraction x of its host hadron’s momentum, a TMD adds a second coordinate: the quark’s momentum in the plane transverse to the hadron’s motion. Together, the pair (x, k-perpendicular) provides what physicists like to call a three-dimensional tomographic image of the hadron — a sort of internal CT scan. For the kaon, the calculation delivers two such functions for each of its valence constituents: the helicity-independent distribution, which is an ordinary number density, and the Boer–Mulders function, which encodes the spin–momentum correlation. Baryon number conservation pins down the normalisation of these functions, and at the hadron scale the valence degrees of freedom must carry all of the meson’s momentum.

The Boer–Mulders function is the more mysterious of the two, and for a fundamental reason: it is odd under time reversal. In a naive picture where quarks move freely and independently, such a function must vanish identically. It can only be nonzero because of the gauge link — a Wilson line threaded through the quark correlator that ensures gauge invariance and, physically, represents the multiple gluon exchanges between the quark struck by the probe and the spectator quark left behind. The new calculation models this gauge link using an eikonal approximation, in which the highly energetic struck quark propagates through the gluon field while interacting with the spectator. Crucially, the two diagrams contributing to the unpolarised TMD cancel in this model, so the gauge link contributes exclusively to the spin-sensitive Boer–Mulders term — a clean separation that makes the physics unusually transparent.

The theoretical engine behind the computation is a symmetry-preserving treatment of a vector-times-vector contact interaction, an approach the group has refined over more than a decade. The contact interaction is not a precision instrument; it deliberately sacrifices some of QCD’s momentum-dependent complexity in exchange for algebraic simplicity, parameter-free predictions and the ability to treat many mesons and baryons within a single framework. Its great virtue here is transparency: every element of the kaon’s Bethe–Salpeter amplitude, its dressed-quark masses and its light-front wave function can be written down explicitly, so the origin of each feature in the final TMDs can be traced by inspection. As a cross-check, the authors rederived their results from the kaon’s light-front wave function via overlap representations and found complete consistency with the diagrammatic calculation — and the light-front route also makes manifest that the helicity-independent TMD, being a modulus-squared of the wave function, is a genuine positive probability density.

One of the most striking results concerns the shape of the kaon’s unpolarised TMD. For the pion, whose two valence quarks are nearly identical in mass, the distribution is symmetric about x = 1/2. The kaon is different: it pairs a light up quark with a much heavier strange antiquark, and the Higgs boson’s coupling generates a strange-quark current mass roughly twenty-seven times larger than the light-quark mass. Naively one might expect the heavier constituent to dominate. Instead, emergent hadron mass — the dynamically generated quark mass that QCD produces through its self-interactions — strongly screens this disparity. The calculated u-in-K distribution peaks at x approximately equal to 0.3 rather than 0.5, a shift the authors attribute directly to Higgs modulation of emergent mass, and the corresponding momentum fractions come out as 0.47 for the up quark and 0.53 for the strange antiquark, a ratio of about 1.13 rather than the naive 27-to-1. In the infrared, where the hadron’s structure is actually built, the running strange and up quark masses differ by only about five to four.

The Boer–Mulders function tells a complementary story. Its magnitude grows with the dressed masses of the valence quarks, which means the strength of spin–momentum correlations inside a pseudoscalar meson is itself a signal and a measure of emergent hadron mass. But the calculation also exposed a subtlety that has plagued the field: the positivity bound. Quantum mechanics requires that the Boer–Mulders function be bounded pointwise by the helicity-independent distribution, a constraint that follows from the positivity of the defining matrix elements. With a momentum-independent gluon in the gauge link, the computed kaon Boer–Mulders function violates this bound. The resolution came from building consistency between the interaction used to bind the meson and the one used to model the gauge link: by giving the exchanged gluon a realistic momentum-dependent propagator that reproduces QCD’s behaviour at ultraviolet momenta, the authors obtained a Boer–Mulders function that is softer in transverse momentum and fully compatible with positivity. The lesson is general — treatments that give the gauge link more support than the quark loop are prone to violating the bound.

Because predictions made at the hadron scale cannot be compared directly with data, the team also studied how their TMDs evolve to higher resolving scales, working with the leading moment in k-perpendicular-squared of the Boer–Mulders function. The evolution equation for the underlying twist-three correlation function contains off-diagonal terms that most practitioners have hitherto neglected. Including them — through a Gaussian ansatz controlled by a width parameter sigma — changes the picture materially: the evolved moments become sensitive to the width, the strange-in-kaon moment is suppressed relative to the up-in-kaon one, and a flavour separation emerges between the evolved kaon moments. After evolution to a scale of 2 GeV, the strange-to-up ratio settles near five to four, again reflecting Higgs modulation of emergent mass rather than the raw current-mass ratio.

The analysis also quantifies the so-called Boer–Mulders shift: the mean transverse momentum, in one direction, of quarks polarised in the perpendicular direction within an unpolarised meson. At the hadron scale, the dimensionless shift is approximately the same for pion and kaon, about minus 0.17 GeV, and after evolution it separates by flavour — minus 0.14 GeV for the up quark in the kaon versus minus 0.10 GeV for the strange antiquark, with the pion in between. The team compared their generalised shift with existing lattice QCD results for a heavy pion and found encouraging qualitative agreement, while noting that more precise lattice calculations will be needed before flavour separation in kaon-like systems can be meaningfully tested. On the experimental side, the situation remains sparse: essentially no data exist from which the pointwise behaviour of kaon TMDs could be extracted, and even the pion Boer–Mulders function is largely unconstrained.

That gap is exactly why this theoretical work arrives at an opportune moment. Next-generation facilities — the Electron–Ion Collider in the United States, the electron–ion collider planned in China, the AMBER experiment at CERN and a luminosity-upgraded Jefferson Lab — are designed to deliver the high-luminosity data needed to reconstruct hadron TMDs and to test whether certain features are universal across hadron species. The new calculation provides precisely the kind of solid, symmetry-respecting prediction such reconstructions will need as a guide. It also raises questions the authors are already pursuing: extending the contact-interaction machinery to the proton’s Boer–Mulders functions, where diquark correlations may leave novel fingerprints, and repeating the analysis with realistic, QCD-connected pion and kaon light-front wave functions that have only recently become available. If the kaon’s hidden spin–momentum architecture can indeed be imaged in three dimensions, this algebraically simple but physically incisive calculation will have drawn the first reliable sketch.

Subject of Research: Theoretical calculation of kaon transverse momentum dependent parton distribution functions, including the Boer–Mulders function, using a symmetry-preserving contact interaction

Article Title: Kaon Boer–Mulders function using a contact interaction

Article References: Cheng, D.-D., Ding, M., Binosi, D., & Roberts, C. D. (2026). Kaon Boer–Mulders function using a contact interaction. The European Physical Journal C, 86(9), Article 1029. https://doi.org/10.1140/epjc/s10052-026-16242-6

Image Credits: AI Generated

DOI: 10.1140/epjc/s10052-026-16242-6

Keywords: kaon, Boer–Mulders function, transverse momentum dependent parton distributions, quantum chromodynamics, emergent hadron mass, contact interaction, gauge link, Higgs modulation, pion, spin–momentum correlations, light-front wave function, parton distribution evolution