Kaonic Deuterium X-Rays Expose the K−n Threshold Interaction

Low-energy interactions between antikaons and nucleons remain difficult to calculate because QCD is non-perturbative in this regime and direct neutron targets are unavailable. Kaonic atoms offer an indirect route: after a negative kaon is captured by an atom, the strong interaction perturbs the lowest atomic levels from their electromagnetic values. Kaonic hydrogen constrained the proton channel, but kaonic deuterium has remained experimentally out of reach because its X-ray transitions have an exceptionally low yield. The SIDDHARTA-2 collaboration reports the first observation of those transitions and extracts the kaonic-deuterium $1s$ shift and width.

Core Contribution

The central result is a direct spectroscopic constraint on the $K^-d$ interaction at threshold: $\varepsilon_{1s}=-810.9\pm24.5\,(\mathrm{stat})\pm2.1\,(\mathrm{syst})\,\mathrm{eV}$ and $\Gamma_{1s}=812\pm97\,(\mathrm{stat})\pm33\,(\mathrm{syst})\,\mathrm{eV}$. The shift compares the measured transition energy with the QED expectation, $\varepsilon_{1s}=E^{\mathrm{meas}}-E^{\mathrm{QED}}$; by the convention used here, its negative value corresponds to a repulsive-type effective interaction and lower binding energy than the electromagnetic calculation.

What is new is not a revised calculation of an already measured quantity, but the experimental signal itself. Deuterium combines a proton and neutron, so this measurement supplies information that cannot be obtained from kaonic hydrogen alone. The authors position the result as the missing experimental input, alongside hydrogen, for determining isospin-dependent $K^-N$ scattering lengths. That makes it relevant to competing descriptions of the $\Lambda(1405)$ and to antikaon behavior in neutron-rich matter, although those downstream quantities are not themselves measured in this work.

Technical Approach

SIDDHARTA-2 operated at the DA$\Phi$NE collider, where $e^-$ and $e^+$ beams produce charged-kaon pairs. Its apparatus combines a cryogenic gaseous-deuterium target with an array of 384 silicon drift detectors for X-ray spectroscopy. Plastic-scintillator counters form a kaon trigger, while three veto systems reject backgrounds; the third is specifically a charged-kaon veto. The apparatus diagram makes clear that the measurement is a timing-and-background-rejection problem as much as a spectroscopy problem.

Figure 2 shows the detector arrangement around the interaction point, with the SDD array surrounding the target and the trigger and veto subsystems defining the event selection.

The selection proceeds through time correlations. The trigger time-of-flight distribution separates the charged-kaon population from minimum-ionizing particles, and the SDD timing distribution defines a region of interest synchronous with kaon triggers. Sidebands characterize electromagnetic background. The paper then models the selected energy spectrum with contributions from kaonic-deuterium transitions, kaonic atoms formed in surrounding materials—primarily Kapton—and fluorescence from Pb shielding and Au, Bi, and Ti. The electromagnetic-background spectrum is independently described by an exponential plus constant and Gaussian fluorescence components, providing a constrained account of line contamination rather than treating the spectrum as a single isolated peak.

Results and Analysis

The fitted X-ray spectrum contains the kaonic-deuterium component alongside the material and fluorescence terms, with the QED prediction for the $K_\alpha$ ($2p\to1s$) transition marked separately. The reported fit quality is $\chi^2/\mathrm{ndf}=1.12$, and the pull panel displays residuals normalized by their statistical uncertainties. That is useful evidence that the composite fit describes the binned spectrum, though it does not eliminate dependence on the assumed background and line-shape model.

Figure 3 presents that decomposition and residual test. The visual separation of components explains why timing selection alone would not suffice: target transitions sit among several apparatus-derived X-ray lines.

The extracted shift has a 24.5 eV statistical uncertainty and a 2.1 eV systematic uncertainty, while the width has 97 eV statistical and 33 eV systematic uncertainties. Statistical uncertainty therefore dominates both observables, especially the width. The result is compared against a set of theoretical calculations in the shift-width plane, including predictions with displayed theory bands where available. The comparison gives the measurement discriminatory value, but the paper does not turn that visual tension into a single selected interaction model or a reported scattering length.

Figure 4 places the measured point, with statistical and systematic uncertainties combined in quadrature for display, against the model predictions. The measurement is precise enough to make model differences experimentally consequential; its larger uncertainty in width means that conclusions should be read as constraints on a coupled shift-and-width prediction, not as an unambiguous verdict on every theory.

Scope and Caveats

This is a compelling experimental milestone because it establishes access to the previously inaccessible neutron-sensitive channel at threshold. The evidence directly supports the reported atomic-level parameters through a documented spectrum fit and timing-based selection. It does not, by itself, establish the underlying $K^-n$ amplitude, resolve the structure of the $\Lambda(1405)$, or predict neutron-star matter. Those inferences require the joint analysis with kaonic hydrogen and theoretical treatment of the deuteron and coupled hadronic channels.