This paper extends a useful trick for calculating how neutrinos change flavor as they travel through matter. The authors generalize the Jacobi diagonalization method so it can handle matter effects that appear in any single diagonal entry of the neutrino interaction matrix. They then apply the result to a type of beyond‑the‑Standard‑Model interaction called scalar non‑standard interactions (SNSI), and they show these can produce an energy‑independent resonant enhancement they call the scalar MSW (SMSW) effect.

The Jacobi diagonalization method works by rotating the 3×3 neutrino Hamiltonian step by step until the off‑diagonal entries are small enough to ignore. The rotations give “effective” mixing angles and mass‑squared differences that describe oscillations in matter. A key point in this work is that the usual parametrization of the neutrino mixing matrix (the PMNS matrix) is not the only convenient choice. The authors pick different Euler‑rotation orders depending on which diagonal of the matter potential is nonzero. That choice makes the Jacobi method usable even when new physics moves the matter effect away from the usual electron (ee) entry.

The authors focus on SNSI cases where only one diagonal coupling is dominant at a time: the ee, μμ, or ττ entry. For each case they derive compact analytic mapping formulas that give the effective mixing angles and mass splittings in matter. These formulas are designed to remain valid even for large matter potentials and to capture resonant effects analogous to the standard Mikheyev–Smirnov–Wolfenstein (MSW) resonance. Because the SNSI resonance is driven by a scalar potential, they call it the scalar MSW (SMSW) effect and note it is energy independent, unlike some other resonances.

There are important caveats. The paper treats only one dominant SNSI coupling at a time. When standard charged‑current matter effects (the usual A_CC potential from electrons) and SNSI act on different diagonal entries, the authors must treat the standard potential as a small perturbation. That approximation works well for the ee SNSI case where both effects sit in the ee entry. It is less accurate for the μμ and ττ cases because the standard potential then appears in a different entry and cannot be kept exactly; the authors apply first‑order perturbative corrections to improve the results.