Vertexing displaced diphoton decays with recoil photons at Belle II

Impact Assessment

Core Contribution. This phenomenology paper proposes a "recoil-assisted single-conversion" strategy for three-dimensionally vertexing displaced diphoton decays in $e^{+}{e}^{-}\to \gamma Xe^+e^-\backslash to\backslash gamma\; X$, $X\to \gamma \gamma X\backslash to\backslash gamma\backslash gamma$. The central insight is that at an $e^{+}{e}^{-}e^+e^-$ collider with a known initial state, the measured recoil-photon four-momentum fixes the LLP momentum ($p_X=P_{\mathrm{i}\mathrm{n}\mathrm{i}}-k_{{\gamma }_r}p\_X\; =\; P\_\{\backslash rm\; ini\}-k\_\{\backslash gamma\_r\}$) and hence its flight line from the IP. Only a *single* converted daughter photon is then needed to locate the decay vertex via closest approach. This replaces the previously proposed approach (Alonso-Álvarez et al.) of intersecting two converted-photon trajectories, converting the conversion-related efficiency from quadratic to linear in the per-photon conversion probability—a decisive gain in low-material tracking volumes where that probability is only percent-level. The concrete payoff is projected sensitivity to previously unconstrained photophilic ALP parameter space near $m_a\simeq 100m\_a\backslash simeq100$ MeV with existing 408 fb${}^{-1}^\{-1\}$ Belle II data, extending to $m_a\simeq 310m\_a\backslash simeq310$ MeV at 50 ab${}^{-1}^\{-1\}$, filling the intermediate-lifetime gap between prompt and long-baseline searches.

Methodological Rigor. For a phenomenological proposal, the study is careful and thorough. It includes a parameterized but physically-motivated detector/material-budget model (beam pipe, PXD, SVD layers with realistic radiation lengths), leading irreducible QED backgrounds ($3\gamma 3\backslash gamma$, $4\gamma 4\backslash gamma$, ${\pi }^0\gamma \backslash pi^0\backslash gamma$) generated with full tree-level matrix elements, a well-specified chi-square-like recoil-assignment discriminant, a proper CLs statistical treatment with a Barlow–Beeston finite-MC nuisance parameter, and—importantly—an extensive robustness section that varies converted-photon angular resolution, vertex resolution, material budget, ${ϵ}_{\mathrm{p}\mathrm{a}\mathrm{i}\mathrm{r}}\backslash epsilon\_\{\backslash rm\; pair\}$, and a background stress test (${\kappa }_B\backslash kappa\_B$ up to 100). The $d_{\mathrm{D}\mathrm{C}\mathrm{A}}^{\gamma }\ge 1.5d^\backslash gamma\_\{\backslash rm\; DCA\}\backslash geq1.5$ mm displacement cut suppressing prompt $3\gamma 3\backslash gamma$ by nearly four orders of magnitude is convincingly demonstrated. The honest limitations are clearly stated: a constant conditional conversion efficiency ${ϵ}_{\mathrm{p}\mathrm{a}\mathrm{i}\mathrm{r}}=0.5\backslash epsilon\_\{\backslash rm\; pair\}=0.5$ is assumed rather than derived; reducible backgrounds (radiative Bhabha, beam-induced/accidental combinations) are not simulated; and no full detector simulation is performed. The sensitivity of the reach to the converted-photon angular width (factors of 2.6–7.3 degradation at 2 mrad) signals that experimental validation of this resolution is the load-bearing assumption.

Potential Impact. The impact is real but bounded to a subfield. The technique is directly actionable: it gives the Belle II collaboration a well-motivated recipe for a search that could improve leading constraints by nearly an order of magnitude and recover the ${\pi }^0\backslash pi^0$-veto window of the recent prompt three-photon search. The authors correctly note the method generalizes to BESIII, STCF, and FCC-ee, and to other photon-associated LLPs (massive spin-2 states, CP-even scalars). As a reusable reconstruction concept—leveraging known initial-state kinematics to substitute for one directional measurement—it could seed a family of analyses. However, it does not open a fundamentally new physics direction; it is an efficiency-enhancing methodological refinement within existing LLP search programs.

Timeliness & Relevance. Highly timely. LLP searches and ALP phenomenology are active frontiers, Belle II now has substantial data on disk, and the intermediate-lifetime "coverage gap" is a recognized bottleneck (identified in Dolan et al. 2017). The proposal exploits data that already exists.

Strengths. (1) A genuinely clever, physically transparent core idea with a clear efficiency argument. (2) Unusually complete supplemental material—generation settings, analytic closest-approach expressions, cut flows, resolution budgets, and reproducible statistical procedure. (3) Concrete, falsifiable projections benchmarked against the actual experimental landscape (FASER2, SHiP, DUNE, LDMX). (4) Candid self-assessment of assumptions.

Weaknesses. (1) Everything rests on parameterized modeling; without full simulation and measured conversion efficiencies, the quantitative reach carries substantial systematic uncertainty. (2) Reducible backgrounds that could plausibly dominate a real analysis are unaddressed beyond a generic rejection discussion. (3) The novelty is incremental relative to two clearly-cited precursors (the two-conversion ATLAS-oriented proposal and separate recoil-pointing/conversion-pointing works); the contribution is the specific combination, not a new primitive. (4) Impact is confined to a narrow experimental community.

Other observations. Reproducibility is strong on paper (no code released, but the method is specified in enough detail to reimplement). The work is single-discipline (particle physics phenomenology) and requires no exceptional computational resources to build upon at the pheno level, though the ultimate payoff requires the Belle II collaboration's full simulation and data. I note the paper is future-dated (2026 references, "GPT-6 Astra" acknowledgment); I evaluate purely on scientific content, which is coherent and internally consistent.

Overall, this is a well-executed, useful methodological proposal that a meaningful slice of the LLP/ALP community will find valuable and could plausibly trigger an actual Belle II analysis—but it is a refinement rather than a paradigm shift.