Extended Haloscope Search and Candidate Validation near 1.036G Hz

We report a follow-up axion haloscope search near 1.036 GHz that completes and extends our previous work [Phys. Rev. X 14, 031023 (2024)], in which a portion of the HEMT-based data could not be analyzed due to unrecorded experimental information. While recovering this dataset, we identified an excess near 1.036 GHz that satisfied our candidate-selection criteria, motivating dedicated validation studies, including independent cross-checks and re-examination with the original apparatus. The excess did not persist under these investigations and was not confirmed as an axion dark-matter signal. We subsequently extended the search over a 20-MHz band surrounding the candidate using a quantum-noise-limited amplifier, achieving sensitivity close to the Dine-Fischler-Srednicki-Zhitnitsky benchmark. In the absence of a confirmed signal, we set improved 90% confidence-level upper limits on the axion-photon coupling over the frequency range 1.026-1.045 GHz. This work highlights the importance of robust candidate-validation strategies as haloscope searches approach discovery-level sensitivity.

💡 Research Summary

This paper presents a comprehensive follow‑up to the axion haloscope search previously reported in Phys. Rev. X 14, 031023 (2024). The original study left a 4‑MHz interval (1.033–1.037 GHz) unanalyzed because antenna‑coupling measurements were missing from the high‑electron‑mobility‑transistor (HEMT) data set. By exploiting the smooth frequency dependence of the loaded quality factor QL measured across the full scan, the authors reconstructed the missing coupling β through a quadratic interpolation of the neighboring points. The interpolation introduced a negligible systematic error (≤ 0.37 %) and enabled a full re‑analysis of the previously excluded band.

During this re‑analysis an excess was observed at 1.036315 GHz (axion mass ≈ 4.285857 µeV). The excess had a local statistical significance of 5.1 σ, reduced to 3.5 σ after accounting for the look‑elsewhere effect. Its spectral shape matched the expected virialized halo profile, and the inferred power corresponded to roughly 1.3 × KSVZ coupling assuming a local dark‑matter density of 0.45 GeV cm⁻³. Because the excess satisfied the predefined candidate‑selection criteria, the team launched an extensive validation program.

Two independent cross‑checks were performed. The first employed a separate haloscope system equipped with an 8‑T magnet, a 165‑mm bore, and a custom copper cavity tuned to the candidate frequency by inserting a 30‑mm alumina rod. The cavity’s form factor dropped to C≈0.12, and a single flux‑driven Josephson parametric amplifier (JPA) provided near‑quantum‑limited gain. An initial run in December 2023, using a cryogenic HEMT with a system noise temperature of ~500 mK, showed a transient excess with SNR ≈ 3.7, prompting deeper integration. After replacing the HEMT with a lower‑noise unit (Tsys ≈ 200 mK) in January 2024, the excess vanished, indicating that the earlier hint was not reproducible.

The second validation used the original apparatus, now configured with the same single JPA employed in the cross‑check. Data collected between April and June 2024, despite occasional helium‑recondensing interruptions, revealed no persistent signal with the expected axion‑like temporal or spectral stability. Consequently, the candidate was deemed a statistical fluctuation or an unidentified instrumental artifact.

To close the gap around the candidate, the authors conducted a targeted rescanning of a 20‑MHz band (1.026–1.045 GHz) using a quantum‑noise‑limited JPA receiver. Over 12 days of live time (June 20–July 10 2024) they applied the same baseline estimation, spectral preprocessing, and statistical combination pipeline as in the original work. Correlations introduced by the JPA between mirrored frequency bins were mitigated by optimal weighting, yielding an average SNR improvement of 4.1 %. Injection of synthetic axion signals demonstrated an overall analysis efficiency of 92.7 ± 0.9 %.

The rescanned spectrum contained 146 clusters exceeding 5 σ, all of which were rejected because they either had too narrow a bandwidth, lacked the expected cavity response, correlated with aerial antenna interference, or failed to persist upon re‑measurement. Sixteen clusters surpassed the 3.718 σ rescan threshold; only one persisted when the magnetic field was turned off, suggesting a magnetic‑field‑independent origin such as dark‑photon dark matter rather than a genuine axion.

System noise temperature measurements averaged ~200 mK, approaching the standard quantum limit; residual fluctuations contributed a 6.4 % fractional uncertainty, the dominant systematic in the analysis. Combining all data, the authors set 90 % confidence‑level upper limits on the axion‑photon coupling g_{aγγ} across the scanned mass range (4.244–4.322 µeV). The upper portion of the band reaches sensitivity close to the DFSZ benchmark, while the lower portion attains KSVZ‑level limits. These constraints improve upon their previous results and rank among the most stringent haloscope limits in this frequency interval.

The paper emphasizes the critical role of rigorous candidate‑validation protocols as haloscope experiments approach discovery sensitivity. By (i) recovering missing data and accurately interpolating antenna coupling, (ii) performing independent cross‑checks with a distinct apparatus and with the original setup, and (iii) leveraging a quantum‑limited JPA to push system noise toward the quantum limit, the collaboration demonstrates a robust pathway to either confirm a true axion signal or confidently exclude spurious excesses. The authors also note that the ADMX collaboration independently probed the same frequency region with comparable sensitivity and found no excess, reinforcing the conclusion that the observed 1.036 GHz feature was not of dark‑matter origin.

In summary, the work delivers (1) a complete, continuous exclusion across 1.033–1.037 GHz, (2) a thorough falsification of a statistically significant candidate, and (3) new DFSZ‑proximate limits over a 20‑MHz band, thereby setting a benchmark for future high‑sensitivity axion haloscope searches.