Search for axions from the Kuo-Sheng nuclear power reactor with a high-purity germanium detector

A search of axions produced in nuclear transitions was performed at the Kuo-Sheng Nuclear Power Station with a high-purity germanium detector of mass 1.06 kg at a distance of 28 m from the 2.9 GW reactor core. The expected experimental signatures were monoenergetic lines produced by their Primakoff or Compton conversions at the detector. Based on $459.0/96.3$ days of reactor ON/OFF data, no evidence of axion emissions were observed and constraints on the couplings $g_{a\gamma \gamma }$ and $g_{aee}$ versus axion mass $m_a$ within the framework of invisible axion models were placed. The KSVZ and DFSZ models can be excluded for ${10}^4\mathrm{eV}\lesssim m_a\lesssim {10}^6\mathrm{eV}$. Model-independent constraints on $g_{a\gamma \gamma }·g_{aNN}^1<7.7\times {10}^{-9}{\mathrm{GeV}}^{-2}$ for $m_a\lesssim {10}^5\mathrm{eV}$ and $g_{aee}·g_{aNN}^1<1.3\times {10}^{-10}$ for $m_a\lesssim {10}^6\mathrm{eV}$ at 90% confidence level were derived. This experimental approach provides a unique probe for axion mass at the keV-MeV range not accessible to the other techniques.

Figure 1

Schematic layout of the HPGe with its anti-Compton detectors as well as inner shieldings and radon purge system.

Figure 2

Schematic diagrams of interactions of axions with matter, via (a) Primakoff and (b) Compton conversions.

Figure 3

The variations of the Primakoff and Compton conversion cross sections with the axion mass at $E_a=2.23\mathrm{MeV}$, using the normalizations of $g_{a\gamma \gamma }=1{\mathrm{GeV}}^{-1}$ and $g_{aee}=1$. The dashed line represents the case where the atomic-screening effects are switched off ($\chi =1$).

Figure 4

(a) The after-cut ON and OFF spectra and (b) the residual spectra for the six specific channels for Period III, with the best-fit Gaussian peaks overlaid.

Figure 5

The variations of (${\Gamma }_a/{\Gamma }_{\gamma }$) with $m_a$ in both KSVZ and DFSZ invisible axion models as parametrized in Ref. [7]. The ${}^{135}\mathrm{Xe}^{*}$ and ${}^{137}\mathrm{Ba}^{*}$ channels have large uncertainties and were discarded. The width represents the variations within the four remaining channels and among the two models.

Figure 6

Exclusion plots of $g_{a\gamma \gamma }$ versus $m_a$ for $g_{aee}=0$. The limits from the KS experiment at 90% CL are denoted by KS. They are derived by fixing $({\Gamma }_a/{\Gamma }_{\gamma }{)}_{pn}$ with the KSVZ and DFSZ invisible axion models. Predicted regions of the PQWW, KSVZ, DFSZ, and HW models on the $(m_a,g_{a\gamma \gamma })$ plane are overlaid. The boundaries defined by Int, Decay, and Kine are constraints due to axions interactions with matter, decays in flights, and kinematics, respectively. The bounds marked Zn are results from Ref. [2], while the region labeled R are from previous reactor axion experiments studying ${\Gamma }_{\gamma \gamma }$ [10]. Results from the other axion experiments using different techniques [23, 24, 25, 26, 29] are displayed as colored blocks. The astrophysical bounds [9] are denoted by the striped region.

Figure 7

Exclusion plots of $g_{aee}$ versus $m_a$ for $g_{a\gamma \gamma }=0$. Similar conventions as in Fig. 6 are adopted. The KS bounds are derived by fixing $({\Gamma }_a/{\Gamma }_{\gamma }{)}_{pn}$ with the DFSZ model. The region marked R was excluded by experiment studying ${\Gamma }_{ee}$ at reactors [11]. Bounds from other experiments [23, 27, 28] are shown as colored blocks.

Figure 8

The model-independent exclusion regions of the KS experiment for (a) $g_{a\gamma \gamma }$ and (b) $g_{aee}$ on the $({\Gamma }_a/{\Gamma }_{\gamma }{)}_{pn}$ and $g_{aNN}^1$ axes. The $g_{a\gamma \gamma }$ and $g_{aee}$ limits from the leading laboratory experiments are also shown for comparisons. The ranges of validity in $m_a$ are indicated.