Search for solar axions produced in the $p(d,{}^3\mathrm{He})\mathrm{A}$ reaction with Borexino detector

A search for 5.5 MeV solar axions produced in the $p+d\to {}^3\mathrm{He}+\mathrm{A}$ (5.5 MeV) reaction was performed using the Borexino detector. The Compton conversion of axions to photons, $\mathrm{A}+e\to e+\gamma$; the axioelectric effect, $\mathrm{A}+e+Z\to e+Z$; the decay of axions into two photons, $\mathrm{A}\to 2\gamma$; and inverse Primakoff conversion on nuclei, $\mathrm{A}+Z\to \gamma +Z$, are considered. Model-independent limits on axion-electron ($g_{Ae}$), axion-photon ($g_{A\gamma }$), and isovector axion-nucleon ($g_{3AN}$) couplings are obtained: $|g_{Ae}\times g_{3AN}|\le 5.5\times {10}^{-13}$ and $|g_{A\gamma }\times g_{3AN}|\le 4.6\times {10}^{-11}{\mathrm{GeV}}^{-1}$ at $m_A<1\mathrm{MeV}$ (90% confidence level). These limits are 2–4 orders of magnitude stronger than those obtained in previous laboratory-based experiments using nuclear reactors and accelerators.

Figure 1

Ratio of the emission probabilities for axions and $\gamma$ quanta (${\omega }_A/{\omega }_{\gamma }$) in the $p+d\to {}^3\mathrm{He}+\gamma$ reaction (curve 1, left-hand scale); cross section of the Compton conversion and axioelectric effect for 5.5 MeV axions on carbon atoms for $g_{Ae}=1$ (curve 2 and 3, right-hand scale).

Figure 2

The expected number of axion decays (l) and inverse Primakoff conversions on ${}^{12}\mathrm{C}$ nuclei (2) in the 100 t per day for KSVZ axion model. Lines (3) and (4) show the corresponding curves under the assumption that axions do not decay during their flight from the Sun.

Figure 3

Energy spectra of the events and effect of the selection cuts. From top to bottom: (1) raw spectrum; (2) with 2 ms muon veto cut; (3) with events within 6.5 s of a muon crossing the SSS removed; (4) events inside FV.

Figure 4

Simulated responses to axion interactions in the Borexino IV: (1) axioelectric effect (5.49 MeV electrons), (2) Compton axion to photon conversion (electrons and $\gamma$-quanta), (3) Primakoff conversion (5.49 MeV $\gamma$-quanta), 4 decay $A\to 2\gamma$. The inset shows the corresponding responses for events reconstructed within the FV.

Figure 5

The fitted Borexino spectrum in the (3.2–8.4) MeV range. Curve 3 is the detector response function for Compton axion-photon conversion at the 90% c.l. upper limit ($S=6.9$ events).

Figure 6

The limits on the $g_{Ae}$ coupling constant obtained by (1) present work, (2) present work for $|g_{Ae}\times g_{3AN}|$, (3) reactor [40, 41] and solar experiments [15, 25], (4) beam dump experiments [42, 43], (5) ortho-positronium decay [44], (6) CoGeNT [45], (7) CDMS [46], (8) solar axion luminosity [47], (9) resonance absorption [48], (10) read giant [49]. The excluded values are located above the corresponding lines. The relations between $g_{Ae}$ and $m_A$ for DFSZ- and KSVZ ($E/N=0$)-models are shown also.

Figure 7

The limits on $g_{A\gamma }$ obtained by (1) present work (a $A\to 2\gamma$, b PC, areas of excluded values are located inside contour), (2) CTF [15], (3) reactor experiment [41], (4) beam dump experiments [42, 43], (5) resonant absorption [50], (6) solar axions conversion in crystals—[51, 52, 53], (7) CAST and Tokyo helioscope [54, 55, 56], (8) telescopes [57, 58, 59], (9) HB Stars [49], (10) expectation region from heavy axion models [10, 11, 12].