750 GeV diphoton resonance in a visible heavy QCD axion model

In this paper, we revisit a visible heavy QCD axion model in light of the recent reports on the 750 GeV diphoton resonance by the ATLAS and CMS experiments. In this model, the axion is made heavy with the help of the mirror copied sector of the Standard Model, while the successful Peccei-Quinn mechanism is kept intact. We identify the 750 GeV resonance as the scalar boson associated with spontaneous breaking of the Peccei-Quinn symmetry, which mainly decays into a pair of axions. We find that the mixing between the axion and $\eta$ and ${\eta }^{\prime }$ plays important roles in its decays and the resultant branching ratio into two photons. The axion decay length can be suitable for explaining the diphoton excess by the di-axion production when its decay constant $f_a\simeq 1\mathrm{TeV}$. We also find that our model allows multiple sets of the extra fermions without causing the domain wall problem, which is advantageous to explain the diphoton signal.

Figure 1

The mixing angles ${ϵ}_{a\eta }$ (blue) and ${ϵ}_{a{\eta }^{\prime }}$ (red) as a function of $m_a$ for $f_a=1\mathrm{TeV}$. The dashed curves indicate the negative mixing angles. Here, we take $\sin \theta \simeq -1/3$ and $f_0\simeq f_{\pi }$.

Figure 2

(Left panel) The relevant partial decay widths of the axion into $\gamma \gamma$ (light blue), $3{\pi }^0$ (dark blue), ${\pi }^0{\pi }^{+}{\pi }^{-}$ (red), $\rho +\gamma$ (black), $\eta +2{\pi }^0$ (yellowish green), $\eta +{\pi }^{+}+{\pi }^{-}$ (green), and ${\gamma }^{\prime }{\gamma }^{\prime }$ (black dashed and gray band) for $f_a=1\mathrm{TeV}$. Here we take the $\eta$–${\eta }^{\prime }$ mixing angle to have $\sin \theta \simeq -1/3$. The blue dashed line indicates $\mathrm{\Gamma }(a\to \gamma \gamma )$ without the $a-\eta$ mixing. Here we can see that the dip around 650 MeV comes from the phase cancellation. The orange dashed curve represents the decay width of the axion into $\gamma \gamma$ only from the intrinsic axion anomalous coupling to $\gamma$ in Eq. (20). (Right panel) The partial decay widths for vanishing intrinsic axion anomalous couplings to $\gamma$ and ${\gamma }^{\prime }$.

Figure 3

The branching ratio of the axion into $\gamma \gamma$ for $f_a=1\mathrm{TeV}$. We set the $\eta$–${\eta }^{\prime }$ mixing angle to be $\sin \theta \simeq -1/3$. The central curve in the band corresponds to $\mathrm{BR}(a\to \gamma \gamma )$ when we estimate the widths into $3\pi$ and $\eta +2\pi$ by using the approximation given in the Appendix, while the upper and the lower curves correspond to the width multiplied by a factor of $1/3$ and 3, respectively. We assume that the $a\to {\gamma }^{\prime }{\gamma }^{\prime }$ width is zero due to the anomaly cancellation.

Figure 4

The boosted decay length of an axion for $f_a=1\mathrm{TeV}$. We assume that the axion is produced from a two-body decay of a resonance with mass 750 GeV. Hence, the boost factor $\gamma \simeq 325\mathrm{GeV}/m_a$. The shaded blue region indicates that the boosted decay width exceeds 50 cm.

Figure 5

The favored decay constant to reproduce the diphoton excess via the di-axion signal. Each colored band corresponds to the decay constant that gives the appropriate cross section times branching ratios in Eq. (12). For each band, we adopt the decay widths of $a\to 3\pi$, which is scaled by a factor of $1/3$ (orange), 1 (blue) and 3 (green), respectively. The shaded regions are excluded from the condition, $c\tau \gamma >50\mathrm{cm}$. The red, yellow and blue regions correspond to where $\mathrm{\Gamma }(a\to 3\pi )$ is multiplied by $1/3$, 1 and 3, respectively.