Experimental Search for $\pi ^-p\rightarrow \gamma \gamma n$

The only previous experimental search for the $\pi ^-p\rightarrow \gamma \gamma n$ reaction was reported by Vasilevsky et al. (11) at the JINR synchrocyclotron. The set-up of the experiment is reproduced in Figure 2.3.

Figure 2.3: JINR experimental set-up: The liquid hydrogen target $T$ is located at the center. Photon-pairs travel through the veto counters $A$ into the lead converter placed in front of counters $G_1$. The $e^+e^-$ pairs produced pass through the counters $G_1$ and $G_2$ and the plastic scintillation counters $C$ that were placed between counters $G_1$ and $G_2$. Counters $A$, $G_1$ and $G_2$ were spark chambers (11).

Negative pions of initial energy of 75 MeV were passed through moderators, a Cerenkov counter C, a pair of plastic scintillators and were stopped in a liquid hydrogen target T. Negative pions and other charged particles passing through the target were vetoed by an anti-coincident counter. The two-photon events recorded in an axially symmetric cylindrical photon-pair detector consisting of a 5 mm thick lead converter, a 17-element plastic scintillator array, and a triple layer of spark chambers A, G$_1$ and G$_2$. The counter A was used to veto charge particles coming from the target.

The experiment observed about 52,000 two-photon events with a small hydrogen target (length $\times$ diameter; 10 cm $\times$ 3 cm) and about 16,000 two-photon events with a large hydrogen target (10 cm $\times$ 8 cm). The measured two-photon spectra corresponding to the small hydrogen target in the region $\cos\theta > -0.87$ ($\theta <$ 150$^o$) is reproduced in Figure 2.4.

Figure 2.4: The opening angle spectrum for two-photon events recorded in the JINR experiment. 105 two-photon events were found in the region $-0.64\leq\cos\theta\leq 0.76$. The large peak at backward angles is due to at-rest $\pi^- p \rightarrow \pi^o n$. The weak continuum at smaller angles is due to random two-photon coincidences. The resulting limit on the $\pi ^-p\rightarrow \gamma \gamma n$ branching ratio is $\leq 5.5\times 10^{-4}$.

The JINR experiment did not record statistically significant $\gamma \gamma n$ events, but to understand the spectra obtained by the experiment, it is important to note that due to kinematical constraints, the $\pi^-p\rightarrow\pi^0n$ capture followed by $\pi^0\rightarrow\gamma\gamma$ decay yields photon-pairs with 155 $^{\circ}<\theta_{}<$ 180$^{\circ}$ whereas the reaction of interest; $\pi ^-p\rightarrow \gamma \gamma n$ capture yields photon-pairs with 0$^{\circ}$ $< \theta<$ 180$^{\circ}$. There were 105 two-photon events in the region 40 $^{\circ} < \theta <$ 150$^{\circ}$ i.e. $0.77 > \cos\theta >-0.87$. These were attributed to random $\gamma-\gamma$ coincidences. These random $\gamma-\gamma$ coincidences result from simultaneous multiple $\pi ^-$ stops. One photon each from either of the randomly coincident pions can contribute to a photon-pair via the reactions $\pi^- p \rightarrow \gamma n$ ( $\Gamma_{\gamma}$ = 0.39), and $\pi^-p\rightarrow\pi^0n$ ( $\Gamma_{\pi^o}$ = 0.61) followed by $\pi^o\rightarrow\gamma\gamma$ ( $\Gamma_{2\gamma}$ = 0.99). Thus, multiple pion stops in one beam pulse can yield a $\gamma$-ray pair with opening angles 0-180$^{\circ}$ and summed energies 106-258 MeV. Based on the observed multiple pion stops, an upper limit of BR $\leq$ 5.5$\times$10$^{-4}$ was obtained on the $\pi ^-p\rightarrow \gamma \gamma n$ branching ratio. This upper limit is an order of magnitude larger than the predicted branching ratio obtained from Beder's calculations (7).

The JINR experiment suffered from insufficient statistics, an overwhelming random coincident multi-pion background, and the inability to measure the photon's energy.