Delayed hadrons in extensive air showers: Evidence for the iron-group nuclei in primary cosmic-ray flux at energies ∼1013-1015eV

The distribution of arrival time of energetic hadrons in the near-core region of air showers of energies ∼${10}^4$-${10}^6$ GeV relative to the shower front has been studied experimentally at mountain altitude. The observed rate of hadron events with (i) energy >50 GeV in the calorimeter, (ii) associated shower particle density >18 ${\mathrm{m}}^{-2\mathrm{}}$, and (iii) a signal ≥5 equivalent particles in a plastic scintillator $T_3$ of area 0.54 ${\mathrm{m}}^2$ placed under 220 g ${\mathrm{cm}}^{-2\mathrm{}}$ of absorber in the calorimeter is found to be 1.85×${10}^{-3\mathrm{}}$ ${\mathrm{m}}^{-2\mathrm{}}$ ${\mathrm{sr}}^{-1\mathrm{}}$ ${\sec }^{-1\mathrm{}}$. Of these events a fraction (0.55±0.05)% have shown the signal from $T_3$ to be delayed by 15 nsec or greater relative to shower particles. Monte Carlo simulations of experimental observations have shown that these requirements on energy and shower density enhance the sensitivity of the observed rate to the contributions due to showers initiated by heavy nuclei. Calculations also show that observed delayed hadrons are mostly associated with showers due to heavy nuclei. For interpretation of observed features two models for primary composition have been considered. In the first model the power-law spectra for protons and lighter nuclei are assumed to have spectral index ${\gamma }_p$ and the heavy (iron group) nuclei the index ${\gamma }_{\mathrm{Fe}}$. An agreement between the expectation and observation requires the values of ${\gamma }_p$ and ${\gamma }_{\mathrm{Fe}}$ to be significantly different as -2.68 and -2.39 in the energy range ∼${10}^3$-${10}^6$ GeV. In the second model the spectral index $\gamma$ is assumed to be the same for all components and the spectra steepen by 0.5 at the same rigidity $R_c$. It is found that the values of $\gamma$ and $R_c$ should be -2.55 and ${10}^5$ GV/c, respectively, to match the observations. It is concluded that a successful understanding of experimental observations requires a relative change between the energy spectra of protons and heavy nuclei in the energy range ∼${10}^4$-${10}^6$ GeV, which would make the proportion of iron-group nuclei about 40% of the primary flux at these energies.