Determination of the proton-to-helium ratio in cosmic rays at ultra-high energies from the tail of the X_(rm max) distribution

We present a method to determine the proton-to-helium ratio in cosmic rays at ultra-high energies. It makes use of the exponential slope, $\Lambda$, of the tail of the $X_{\rm max}$ distribution measured by an air shower experiment. The method is quite robust with respect to uncertainties from modeling hadronic interactions and to systematic errors on $X_{\rm max}$ and energy, and to the possible presence of primary nuclei heavier than helium. Obtaining the proton-to-helium ratio with air shower experiments would be a remarkable achievement. To quantify the applicability of a particular mass-sensitive variable for mass composition analysis despite hadronic uncertainties we introduce as a metric the `analysis indicator' and find an improved performance of the $\Lambda$ method compared to other variables currently used in the literature. The fraction of events in the tail of the $X_{\rm max}$ distribution can provide additional information on the presence of nuclei heavier than helium in the primary beam.