Degradation of resolution in a homogeneous dual readout hadronic calorimeter

Abstract

If the response to a hadronic shower in a semi-infinite uniform calorimeter structure is $S$ relative to the electronic response, then $S/E = [\fem + (1-\fem)(h/e)]$, where $E$ is the incident hadron energy, $\fem$ is the electronic shower fraction, and $h/e$ is the hadron/electron response ratio. In conventional calorimeters the energy resolution is dominated by the stochastic variable $\fem$, whose broad, skewed pdf has an energy-dependent mean. The slow increase of the mean with $E$ is responsible for response nonlinearity and the skewness results in a non-Gaussian response. If the cascade is observed in two channels with different values of $h/e$ (typically scintillator($S$) and Cherenkov ($C$)), $\fem$ can be eliminated. An energy estimator, linear in $C$ and $S$, is obtained which is proportional to the incident hadron's energy. The resolution depends upon the contrast in $h/e$ between the two channels. The Cherenkov $h/e$ will be 0.20--0.25. In sampling calorimeters, $h/e$ can be increased to about 0.7 by arranging for preferential absorption of the electromagnetic (EM) shower energy in the absorber (decreasing $e$) and using a hydrogenous detector (organic scintillator) to enhance $h$ through the contribution of recoil protons in $n$--$p$ scattering. Neither mechanism is available in a homogeneous crystal or glass scintillator,\rm where $h/e$is expected to be in the vicinity of 0.4 because of invisible hadronic energy loss and other effects. The $h/e$ contrast is very likely too small to provide the needed energy resolution. We support this conclusion with simple Monte Carlo simulations.