Mass and width of sigma(750) scalar meson from measurements of piN->pi(-)pi(+)N on polarized targets

Abstract

The measurements of reactions $\pi^- p_\uparrow \to \pi^- \pi^+ n$ at 17.2 GeV/c and $\pi^+ n_\uparrow \to \pi^+ \pi^- p$ at 5.98 and 11.85 GeV/c made at CERN with polarized targets provide a model-independent and solution-independent evidence for a narrow scalar state sigma(750). The original chi^2 minimization method and the recent Monte Carlo method for amplitude analysis of data at 17.2 GeV/c are in excellent agreement. Both methods find that the mass distribution of the measured amplitude $|\overline S |^2\Sigma$ with recoil transversity ``up'' resonates near 750 MeV while the amplitude $|S|^2\Sigma$ with recoil transversity ``down'' is large and nonresonating. The amplitude $|S|^2\Sigma$ contributes as a strong background to S-wave intensity I_S = (|S|^2 + |\overline S |^2)\Sigma$ and distorts the determinations of $\sigma$ resonance parameters from $I_S$. To avoid this problem we perform a series of Breit-Wigner fits directly to the measured distribution $|\overline S |^2\Sigma$. The inclusion of various backgrounds causes the width of sigma(750) to become very narrow. Our best fit with $t$-averaged coherent background yields $m_\sigma = 753 \pm 19$ MeV and $\Gamma_\sigma = 108 \pm 53$ MeV. These values are in excellent agreement with Ellis-Lanik theorem for the width of scalar gluonium. The gluonium interpretation of $\sigma(750)$ is also supported by the absence of $\sigma(750)$ in reactions $\gamma\gamma \to \pi\pi$. We also show how data on polarized target invalidate essential assumptions of past determinations of $\pi\pi$ phase shifts .