The PHENIX experiment has measured the suppression of semi-inclusive single high transverse momentum pi^0's in Au+Au collisions at sqrt(s_NN) = 200 GeV. The present understanding of this suppression is in terms of energy-loss of the parent (fragmenting) parton in a dense color-charge medium. We have performed a quantitative comparison between various parton energy-loss models and our experimental data. The statistical point-to-point uncorrelated as well as correlated systematic uncertainties are taken into account in the comparison. We detail this methodology and the resulting constraint on the model parameters, such as the initial color-charge density dN^g/dy, the medium transport coefficient <q^hat>, or the initial energy-loss parameter epsilon_0. We find that high transverse momentum pi^0 suppression in Au+Au collisions has sufficient precision to constrain these model dependent parameters at the +/1 20%-25% (one standard deviation) level. These constraints include only the experimental uncertainties, and further studies are needed to compute the corresponding theoretical uncertainties.
$\pi^0$ $0-5\%$ centrality
A three-dimensional (3D) correlation function obtained from mid-rapidity, low pT pion pairs in central Au+Au collisions at sqrt(s_NN)=200 GeV is studied. The extracted model-independent source function indicates a long range tail in the directions of the pion pair transverse momentum (out) and the beam (long). Model comparisons to these distensions indicate a proper breakup time \tau_0 ~ 9 fm/c and a mean proper emission duration \Delta\tau ~ 2 fm/c, leading to sizable emission time differences (<|\Delta \tau_LCM |> ~ 12 fm/c), partly due to resonance decays. They also suggest an outside-in 'burning' of the emission source reminiscent of many hydrodynamical models.
1D correlation function. Systematic errors are less than the statistical errors.
Experimental correlation moments $R^0(q)$ Data. Systematic errors are less than the statistical errors.
Experimental correlation moments $R^0(q)$ Fit. Systematic errors are less than the statistical errors.
Forum