Polarized SIDIS: comment on purity method for extraction of polarized quark distributions

The role of hadronization mechanism in polarization phenomena in semi inclusive deep inelastic scattering (SIDIS) and a purity method for extraction of polarized distribution functions are discussed. According to the Monte Carlo (MC) event generator, on which this method is based, hadrons can be produced via quark (diquark) fragmentation or light cluster decays. In contrast, the purity method assumes that only quark fragmentation gives contribution to hadron production in the current fragmentation region. The ignorance of contributions from diquark fragmentation and cluster decays to asymmetry can be source of incorrect values of polarized quark distributions extracted by the purity method. The investigation of deep inelastic scattering (DIS) of leptons is an excellent source to study the internal structure of nucleon. We know with good precision the unpolarized parton distribution functions. Many efforts has been done to investigate also the polarized parton content of nucleon. In this case, for example, to separate the individual polarized parton distributions in the LO we still need more data (g1 measurement at very high Q, or asymmetry measurements in neutrino DIS on polarized target) or some additional input like information from hyperon β-decay. Experimental progress in recent years allows to investigate SIDIS. There is hope that, for example, the measurement of different hadron production asymmetries on proton and neutron targets will allow a further flavor separation of polarized quark distributions. Recently the new preliminary data from HERMES on quark flavor separation has been presented [2]. The LO analysis of semi inclusive DIS has been done by using the purity method and suggests that “the strange sea appears to be positively polarized” in contrast to generally accepted negatively polarized strange sea scenario at LO (see discussion in [3]). An “explanation” of HERMES result of a positively polarized strange sea is ∗The main idea of this note has been presented in [1].