The elemental composition of solar wind with implications for fractionation processes during solar wind formation

Introduction. The elemental composition of the solar wind (SW) is known to be different from photospheric abundances. Fractionation processes (between different elements as well as between SW regimes) have to be understood and quantified to infer solar abundances from SW data that is one important objective of the Genesis mission. The Genesis spacecraft collected bulk SW and, on separate collectors, the main SW regimes: fast and slow SW and matter from coronal mass ejections (CME) [1]. Here, we present abundances of elements measured in the bulk SW and the three SW regimes that comprise a wide range of masses and ionization properties and give an overview about fractionation processes in the bulk SW as well as between the SW regimes. We then discuss implications for solar abundances of selected elements. Experimental. Na, Mg, Al ,Ca, and Cr were measured in the bulk SW and the SW regimes (except Cr) in Si collectors by backside depth profiling using secondary ion mass spectrometry (for analytical details see [2]). Fluences were calibrated against reference ion implants. The fluence of the Mg reference implant was absolutely calibrated relative to the Mg concentration in the NIST 617 glass [3]. The same technique will be applied to Ca and Al and final numbers will be available at the conference. Details on the analyses of C, N and O are given in [4, 5]. C and O fluences of the reference ion implants were calibrated against ion implants elaborately conducted at CSNSM (Orsay). Isotopically enriched source gases were used for three separate implantations of C and O as molecules or single ions at 40keV into Si. The beam intensity was monitored with a precision of a few % by Faraday cups and induced current measurement at the focal plane and at the target position. Potential interfering isotopes and molecules were shown to be negligible by high resolution mass scans performed before and after each implantation procedure. Fluences of the three implants of C and O agreed within 1%. The accuracy of the O fluence in four of the Orsay implants was further confirmed by Nuclear Reaction Analysis using a resonance in the O(p,α)N reaction that yielded O fluences in agreement within 1-2% with the expected value. Tentatively, we assume that the real N fluence of our reference implant does not differ more than ~5% from its nominal value as an external calibration is still pending. This is supported by the fact that N and O reference implants were subsequently implanted in the same target and the difference in the O fluence between the Orsay calibration and the reference implant was 5.5%. Presented noble gas data are from [6, 7] for He, Ne and Ar; Kr and Xe data are averages from [8-10].