RESOLVING THE ξ BOO BINARY WITH CHANDRA, AND REVEALING THE SPECTRAL TYPE DEPENDENCE OF THE CORONAL “FIP EFFECT”

Abstract

On 2008 May 2, Chandra observed the X-ray spectrum of ξ Boo (G8 V+K4 V), resolving the binary for the first time in X-rays and allowing the coronae of the two stars to be studied separately. With the contributions of ξ Boo A and B to the system's total X-ray emission now observationally established (88.5% and 11.5%, respectively), consideration of mass loss measurements for GK dwarfs of various activity levels (including one for ξ Boo) leads to the surprising conclusion that ξ Boo B may dominate the wind from the binary, with ξ Boo A's wind being very weak despite its active corona. Emission measure (EM) distributions and coronal abundances are computed for both stars and compared with Chandra measurements of other moderately active stars with G8-K5 spectral types, all of which exhibit a narrow peak in EM near log T = 6.6, indicating that the coronal heating process in these stars has a strong preference for this temperature. As is the case for the Sun and many other stars, our sample of stars shows coronal abundance anomalies dependent on the first ionization potential (FIP) of the element. We see no dependence of the degree of "FIP effect" on activity, but there is a dependence on spectral type, a correlation that becomes more convincing when moderately active main-sequence stars with a broader range of spectral types are considered. This clear dependence of coronal abundances on spectral type weakens if the stellar sample is allowed to be contaminated by evolved stars, interacting binaries, or extremely active stars with log L X >29, explaining why this correlation has not been recognized in the past.