2000ApJ...544..242B

The existence of large quantities of gas that have cooled and dropped out of the hot phase in massive elliptical galaxies, groups, and clusters is the key prediction of the inhomogeneous cooling flow scenario. Using spatially resolved, deprojected ROSAT Position Sensitive Proportional Counter (PSPC) spectra of 10 of the brightest cooling flow galaxies and groups with low Galactic column densities, we have detected intrinsic absorption over energies ∼0.4-0.8 keV at the 2 σ/3 σ level in half of the sample. Since no intrinsic absorption is indicated for energies below ∼0.4 keV, the most reasonable model for the absorber is collisionally ionized gas at temperatures T=10^5–6 K with most of the absorption arising from ionized states of oxygen but with a significant contribution from carbon and nitrogen. The soft X-ray emission of this warm gas can also explain the sub-Galactic column densities of cold gas inferred within the central regions of most of the systems. (This could not be explained by an absorber composed only of dust.) Attributing the absorption to ionized gas reconciles the large columns of cold H and He inferred from Einstein and ASCA with the lack of such columns inferred from ROSAT.

Within the central ∼10-20 kpc, where the constraints are most secure, the mass of the ionized absorber is consistent with most (perhaps all) of the matter deposited by a cooling flow over the lifetime of the flow. Since the warm absorber produces no significant H or He absorption, the large absorber masses are consistent with the negligible atomic and molecular H inferred from H I and CO observations of cooling flows. It is also found that if T≳2x10⁵ K, then the optical and far-ultraviolet emission implied by the warm gas does not violate published constraints. An important theoretical challenge is to understand how the warm temperature is maintained and how the gas is supported gravitationally, and we discuss possible solutions to these problems that would require fundamental modification of the standard cooling flow scenario. Finally, we discuss how the prediction of warm ionized gas as the product of mass dropout in these and other cooling flows can be verified with new Chandra and XMM observations.