The Basic Atomic Physics of Warm Absorbers

Introduction

The term "Warm Absorbers" is simply short-hard for gas which lies along the line of sight and which is photoionized by intense UV/X-ray radiation field. It is commonly used when the ionization structure of the illuminated gas is such that a significant fraction of the the abundant elements which have K-shell photoelectric absorption edges in the X-ray band (i.e. C through to Fe) have lost one or more of their outer electrons (i.e. are ionized). Under suitable conditions this warm absorber will therefore imprint characteristic spectral features on the observed UV/X-ray continuum during the passage of this radiation through the gas. For historical (primarily observational) reasons the characteristic spectral features imprinted on the observed spectrum are often distinguished as being of two types: photoelectric (or bound/free) absorption "edges" (as mentioned above) resonant absorption/emission lines (or resonant scattering lines) However it is important to realise that these are actually simply different aspects of the same process: the absorption of a photon resulting in an electron being raised to a higher energy level

The term "warm absorber" was first introduced by Halpern (1984) as a shorthand to denote the fact that the electron temperature (e.g. at the illuminated face of the material; typically <105K) is much lower than in the case of collisionally-ionized ("thermal") gas with a similar level of ionization. The term "Lukewarm Absorber" has also been introduced (first by Warwick, Done & Smith 1995; but also by Kraemer et al 2000), again simply as short-hand, to denote photoionized gas was a lower degree of mean ionization than a "classic" Warm Absorber. However there is no concensus on the use of the term. For instance, somewhat confusingly, the electron temperature at the illuminated face of the photoionized material in NGC 4151 considered by Warwick, Done & Smith (1995) is also ~105 K (but the mean ionization state is lower since the total column density is larger), whilst the electron temperature at the illuminated face of the photoionized material in NGC 3227 considered by Kraemer et al (2000) is only ~2x104 K (but the column density, ~2x1021 cm-2, within the range covered by "classic" warm absorbers).

In the case the bound/free transitions, the electron is raised to a sufficiently large energy that it is able to escape from the ion. The ion now has one fewer electron (& hence the level of ionization increases by one; e.g. OVI becomes OVII etc). As there can be considered to be an infinite number of levels above the photoelectric absorption threshold energy the absorption of any photon with an energy greater than this threshold will result in this process. Thus there will be a "sharp" drop in the number of photons seen in the observed spectrum at the theshold energy. The probability of such a transition is a fairly strong function of energy, with the cross-section for photoelectric absorption decreasing as one moves to higher and higher energies, the end result is an "edge" (a sharp drop in the observed spectrum followed by a gradual recovery as one moves to higher energies) imprinted on the observed spectrum.

In the case of resonant absorption/scattering, the electron is retained by the ion, and is simply raised to a higher ("vacant") energy level (the ion is "excited"). Thus in this case only photons of exactly the right energy (the difference between the lower/initial and upper/final energy levels of the electron) will be absorbed. However since there are a large number of upper/final energy levels available, a series of narrow/discrete absorption "lines" (of increasing energy) will be imprinted on the observed spectrum. However, as above, for a given lower/initial energy level, the cross-section for the transitions decreases as one considers transitions to higher and higher upper/final energy levels. Thus the equivalent width (effectively the number of incident photons removed while traversing the warm absorber) of each of the narrow/discrete absorption lines in the series will decrease as one moves towards higher energies.

Again it should be noted that these two processes are essentially the same. The cross-section for an electron being raised to some upper energy levels decreases the larger the difference between the initial and final level. This effect continues when the final level is at and beyond the (bound/free) ionization energy. However, despite increasingly small cross-sections for any given final level, as noted above, at energies above the threshold there are an infinite number of possible upper/final energy levels. This is the underlying cause of the position and shape of the absorption edge. Indeed under certain circumstances, the superposition of a large number of resonant absorption lines can "eat into" the edge, making it appear to be at a lower energy than it actually is.

The Characteristics of Edges

More on Resonant Absorption/Scattering Transitions

incomplete