Next Contents Previous

5. MID-INFRARED DIAGNOSTICS TO IDENTIFY THE NATURE OF GALAXIES

The first such tool is one that combines ISO and IRAS information in the so-called ISO-IRAS color diagram. It compares the [6.75 µm] / [15 µm] ratio from ISOCAM to the [60 µm]/[100 µm] ratio from IRAS (see fig. 3a). It is a first step in assessing the nature of ISO galaxies. For a large range of [60 µm] / [100 µm] colors, the [6.75 µm] / [15 µm] color is roughly constant. This is the space occupied by normal star forming galaxies. It is only beyond an IRAS color geq -0.2 that the [6.75 µm] / [15 µm] color decreases. Only blue compact, interacting or starburst galaxies occupy that part of the diagram, an expected fact from the previous discussion: the radiation field is high enough that the small grain continuum has been shifted into the ISOCAM band.

Figure 3

Figure 3. (a-top left) The ISO-IRAS color diagram, that plots the [6.75 µm] / [15 µm] ratio versus the [60 µm]/[100 µm] ratio. Asterisks are for starburst and active galaxies, diamonds for Virgo spirals, filled circles for barred spirals and open squares for blue compact galaxies. (b-right) The Genzel et al. (1998) diagram that combines the [OIV] / [NeII] ratio with the 7.7 µm line-to-continuum ratio. AGN-dominated galaxies have small 7.7 µm L/C ratios and large [OIV] / [NeII] ratios. Known starburst are plotted as triangles, known AGNs as crossed squares and their sample of ULIRGs as filled circles. (c-bottom left) The Laurent et al. (2000) diagram for ISOCAM data, that plots the ratio of the 15 µm band to the 6.75 µm band versus the ratio of the 6.75 µm band to the 6 µm band. Galaxies from the ISOCAM central program are placed in that diagram. Large symbols represent objects with a known AGN that, expectedly, fall in the AGN corner of the diagram.

This region of the ISO-IRAS color diagram is obviously of high interest as it hosts galaxies providing most of the IR energy collected in the Universe, and it has therefore been explored in more details. Genzel et al. (1998) were very succesful in arranging a sample of 13 ULIRGs on a plot representing the ratio [OIV] / [NeII] versus the relative strength of the 7.7 µm infrared band (fig. 3b). In this diagram, the ULIRGs tend to lie close to the starburst region, but some clearly contain an energetic AGN. Similar analyzes, using only the 7.7 µm line-to-continuum (L/C) tool on larger samples were presented by lsrmg98 and Rigopoulou et al. (2000) from which they conclude that indeed the fraction of ULIRGs powered by an AGN increases with the infrared luminosity, but also that ULIRGs are predominantly (~ 70-80%) starburst powered.

This 7.7 µm L/C tool is however ambiguous as some starbursts have no infrared bands while some AGN exhibit strong bands. Thus the Laurent et al. (2000) diagnostic (fig. 3c), working on the broader ISOCAM band, uses the flux ratio of the broad 6.75 µm to the broad 15 µm band compared to the flux ratio of the broad 6.75 µm band to the narrow 6 µm band to make a finer distinction between AGN, starburst and normal star-forming regions. Given that AGNs and starbursts have very different continuum shapes, this tool is very successful in distinguishing one from the other. Since AGNs have much more flux in the 6 µm range than starburst or star-forming galaxies, a "band-less" starburst is not mistaken for an AGN. Applying this tool to ULIRGs show that a fraction larger than that identified by e.g. Rigopoulou et al. (2000) is AGN-powered. This type of method has great potential for future studies with the NGST, if its wavelength range is sufficiently extended.

Next Contents Previous