6. INTRACLUSTER PLANETARY NEBULAE

The discovery of intracluster planetary nebulae (ICPNe) in the Virgo and Fornax clusters is one of the most interesting new developments in the study of extragalactic PNe. ICPNe were first discovered by Arnaboldi et al. (1996; A96) in a survey of the Virgo galaxy NGC 4406. Spectra were taken of PNe candidates (Jacoby et al. 1990; J90) in two fields that were centered 134"E and 134"W of NGC 4406. Sixteen of the PNe candidates in the two fields have radial velocities near N4406's systemic velocity of -227 km s^-1. Three of the candidates in the W field have velocities of 1729 km s^-1, 1651 km s^-1, and 1340 km s^-1, provided the observed emission line is [OIII] 5007. At these redshifts, the [OIII] 4959 emission line falls in J90's on-band filter (_c ~ 4998 Å, FWHM ~ 30 Å), accounting for the detection of the high velocity PNe. Because the nebulae were identified with a filter centered at 4998 Å, the emission lines detected spectroscopically at ~ 5033 Å cannot be Ly from galaxies with redshifts z ~ 3.14. A96 suggested that these three nebulae are PNe from a Virgo intracluster stellar population.

Subsequent surveys of the Virgo and Fornax clusters have identified 300 emission line sources that may be ICPNe. The surveys are summarized in Table 3. Column 5 lists the survey authors' estimated fraction of the cluster's total stellar mass that is in the intracluster medium. These estimates suggest that 20% to 50% of a cluster's stellar mass may be in the intracluster medium!

Depending on the depth of the survey, 25% or more of the PNe candidates may be starbursting galaxies at z ~ 3.1 with Ly emission redshifted into the [OIII] 5007 on-band filter (cf Kudritzki et al. 2000; Krelove et al. 1999; Freeman et al. 2000). Provided there is a sufficient signal-to-noise ratio in the spectra, PNe can be distinguished from starbursting galaxies and QSOs at z ~ 3.1 by the presence of [OIII] 4959 and the absence of a continuum. At spectral resolutions of ~ 10 Å or higher, the Ly emission line will be resolved, whereas [OIII] 5007 in a PN with an expansion velocity of ~ 20 km s^-1 will be unresolved.

Several arguments suggest that a large fraction of the ICPNe candidates are in fact intracluster planetary nebulae. Freeman et al. (2001) spectroscopically confirmed 23 ICPNe in the Virgo cluster by detecting [OIII] 4959 and [OIII] 5007 at the expected wavelengths and intensity ratio. The anomalous PNLF in M87's halo has PNe that are brighter than the PNe in M87's main body, and thus are likely foreground ICPNe (Ciardullo et al. 1998). SN 1980I occurred midway between NGC 4374 and NGC 4406, showing that there are intracluster stars in the Virgo cluster (Smith, 1981). Ferguson, Tanvir, & von Hippel (1998) used HST images to detect faint intracluster stars in isolated Virgo fields. This population of (old) stars will produce planetary nebulae.

ICPNe are important for many reasons. They may provide thousands of test particles for detailed studies of the mass distribution in clusters. Their kinematics may reveal otherwise unobservable tidal streams that record tidal interactions over the last few Giga-years.

Little is known about the presence of intragroup stars in small groups of galaxies. If groups of galaxies have the same fraction of luminous intergalactic material as estimated for Virgo and Fornax, there could be hundreds of intergalactic PNe. The kinematics of these nebulae could be used to investigate the distribution of mass in sparse groups, and to study the history of tidal interactions.