Dust outflows from quiescent spiral disks

We have conducted a search for "dust chimneys" in a sample of 10 highly-inclined spiral galaxies (i = 86 - 90 degrees) which we had previously observed in the H alpha emission line (Rand 1996). We have procured B-band CCD images for the purpose and employed unsharp-masking techniques to accentuate the structure of the dust lane. A scattering+absorption radiation transfer model enabled us to separate 5 galaxies from the sample which are sufficiently inclined (i > 87 degrees) for us to reliably identify and quantify dust clouds residing at over 2 scale-heights above the disk. Three of these galaxies possess numerous curvi-linear chimney structures stretching up to 2 kpc from the midplane and the fraction of total galactic dust contained in such structures is of order 1%. Optical extinction offers a lower limit to the amount of dust contained in the extraplanar layer but, by examining the transparent submm thermal emission from NGC 891, we fix an upper limit of 5%. Our results are consistent with a similar recent study by Howk & Savage (1999) which indicates that about half of quiescent spiral disks possess detectable dust chimneys. We have compared our optical images with the corresponding H alpha emission-line radiation. We do not find a detailed spatial correspondance between dust chimneys and either sites of recent star-formation or the extraplanar diffuse ionized gas. This is somewhat surprising given that FIR-bright galaxies, such as M 82, are known to entrain dust at the working surface of the starburst-driven outflow (traced in H alpha). It is possible a global correlation exists, with disks experiencing overall higher rates of star-formation also possessing the greatest number of chimneys. This may indicate a timescale difference between the two phenomena with the H alpha phase lasting similar to 10(6) yr but chimneys requiring similar to 10(7) yr to form. Additionally, we have investigated the edge-on disk NGC 55 which, being ten times closer than galaxies in our main sample, allows us to examine in greater spatial detail the relationship between chimneys and recent star-formation. Our discussion touches upon high latitude dust and supershells observed in the Milky Way. We rule out quiescent disks as prolific sources of intergalactic grains and metals but note that the rate at which dust is expelled from the main dust layer is comparable to the rate at which it is produced by disk stars (suggesting that it may be an important regulatory process).