Dust-gas separation in the planetary nebula NGC 6445

Peter A.M. van Hoof ^1,3, Griet C. Van de Steene ², Peter G. Martin ³, Stuart R. Pottasch ⁴, & Gary J. Ferland ^1,3

¹ University of Kentucky, Dept. of Physics and Astronomy, 177 CP Building, Lexington, Kentucky, USA
² Mount Stromlo Observatory, Australian National University, Private Bag, Weston Creek P.O., ACT 2611, Australia
³ Canadian Institute for Theoretical Astrophysics, McLennan Labs, University of Toronto, 60 St. George Street, Toronto, ON M5S 1A1, Canada
⁴ Kapteyn Astronomical Institute, Postbus 800, NL-9700 AV Groningen, The Netherlands

In this paper we show that the ionized region of NGC 6445 is mostly or completely free of dust grains. We argue that this cannot be due to grain destruction, but more likely resulted from the separation of the grains from the gas. We discuss our results in the context of central star evolution.

The planetary nebula NGC 6445 is classified as a Type I nebula, which are assumed to originate from massive progenitors ( ). Based on its low density, the nebula appears well evolved. Preliminary models show that the central star temperature is around 150,000 K. Both the ionized and neutral material show a high expansion velocity of 35 to 38 kms^-1. These properties are consistent with a luminous and hence high-mass central star.

Both SWS01 and LWS01 data for this nebula show evidence for the presence of cool dust. The temperature of  K is lower than what is typically observed for planetary nebulae. The ratio of the total infrared to H flux is between 15 and 35, depending on the adopted value for the extinction. The presence of strong neutral lines in the spectrum indicates that the nebula is ionization bounded. We therefore assume that no significant fraction of the Lyman continuum or Lyphotons escape the nebula. This implies that most of the grain heating can be accounted for by Ly photons. Consequently Lyman continuum photons contribute little to grain heating and most of the ionized region must be dust free.

This argument was confirmed by photo-ionization modeling of the nebula using CLOUDY. To obtain the lowest possible value for the dust temperature, without invoking the presence of dust-free gas, we assumed that all nebular material was in a thin shell near the Strömgren radius. However the resultant color temperature was still higher than the observed value. Hence dust-free material must be present in the inner regions of the nebula.

Assuming that dust formed throughout the AGB mass loss period, there are two possible explanations for this result: either the dust grains were destroyed, or the grains were separated from the gas in the inner region. If the grains were destroyed, this would affect the abundance pattern in the gas, returning it to solar values. Based on the Si II 34.8 m line, we derived a silicon abundance which shows that silicon is strongly depleted. Hence no significant grain destruction can have occurred in the inner parts of the nebula.

Consequently the dust must have been separated from the gas. However, models show that the current grain drift velocity in this nebula is less than 1 kms^-1, which is much too low to cause the dust-gas separation, even in a very old nebula. This implies that the separation must have occurred in an earlier phase of the evolution, most likely during the AGB-PN transition phase.