Attenuation from the optical to the extreme ultraviolet by dust associated with broad absorption line quasars: the driving force for outflows

TL;DR

This study derives the EUV attenuation curve for BALQSO dust, revealing a steep, SMC-like shape that influences AGN outflows, with implications for understanding dust properties and radiation-driven gas acceleration.

Contribution

It provides the first detailed EUV attenuation curve for BALQSO dust, showing its SMC-like shape and impact on AGN outflows, challenging previous assumptions about BALQSO phases.

Findings

BAL dust causes steep, SMC-like EUV attenuation

EUV attenuation is sufficient for radiative acceleration of gas

LoBALs show negative optical attenuation, indicating scattering effects

Abstract

We derive a mean attenuation curve out to the rest-frame extreme ultraviolet (EUV) for 'BAL dust' - the dust causing the additional extinction of active galactic nuclei (AGNs) with broad absorption lines (BALQSOs). In contrast to the normal, relatively flat, mean AGN attenuation curve, BAL dust is well fit by a steeply rising, SMC-like curve. We confirm the shape of the theoretical Weingartner & Draine SMC curve out to 700 Angstroms but the drop in attenuation at still shorter wavelengths is less than predicted. The similar SMC-like attenuation curve for low-ionization BALQSOs (LoBALs) does not support the idea that they are an early phase in the life of an AGN when it is breaking out of a cocoon of star-forming dust. Although the attenuation is only E(B - V) ~ 0.03 - 0.05 in the optical, it rises to one magnitude in the EUV, which is an optimum value for radiative acceleration of dusty gas. Because the spectral energy distribution of AGNs peaks in the EUV, the force on the dust dominates the acceleration of BAL gas. Although the shape of the attenuation curve for LoBALs is similar to the shape for HiBALs, the LoBALs on average show negative attenuation in the optical. This is naturally explained if there is more light scattered into our line of sight in LoBALs compared with non-BALQSOs. We suggest that this and partial covering are causes when attenuation curves appear to be steeper in the UV than an SMC curve.