[Paper Review] Towards a physical model of dust tori in Active Galactic Nuclei - Radiative transfer calculations for a hydrostatic torus model

M. Schartmann, Klaus Meisenheimer+3 more|ArXiv.org|Apr 5, 2005

Astrophysics and Star Formation StudiesPhysics and Astronomy52 references91 citations

TL;DR

This paper presents a physically self-consistent model of dusty tori in Active Galactic Nuclei by deriving the dust distribution from hydrostatic equilibrium between gravitational, centrifugal, and turbulent pressure forces. Using 3D radiative transfer with the MC3D code, it shows that dust grain size and composition variations—especially differing sublimation radii—significantly suppress the 9.7 μm silicate feature and enable accurate fitting of both type I and type II AGN SEDs, while high-resolution observations of NGC 1068 and Circinus reveal distinct temperature structures in the inner torus.

ABSTRACT

We explore physically self-consistent models of dusty molecular tori in Active Galactic Nuclei (AGN) with the goal of interpreting VLTI observations and fitting high resolution mid-IR spectral energy distributions (SEDs). The input dust distribution is analytically calculated by assuming hydrostatic equilibrium between pressure forces - due to the turbulent motion of the gas clouds - and gravitational and centrifugal forces as a result of the contribution of the nuclear stellar distribution and the central black hole. For a fully three-dimensional treatment of the radiative transfer problem through the tori we employ the Monte Carlo code MC3D. We find that in homogeneous dust distributions the observed mid-infrared emission is dominated by the inner funnel of the torus, even when observing along the equatorial plane. Therefore, the stratification of the distribution of dust grains - both in terms of size and composition - cannot be neglected. In the current study we only include the effect of different sublimation radii which significantly alters the SED in comparison to models that assume an average dust grain property with a common sublimation radius, and suppresses the silicate emission feature at 9.7 micron. In this way we are able to fit the mean SED of both type I and type II AGN very well. Our fit of special objects for which high angular resolution observations (less than 0.3 arcseconds) are available indicates that the hottest dust in NGC 1068 reaches the sublimation temperature while the maximum dust temperature in the low-luminosity AGN Circinus falls short of 1000 K.

Motivation & Objective

• To develop a physically grounded, self-consistent model of dust tori in AGN based on hydrostatic equilibrium, rather than free-form geometry.
• To address the limitations of previous models that assume homogeneous dust distributions and uniform sublimation radii.
• To improve the fit of observed mid-infrared SEDs of both type I and type II AGN by incorporating realistic dust grain size and composition stratification.
• To interpret high-resolution VLTI/MIDI observations of NGC 1068 and Circinus by modeling temperature and structure in the inner torus.
• To reduce ambiguity in torus modeling by using physical constraints on dust distribution and radiative transfer, rather than adjustable free parameters.

Proposed method

• Derive the 3D dust density distribution analytically from hydrostatic equilibrium balancing gravitational, centrifugal, and turbulent pressure forces from the central black hole and stellar distribution.
• Use the 3D Monte Carlo radiative transfer code MC3D to simulate radiative transfer through the torus, accounting for anisotropic radiation and dust absorption/re-emission.
• Incorporate variable dust grain properties—specifically different sublimation radii based on grain size and composition—instead of assuming a single average grain model.
• Fix key model parameters (e.g., dust mass, luminosity, geometry) using observational constraints from SEDs and interferometric data.
• Perform a parameter study varying dust properties, masses, and viewing angles to assess model robustness and fit quality.
• Compare model results with high-resolution MIDI interferometric data and published SEDs of NGC 1068 and Circinus to validate predictions.

Experimental results

Research questions

• RQ1 How does a physically derived dust distribution from hydrostatic equilibrium affect the predicted mid-infrared SED of AGN?
• RQ2 To what extent does dust grain size and composition stratification—especially differing sublimation radii—alter the silicate emission feature at 9.7 μm compared to homogeneous models?
• RQ3 Can the model reproduce the observed SEDs of both type I and type II AGN using physically motivated dust properties and geometry?
• RQ4 What are the implications of high-resolution VLTI/MIDI observations for the temperature structure and inner torus geometry in NGC 1068 and Circinus?
• RQ5 Why do standard homogeneous dust models fail to reproduce the silicate feature depth observed in interferometric correlated fluxes?

Key findings

• The observed mid-infrared emission in homogeneous dust models is dominated by the inner funnel, even when viewed along the equatorial plane, indicating that dust stratification is essential.
• Incorporating different sublimation radii for dust grains of varying size and composition significantly suppresses the 9.7 μm silicate emission feature, matching observations better than models with a single average grain.
• The model successfully fits the mean SEDs of both type I and type II AGN by accounting for dust property variations, improving on previous homogeneous models.
• For NGC 1068, the hottest dust reaches sublimation temperature, consistent with VLTI observations showing a central component hotter than 800 K.
• In the low-luminosity AGN Circinus, the maximum dust temperature remains below 1000 K, indicating cooler inner torus conditions than in luminous AGN.
• The model's inability to reproduce the deep silicate absorption in MIDI correlated flux data suggests a need for a shallower temperature gradient, possibly requiring clumpy dust structures or modified radiation anisotropy.