2. DUST REGULATING THE THERMODYNAMIC STATE OF THE GAS

The intergalactic medium

One of the biggest puzzles concerning galaxy evolution and baryonic physics on scales of hundreds of kpc is the mechanism by which galaxies can accrete gas to fuel the ongoing star-formation in their disks. One observes in the Milky Way halo and around other massive galaxies cold infalling HI clouds but these galaxies are supposed to reside in hot virialized haloes and the nature of the thermal instabilities needed to cool gas from temperatures of 10⁶ - 10⁷ K (for the more massive haloes) is unknown. However it is sometimes forgotten that the most efficient way of cooling such a medium is not through bremsstrahlung or line emission in the Xrays, but through inelastic collisions of the hot electrons and ions with dust grains, providing there are some small quantities of grains in this medium (see Popescu et al. (2000a)). Since collisional heating of dust grains is such an efficient mechanism, even trace quantities of grains can make a difference to the cooling for the gas, or at least trigger further cooling. These mechanisms have long been studied in the context of shocked interstellar gas (e.g. Dwek & Werner (1981)) but it is only recently that it was realised that the effect of dust cooling should be considered in the context of structure formation.

This process has been recently investigated by Montier & Giard (2004), who derived cooling curves for the gas in the IGM with and without the presence of dust grains. Fig. 1 (left panel) shows their curves for different grain sizes and for a dust-to-gas ratio of 10^-4. At low temperatures, between 10⁴ - 10⁶ K, where the medium is not fully ionised, the cooling is dominated by atomic processes. For higher temperatures, T > 10⁶ K, where up to now bremsstrahlung was considered to be the most efficient cooling mechanism, it is shown that the dust is the dominant cooling factor. Montier & Giard (2004) also calculated how the ratio between infrared to X-ray dominated cooling changes as a function of the dust-to gas ratio. Fig. 1 (right panel) shows that for a dust-to gas ratio of 10^-4 infrared cooling is more efficient for temperatures above 10⁶ K, while for smaller dust-to gas ratios this condition applies at increasingly high temperatures.

Figure 1. Left: Cooling curves for gas in the IGM without dust (solid line) and with dust (dotted and dashed curves for grain sizes of 0.5; 0.025 and 0.001 µm) from Montier & Giard (2004). The intergalactic gas is modelled by a mixture of hydrogen and helium in cosmic proportion X = 0.75 and Y = 0.25, respectively, with a dust-to-gas ratio of Zd = 10-4 and without a background UV radiation field. The cooling is due to recombination, collisional ionisation, collisional excitation, bremsstrahlung and dust emission. Right: Infrared to X-ray dominated cooling in the parameter space dust-to-gas ratio Zd and temperature of the gas, taken from Montier & Giard (2004).

The cooling functions described above have been coupled with numerical simulations for galaxy formation by Pointecouteau et al. (2009). Their simulations showed that although the differences between the two cases (in the absence and in the presence of dust) are not striking, the concentration of gas in the cool and high-density phase is higher when dust is included than in the purely gas cooling case.

To check whether these processes are operating in reality one should search for dust emission associated with X-ray emission in the transition regions between the cosmic web filaments and the star-forming disks of galaxies and ideal laboratories for this are clusters and groups. Thus far, the best and perhaps only example of a correspondence between FIR dust emission and X-ray emission from an IGM around galaxies is in Stephan's Quintet (SQ) (see Guillard et al. (2010), Natale et al. (2010)), though it is unclear to what extent this correspondence is due to cooling instabilities in the IGM like those predicted in Montier & Giard (2004). Recent spectroscopic observations of the SQ (Cluver et al. (2010)) have also shown that the pure rotational lines of hydrogen can be effective in cooling the IGM in the shock region of SQ. This was modelled as a multi-phase medium by Guillard et al. (2009).

The interstellar medium

Once inside galaxies, the subsequent evolution of the gas can be affected by grains in other ways. As is well known, dust grains heat the diffuse ISM via the photoelectric effect and the thermodynamical balance is maintained through an equality between the photoelectric heating and the FIR cooling lines which are powered by inellastic collisions with gas particles. Once the UV radiation field is suppressed, either by turning the sources off or by self-shielding by grains, the cooling is no longer balanced by heating, and the gas will condense further into denser structures, setting the seeds for condensation into dense molecular clouds.

In the simulation of Juvela et al. (2003) from Fig. 2 one can see the predicted ratio between [CII] and FIR within an individual simulated cloud, where the darker regions delineate more optically thick regions of the cloud where the UV photons can't penetrate but dust is still heated by the optical photons. Thus thermal pressure support can be reduced in density enhancement, potentially leading to further development of density contrast and ultimately affecting the propensity of gas to condense into stars. This shows again how dust is shaping the density structure of the gas, this time in the ISM.

Figure 2. Simulated image of the ratio between [CII] and FIR within an individual cloud, from Juvela et al. (2003).