Probing molecular clouds with supermassive black hole X-ray flares
The centre of the Milky Way is a very special place, harboring many exotic objects, such as the supermassive black hole Sagittarius A* and giant molecular clouds. Some of these clouds, despite being cold, are sources of high energy photons. It is believed that the clouds are not producing these photons themselves, but rather scatter the X-ray radiation coming from outside. Even though Sgr A* is currently very faint in X-rays, it is considered as the main culprit of this radiation, in the form of short but intense flares, which happened over the past few hundred years. The time delay caused by light propagation from Sgr A* to the clouds and then to us, allows one to study Sgr A*’s past activity. At the same time, flares serve as an extremely powerful probe of molecular gas properties. In particular, the full 3D structure of molecular clouds and their density distribution on small scales can be reconstructed.
Although our Galaxy’s supermassive black hole Sgr A*, which has 4 million times the mass of our Sun, is currently very dim, there are indications that it experienced powerful flares in the not very distant past. In particular, reflection of Sgr A*’s X-ray emission on molecular clouds surrounding it provides evidence for such recent flares.
In reconstructing this history, there are two effects that have to be taken into account. First, the reflected emission is proportional to both the intensity of the illuminating radiation and the density of the gas. Second, the time delay attained during light propagation from the primary source (i.e. Sgr A*) to a reflector (i.e. a molecular cloud), and then from the reflector to an observer amounts to hundreds of years. From this, the history record of Sgr A* activity can be reconstructed, provided that the relative positions of the source and the reflector are known with sufficient precision. This is informally known as X-ray archaeology (see, e.g. Highlight: Neutral iron Kalpha diagnostic -- X-ray archaeology). Unfortunately, the line-of-sight distances are poorly known, so one has to look for some additional ways to break down degeneracies associated with the simple time-delay arguments.
A series of recent papers has shown that exploring spatial and temporal variations of the reflected emission can lift these degeneracies. Indeed, data collected by the space telescopes Chandra and XMM-Newton over more than 15 years show that the reflected X-ray emission is variable on timescales on the order of years and on spatial scales of less than one parsec (see Fig.2).
The observed variability implies that the original flare itself must have been shorter than few years. With this in mind, one may take a more rigorous look at the statistical properties of the variability in time and space, which should be closely related to each other. Indeed, in the short flare scenario, variations in the space domain simply reflect density fluctuations in a thin slice of the reflecting medium projected on the picture plane (Fig. 1). On the other hand, variations in the time domain (at a given sky position) arise from similar density fluctuations but sampled along the line-of-sight, i.e. with slightly different time delays. If the statistical properties of the underlying density field are isotropic on small scales, there is a straightforward transformation connecting the two variability patterns. The parameters of this transformation are being determined by the relative 3D positions of the primary source and the reflector.
If one compares the X-ray flux variability in the time and space domains, these variability patterns match each other if one assumes that the light front propagates along the line of sight with a of velocity 0.7 the speed of light. This value immediately gives the position of the cloud with respect to Sgr A* and the age of the flare as about 110 years. Most likely the flare lasted less than one year, and is now reflected by the molecular cloud known as the 'Bridge complex' some 30 pc away from Sgr A*.
Using data on the emission of the same region in various molecular lines, the average density of reflecting the gas can be estimated and from this, the integrated X-ray flux provided by the flare can be inferred. Such an analysis suggests that the flare might have been the result of a tidal disruption of a planet (or the partial disruption of a star) being careless enough to come too close to the supermassive black hole.
Knowing the age of the outburst, it is straightforward to reconstruct the 3D density distribution of the molecular gas (see Fig. 3). So far, using the data of 15 years of monitoring, only a thin ~3.5 parsec slice can be reconstructed. This is certainly not the end of the story, since the molecular complex, being bright at the moment, will eventually fade away when the illumination front will have completely passed through it. At the same time other molecular clouds in the Central Molecular Zone might come into the spotlight, with ‘X-ray echoes’ of a single flare being potentially observable over several hundred years, the light-crossing-time of the entire Central Molecular Zone (CMZ). A movie illustrating the possible evolution of the CMZ X-ray map over the next 500 years is shown below.
Flare illuminates Central Molecular Zone
Interestingly enough, not only studies of Sgr A* activity do benefit from the observations of its ‘X-ray echoes’. The properties of the gas density field can be studied in detail, without being hindered by projection effects or by the sensitivity to the chemical abundance of a particular molecular species as is commonly the case for molecular emission lines data.
In the short flare scenario, the illuminated region is just a thin slice of molecular gas and the intensity of the reflected X-ray emission is simply proportional to the number density of the gas (in the optically thin limit). The probability distribution function of the gas density measured in this way appears to be well described by a log-normal shape (see Fig. 4), in line with the theoretical and numerical predictions for supersonic turbulence, which is believed to shape the structure of molecular gas on the scales probed.
However, a number of effects could mimic such a shape of the distribution function, namely high opacity even for X-rays for the high end or low count statistics on the low end. These issues can partly be addressed with sufficiently deep Chandra observations complemented by realistic simulations of the molecular clouds. In principle, with the angular resolution provided by Chandra, it is possible to study scales down to 0.05 pc, where self-gravity starts to become dominant and which effectively seed the star formation process.
Thus, X-raying molecular clouds might become useful for solving the long-standing problem of suppressed star-formation efficiency in the Central Molecular Zone. Next generation of X-ray observatories equipped with micro-calorimeters, like ATHENA and Lynx, will be capable of probing also the velocity field in the reflecting gas. The full picture of the turbulent inner life of the Galactic Center molecular clouds could then be reconstructed. Equally important are future X-ray polarimetric observations that will provide solid proof that the source of illuminating photons is indeed Sgr A* by measuring the polarization angle, while the degree of polarization will provide an independent way of measuring the line-of-sight position of the cloud.
E.Churazov, I.Khabibullin, R.Sunyaev