X-Rays from a Central "Exhaust Vent" of the Galactic Center Chimney

The features of interest within the IBDF are bright, linear ridges of X-ray emission oriented roughly perpendicular to the Galactic plane. They are characterized by relatively soft X-rays and do not appear outside of the 0.5–2.0 keV band. We visually define seven regions across the IBDF field denoted by Greek letters α through η going from positive to negative Galactic longitudes (Figure 3).

Zoom In Zoom Out Reset image size

Feature β is a narrow, bright ridge that runs parallel to lines of constant Galactic longitude and points toward the Sagittarius A complex at the Galactic center. Features γ, δ, and are parallel to β, and while they are noticeably dimmer than β, they are regularly spaced in longitude and alternate in surface brightness, proceeding from β to ζ, so that feature δ, situated at the center of the seven regions, appears brighter than the regions on either side of it. Regions ζ and η combine into a distinct feature of notable complexity, as can best be seen in Figure 3. Its orientation is tilted with respect to that of the other parallel features by a small angle (∼15°) and it features a brighter component (ζ) on its eastern side bordering and a dimmer component (η) on its western side.

The IBDF X-ray features are located within the southern Galactic chimney, which itself is nestled within the southern radio lobe (Ponti et al. 2019, 2021; Heywood et al. 2019), as shown in Figure 4. As was noted by Ponti et al. (2021), the radio emission arises from the periphery of the southern X-ray chimney, which evokes a structure in which the X-rays arise from a confined hot plasma, while the radio emission arises in a thick, higher-density surrounding shell where the hot plasma interacts with the ambient interstellar medium and the predominantly vertical magnetic field of the Galactic center.

Zoom In Zoom Out Reset image size

The hardness ratio comparing the 1.2–2.0 keV band to the 0.5–1.2 keV band across the field is shown in Figure 5. We define the hardness ratio as

$$\begin{eqnarray}&&\mathrm{Hardness}\ \mathrm{Ratio}=\displaystyle \frac{{C}_{\mathrm{high}}-{C}_{\mathrm{low}}}{{C}_{\mathrm{high}}+{C}_{\mathrm{low}}},\end{eqnarray} \tag{ 1 }$$

where C_high and C_low are the exposure-corrected observed counts in the higher and lower energy bands, respectively, and a higher hardness ratio indicates harder emission.

Zoom In Zoom Out Reset image size

The error bars for latitude and longitude represent the uniform 001-wide segments over which the counts were summed, while the hardness ratio error comes from the statistical error of the observed counts in these segments. Progressing across the field in Galactic longitude, the brighter linear features β, δ, and ζ each roughly correspond to lower hardness ratios, while the overall minimum hardness ratio occurs in region α. In the progression with Galactic latitudes, the hardness ratio shows a broad minimum at the bright centers of the IBDF, with a small peak at b = − 1448 ± 0005 and a broad minimum at b = −1469 ± 0005. Panel (c) in Figure 5 demonstrates differences in X-ray hardness within the ζ/η feature, with the softer eastern edge corresponding to the bright ridge seen in Figure 2, and relatively hard emission trailing off to the west.

The spectra from the five regions were fit simultaneously to determine the most likely physical model and set of parameters for each region (Table 2). When fitting the data, we find that there is little variation in the recombination timescale τ and the initial recombining plasma temperature across the seven regions, and we therefore link the regions together for these parameters. We also fix the metal abundance to 1.0 in solar units. We evaluate the goodness-of-fit using the reduced chi-squared value (i.e., χ²/degrees of freedom (dof)), where a lower value of χ²/dof indicates a better fit. For each region, the model with the lowest χ²/dof consists of a combination of RP and APEC components, each convolved with a Tuebingen–Boulder interstellar medium grain absorption model (Wilms et al. 2000), i.e., tbabs*(rnei+apec). This model yields a χ²/dof value of 1160/1062 = 1.092, which is significantly better than the other attempted models, such as a double APEC model (1921/1064 = 1.805) or a model replacing the recombining plasma with a nonthermal power-law component (1908/1076 = 1.773). A comparison of the spectra for features β, δ, and ζ fit using the RP+APEC model is shown in Figure 6.

Zoom In Zoom Out Reset image size

Table 2. Simultaneous Fit

Parameter	α	β	γ	δ		ζ	η
nH (1022 cm−2)	0.79 ± 0.04	0.86 ± 0.04	0.92 ± 0.07	0.75 ± 0.07	0.95 ± 0.04	0.92 ± 0.03	0.93 ± 0.03
RP kTinit (keV)	0.35 ± 0.01 (linked)
RP kT (keV)	0.183 ± 0.009	0.18 ± 0.01	0.20 ± 0.02	0.19 ± 0.01	0.24 ± 0.02	0.148 ± 0.009	0.15 ± 0.01
RP τ (1010 s cm−3)	9 ± 3 (linked)
APEC kT (keV)	2.4 ± 0.4	2.8 ± 0.8	1.8 ± 0.4	2.8 ± 0.8	1.05 ± 0.06	1.22 ± 0.09	1.49 ± 0.08
APEC EM (1052 cm−3)	0.31 ± 0.02	0.36 ± 0.03	0.27 ± 0.04	0.31 ± 0.03	0.42 ± 0.05	0.60 ± 0.06	0.94 ± 0.05
RP EM (1052 cm−3)	8.9 ± 0.8	12 ± 1	7 ± 1	6 ± 1	3.9 ± 0.9	35 ± 3	25 ± 2

Download table as:  ASCIITypeset image

The IBDF feature contains two prominent bright ridges, regions β and ζ, which border a column of relatively low X-ray surface brightness (regions γ, δ, and ). The linear morphology of these quasi-parallel ridges, coupled with the fact that ridge β has a sharp western edge, but falls off less abruptly toward the east into region α, leads us to hypothesize that the IBDF feature is a cylindrical, edge-brightened tunnel lying along the central axis of the southern chimney. In-keeping with the notion that the chimneys represent collimated plasma outflows from the Galactic center (Ponti et al. 2019), we regard this tunnel as the channel, or vent, along which the outflowing plasma is moving. While no measurement yet exists of the outflowing plasma velocity within the putative tunnel, we argue that it must be at least as large as the velocity of the wind emanating from the overall Galactic center region, ≳1000 km s⁻¹ (Carretti et al. 2013; Fox et al. 2015; Fujita 2023). We assume that the outflow velocity of the plasma is highest along the central axis of the tunnel because the flow along the edges would be somewhat slowed by turbulent interactions with the adjacent ambient medium. The relatively low X-ray brightness of the interior of the tunnel can then be attributed to a lower plasma density there via the continuity equation, coupled with the density-squared dependence of the X-ray emissivity of the plasma.

The relatively high emissivity of the tunnel walls, manifested as the ridges β and ζ, can be attributed to shocks that occur where the outflowing plasma impacts and compresses the surrounding ambient gas. The particularly high brightness of ridge ζ is readily attributable to its orientation; it is slightly askew relative to the axis of the tunnel, so the outflowing plasma strikes this portion of the tunnel wall at a more direct angle, and therefore with more force, than it does elsewhere where the velocity vector of the outflowing plasma is presumably almost parallel to the walls of the tunnel.

We note that the detailed morphology of the tunnel appears to be somewhat more complex than a perfectly uniform cylinder, given the presence of a fainter ridge, δ, projected near the center of the tunnel.

The brightest regions in the IBDF coincide with the smallest hardness ratios, as shown by a comparison of the 1.2–2.0 keV and 0.5–1.2 keV emissions. If the IBDF feature is indeed attributable to shocks propagating into the ambient medium, then this shock compression of the "tunnel walls" (i.e., β and ζ) would naturally lead to higher densities. This high density would in turn give rise to greater emissivity and therefore more rapid cooling, ultimately producing the bright but soft X-ray emission that we observe. A plausible candidate for energy injection into this region is the series of "hundred-year events" that have been inferred from moving X-ray echoes observed primarily in the fluorescent 6.4 keV iron line (Ponti et al. 2010; Clavel et al. 2013; Ponti et al. 2013; Churazov et al. 2017; Chuard et al. 2018; Marin et al. 2023). If the cooling time of the IBDF plasma is greater than a few hundred years, then this sequence of hundred-year events, presuming that they occur with some regularity on few-hundred-year timescales, may effectively act as a continuous flow that continually replenishes the region.

Using the fit parameters for density and temperature from Table 2, we can estimate the cooling timescale of the shocked plasma for comparison to this few-hundred-year timescale. The characteristic cooling timescale of the plasma can be calculated using the thermal energy of the plasma via the relation

$$\begin{eqnarray}&&{\tau }_{\mathrm{cool}}=\displaystyle \frac{\tfrac{3}{2}\left(1+\tfrac{{n}_{{\rm{i}}}}{{n}_{{\rm{e}}}}\right){kT}}{{\rm{\Lambda }}(t){n}_{{\rm{H}}}}.\end{eqnarray} \tag{ 2 }$$

Here Λ(t) is the cooling function, which, in this case, essentially amounts to bolometric power normalized by emission measure. Using cooling curves provided by Sutherland & Dopita (1993), we estimate Λ(t) ∼ 5 × 10⁻³⁰ W cm³ based on temperature and approximate metal abundance. Then, assuming n_e = n_i = n_H, we can estimate the cooling timescale for various possible geometries. To get a lower limit, we first assume that emission is confined to a volume whose depth is equal to its angular width so that the dimensions of region β are 01 × 001 × 001, giving us an estimated cooling timescale in this region of ≳200 yr. This estimate assumes a filamentary feature in the simplest possible case. If, however, we are looking at a cylindrical tunnel with a diameter of about 01, then the depth of region β is represented by a chord on that cylinder lying approximately 0025 from the center. Thus, the depth is ∼0087, and we calculate an estimated cooling timescale of ≳2000 yr.

To maintain the observed structure, the nonequilibrium plasma of the IBDF must be continually replenished or heated on a timescale shorter than this cooling time, and the Sgr A^* hundred-year events therefore provide a plausible mechanism for sustaining the heating and outflow of the IBDF plasma and counteracting this relatively short cooling timescale. Using the temperature, emission measure, and cooling timescale of the region, we estimate that the lower limit on the power is ∼10³⁶ erg s⁻¹. This estimate does not account for the kinetic energy of the plasma since we do not have sufficient information on its velocity. Nevertheless, this lower limit is comparable to or lower than the energies associated with the hundred-year events, which could easily sustain the observed IBDF emission assuming an average outburst luminosity of L_{2–10 keV} ∼ 10³⁹ erg s⁻¹ lasting at least 1 yr (Ponti et al. 2013).

Supernovae are an additional possibility for contributing to the powering of the plasma in the IBDF. Zhang et al. (2021) recently presented a dynamical model showing that the chimneys could be produced by supernovae taking place in the central molecular zone (CMZ), with a mean occurrence rate of one per 10³ yr. In their model, the collimation into a cylindrical volume is effected by imposing a vertical magnetic field strength of 80 μG. It remains to be seen whether their assumed supernova rate is consistent with the overall star formation rate in the CMZ of ∼0.1 M_⊙ yr⁻¹ (Henshaw et al. 2023), but supernovae might at least contribute substantially to the production of outflowing plasma on timescales comparable to the cooling time of the IBDF.

Nakashima et al. (2013) report Suzaku results indicating the presence of a recombining plasma south of the Galactic plane in a larger region that contains the IBDF. Our spectral analysis of the IBDF is indeed consistent with this result as a model containing a recombining plasma and a thermal APEC component produced acceptable fits with the lowest χ² values. We find that no single-plasma model alone is able to sufficiently fit the data, which leads us to introduce the additional APEC component in our RP+APEC and APEC+APEC models. The improvement to the fit afforded by the additional APEC component to the RP model suggests that there may be an associated thermalized plasma mixed in with or adjacent to the bulk recombining plasma, or that the plasma may have a more-or-less continuous distribution of temperatures. Comparing the temperatures in Table 2, we see that the APEC component is at a higher temperature than the RP component for all seven of the subregions in the fit. This may be an indication that the plasma responsible for the observed bright features is distinct from another plasma component sitting within or moving through the tunnel, or that the plasma of the IBDF has a highly variable temperature distribution. Further spectral analysis of the IBDF will be required to provide a more accurate picture. Follow-up high-resolution spectroscopy with XRISM may be especially helpful in this effort.

The presence of a recombining plasma in the IBDF is also consistent with the Galactic outflow hypothesis. The morphology of the IBDF features, along with the trends in hardness ratio, suggest the existence of shocks in the region. Strong shocks stemming from the outflow would propagate into the ambient interstellar medium at high velocity, compacting the gas and sustaining the recombining plasma environment.

A number of features observed on different scales merit consideration as possibly being related to the X-ray features in the IBDF. As mentioned above, the X-ray chimneys, in which the IBDF is embedded, have been invoked as the channel through which plasma and cosmic rays generated in the vicinity of the Galactic center transit out to the Galactic halo, potentially provoking the gamma- and X-ray emission arising in the large-scale Fermi and eRosita Bubbles (Ponti et al. 2019, 2021).

We also note that on the scale of a few parsecs, a linear X-ray feature close to the Galactic black hole has been identified and interpreted as a jet from the black hole (Baganoff et al. 2003; Muno et al. 2008; Li et al. 2013; Zhu et al. 2019). That putative jet is oriented in the same direction as the southern chimney and the linear features in the IBDF, but it has not been detected at distances exceeding a few parsecs from the black hole, presumably because its nonthermally emitting particles have lost sufficient energy beyond that point to continue emitting a detectable flux of X-rays (Zhu et al. 2019). However, at very low radio frequencies, Yusef-Zadeh et al. (1986; see also Kassim et al. 1986) reported the presence of a ridge of radio continuum emission extending from the Galactic center out to a distance of about 25 pc normal to the Galactic plane, ⁸ that is, in the same direction as the hypothetical X-ray jet, but far short of the 220 pc distance of the IBDF from the Galactic black hole. Continued energy loss by the electrons in the hypothetical jet could lead to a fossil jet of low-energy particles responsible for the observed low-frequency synchrotron emission. Eventually, if those particles are moving through the plasma in the chimney, they will thermalize with the plasma and thus have no observable excess manifestation of their presence at the distance of the IBDF.

While the placement and morphology of all the features mentioned here are very suggestive, further investigation is clearly needed to establish and elucidate any physical links between them.