THE GALACTIC CENTER IN THE FAR-INFRARED

The central ∼1 pc of the galaxy is a low density central cavity that surrounds Sgr A*, the compact variable radio source located close to the dynamical center of the Galaxy—a black hole of ∼4.5 × 10⁶ M_☉ at a distance of 8.3 ± 0.4 Kpc (Kerr & Lynden Bell 1986; Ghez et al. 2008; Gillessen et al. 2009). The central cavity is characterized by emission of low ionization atomic species. The cavity contains a cluster of Wolf-Rayet (WR) stars and O and B stars. Using far-infrared (FIR) and submillimeter continuum data, Zylka et al. (1995) estimated that the central cavity (radius ∼1 pc) contains ∼400 M_☉ of dust at T_kin ∼ 40 K, 4 M_☉ at ∼170 K, and ∼0.01 M_☉ at ∼400 K. The luminosities emitted by these three phases are ∼2 × 10⁶ L_☉, 4 × 10⁶ L_☉, and 5 × 10⁵ L_☉, respectively. Latvakoski et al. (1999) studied the region from the Kuiper Airborne Observatory (KAO) in the 30 μm band with a resolution of 85 and reported that structures in the region have temperatures ranging from about 70 K to 100 K. They found the total luminosity of the components in the inner region (from their Table 1) to be ∼6 × 10⁶ L_☉ and its mass ∼2000 M_☉.

Surrounding the central cavity, extending from 1 pc to ∼5 pc, is the circumnuclear disk (CND). The CND has a mass of several thousand solar masses and rotates with a velocity ∼100 Km s⁻¹. The neutral material in this ring is clumpy, dense (n(H₂) ∼10³–10⁷ cm⁻³), and fragmented (Genzel et al. 1982; Genzel & Townes 1987) having a filling factor ∼10% (Zylka et al. 1995; McGary et al. 2001; Wright et al. 2001). There is a clumpy and filamentary structure known as the mini-spiral that connects the CND with the central ionized cavity, allowing ultraviolet radiation from the central ionized cavity to penetrate into the CND, heating and photoionizing the gas (Yusef-Zadeh et al. 2001).

The kinetic temperatures of the gas in the CND and central cavity range as high as 200 K–300 K in CO(J = 7–6) (Bradford et al. 2005; Martin et al. 2004) and ∼400 K in NH3 (6,6) (Montero-Castaño et al. 2009; Herrnstein & Ho 2005). The source of heating remains unclear. One explanation suggested is that the ionized central cavity is excited by the radiation from a central cluster of O and B stars, and that at the cavity edge the surrounding material is likely to be shock compressed by the expanding gas around the photodissociation regions. Other possibilities include C-shock heating, perhaps by magnetic viscous heating, or possibly turbulent dissipation. Bradford et al. (2005) suggest magnetohydrodynamic (MHD) shock heating with a shock velocity of order 20–30 Km s⁻¹ and a magnetic field of 0.3–0.5 mG. If so, the material in the CND will dissipate its orbital energy within a few revolutions, and they suggest that its apparent longevity may result from refueling by infalling material from the surrounding clouds. Several authors suggest that the ionization of the molecular clouds in the Galactic center is due to a large flux of cosmic rays heating the region (Guesten et al. 1981; Huettemeister et al. 1993; Yusef-Zadeh et al. 2007).

The CND is surrounded by a number of dense molecular clouds and filamentary structures (see Figure 1). At a distance of ∼39 pc to the northeast of Sgr A* is the Radio Arc consisting of thermal and nonthermal structures aligned almost perpendicularly to the Galactic plane (Yusef-Zadeh et al. 1984). The Radio Arc seems to connect to Sgr A* via two concentrations of ionized and molecular gas: the Sickle and the Arched Filaments. Two clusters of WR and O-type stars, the Quintuplet and the Arches Cluster, are responsible for the ionization of the Sickle and the Arched Filaments, respectively. The Arches Cluster is slightly in the foreground, about 20 pc from the Arched Filaments, and irradiates the several filaments approximately uniformly (Lang et al. 2001).

The Midcourse Space Experiment (MSX) image at 21.3 μm and the PACS image at 70 μm trace a shell surrounding the Arches and Quintuplet clusters. The shell goes through the Arched Filaments, the Lima bean, and passes through the south of the G0.11-0.11 molecular cloud forming a ring of ∼44 pc in diameter (Figure 1). Bally et al. (2010) suggest that the radiation and the stellar winds from the Arches and the Quintuplet clusters and the black hole in Sgr A* may be responsible for the formation of this shell, which can clearly be seen in the Spitzer 24 μm image (Yusef-Zadeh et al. 2009), and which has also been reported by Moneti et al. (2001).

Zoom In Zoom Out Reset image size

Recent studies by Molinari et al. (2011), using far-infrared images obtained with the Herschel PACS and SPIRE instruments, reveal the presence of a 100 × 60 pc elliptical loop shaped by a chain of cold and high column density clumps (T_dust ⩽ 20 K; N(H) > 2 × 10²³ cm⁻²) orbiting around the Galactic center. The loop is shifted with respect to the Galactic center with Sgr B and Sgr C placed close to the two ends of the 100 pc semi-major axis, and Sgr A* at a projected distance of 24 pc toward the negative semi-major axis. Herrnstein & Ho (2005) suggested that Sgr A* is interacting with the M-0.02-0.07 molecular cloud (Figure 1) which indicates that Sgr A* is probably closer to the front of the loop than to the back (Molinari et al. 2011).

The objective of this paper is to further constrain the physical conditions in the central 2 pc of the Galactic center and to help clarify the excitation conditions. We use archival Herschel PACS and SPIRE photometric observations of the region to determine the dust properties more accurately; many earlier data sets at shorter wavelengths (Latvakoski et al. 1999) would have been insensitive to a cold dust component. We also incorporate previously unpublished infrared images from the Infrared Space Observatory Long Wavelength Spectrometer (ISO-LWS), operating in parallel mode; its grating sampling allows us to obtain a more precise analysis of the character and structure of dust in the central region and the CND.

Herschel PACS (Poglitsch et al. 2001) data at 70 μm and 160 μm, and SPIRE (Griffin et al. 2006) data at 250 μm, 350 μm, and 500 μm, were used to measure the dust temperature and the column density distribution across the Galactic center region. The basic data were obtained from the public Herschel archive and were reprocessed using HIPE, the Herschel Interactive Processing Environment. The data were collected in parallel mode and are still rather preliminary in absolute accuracy, suffering from calibration errors. PACS fluxes off the bright Galactic plane are negative, indicating that the pipeline corrections are inaccurate; moreover, there are indications that in this extremely bright region some nonlinear detector responses are uncorrected. We do not apply any across-the-board offsets to get positive flux values. As a result, the derivative images we discuss below are reliable over most but not all of the area, but the spectral energy distributions (SEDs) we fit to the warm central region (see Section 4) are accurate. Images are made in Jy pixel⁻¹ at 20 pixel⁻¹ and 30 pixel⁻¹ for PACS 70 μm and 160 μm, respectively, and in Jy beam⁻¹ for SPIRE data at 250, 350, and 500 μm, with beam areas, respectively, of 3950², 7400², and 15710² (PACS and SPIRE pixel and beam sizes are given in Table 1). Figure 1 covers an area of 28.46 × 29.4 arcmin² centered at R.A.(2000) = 17^h45^m39.77, decl.(2000) = −29°00'430. For a distance of 8.5 kpc, the covered area is ∼70.4 × 72.7 pc².

The image is color-stretched to show more of the faint or cool structures, which are labeled. The most obvious places where the photometry issues make analyses difficult are in the coldest, dark regions, for example, in G0.11-0.11. We white-out the corresponding area in the derivative column density and spectral index images because of the uncertain correction to the negative fluxes reported by the current Herschel pipeline.

The MSX surveyed the Galactic plane at six bands from 4.2 to 26 μm at a spatial resolution of 18'' (Egan et al. 1999). The MSX data can also be used to study the morphology, dynamics, and physics of the interstellar medium providing the thermal emission from dust grains at mid-infrared wavelengths (Cohen 1999). We used the MSX Band E data at 21.3 μm to trace the warm dust across the Galactic center.

We also report in this paper previously unpublished observations of Sgr A*, taken by the LWS (spectral range 46–180 μm) on the ISO satellite in parallel mode. ISO was equipped with two spectrometers, the Short Wavelength Spectrometer and LWS, a camera, ISOCAM, and an imaging photopolarimeter, ISOPHOT, to cover wavelengths from 2.5 to 240 μm with spatial resolutions of 15 at the shortest wavelengths and 85''–90'' at the longer wavelengths. Two instruments, LWS and ISOCAM, were able to operate simultaneously in parallel mode. More than 100,000 individual pointings were made in parallel mode in over 17,000 individual observations with a total sky coverage of about 1%. Key to our study, the LWS operated with 10 wavelengths in the range 46–180 μm (Lim et al. 2000), each with bandwidths of about 0.3 μm for the five short wavelength (SW) detectors and 0.6 μm for the five long wavelength (LW) detectors. Parallel mode maps are generated by combining different raster scans. Several data reduction tools were developed in the IDL programming language and these form the LWS parallel interactive analysis package that can be found at http://www.ipac.caltech.edu/iso/lws/lia/lia.html.

Now that the Herschel observations have obtained longer wavelength coverage, out to 500 μm, the earlier ISO data can reliably be combined with them to prepare a more comprehensive picture. In particular, the spectral information obtained with ISO-LWS allows us to determine the peak wavelength and flux of the dust emission with much greater precision than is possible with broadband photometry; this in turn allows us to specify the conditions of excitation much more accurately.

Herschel PACS and SPIRE photometry have a much higher spatial resolution in the FIR than any previous mission. When analyzed, the maps provide a much improved spatial distribution of the fluxes, temperature, spectral index, and H₂ column density across the Galactic center region. The MSX 21.3 μm images provide an important measure of the warmer material in the region. Figure 1 shows the Galactic center as seen by combined images of MSX-E 21.3 μm (blue), PACS 70 μm (green), and SPIRE 350 μm (red). The Arched Filaments, the Radio Arc, and Sgr A* are clearly visible at 21.3 μm and 70 μm, while SPIRE 350 μm traces the very cold and dense molecular clouds. Herschel resolves the central cavity and the surrounding CND at 70, 160, and 250 μm; the lower SPIRE 350 and 500 μm resolutions (FWHM = 253 and 369, respectively) are not quite able to resolve the cavity.

The observed continuum emission from the Galactic center region peaks at about 66 μm, in the range of ISO-LWS detector SW3 (Figure 2). As shown in Figure 2, the shape of the SED cannot be fit by a single gray body curve. In the ensuing discussion, we present our procedure for decomposing the shape, which depends on the temperature, the number of grains along the line of sight, and their emissivity and spectral index, β. At the longest wavelengths, the grain properties determine the shape of the SED. We assume a grain size of a ∼ 0.1 μm. The temperature and the spectral index, β, distribution across the Galactic center were then obtained by fitting the PACS and SPIRE wavelengths with a gray body curve:

$$\begin{equation} F_{\nu }=\Omega \times B_{\nu }(T)\times (1-e^{-\tau _{\nu }}), \end{equation} \tag{ 1 }$$

where Ω is the solid angle, T is the temperature, and τ_ν is the optical depth for a frequency ν, given as

$$\begin{equation} \tau _{\nu }=\tau _0 \left(\frac{\nu }{\nu _0}\right)^{\beta }, \end{equation} \tag{ 2 }$$

with τ₀ the reference optical depth at ν₀, the reference frequency, which was chosen to be c/100 μm. The shape of the curve is only mildly sensitive to the choice of ν₀ (Lewis & Chapman 2000).

Zoom In Zoom Out Reset image size

The SED for each pixel in the Herschel image was fitted after smoothing the images and adjusting the pixel sizes uniformly to the resolution of the SPIRE 500 μm channel band. The H₂ column density was then derived assuming a gas-to-dust-ratio of 100. We did not include the PACS 160 μm image data with the other four Herschel bands in the calculation of the temperature and the column density maps.

At the brightest regions, the flux values from this detector did not fit any physical SED, perhaps due to a poor correction of low level stripes in the data and/or errors on the background subtraction. Moreover, in the low intensity regions the detector map has negative values possibly due to uncorrected gain adjustments (Bernard et al. 2010). Its omission will not affect the conclusions significantly; subsequent pipeline reductions will hopefully address these issues.

Figures 3–5 show the resultant temperature, molecular hydrogen column density, and β maps, respectively; north is up in all cases. The white circle on the temperature map, centered at Sgr A*, shows the average size of the ISO-LWS beam which covers, approximately, the central cavity and the inner 1 pc of the CND, for a distance to the region of 8.5 kpc. Figure 7 shows the PACS 160 μm map; although the absolute calibration is defective, as noted, its higher resolution provides an improved picture of the CND than does either the temperature or column density maps, which have been reduced to the SPIRE 500 μm beam size and are therefore unable to resolve the CND and cavity structure.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

3.1. The Temperature Map

3.2. The Column Density Map

3.3. The Spectral Index Map

3.4. The Distribution of the Very Small Grains

The O and WR stars from the Arches and the Quintuplet clusters (Figure 1) provide enough photons to ionize and heat the Sickle and the Arched Filaments (Cotera et al. 1996) and also generate strong winds that could contribute to the depletion of the largest dust grains whose population is traced by the N(H₂) map. The largest dust grains can be destroyed due to the strong winds and the UV radiation from the massive O and WR stars from the Arches and Quintuplet clusters.

The MSX-E map at 21.3 μm (Figure 6) can be used to trace the population of very small grains (VSGs) throughout the Galactic center region (this band is more reliable than the features observed at shorter infrared wavelengths which can be affected by extinction). VSGs are small (with sizes of ∼0.01 μm), and absorption of one UV photon (Pomarés et al. 2009) can increase their temperature up to 80 K, prompting them to re-emit at these shorter wavelengths (λ ∼ 24 μm; Desert et al. 1990; Boulanger et al. 1985). The MSX-E map correlates well with the temperature map, showing that the population of the small grains is distributed through the regions where the population of the largest grains has been depleted, i.e., the Sickle and the Arched Filaments. This effect can also be observed in Sgr A*: PACS 160 μm, Figure 7, shows very low emission at this far-infrared wavelength in the central cavity of Sgr A*, but it is surrounded by bright emission from the CND. The emission in the inner parsec, as traced by MSX-E at 21.3 μm, appears confined to the central cavity where the O and B stars from the central cluster produce a strong radiation field, and where stellar winds heat up the small grains and deplete large grains. Within the central 2 pc, the largest grains are found remaining in the CND itself, where they are shielded from the central strong UV radiation; their emission is traced by Herschel PACS+SPIRE at far-infrared wavelengths λ > 60 μm.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

ISO-LWS obtained images of the central region of the galaxy operating in its parallel mode, one image simultaneously from each of the 10 ISO-LWS detectors, with the spectrometer's grating set at a fixed position near the center of its scan range. The ISO-LWS parallel mode maps provide us with complementary wavelength coverage just where it is most needed, at the peak of the dust emission, in the range of 46–180 μm that can be used to defined the SED of Sgr A* at far-infrared wavelengths. Unfortunately, with a beam of ∼85'' and a pixel size of 42'', these ISO maps are unable to resolve as many structures as do higher resolution Herschel PACS and SPIRE data.

Figure 8 shows the resultant Galactic center images in each of the 10 bands. The images are centered at R.A. (2000) = 17^h45^m5638, decl. (2000) = −28°55'1524 and cover an area of ∼22.8 × 29.6 arcmin² (∼ 59 × 77 pc²). In order to obtain accurate photometry from these images, an extended source flux correction factor is required, as described in the ISO Users Manual. Fortunately, we also had pointed observations with which to confirm the correction procedure (see G. J. White et al. 2011, in preparation for a description of the ISO pointed photometric and spectroscopic observations). The observed flux in Jy is

$$\begin{equation} S_{\nu }({\rm Jy})= F_{\nu }\times f / (\Omega \times 10^6)({\rm MJy\, sr^{-1}}), \end{equation} \tag{ 3 }$$

and the extended source correction factors ( f ) and the solid angle Ω are given in Table 2. Besides, Casassus et al. (2008) reported that the ISO-LW1 at 102.42 μm intensities were 27% higher than IRAS 100 μm for intensities bellow 1000 MJy sr⁻¹ and 40% higher for intensities above 1000 MJy sr⁻¹. This was in agreement with Chan et al. (2003). This offset on the ISO parallel mode data was calculated by comparing the fluxes at different position along the Galactic center observed in primary mode with the fluxes observed in parallel at the same position for the 10 ISO-LWS detectors. The correction factors, $\mathcal {Q}$(Flux_parallel/Flux_primary), are also given in Table 2.

Zoom In Zoom Out Reset image size

Table 2. Table of Extended Source Correction ( f ) and Effective Solid Angle of the Beam (Ω) for the Different LWS Detectors

Detector	λ	f	Ω × 106	$\mathcal {Q}$
	(μm)		(sr)
SW1	46.22	0.8704	0.1140	1.082
SW2	56.20	0.8677	0.1321	0.899
SW3	66.11	0.8421	0.1410	0.889
SW4	75.69	0.7334	0.1223	0.823
SW5	84.79	0.6878	0.1163	0.878
LW1	102.42	0.6758	0.1111	0.930
LW2	122.21	0.6734	0.1128	0.823
LW3	141.80	0.6035	0.0935	0.948
LW4	160.55	0.5411	0.0904	0.992
LW5	177.97	0.4596	0.0833	1.004

Download table as:  ASCIITypeset image

Despite the correction factors were applied, the LW2 fluxes still seem to be offset in our data by 10% above the continuum level. Due to this problem, the LW2 flux was omitted when fitting the ISO-LWS SED in Figure 2.

Sgr A* is very bright at the shortest wavelengths and clearly seen in Figure 8. The SW detectors (46.2–84.79 μm) also show the northern part of the Radio Arc. Detectors SW4 75.69 μm, SW5 84.79 μm, LW1 102.42 μm, and LW2 122.2 μm show Sgr A* as a single object about ∼4 pc in radius that includes Sgr A* and extends to the M-0.02-0.07 molecular cloud (Figure 1) in the south. At 160.55 μm (LW4) and 177.97 μm (LW5), the most conspicuous feature is the cold and dense molecular cloud M-0.02-0.07 (see also Figure 1).

The SED of Sgr A* within the ∼85'' ISO-LWS beam is shown in Figure 2. Each data point in Figure 2 represents the integrated flux on the ISO-LWS beam centered at the position of Sgr A* at each wavelength. We note that at a distance to the region of about 8.5 kpc, the ISO-LWS beam size of 3.5 pc includes the central cavity and the inner ∼0.8 pc of the CND, not quite large enough to completely include the full CND observed in the Herschel maps. In order to model the ISO-LWS continuum spectrum across the wavelength range ∼40–200 μm, we assumed typical dust properties: dust grains of 0.1 μm in radius and a bulk density of 2.5 (g cm⁻³). These values are characteristic of silicate grains that contain H₂O inside their structures (e.g., Pollack et al. 1994). The dust opacity is defined as

$$\begin{equation} \kappa _{\lambda }\cdot \rho =\frac{\kappa _{{\rm const}}}{\lambda ^{\beta }}, \end{equation} \tag{ 4 }$$

where κ_λ(cm² g⁻¹) is the dust opacity, ρ(g cm⁻³) is the bulk density of the grains, and κ_const is a proportionality constant. For λ = 100 μm and κ_const = 0.8 the dust opacity in the central cavity is κ_100 μm ∼ 92 cm² g⁻¹, characteristic of dust grains with thin ice mantles, while the opacities at the CND are ∼160 and 200 cm² g⁻¹ for the hot and the cold component, respectively. These values are typical of dust grains with thick ice layers (Ossenkopf & Henning 1994).

The best model that we found able to reproduce the LWS spectral continuum consists of three temperature components: an SED arising from the central cavity of r = 1 pc in radius, and a combination of a warm plus cold SED arising from the inner 0.86 pc of the CND. Figure 4 shows the steep gradient in column density at the position of Sgr A*, from values of NH₂ ∼ 3 × 10²² cm⁻² at the center of the cavity to higher values ∼1 × 10²³ cm⁻² at the CND. The temperature map (Figure 3) likewise shows the largest values just on the center, T_dust ≳ 60 K, surrounded by colder dust. The SED from the cavity can be fitted by a gray body curve defined by a dust temperature of T_dust = 90 K, β = 1.23, and a low column density of N(H₂) = 8.73 × 10²¹ cm⁻².

As noted, the SED of the CND requires two components in order to fit the whole ISO-LWS continuum: a hot component of T_dust = 44.5 K, β = 1.35, and a column density of N(H₂) = 3.18 × 10²² cm⁻². Significantly, a cold component also has to be introduced in order to fit the longest wavelengths of the ISO-LWS data. The cold component can be fitted by a T_dust = 23.5 K modified blackbody with β = 1.40 and a higher column density N(H₂) = 6.39 × 10²² cm⁻². This massive new cold component, previously unreported in observations done only at shorter wavelengths, is reconfirmed by the longer wavelength Herschel photometry. This simple, three-component model reproduces very well the observed continuum emission at the ISO-LWS spectrum (Figure 2) and the SPIRE and PACS 70 μm data. The luminosity of each component is L_cavity ∼ 2.2 × 10⁶ L_☉, L^hot_CND ∼ 1.4 × 10⁶ L_☉, and L^cold_CND ∼ 1.0 × 10⁵ L_☉, and the total gas mass: M_cavity ∼ 1.7 × 10³ M_☉, M^hot_CND ∼ 1.56 × 10⁴ M_☉, and M^cold_CND ∼ 3.15 × 10⁴ M_☉. Table 3 lists the parameters from the fit to the ISO-LWS continuum.

Table 3. Parameters Used to Fit the ISO-LWS Continuum

Physical Parameters	Cavity	CNDhot	CNDcold
Radius (pc)	1.0	0.86	0.86
β	1.23	1.35	1.40
κconst.	0.8	0.8	0.8
κ100 μm	92.28	160.38	200.
Tdust (K)	90.0	44.5	23.5
ρdust (cm−3)	9.45 × 10−23	4.0 × 10−22	8.05 × 10−22
τ100 μm	5 × 10−3	3 × 10−2	8 × 10−2
$N_{{\rm H_2}}$ (cm−2)	8.73 × 1021	3.18 × 1022	6.39 × 1022
$n_{{\rm H_2}}$ (cm−3)	2.8 × 103	1.19 × 104	2.41 × 104
χdust	0.01	0.01	0.01
L(L☉)	2.2 × 106	1.4 × 106	1.0 × 105
M(M☉)	1.7 × 103	1.56 × 104	3.15 × 104

Download table as:  ASCIITypeset image

The flux errors on the ISO-LWS data are 10%–15% (Gry et al. 2003). Assuming an error of ∼12% for each detector, the goodness of the fit in Figure 2 is χ² = 0.89 (LW2 flux was not taken into account on the estimation of χ²). The CND cold component has to be included in the model in order to fit accurately the SPIRE fluxes at wavelengths ⩾250 μm. Models with just one CND component can also reproduce the ISO-LWS continuum very precisely. However, the model must satisfy another constraint set by the fluxes of the atomic and molecular lines detected by ISO-LWS in this region. We have used a radiative transfer code as described in González-Alfonso et al. (2002) to simultaneously fit the continuum and the spectral lines, obtaining the best results for the case of two CND components. We present the discussion of the gas properties through Sgr A* in a separate paper (G. J. White et al. 2011, in preparation).

We present new Herschel PACS and SPIRE maps of the Galactic center, together with previously unpublished ISO-LWS 10 band parallel mode maps that enable us to obtain an accurate calibration for the central parsecs and a detailed spectral shape around the wavelength peak of the continuum emission. Photometry from the data set allows us to model dust in the central 2 pc of the Galactic center. Combined with MSX-E 21.3 μm images, we use these new results to trace the dust properties throughout the Galactic center region.

The ISO-LWS continuum in the direction of Sgr A*, extended with Herschel data out to 500 μm, is best fit by a model of a central cavity of 1 pc radius emitting like a gray body with T = 90 K and β = 1.23 and with low density n(H₂) = 2.8 × 10³ cm⁻³. The region is surrounded by the CND represented by two different components: a hot component with a dust temperature of 44.5 K and n(H₂) = 1.19× 10⁴ cm⁻³ with β = 1.35, and a cold component with T_dust = 23.5 K, n(H₂) = 2.41× 10⁴ cm⁻³, and β = 1.4. The two components indicate that the CND is not a single uniform body, but (as suggested by the images and previous authors) has numerous clumps, gaps, and substructures but whose average properties manifest these two basic behaviors. This simple model is able to reproduce the observed ISO-LWS continuum as well as the SPIRE data at 250 μm, 350 μm, and 500 μm. The MSX-E map traces the emission of the VSG population through the Radio Arc, the Sickle, and the Arched Filaments, and that confined in the central cavity in Sgr A*. These spatial structures are now directly detected in the far-infrared, and, with the full SED available, their dust features are all characterized. As expected, the largest grains are distributed throughout the CND and correspond to the densest and coolest molecular clouds. Unresolved issues with the Herschel photometry in the current data set leave the absolute accuracy of the coldest, densest regions (like G0.11-0.11) approximate.

Dust opacity properties in the central cavity are typical of grains with a thin ice layer while the dust properties in the CND, with lower temperatures and the highest densities, are more typical of dust grains with thick ice mantles.

The total integrated continuum flux around the inner edge of the CND and the central cavity provides a total mass of M ∼4.9 × 10⁴ M_☉. The luminosities obtained are L_cavity ∼ 2.2 × 10⁶ L_☉ and L_CND ∼ 1.5 × 10⁶ L_☉ in the central ∼2.0 pc in Sgr A*.

We thank the referee for his or her careful reading of the paper, and for the useful comments made. We are very grateful to Tanya Lim for her suggestions on the treatment of the ISO-LWS parallel mode data. This work was supported in part by NASA Grant NNX10AD83G. The ISO-LWS LIA is a joint development of the ISO-LWS Instrument Team at Rutherford Appleton Laboratory and the Infrared Processing and Analysis Center (IPAC/Caltech, USA). HIPE is a joint development by the Herschel Science Ground Segment Consortium consisting of ESA, the NASA Herschel Science Center, and the HIFI, PACS, and SPIRE consortia.