MID-INFRARED POLYCYCLIC AROMATIC HYDROCARBON AND H₂ EMISSION AS A PROBE OF PHYSICAL CONDITIONS IN EXTREME PHOTODISSOCIATION REGIONS

Dense (>10⁴ cm⁻³) and high UV field (>10⁴ times the standard interstellar radiation field in units of the Habing field, written G₀ hereafter) photodissociation regions (PDRs) rule the energy balance, and thus evolution, of some of the most fundamental astrophysical objects such as protoplanetary disks, planetary nebulae, and starburst galaxies. One of the best tools to probe this UV-illuminated matter is spectroscopy in the mid-infrared (hereafter mid-IR; 5–15 μm) because it provides information on both the gas and small dust grain properties (polycyclic aromatic hydrocarbons and very small grains, hereafter PAHs and VSGs). Ideally, one would like to spatially resolve the emission of these components in order to study their variations as the UV field is attenuated. Unfortunately, such observations are very hard to achieve (for the moment), on one hand because large ground-based telescopes, providing arcsecond angular resolution, are restrained in wavelength coverage, while on the other hand, space-borne telescopes have diameters that are usually too small to resolve the sources. Because it is the closest ultracompact H ii region at a distance of 850 pc, Monoceros R2 (Mon R2; see Wood & Churchwell 1989; Howard et al. 1994) constitutes one of the rare exceptions where one can resolve the PDR between the H ii region and molecular cloud. In this Letter, we present and analyze the mid-IR PAH and molecular hydrogen emissions in Mon R2 based on spatially resolved Spitzer spectral mapping.

Mon R2 was observed using the Infrared Spectrograph (IRS) on board Spitzer, in the low-resolution mode (λ/Δλ = 60–127) as part of the "SPECHII" program (PI: C. Joblin). The data were obtained in the spectral mapping mode. The full spectral cubes of the SL1 and SL2 modules were built using the CUBISM software (Smith et al. 2007) from the basic calibrated files retrieved from the Spitzer archive (ver. S19 pipeline). The two cubes were then assembled to provide the full SL cube of 26 × 36 positions in space and ∼170 points in wavelength, ranging from 5 to 14.5 μm.

An overview of our observations is presented in Figure 1. The ionized gas in the H ii region, traced by the [Ne ii] line at 12.8 μm, appears confined in a spherical region, as seen by radio continuum observations (Wood & Churchwell 1989) and in higher angular and spectral resolution [Ne ii] ground-based observations (Takahashi et al. 2000; Jaffe et al. 2003). The cometary shape of the H ii region is well seen in our [Ne ii] map and peaks at about 0.12 erg s⁻¹ cm⁻² sr⁻¹ in the surroundings of the B1 star IRS1. The fact that [Ne ii] maximum contours do not include the exact position of IRS1 is due to the saturation of the IRS detectors at this position, implying that the intensity could not be measured, though likely peaking there as shown by Takahashi et al. (2000) and Jaffe et al. (2003). PAH emission is present everywhere in the region, but for clarity in Figure 1 we only present the region where it is the brightest in the 11.3 μm band (1–4 × 10⁻² erg s⁻¹ cm⁻² sr⁻¹), forming a filamentary/shell structure that surrounds the H ii region. The H₂ 0–0 S(3) rotational line at 9.7 μm peaks in a filament that lies between the H ii region and the cold and dense molecular gas traced by CS J = 5–4 (Figure 1) transition. The intensity of the H₂ 0–0 S(3) line is of the order of (1–4) × 10⁻⁴ erg s⁻¹ cm⁻² sr⁻¹. The H₂ 0–0 S(2) line at 12.3 μm follows a similar spatial distribution. As in the Orion bar (Tielens et al. 1993), the spatial distribution of these H₂ lines is not correlated with the PAH emission contrary to what is seen in lower UV field PDRs like the Horsehead nebula (Habart et al. 2003; Compiègne et al. 2007) or the ρ-Ophiucus filament (Habart et al. 2005). In the following, we investigate the origin of these structures from a physical point of view, using PAH emission and modeling the H₂ excitation in the PDR. To simplify the task, we select three different zones named PDR 1, PDR 2, and PDR 3 as displayed in Figure 1. PDR 1 was chosen to be representative of the region where the maximum of H₂ emission is found (PAH emission is weaker but present). PDR 2 is the transition from the H ii region to the molecular gas. At these two positions, it is well seen that H₂ gas is found further from the star than the PAH emission (see cut in Figure 1). Finally, we positioned PDR 3 in a region situated further from IRS1, which is a smoother transition from ionized to neutral and molecular gas. For comparison, we also consider the mid-IR spectrum obtained by integrating the whole IRS cube over the area imaged.

Zoom In Zoom Out Reset image size

4.1. H₂ Emission as a Probe of Gas Density and Temperature

In dense and high-UV PDRs like Mon R2, the excitation of the H₂ pure rotational lines is dominated by inelastic collisions (Le Bourlot et al. 1999). To quantitatively probe the role of density n_H and radiation field G₀ on the excitation of the lowest energy rotational levels of H₂, we use the revised Meudon PDR code (Le Petit et al. 2006; Goicoechea & Le Bourlot 2007) for a large grid of radiation fields and densities above 10³ cm⁻³. Since the critical density of the S(3) line is ∼5 × 10⁴ cm⁻³ (Le Bourlot et al. 1999), the lowest energy rotational levels are thermalized when n_H ≳ 5 × 10⁴ cm⁻³. As a consequence, the S(3)/S(2) line ratio scales with the gas temperature and T_rot ≃ T, where T_rot is the rotational temperature of the associated rotational transition. In high-UV high-density PDRs, the main mechanism heating the gas is the photoelectric heating by electrons ejected from very small dust grains and PAHs. Thus, the photoelectric heating efficiency will depend on the ability to eject electrons from the grain surface (less efficient as grains become positively charged) and on the density of electrons in the gas with which charged grains recombine and neutralize. In PDRs, low-energy electrons are provided by the ionization of carbon atoms, and as essentially all the carbon is ionized, the electron density depends on gas density and carbon abundance. Overall, as the elemental abundance [C]/[H] is constant, an increase of the gas density (thus of electron density) reduces the grains charge and increases the photoelectric heating efficiency (i.e., the gas temperature). This explains the dependence of the S(3)/S(2) ratio with density seen in Figure 2.

Zoom In Zoom Out Reset image size

4.2. PAHs as a Probe of Radiation Field

4.2.1. Mid-IR Emission Model

In order to analyze the full Spitzer mid-IR cube, we use a model adapted from Joblin et al. (2008) and Berné et al. (2009). This emission model includes the four major dust components, introduced in these previous works: VSGs (small carbonaceous dust grains), free neutral PAHs, positively ionized PAHs (PAH⁺), and finally PAH^x that are large (with N_C > 100 carbon atoms) charged PAHs. We decompose the observed mid-IR emission using the four template spectra presented in Berné et al. (2009). In addition, we consider continuum emission that is simply modeled with two slopes. Furthermore, we include the effect of extinction by multiplying the mid-IR spectrum by $({1-e^{-\tau _{\lambda } }})/{\tau _{\lambda }}$, where τ_λ = C_ext(λ) · N_l, C_ext(λ) being the extinction cross section per nucleon, taken from Weigartner & Draine (2001) with R_V = 5.5 and N_l, the column density on material on the line of sight, is left as a free parameter in the fit. The values A_V of visual extinction corresponding to the derived N_l (related by N_l = 1.8 × 10²¹A_V) for each PDR, and used to correct the measured H₂ lines intensities as a function of their wavelength, are given in Table 1. We fit all the spectra of the cube using this technique, thus providing the spatial distributions of the emission of each component. An example of these fits is provided in Figure 1, and spatial distributions of PAH⁺ and PAH⁰ populations are presented in Figure 3.

Table 1. Observational Diagnostics in PDR 1, PDR 2, and PDR 3

Position	H2 Lines (Corrected from Extinction)				Fit of PAH Bands
	IS(2) (erg cm−2 s−1 sr−1)	IS(3) (erg cm−2 s−1 sr−1)	IS(3)/IS(2) (K)		AV	I6.2/I11.3	$\frac{[{\rm PAH}^+]}{[{\rm PAH}^0]}$	$\frac{G_0\sqrt{T/10^3}}{n_{\rm H}}$ (K1/2 cm−3)
PDR 1	2.3±0.15 × 10−4	9.6±0.1 × 10−4	4.2±0.30		16	1.1	0.03	0.28
PDR 2	1.6±0.1 × 10−4	2.1±0.1 × 10−4	1.3±0.15		12	1.8	1.60	0.49
PDR 3	1.8±0.1 × 10−4	2.0±0.1 × 10−4	1.1±0.15		19	2.0	2.44	0.56
Mon R2a	8.0±0.1 × 10−5	10.3±0.1 × 10−5	1.3±0.10		9	1.9	1.9	0.52

Notes. Observational errors are given for H₂ lines. ^aSpectrum obtained integrating the IRS cube over area imaged.

Download table as:  ASCIITypeset image

Zoom In Zoom Out Reset image size

4.2.2. Estimating G₀ Using PAH Ionization Ratio

As shown by models (Tielens 2005) and observations (e.g., Bregman & Temi 2005; Flagey et al. 2006; Galliano et al. 2008), the PAH ionization ratio ([PAH⁺]/[PAH⁰]) depends on the parameter $G_0 \sqrt{T}/n_e$, where T is the gas temperature and n_e is the electron density. Ionization of PAHs will influence the emission cross section of these molecules so that the ratio between the 6.2 and 11.3 μm bands (I_6.2/I_11.3) will increase with increasing $G_0 \sqrt{T}/n_e$. Assuming that most electrons are provided by the ionization of carbon, we can write that n_e ≃ x(C⁺)n_H ≃ [C]/[H]n_H, with [C]/[H] the elemental abundance of carbon assumed to be 1.6 × 10⁻⁴ (Sofia et al. 2004). Finally, using the empirical law of Galliano et al. (2008), we can relate the I_6.2/I_11.3 to G₀(T/10³)^1/2/n_H by

$$\begin{equation} G_0(T/10^3)^{1/2}/n_{\rm H} \simeq (1990 [{\rm C}]/[{\rm H}]) \times ((I_{6.2}/I_{11.3})-0.26). \end{equation} \tag{ 1 }$$

Using the ratio between the 6.2 and 11.3 μm bands for neutral and ionized PAHs found in Table 1 in Rapacioli et al. (2005), we relate I_6.2/I_11.3 to the ionized to neutral PAH density ratio $\frac{[{\rm PAH}^+]}{[{\rm PAH}]^0}$, using Equation (2) in Joblin et al. (1996):

$$\begin{equation} I_{6.2}/I_{11.3}=1.12 \times \frac{1+\frac{[{\rm PAH}^+]}{[{\rm PAH}]^0}\times 0.85}{1+\frac{[{\rm PAH}^+]}{[{\rm PAH}]^0}\times 0.29}. \end{equation} \tag{ 2 }$$

The high value of I_S(3) we observe requires both G₀ > 1 × 10⁴ and n_H > 1 × 10⁴ cm⁻³ (Figure 2 and Burton et al. 1992; Kaufman et al. 2006). This implies that the observed I_S(3)/I_S(2) ratio, after correction for extinction (Table 1), allows the derivation of the approximate densities of PDRs 1, 2, and 3 in Figure 2(b). In addition, since these lines are thermalized, the gas temperature, T, is coincident with the rotational temperature, T_rot, and we can infer T for the different PDRs (Table 2). To calculate the rotational temperature, we have assumed an ortho-to-para ratio of 3, that is the equilibrium value for temperatures larger than 100 K. Finally, using the value of G₀(T/10³)^1/2/n_H estimated with the PAH ionization fraction (or I_6.2/I_11.3), and the above derived T and n_HH, we can derive the intensity of the radiation field that illuminates the three PDRs and precisely position PDRs 1, 2, and 3 in Figure 2. The found values for n_H, T, andG₀, (Table 2) are consistent with Mon R2 being a dense and highly UV irradiated PDR as estimated from other molecular lines (Choi et al. 2000; Rizzo et al. 2003, 2005). The found values for density are lower than in these previous works because the present tracers (PAHs and H₂) probe the outermost cloud regions directly exposed to the UV radiation field that are usually warmer (>500 K) and less dense than regions situated deeper into the molecular cloud (see cut in Figure 1). Finally, the estimated parameters using this technique, for the entire Mon R2 region (see Table 2), are more consistent with the ones derived for spatially resolved PDRs 2 and 3.

The main results we have presented in this Letter are (1) Mon R2 is a unique template of high-UV/high-density PDRs that is spatially resolved, similar to the Orion bar (where the UV field is slightly lower); (2) our observations are consistent with PDR models that predict that in high UV and dense PDRs, the excitation of pure rotational H₂ lines is collisional and depends on gas temperature; (3) for this reason, the spatial distributions of PAHs and warm H₂ are different, contrary to what is seen in cool PDRs; (4) however, because PAH emission is present in the extended region illuminated by the radiation, their ionization fraction can be used to probe the intensity of the UV radiation field, even where H₂ emission peaks. Thus, the mid-IR spectrum of dense and highly irradiated PDRs appears as an efficient probe of the physical conditions in these environments. The ratio between the molecular H₂ 0–0 S(3)/S(2) line intensities allows to directly compute the density and temperature of the gas, while the measurement of the ionization fraction of PAHs allows to derive the intensity of the radiation field. We show that the derived parameters for the entire Mon R2 region are consistent with those found for the spatially resolved PDRs 2 and 3. This is likely because the dense PDR 1 occupies only a small part of the whole region, and therefore dilution effects imply that the spatially averaged spectrum is dominated by emission from lower density PDRs. Thus, we suggest that this spectral methodology could be useful to derive the dominant physical conditions in PDRs that are not spatially resolved in the mid-IR. In particular, the PDRs at the surface of protoplanetary disks around Herbig Ae/Be stars (Berné et al. 2009) or in the inner rim of T Tauri disks (Agúndez et al. 2008) are expected to be the targets where such an analysis can be applied. The mid-IR spectrum of starburst galaxies is probably dominated by the emission of PDRs having similar conditions as those described in this Letter (see Fuente et al. 2008 for the case of M82). Thus, the present methodology could be a useful tool to derive global properties of galaxies, in connection with massive star formation activity, using the forthcoming James Webb Space Telescope and SPICA telescope, which will observe these emission features at low and high redshifts.

We thank the anonymous referee for constructive criticism and a careful reading of the manuscript. This work is based on observations made with the Spitzer Space Telescope, which is operated by the Jet Propulsion Laboratory, California Institute of Technology under NASA contract 1407. O.B. is supported by JAE-Doc CSIC fellowship. J.R.G. was supported by a Ramon y Cajal research contract from the spanish MICINN and co-financed by the European Social Fund. O.B., P.P., and C.J. acknowledge the french national program PCMI.