The first CO⁺ image: I. Probing the HI/H₂ layer around the ultracompact HII region Mon R2

Abstract

The CO⁺ reactive ion is thought to be a tracer of the boundary between a HII region and the hot molecular gas. In this study, we present the spatial distribution of the CO⁺ rotational emission toward the Mon R2 star-forming region. The CO⁺ emission presents a clumpy ring-like morphology, arising from a narrow dense layer around the HII region. We compare the CO⁺ distribution with other species present in photon-dominated regions (PDR), such as [CII] 158 µm, H₂ S(3) rotational line at 9.3 µm, polycyclic aromatic hydrocarbons (PAHs) and HCO⁺. We find that the CO⁺ emission is spatially coincident with the PAHs and [CII] emission. This confirms that the CO⁺ emission arises from a narrow dense layer of the HI/H₂ interface. We have determined the CO⁺ fractional abundance, relative to C⁺ toward three positions. The abundances range from 0.1 to 1.9 ×10⁻¹⁰ and are in good agreement with previous chemical model, which predicts that the production of CO⁺ in PDRs only occurs in dense regions with high UV fields. The CO⁺ linewidth is larger than those found in molecular gas tracers, and their central velocity are blue-shifted with respect to the molecular gas velocity. We interpret this as a hint that the CO⁺ is probing photo-evaporating clump surfaces.

1. Introduction

Reactive ions are destroyed in almost every collision with H and H₂ and recombine rapidly with e⁻. These compounds present enhanced abundances toward the hot layers of the photon-dominated regions (PDRs), where the far ultraviolet (FUV) field is only partially attenuated and maintains high abundances of the parent species C⁺ and S⁺ (Sternberg & Dalgarno 1995). In particular, Sternberg & Dalgarno (1995) predict a high CO⁺ abundance at the HI/H₂ interface (A_V ≈ 1 mag) of dense PDRs, where it is mainly produced by the C⁺ + OH → CO⁺ + H reaction. So far, the CO⁺ ion has been detected in several PDRs; such as the M17SW, Orion Bar, NGC7027, NGC7023 (Latter et al. 1993; Stoerzer et al. 1995; Fuente & Martín-Pintado 1997; Fuente et al. 2003), G29.96, MonR2 (Rizzo et al. 2003) and S140 (Savage & Ziurys 2004). However, all the detections have been obtained after long integrations toward a single position and they lack the information on the spatial distribution.

The Mon R2 star-forming region, located at 830 pc (Herbst & Racine 1976), contains an ultracompact (UC) HII region surrounded by a series of PDRs with different physical conditions (Pilleri et al. 2013; Treviño-Morales et al. 2014). The main PDR, corresponding to IRS 1 (hereafter IF), is irradiated by a high UV field of G₀ > 10⁵ (in units of the Habing flux; Habing 1968), and presents high densities (> 10⁵ cm⁻³) and kinetic temperatures (T_k ≈ 600 K; Berné et al. 2009). A second PDR, associated with the molecular peak MP2, is detected 40″ north from IF, and shows chemical properties similar to those found in low- to mid-UV irradiated PDRs (Ginard et al. 2012). Due to its proximity and physical conditions, Mon R2 turns to be an excellent candidate to study the HI/H₂ interface. CO⁺ is thought to be a good PDR tracer, and its distribution is potentially an excellent diagnostic tool to learn about the physical structure of these regions.

In this paper, we present a study of the CO⁺ (J = 2–1) transition line toward Mon R2 and compare its spatial distribution with Spitzer data reported by Berné et al. (2009), Herschel data from Pilleri et al. (2012) and Ossenkopf et al. (2013) and the HCO⁺ and H¹³CO⁺ molecules from Treviño-Morales et al. (2014).

2. Observations and data reduction

We observed 2′ × 2′ maps of CO⁺ transition lines (J = 2–1 at 235.380 GHz, 235.789 GHz and 236.062 GHz) using the IRAM-30m telescope (Pico Veleta, Spain). The observations were performed using the EMIR receiver with the Fast Fourier Transform Spectrometer (FTS) at 200 kHz of resolution. Throughout this paper, we use the main-beam brightness temperature (T_MB) as intensity scale. The data were reduced using standard procedures with the CLASS/GILDAS package (Pety et al. 2005). The three lines were detected, but only the 236.0625 GHz has sufficient signal-to-noise ratio (S/N) for good imaging. In order to improve the S/N, we smoothed the native observation to an angular resolution of 16″ (see Fig. 1) and to a spectral resolution of 1 km s⁻¹. In the final datacube, CO⁺ has linewidths of 6–8 km s⁻¹ and intensity peaks of 60–200 mK (with rms~ 20 mK). The main CO⁺ line is located close to a bright ¹³CH₃OH line (at 236.0628 GHz). However, we dismiss the idea of a possible blending in the CO⁺ line, as in the spectral line survey conducted toward Mon R2 (Treviño-Morales 2016) we do not find ¹³CH₃OH emission at any frequency. Moreover, the main compound CH₃OH is not detected at the positions where CO⁺ is bright.

Fig. 1.

Color image shows the integrated emission (in K km s⁻¹) of the CO⁺ line, with the original (11″, Panel A) and smoothed (16″, Panel B to D) angular resolution. In Panels A to D, the gray contour levels range from 40% to 100%, in steps of 10% of the intensity peak, where the lower contour level corresponds to a S/N= 5σ. The yellow contour indicates the 3σ emission. The blue squares indicate the IF and MP2 positions, where IF corresponds to α(J2000)= 06h07m46.2s, δ(J2000)= −06°23′08.3″. Panel A shows the H¹³CO⁺ (3–2) emission (black contours) tracing the molecular gas (Treviño-Morales et al. 2014). Panel B shows the [NeII] emission (red contours) tracing the HII region and the emission of the H₂ S(3) rotational line at 9.7 µm (black contours). Black contours in Panel C show the PAHs (11.3 µm) emission. Panel D shows the [CII] emission at 158 µm (black contours, Pilleri et al. 2014). The Spitzer data are explained in Berné et al. (2009).

3. Results

In Fig. 1 we compare the CO⁺ spatial distribution with other species. Panel A shows the H¹³CO⁺ (3–2) line emission (black contours) tracing the molecular gas (Treviño-Morales et al. 2014), that is distributed around the CO⁺ emission. CO⁺ presents a clumpy structure, where the main CO⁺ clumps seem to have a counterpart in the H¹³CO⁺ (3–2) emission. However, the peaks of CO⁺ are located ≈ 5–10″ closer to the HII region, likely tracing an inner layer of the region. We find that CO⁺ emission appears surrounding the HII region with its intensity peak at the offset [0″, −7″], very close the IF position (see Panel B of Fig. 1). Moreover, the two most intense CO⁺ clumps are correlated with the H₂ emission, in an area where the density is presumably larger. Panels C and D show the PAHs and the [CII] emission, respectively (black contours). The CO⁺ emission present a clumpy ring-like distribution, spatially coincident with the PAHs emission. The CO⁺ secondary clumps are associated with the PAHs emission peaks, but not the most intense one. Panel D shows a comparison of the CO⁺ spatial distribution with the emission of its chemical precursor [CII]. These species are spatially associated, with the main difference being the location of the peaks: CO⁺ has its intensity peak to the south of the IF position, while the [CII] peak is located to the west. Therefore, we interpret that the CO⁺ is found toward the densest area of the region (where all the molecular gas piles up, e. g., the H¹³CO⁺ spatial distribution) as expected since the critical densities of H¹³CO⁺ (n_cr ~ 10⁵ cm⁻³) and CO⁺ (n_cr ~ a few 10⁵ cm⁻³; Stäuber & Bruderer 2009) are larger than that of [CII] (n_cr ~ a few 10³ cm⁻³; Goldsmith et al. 2012). It is worth noting that the CH⁺ molecule is also related to the CO⁺ and [CII] chemistry. When comparing their spatial distribution, we find that the CO⁺ emission also coexist with CH⁺ but, as [CII], its intensity peak is located to the west of the IF position (Pilleri et al. 2014).

In order to better understand the spatial distribution, we did intensity cuts with a position angle of 45° throughout the IF position and cutting the ring-like structure seen in the CO⁺ emission at the south-east and north-west (pink dashed line in left panel of Fig. 2). The intensity cuts for the species CO⁺, H¹³CO⁺, [CII], H₂ and PAHs are shown in Panels A to F of Fig. 2. The molecular gas as traced by H¹³CO⁺ shows emission between the offset–20″ and 0″(corresponding to the area between [20″,−20″] and [0″,0″] in the map) with no emission associated with the HII region. The most intense emission of CO⁺ comes from this region, but there is also 3σ-level emission associated with the HII region (gray-shaded area in the map). Note, however, that the CO⁺ peak is closer to the HII region than the H¹³CO⁺ peak. The [CII] emission is very intense and extended in the whole area, and its intensity cut shows two emission peaks: one coincident with the CO⁺ peak, and a second one (the brightest one) on the opposite edge of the ring-like structure (to the north-west). The PAHs emission is weaker than [CII] but their intensity cuts are quite similar. Finally, the H₂ S(3) line only shows emission over 3σ between the offset–10″ and 0″. Summarizing we can see a trend of spatial segregation, with the H¹³CO⁺ tracing the outer layers (far from the HII region), then CO⁺ and [CII] and PAHs peaking closer to the ionized gas, and finally the H₂ emission tracing a hotter layer close to the HII region (Panel F of Fig. 2).

Fig. 2.

Left: CO⁺ integrated emission (see Fig. 1). The pink-dashed line indicates the direction of the intensity cuts. The color circles indicate the positions where spectra were extracted. Right: Panel A to E show the intensity cuts (red continuum lines) of CO⁺, H¹³CO⁺, [CII], H₂ and PAHs. In these panels, the gray continuum lines indicates the errors in the cuts. Panel F shows the comparison of the intensity cuts scaled to unity. In Panels A to F, the pink dot-dashed indicate the 3σ level for each species and the gray area indicates the position of the UC HII region. Panel G shows the CO⁺, [CII], H¹³CO⁺ and HCO⁺ spectra at IF, MP2 and PDR3. The color of the boxes are related to this positions (circles in left panel).

The circles in the left panel of Fig. 2 mark the positions where we compare the spectral profiles of CO⁺, [CII], H¹³CO⁺ and HCO⁺ (Panel G). The positions correspond to (i) the ionization front (IF at offset [0″, 0″], red), (ii) the offset [−10″, 24″] (close to the PDR3 in Berné et al. 2009, black), and (iii) the MP2 position (offset [0″, 40″], blue) that is known to be a PDR with a low UV field and high density (Ginard et al. 2012). In Fig. A.1, we present a comparison of the CO⁺ line profile (in red) with the [¹³CII] (Ossenkopf et al. 2013), ¹³CO (10–9) and CO (9–8) lines at IF (Pilleri et al. 2012). We find that CO⁺ presents larger linewidth than ¹³CO, H¹³CO⁺ and HCO⁺ (≈ 7±0.7 km s⁻¹ vs 3– 5 km s⁻¹). The CO⁺ linewidth, however, is comparable to that of carbon recombination lines (≈6 km s⁻¹; Treviño-Morales 2016). Similarly, the CO⁺ line has a similar profile to that of the [¹³CII] and the high excitation CO (9–8) line. The CO⁺ and [¹³CII] spectra are closely similar in the line shape and line widths; this similarity may be well explained if CO⁺ lies in the dissociation front, between the ionized gas and the molecular gas. We find weak CO⁺ emission toward the HII region that might come from the back and front walls of the PDR.

On the basis of reactive ions (CH⁺, OH⁺, and H₂O⁺) observations, Pilleri et al. (2014) constructed a schematic view of the Mon R2 geometry. They find that the emission of the high density molecular gas seems to come from the back side (relative to the observer) of the HII region. This is confirmed by the detection of the cluster at infrared wavelengths (i. e., not obscured by in-front molecular gas). In this scenario, one expects that an expansion of the HII region is indicated by an excess of redshifted emission in the observed lines. However, while we find hints of the molecular tracers (H¹³CO⁺, CO) to be skewed toward red-shifted velocities, the PDR tracers (CO⁺, [¹³CII]) seem to have a spectral profile skewed toward blue-shifted velocities (Figs. 2 and A.1). In order to quantify this, we have measured the amount of emission in each spectral line that is blue-shifted and red-shifted with respect to the systemic velocity (10 km s⁻¹ for IF and PDR3, and 8 km s⁻¹ for MP2; Treviño-Morales et al. 2014). For the IF position, we find 55 ± 11% of the emission of PDR tracers to be blue-shifted, and only 33 ± 6% for the molecular tracers. Similarly, for MP2 and PDR3 we find 44±9% and 55 ± 11% of blue-shifted emission for the PDR tracers and 18 ± 4% and 39 ± 8% for the molecular tracers. In general, we have 51 ± 10% of the PDR-tracer emission to be blue-shifted, and only 30 ± 6% for the molecular tracers. Considering that the molecular gas is located behind the UC HII region, the difference between the PDR and molecular tracers can be explained if the PDR is formed by dense condensations that are being photo-evaporated. In this case, the photo-evaporated gas (PDR tracers) would be ejected toward us, and therefore, blue-shifted with respect to the molecular gas. This effect is also visible in Fig. A.1, where [¹³CII] presents its velocity peak at ~9 km s⁻¹ (blue-shifted) and the emission associated with the molecular gas is red-shifted. Finally, we do not find significant velocity gradients (< 1 km s⁻¹) between the CO⁺ clumps, suggesting low levels of turbulence between them, and consistent with the low expansion velocity found by Fuente et al. (2010) and Pilleri et al. (2012).

4. CO⁺ and HCO⁺ fractional abundances

On the basis of the spectra presented in Figs. 2 and A.1, we calculated the CO⁺, HCO⁺ and C⁺ column density (N) in ranges of 1 km s⁻¹ at the three selected positions (see Appendix B). We assume optically thin emission and a beam filling factor of 1 for all the species. The N[CO⁺] values have been calculated assuming a Boltzmann distribution of rotational levels with T_ex = 18 K. This value of the excitation temperature is based on calculations by Stäuber & Bruderer (2009) and the detection of several CO⁺ lines towards the Orion Bar (Cuadrado et al. 2015, S. Cuadrado, priv. communication). We used the MADEX large velocity gradient (LVG; Cernicharo 2012) and RADEX code (van der Tak et al. 2007), to derive N[H¹³CO⁺] and N[C⁺]. According to the physical conditions derived by Berné et al. (2009) from the H₂ ground state rotational lines, we assume T_k = 600 K, n_H = 4 × 10⁵ cm⁻³ for the IF and T_k = 300 K, n_H = 4 × 10⁴ cm⁻³ for PDR3. Unfortunately, the MP2 position is out of the Spitzer map; for this position, we make a reasonable guess of T_k = 300 K and n_H = 2 × 10⁵ cm⁻³. The assumption of optically thin emission is not valid for the HCO⁺ (3–2) line at velocities close to cloud systemic velocity. Thus, between 8–11 km s⁻¹, we derived the N[HCO⁺] using the rarer isotopologue line H¹³CO⁺ (3–2) and assuming ¹²C/¹³C= 50 (Treviño-Morales et al. 2014). The derived H¹³CO⁺ excitation temperatures (~10 –16 K) toward the IF and MP2 are in agreement with those measured by Treviño-Morales et al. (2014). The collisional rate coefficients for HCO⁺ and H¹³CO⁺ are taken from Flower (1999). To derived the N[C⁺] at the IF position, between 6–12 km s⁻¹, we used the rarer isotopologue line [¹³CII]. A significant fraction of the [CII] emission is expected to come from the atomic layer. The calculated C⁺−H collisional rates are similar to those with H₂ within a factor 1.3 (Wiesenfeld & Goldsmith 2014, Barinovs et al. 2005). Hence, the relevant parameter regarding collisional excitation is the number of particles, either H or H₂. In our calculations we assume that the hydrogen is in molecular form which implies an uncertainty of a factor of 2 in the assumed density. Taking into account that we are well over the critical density of the [CII] 158 µm line, this translates into an uncertainty of <30% in the N[C⁺]. The obtained N[C⁺] are in good agreement with those obtained by Ossenkopf et al. (2013).

We have computed the N[CO⁺]/N[HCO⁺] ratio. We find values between 0.01–0.1 toward IF, with the highest values in the velocity wings. Note that we have not detected CO⁺ toward the PDR with lower UV field (MP2) with a significant upper limit of N[CO⁺]/N[HCO⁺]<0.008. Toward PDR3, we obtain values of N[CO⁺]/N[HCO⁺]~ 0.006, i. e., a factor of ≈2 lower than IF. C⁺ is known to be a good probe of the skin (A_v < 4 mag) of PDRs. Because of the similar spatial distribution and velocity profiles between the CO⁺ and C⁺, we used C⁺ to estimate the absolute fractional abundance of CO⁺. We can safely assume that almost all the carbon is in C⁺ in the region from which CO⁺ is coming. Assuming a carbon elemental abundance of 10⁻⁴ (Ossenkopf et al. 2013, Wakelam & Herbst 2008), we derive X[CO⁺] between a few 10⁻¹¹ to ~ 1.9 × 10⁻¹⁰ toward both, IF and PDR3. It is worth noting that the beam of CO⁺, HCO⁺, H¹³CO⁺, [¹³CII] and [CII] are quite similar, thus the calculated N[CO⁺]/N[HCO⁺] and X[CO⁺] values are not affected by beam filling factors. Our results are in good agreement with the model predictions presented by Sternberg & Dalgarno (1995), for G₀ ~ 5 × 10⁵ and n_H ~ 10⁶ cm⁻³. More recently, the models of Spaans & Meijerink (2007) predict the X[CO⁺] in PDRs, for n_H = 10⁵ cm⁻³ and G₀ = 10^3.5. However, the high cosmic ray ionization rate (about 100 times larger than the Galactic value) prevents us from a direct comparison with Mon R2. In general, the production of CO⁺ seems to depend on the temperature of the gas. Stäuber & Bruderer (2009) suggest that X[CO⁺] of about 10⁻¹¹ are only reached in gas with T_k ≥ 300 K. As T_k depends on G₀ and n_H, it is expected that the production of CO⁺ only occurs in regions with n_H ≥ 2 × 10⁴ cm⁻³ and G₀ ≥ 10³. The densities and G₀ measured in IF and PDR 3 are in good agreement with these values, as IF presents G₀ ~ 5 × 10⁵ and n_H ≥ 5 × 10⁴ cm⁻³ (Rizzo et al. 2003; Fuente et al. 2010) and PDR3 presents G₀ ~ 3.7 × 10⁴ and n_H ~ 3.7 × 10⁴ cm⁻³ (Berné et al. 2009).

5. Discussion and summary

We present a CO⁺ map toward Mon R2 star-forming region. This is the first map ever reported of this reactive ion. The spatial distribution of CO⁺ consists of a ring-like structure (similar to PAHs), tracing the layer between the HII region and the molecular gas. The maps reveal a clumpy structure in the hot layer of the mainly atomic gas. Previous works (Young et al. 2000; Goicoechea et al. 2016) suggest that there exist fragmentation in the photodissociation front. This is, there are not uniform layers between the HII region and the molecular cloud, but they present clumps that allow the radiation to penetrate deeper into the cloud. In this scenario where the PDR is conformed by a series of clumps, the emission of PDR tracers would be related to the external layers of dense clumps being photo-evaporated by the UV radiation. Despite the moderate angular resolution of our observations, we find hints that favor this scenario: (a) the spatial distribution of the CO⁺ as observed in the higher-angular resolution (11″) map (Panel A of Fig. 1) suggests that the CO⁺ emission is coming from the illuminated surface of the H¹³CO⁺ clumps, and b) the excess of blue-shifted emission seen for the PDR tracers in comparison with the molecular tracers. Considering that the molecular gas is located behind the UC HII region (Pilleri et al. 2014) and the chemical segregation, the difference in velocity between tracers can be explained if the PDR is formed by dense condensations that are being photo-evaporated. In this case, the photo-evaporated gas (PDR tracers) would be ejected toward us, and therefore, blue-shifted with respect to the molecular gas. Future higher angular resolution observations will help to confirm or discard this scenario.

Finally, we have determined X[CO⁺] in three positions. Toward IF we derive an abundance of a few 10⁻¹¹, in agreement with chemical model predictions (Sternberg & Dalgarno 1995) for n_H ~ 10⁶ cm⁻³ and G₀ ~ 5 × 10⁵. Toward MP2 we do not detect CO⁺ emission with an upper limit to the CO⁺ abundance of < 4 × 10⁻¹¹. Abundances of 10⁻¹¹–10⁻¹⁰ had been previously observed in PDRs with G₀ > 10³ Habing field (M17SW: Latter et al. 1993, Stoerzer et al. 1995; Orion Bar: Fuente & Martín-Pintado 1997; NGC7023: Fuente et al. 2003; G29.96–0.02: Rizzo et al. 2003). The non-detection of CO⁺ in MP2, together with the abundances found in the other PDRs, suggest that the production of CO⁺ only occurs in dense regions with high radiation fields. High UV fields (G₀ > 10³) and n_H (> 2 × 10⁴ cm⁻³) are required to achieve gas temperatures ≥ 300 K that are necessary to produce high abundances of OH in the external layer of the PDR (A_V ~ 1 mag; Stäuber & Bruderer 2009). This is also consistent with the non-detection of CO⁺ in the Horsehead PDR (Goicoechea et al. 2009), where the UV field is G₀ ~ 100 and presents chemical properties similar to MP2 (Ginard et al. 2012). A counter-example that challenge this interpretation could be the detection of CO⁺ toward S140 where the incident UV field is estimated to be G₀ ~ 100–300 (Savage & Ziurys 2004). However, the number of CO⁺ detections is scarce and the statistics is not enough to draw firm conclusions between the relation of the CO⁺ and the physical properties (n_H and G₀) of PDRs. A larger sample of objects need to be studied, including maps to characterize and understand the spatial distribution of the CO⁺ in different environments.