EDP Sciences
Free Access
Issue
A&A
Volume 608, December 2017
Article Number A12
Number of page(s) 11
Section Interstellar and circumstellar matter
DOI https://doi.org/10.1051/0004-6361/201730504
Published online 29 November 2017

© ESO, 2017

1. Introduction

Solid carbon dioxide (CO2) is an important constituent of icy grain mantles in the interstellar medium (ISM) with an abundance of 10–50% with respect to water ice (H2O; Gerakines et al. 1999; Nummelin et al. 2001; Boogert et al. 2015).

Interstellar solid CO2 has been detected in the mid-infrared through its two active fundamental vibration modes, e.g. the ν3 stretching mode at 4.27 μm (2340 cm-1) and the ν2 bending mode at 15.2 μm (660 cm-1). The bending mode feature is especially sensitive to the ice temperature and composition (polar and apolar mixtures, e.g. Ehrenfreund et al. 1996), but it has a weaker band strength than the stretching mode. The detection of the combination modes ν1 + ν3 and 2ν2 + ν3 at 3707 cm-1 and 3600 cm-1 toward the protostellar object S140:IRS1 was reported by Keane et al. (2001). More than two decades ago, interstellar CO2 was first detected in the solid phase by the low-resolution (λ/ Δλ in the range 10 to 60) Infrared Astronomical Satellite (IRAS) by the vibrational bending mode (d’Hendecourt & Jourdain de Muizon 1989). With the advent of the short-wavelength spectrometer (SWS) on board the Infrared Space Observatory (ISO) in 1995, detection of CO2 was made easier with a better resolution (λ/ Δλ ~ 500 and 1000). Soon solid CO2 was found to be ubiquitous in a variety of different astronomical environments. In particular, it was found toward the Galactic Centre (de Graauw et al. 1996), massive protostars (Gerakines et al. 1999; Gibb et al. 2004), low-mass young stellar objects (YSOs; Nummelin et al. 2001), and toward some background stars (Whittet et al. 1998). ISO observations toward cold, quiescent dark clouds were limited, however, because of the characteristic faintness of the sources and the sensitivity of the SWS instrument (Valentijn & Thi 2000). In addition, fringing was often present in standard processed SWS data (Nummelin et al. 2001). More recently, observations of the CO2 bending mode band obtained with the Infrared Spectrograph (IRS) aboard the Spitzer Space Telescope allowed the identification of CO2 towards faint sources in dark quiescent dense molecular clouds (Werner et al. 2004).

Despite the ubiquitous presence of CO2 in the interstellar medium, its formation route is highly debated, especially in cold, dark quiescent dense molecular clouds. The CO2 molecule is more abundant in the solid phase than in the gas phase (van Dishoeck et al. 1996; Boonman et al. 2003), which indicates that it is formed on the surface of dust grains by grain surface reactions and/or energetic processing (Nummelin et al. 2001). Potential non-energetic formation routes in the solid phase have been discussed in many laboratory experiments, theoretical models, and observations. Whittet et al. (1998) suggested that observation of CO2 in the quiescent dark cloud towards the line of sight of Elias 16, a field star, is solely due to chemical reactions occurring without the input of energy. They further argued that CO2 formation can occur without the radiation from a local embedded protostar and since the line of sight is a field star, which is behind a dense molecular cloud, they assumed that local chemical reactions around ice mantles are not triggered by energetic processing.

Theoretical models and laboratory experiments are primarily based on these two reaction schemes: Reaction (1) has been suggested as the simplest route to CO2 formation by Tielens & Hagen (1982). Furthermore, there is a high reaction barrier for the formation of CO2 in the gas phase, which can be lowered on surfaces via hot O-atom or Eley-Rideal mechanism (Talbi et al. 2006; Goumans et al. 2008). Laboratory experiments also suggest that the reaction is feasible upon heating, or via reaction in water pores or under a water ice cap (Roser et al. 2001). However, Raut & Baragiola (2011) have shown by means of infrared spectroscopy and micro-gravimetry analysis that only a small quantity of CO2 is formed upon co-deposition of CO and thin films of O and O2 at 20 K. The reaction of O atoms to form O3 is found to be more efficient in the above reaction. Minissale et al. (2013) have recently suggested that there is a competition between reactions (1) and (2) for efficient CO2 formation in space. The efficiency of CO2 formation via reaction (1) depends on the temperature of the grain surface and the abundance of O atom in the observed region. On the other hand, theoretical models and laboratory experiments both suggest that reaction (2) can be a reaction pathway to the formation of CO2 without energetic processing. In their theoretical model, Goumans et al. (2008) suggested CO2 formation via an HO-CO intermediate. Independent experimental studies (Oba et al. 2010; Ioppolo et al. 2011; Noble et al. 2011) further strengthen this reaction pathway. More recently, the laboratory work of Ioppolo et al. (2013a) has shown that reaction (2) is ten times more efficient than reaction (1) under the same interstellar relevant conditions. Minissale et al. (2015) have suggested an efficient CO2 formation via the following reaction scheme in the solid phase: (3)The role of energetic processing such as irradiation by UV photons, ions, and electrons in the formation of CO2 in the solid phase has been extensively studied in several laboratory experiments (d’Hendecourt et al. 1986; Moore et al. 1991; Bernstein et al. 1995; Gerakines et al. 1996; Ehrenfreund et al. 1997; Palumbo et al. 1998; Watanabe & Kouchi 2002; Mennella et al. 2004; Mennella et al. 2006; Loeffler et al. 2005; Jamieson et al. 2006; Ioppolo et al. 2009; Ioppolo et al. 2013b; Laffon et al. 2010; Garozzo et al. 2011). Quiescent dark molecular clouds are shielded from the external UV field. However, cosmic rays can penetrate dense clouds and therefore induce energetic processing in these regions. Moreover, cosmic rays can induce an internal UV field through the interaction with molecular hydrogen gas in dense clouds (Prasad & Tarafdar 1983; Jenniskens et al. 1993; Mennella et al. 2003; Shen et al. 2004).

In this manuscript first we present the experimental results relative to CO2 formation in three different ice mixtures by ion irradiation with 200 keV H+ at low temperature. Then we use laboratory spectra to fit the profile of the CO2 bending mode band observed by the Spitzer Space Telescope in the line of sight to quiescent dark cloud regions to understand the chemical processes in the earliest stages of star formation. In the fitting procedure, in addition to the new experiments here presented, we used some selected laboratory spectra, which are available in the Catania database1. Furthermore, we discuss the quality of the fits obtained and the quantitative contribution of each ice component to the fit including solid CH3OH in the formation of CO2 upon energetic processing of the ice.

2. Experimental methods

The experimental results presented here were obtained in the Laboratorio di Astrofisica Sperimentale at INAF – Osservatorio Astrofisico di Catania (Italy). The experiments were performed in a stainless steel ultra-high vacuum (UHV) chamber with a base pressure better than 10-9 mbar. A closed-cycle helium cryostat is attached to the UHV chamber and is used to cool a KBr substrate which is in thermal contact with the sample holder of the cryostat. The substrate temperature can be varied between 16 and 300 K. The chamber is also connected to a 200 kV ion implanter (Danfysik 1080) which can produce ions in the energy range between 30 and 200 keV (or 400 keV for double ionization). The ion implanter is optimized to accelerate ions from gas phase species (such as H, He, Ar, C, O, and N). In the present study all samples were irradiated with 200 keV H+.

Table 1

New experiments performed in this work.

Pure gases or mixtures are admitted into the main chamber by means of a needle valve and deposited onto the KBr substrate at low temperature. A He-Ne laser (λ = 543.5 nm) is used to monitor the ice thickness on the substrate during accretion. More details on this procedure can be found in Baratta & Palumbo (1998) and Urso et al. (2016). Furthermore, ices are studied by using a Fourier transform infrared (FTIR) spectrometer (Bruker Vertex 70) which works in the spectral range between 7500 and 400 cm-1 (1.33–25 μm) with a resolution of 1 cm-1. The substrate forms an angle of 45° with respect to both the ion beam and the IR beam. This allows the acquisition of transmittance spectra without needing to turn the sample. Spectra are taken before and after irradiating the sample at low temperature (17 K) and also at selected higher temperatures after annealing the substrate up to 90 K. A schematic depiction of the UHV chamber is presented in Fig. 1.

thumbnail Fig. 1

Schematic depiction of the UHV chamber.

Open with DEXTER

The experiments performed in this work are listed in Table 1. Spectra obtained in transmittance units (If) are converted into optical depth units as τ(ν) = ln(I0/If), where I0 is the normalization continuum. Furthermore, the column density (N, molecules cm-2) of each species is calculated from the respective band area τ(ν)dν of each IR band divided by the band strength (A) value obtained from literature. All band strength values used in the current work are listed in Table 2.

The ion beam is electrostatically swept and produces a spot on the sample which is larger than the area probed by the IR beam. The penetration depth of the impinging ions (200 keV H+) is greater than the sample thickness (about 1 μm) as verified using the Srim code (Ziegler et al. 2008). Macroscopic heating of the substrate is avoided by keeping the ion beam current density lower than 1 μA cm-2.

A rotatable polarizer is placed in the path of the IR beam, which enables us to subsequently record spectra in P (i.e. the electric vector is parallel to the plane of incidence) and S (i.e. the electric vector is perpendicular to the plane of incidence) mode where the plane of incidence is the plane of the paper in Fig. 1. Background spectra of the bare cold substrate are recorded before ice deposition in both P and S modes and are subsequently subtracted from each of the following P and S spectra of the ice.

The energy released to the sample by the impinging ions (dose) is given in eV/16u, where u is the unified atomic mass unit defined as 1/12 of the mass of an isolated atom of carbon-12. This dose is obtained by calculating the average stopping power (eV cm2/molecule) using the Srim software and multiplying it by the ion fluence (ions cm-2) measured during each experiment. As discussed by Strazzulla & Johnson (1991), the dose given in units of eV per small molecule (16u) is a convenient way to characterize chemical changes and to compare the results obtained by irradiating different samples.

As discussed in previous studies (Ioppolo et al. 2009, 2013b), the profile of the CO2 bending mode band produced after ion bombardment does not depend on the polarization mode. When the band profile does not depend on the polarization mode, laboratory transmittance spectra directly correlate with the variation of the absorption coefficient in the solid sample (Baratta et al. 2000; Palumbo et al. 2006). Hence, a direct comparison between laboratory spectra and astronomical observations is possible. In this work, only spectra in P mode are taken into account owing to their better signal-to-noise ratio.

In addition to the new experiments presented here, we used some selected laboratory spectra which are available in the Catania database to fit the astronomical observations (Ioppolo et al. 2009, 2013b). A full list of the laboratory spectra considered in this work is given in Table 3. As an example, the spectra relative to pure CO after bombardment with 200 keV H+ at 16 K, a mixture CO:N2 = 1:1 after bombardment with 200 keV H+ at 16 K, and pure CH3OH after bombardment with 200 keV H+ at 16 K are shown in Fig. 2. The main features formed after bombardment are labelled. The vertical grey arrows indicate the spectral range where most of the features assigned to carbon chain oxides and nitrogen oxides, also formed after bombardment, occur. A detailed description of the spectra listed in Table 3 is given by Palumbo et al. (2008), Modica & Palumbo (2010), Sicilia et al. (2012), Ioppolo et al. (2013b), and Kaňuchová et al. (2016).

Table 2

List of the band strength (A) values used.

Table 3

Laboratory spectra used to fit observations.

thumbnail Fig. 2

Selected infrared spectra, in the range 3800–560 cm-1, used in the fitting procedure. From top to bottom the spectra refer to pure CO after bombardment with 200 keV H+ at 16 K, a mixture CO:N2 = 1:1 after bombardment with 200 keV H+ at 16 K, and pure CH3OH after bombardment with 200 keV H+ at 16 K. Spectra are normalized to their maximum optical depth value and are shifted for clarity. The main features formed after bombardment are labelled. The vertical grey arrows indicate the spectral range where most of the features assigned to carbon chain oxides and nitrogen oxides occur. A detailed description of these spectra is given by Palumbo et al. (2008), Sicilia et al. (2012), and Modica & Palumbo (2010).

Open with DEXTER

3. Results

3.1. Experimental results

3.1.1. H2O:CO mixture

A H2O:CO = 8:1 mixture is deposited onto a cold KBr substrate at 17 K and then irradiated with 200 keV H+ to study the formation of CO2. Figure 3 shows the column density of CO2 divided by the initial column density of CO, N(CO2)/N(CO)i, as a function of irradiation dose. The column density of CO is calculated using the band area of the CO stretching mode at 2139 cm-1. The column density of CO2 is calculated using the band area of the C=O stretching mode at 2345 cm-1. The corresponding band strength values (A) are listed in Table 2. As shown in Fig. 3, the CO2 column density rapidly increases at low doses and then almost reaches an equilibrium value at the highest investigated dose. The value of the stopping power used to calculate the dose is given in Table 1.

thumbnail Fig. 3

Column density of CO2 divided by the initial column density of CO after ion bombardment with 200 keV H+ of a H2O:CO = 8:1 mixture at 17 K. Experimental data are fitted by an exponential curve.

Open with DEXTER

The experimental data relative to the formation of CO2 are fitted by the exponential curve (4)where y is the N(CO2)/N(CO)i ratio, Y is the asymptotic value, σ is a parameter (in 16u/eV), and D is the dose (in eV/16u). The parameter σ is often referred to as the cross-section (see Rothard et al. 2017, for a recent review) and according to chemical kinetics it can be regarded as a rate constant in units of dose-1 instead of time-1. The fitting procedure gives Y = 0.64 ± 0.01 and σ = 0.08 ± 0.01 16u/eV. The value of Y obtained in this experiment is comparable to the value obtained previously by Ioppolo et al. (2009) for the same mixture irradiated with 30 keV He+, which suggests that the equilibrium value of CO2 produced does not depend on the energy and mass of the ion used within experimental uncertainties.

3.1.2. CO:CH4:N2 mixture

A mixture CO:CH4:N2 = 1:1:1 deposited at 17 K and irradiated by 200 keV H+ is used to retrieve information on the formation of CO2 in an apolar environment. Figure 4 shows the column density of CO2 divided by the initial column density of CO, N(CO2)/N(CO)i, as a function of irradiation dose. The column densities of CO and CO2 are calculated following the same procedure described above (Sect. 3.1.1). The experimental data are fitted by Eq. (4) in this case as well. The fitting procedure gives Y = 0.080 ± 0.001 and σ = 0.027 ± 0.001 16u/eV. The value of Y obtained in this experiment is similarly comparable to the value obtained previously by Ioppolo et al. (2009) for the same mixture irradiated with 30 keV He+. This again suggests that the equilibrium value of CO2 does not depend on the ion used.

Table 4

Observed column densities (in units of 1017 cm-2) of CO, H2O, CO2, and CH3OH and AV values for the lines of sight studied here.

thumbnail Fig. 4

Column density of CO2 divided by the initial column density of CO after ion bombardment of a CO:CH4:N2 = 1:1:1 mixture at 17 K. Experimental data are fitted by an exponential curve.

Open with DEXTER

3.1.3. H2O:CH4:NH3 mixture

In order to investigate the formation of CO2 in an ice mixture that does not contain CO, we considered a mixture of H2O:CH4:NH3 = 1:1:1 at 17 K. Figure 5 shows the column density of CO2 after ion bombardment of 200 keV H+ divided by the initial column density of CH4. The column densities of CH4 and CO2 are calculated using the band area of the C-H stretching mode of CH4 at 3010 cm-1 and the C=O stretching mode of CO2 at 2345 cm-1 together with the band strength values (A) listed in Table 2. We see that there is a clear difference in the formation trend of CO2 at low doses with respect to higher doses. The experimental data are therefore fitted by two exponential curves with a threshold dose (D0) to take into account the change in the CO2 production rate. The variation of CO2 formation rate at different doses has been observed previously by Ioppolo et al. (2009) in a H2O:CH4 = 4:1 mixture irradiated with 30 keV He+, which suggests that regardless of the type of ion used during irradiation, this trend generally occurs for CH4-containing ice mixtures. More hydrocarbons are formed at low doses, as shown by Baratta et al. (2002) for pure CH4 ice, while column densities of CO and CO2 are low. At higher doses, the production rate of CO and CO2 increases, while the yield of hydrocarbons decreases. In particular, we found for DD0(5)and for D >D0(6)where D0 is the threshold dose. The fitting procedure gives D0 = 102 ± 12 eV/16u, Y1 = 0.02 ± 0.02, σ1 = 0.001 ± 0.001 16u/eV, Y2 = 0.14 ± 0.04, and σ2 = 0.0013 ± 0.0004 16u/eV.

thumbnail Fig. 5

Column density of CO2 divided by the initial column density of CH4 after ion bombardment of a H2O:CH4:NH3 = 1:1:1 mixture at 17 K. Experimental data are fitted by two exponential curves.

Open with DEXTER

thumbnail Fig. 6

Three-component fit. Laboratory spectra are: H2O on C-grains +200 keV H+, CO (pure or mixed) +200 keV H+ (16 K), and CO (pure or mixed) +200 keV H+ (40 K).

Open with DEXTER

3.2. Fit to observations

Previous studies compared the profile of the CO2 bending mode observed towards high-mass (Ioppolo et al. 2009) and low-mass (Ioppolo et al. 2013b) YSOs and one field star (CK2) with the laboratory spectra of CO2 formed after ion bombardment of C- and O-bearing ice samples. It was also extensively discussed that the CO2 bending mode band profile is strongly sensitive to ice mixture and temperature. Here, we extend on the work of Ioppolo et al. (2009, 2013b) by selecting several sources along quiescent dense molecular clouds (i.e. background sources) and using the CO2 vibrational bending mode band as observed by the IRS of the Spitzer Space Telescope (Bergin et al. 2005; Knez et al. 2005; Whittet et al. 2007, 2009) to characterize the formation of CO2 after energetic processing of interstellar relevant ices.

thumbnail Fig. 7

Comparison between the band profile of the CO2 bending mode in H2O on C-grains +200 keV H+ and in a H2O:CO (8:1) mixture +200 keV H+.

Open with DEXTER

The selected lines of sight are listed in Table 4. As discussed by Whittet et al. (2007, 2009), the profiles of the CO2 bending mode band towards these lines of sight are very similar to each other and there is no clear evidence for structures that could be indicative of thermal processing of ices. We followed a systematic fitting procedure of the observational spectra with laboratory spectra using the software Polyfit, which we developed in-house, which uses a linear combination of up to five different laboratory spectra to obtain the best fit. In all the figures which show the results of the fitting procedure, observational data are always represented by a step horizontal centre light grey line, the black solid line is the best fit result, and different laboratory components are represented by solid thick coloured lines. It is important to stress that the computed fits are not unique, and only represent a possible solution that can reproduce the observed CO2 vibrational bending mode profile towards quiescent dark clouds. As shown in Fig. 6, a reasonable agreement is reached between the laboratory spectra and the observed spectra by using three fit components of the CO2 bending mode band profile obtained after irradiation of relevant ice mixtures.

Our choice of fit components is based on observational evidence. In dense molecular clouds, H2O, CO, and other molecules appear as a frozen mantle on the surface of dust grains, as shown by the 3.1 μm and 4.67 μm absorption features of solid H2O and CO ices (Whittet et al. 1988; Chiar et al. 1995), respectively. Cosmic rays can penetrate quiescent dark molecular clouds that induce processing of the ices and produce CO2; this molecule has a comparable formation threshold to that of H2O (Bergin et al. 2005; Boogert et al. 2015), which suggests that there is a close chemical connection between these two species. Moreover, in the initial stages of cold quiescent dark clouds, H2O is formed on the grain surface (van Dishoeck et al. 2013) and previous laboratory experiments have shown efficient CO2 production by ion irradiation of H2O on C-grain surfaces (Mennella et al. 2004; Raut et al. 2012; Ioppolo et al. 2013b). Taking into consideration all these factors, we used CO2 formed after irradiation of H2O on C-grains as a common component that is used to fit a broad contribution to the band profile, as discussed by Ioppolo et al. (2013b). Moreover, some CO may be trapped in a water-rich ice before its “catastrophic” freeze-out at higher AV. Therefore, we also included new experimental results of CO2 formation by proton irradiation of a H2O:CO = 8:1 mixture. Figure 7 shows that the CO2 vibrational bending mode profile obtained by irradiating a H2O:CO = 8:1 mixture is very similar to ion irradiation spectra of H2O on C-grains. Both profiles are broad and either of them can be used as the polar component. The other two components are a combination of either pure CO ice or CO mixed with H2O or N2 ice in different proportions irradiated with 200 keV H+. In all these fits we have always considered the spectra of CO2 formed after irradiation of either pure CO or CO containing mixtures at 16 K (cold component), which fits the narrow feature of the observed band profile at 660 cm-1 and the spectra of the same irradiated ice samples warmed up to 40 K (warm component), which fits the weak shoulder of the observed profile at about 655 cm-1. The laboratory spectra used for the systematic fit of the observed spectra are listed in Table 3.

CH3OH ices.

Solid methanol (CH3OH) has been detected in the line of sight to YSOs and background sources (e.g. Boogert et al. 2015). Its abundance with respect to solid H2O ranges between <1% and 31% towards YSOs and between <1% and 12% towards background sources, with a median value of 69%, which seems to be independent of the nature of the source. It is generally accepted that CH3OH forms on grain surfaces after CO hydrogenation (Watanabe et al. 2004; Fuchs et al. 2009). With this in mind, we also present four-component fits computed by adding the CO2 bending mode profile obtained after irradiation of pure CH3OH ice at 16 K as an additional component to the three-component fits reported in Fig. 6. The four-component fits, shown in Fig. 8, have been computed only for the three observational spectra which have the best signal-to-noise ratio and have an observed CH3OH column density given in Table 4.

thumbnail Fig. 8

Four-component fit. Laboratory spectra are: H2O on C-grains +200 keV H+, CO mixtures +200 keV H+ (16 K), and CO mixtures +200 keV H+ (40 K) and CH3OH +200 keV H+ (16 K).

Open with DEXTER

Table 5

Expected column density (in units of 1017 cm-2) of CO2 obtained after irradiation of three components – H2O on C-grains +200 keV H+, CO (pure or mixed) +200 keV H+ (16 K), and CO (pure or mixed) +200 keV H+ (40 K) – and column density of CO2 obtained from the fit for some selected sources.

Table 6

Expected column density (in units of 1017 cm-2) of CO2 obtained after irradiation of four components – H2O on C-grains +200 keV H+, CO mixtures +200 keV H+ (16 K), CO mixtures +200 keV H+ (40 K), and CH3OH +200 keV H+ at 16 K – and column density of CO2 obtained from the fit for some selected sources.

3.3. Observational constraints

In Table 4, we have listed all the sources along with the observed column densities of H2O, CO, CO2, CH3OH, and the visual extinction AV for each selected line of sight. In order to check that our fitting analysis provides us with the best fit and also a physically meaningful comparison between observational and laboratory spectra, we calculated the expected column density of CO2 after cosmic ion bombardment of H2O on C-grains, pure and mixed CO, and CH3OH using Eqs. (7), (8a), and (9), respectively, (7)

(9)where Nobs is the observed column density.

These equations are based on the results reported by Mennella et al. (2006) and Ioppolo et al. (2009, 2013b) assuming that the interstellar cloud lifetime is 3 × 107 yr. To derive the Ntot(CO), we correlate it with the column density of H2 using Eq. (8b) as suggested by Liu et al. (2013). The column density of H2 has been computed using Eq. (8c) following Knez et al. (2005). The CO2 column density obtained after the fitting procedure and the expected CO2 column density (Nexp(CO2)) are listed in Tables 5 and 6. As discussed by Ioppolo et al. (2009, 2013b), the calculated expected column density values have to be regarded as upper limits.

4. Discussion

The bulk of interstellar ices comprises H2O, CO, CH3OH and CO2. Other species, such as CH4 and NH3, have been detected in the line of sight to young stellar objects and although less abundant they still represent a significant fraction of the ice (between 1% and 10% with respect to H2O; Boogert et al. 2015). With this in mind, we present the proton irradiation of the CO:CH4:N2 = 1:1:1 mixture and H2O:CH4:NH3 = 1:1:1 to investigate CO2 formation in such ice mixtures. Although it can be argued that the ice mixtures in these ratios are not relevant in dense molecular clouds, nonetheless such experiments can provide the values of the formation cross-section of CO2 in proton-irradiation experiments in these mixed ices (e.g. Rothard et al. 2017), and on the other hand reveal the variation in the CO2 bending mode band profile, due to the presence or absence of these species.

In this work we used the new experiments presented here and those from the Catania database to systematically fit the observed CO2 band profile using the vibrational bending mode band at 15.2 μm (660 cm-1). Similar fits have been computed by Ioppolo et al. (2009, 2013b) for YSOs and by Mennella et al. (2006) and Ioppolo et al. (2013b) for some field stars. Here we have extended this study to nine field sources.

Previous studies (Knez et al. 2005; Bergin et al. 2005; Whittet et al. 2009) have already compared the profile of the CO2 bending mode band observed towards background sources with laboratory spectra of co-deposited species. In these investigations laboratory spectra of polar (e.g. H2O:CO2) and apolar (e.g. CO:CO2) mixtures are considered. From a qualitative point of view the fit obtained in the present work are equivalent to those obtained in previous studies.

The comparison between laboratory spectra shows that the profile of CO2 bands formed after ion irradiation of either CO or CH3OH is similar to the profile of CO2 bands when CO2 is co-deposited with CO or CH3OH. Similarly, the profile of the CO2 bands formed after ion irradiation of H2O on C-grains or H2O:CO mixtures is similar to the profile of CO2 bands when CO2 is co-deposited with H2O. However, the advantage of fitting the astronomical observations with the spectra of ion irradiated mixtures over the co-deposited mixtures is that the fit not only provides information on the chemical and physical environment in which CO2 is present, but also gives clues to the formation process of CO2 in ice grain mantles and allows a quantitative estimation of the reliability of the results. To the best of our knowledge, the formation of interstellar CO2 after cosmic-ion irradiation is the only process that has been checked to give a good agreement, from a qualitative and a quantitative point of view, with observations of solid CO2 towards high-mass (Ioppolo et al. 2009) and low-mass (Ioppolo et al. 2013b) YSOs and background sources (this work). Furthermore it can justify the presence of CO2 concurrently in polar and apolar environments along the same line of sight.

Here we present two sets of reasonable fits, one with three components (Fig. 6) and another with four components (Fig. 8). Close examination of the CO2 bending mode band profiles for the observed sources show that they look identical and have a single asymmetric peak (at 660 cm-1) with a broad shoulder on the lower wavenumber side of the band at about 650 cm-1. Since our sources are cold, dark quiescent clouds and our observed CO2 band profile lacks a double peak, we tried our fits with cold and slightly warm components. Two-component fits obtained by irradiated H2O-dominated (polar) and CO-dominated (apolar) component at 16 K are not sufficient to produce reasonable fits for all the sources, as suggested above. So we tried the irradiated warm CO-dominated (apolar) component at 40 K, which reproduces the weak shoulder (at about 655 cm-1) quite nicely. Reliable results with a 30 K warm component were previously obtained for the background sources CK2 and Elias 16 by Knez et al. (2005) to fit the observed CO2 bending mode feature, and also by Chiar et al. (1994) to fit the CO stretching mode feature of the field star CK2. Hence, the choice of a slightly warm component to obtain a good fit is not unreasonable. Furthermore, if we imagine the long line of sight towards the field star then we might find a small temperature gradient along the observed line since the edges are not shielded and are more open to thermal processing. A small contribution from a warm apolar component improves the fit. In addition, we also tried to replace this component with a spectrum of irradiated CH3OH (polar) because the shoulder can be caused by the interaction between solid CO2 and alcohols as discussed in earlier works (e.g. Dartois et al. 1999; Gerakines et al. 1999; Pontoppidan et al. 2008). However, after taking into account the entire experimental database, we obtained better fits when we used a small contribution from both a warm apolar and a polar CH3OH component as shown in Fig. 8. We chose the sources with the best signal-to-noise ratio to compare the three- and four-component fits.

A reasonable fit cannot be the only criterion for our assumption, so we tried to estimate quantitatively the column density of CO2 from the fits as discussed in Sect. 3.3. The column densities of CO2 obtained by the different components are separately calculated and are compared with the expected CO2 column density in Tables 5 and 6 for three- and four-component fits, respectively. The fits in Figs. 6 and 8 are almost identical from the qualitative point of view, but from Tables 5 and 6, we can conclude that the fits with a CH3OH component give a better quantitative agreement. Within the error of over simplification in estimating the expected column densities of CO2, we found that the values for our three sources with the best signal-to-noise ratio is compatible with the H2O, CO (apolar), and CH3OH components as obtained from the fit; instead, for our three-component fit we find a worse agreement with the expected abundance of the polar component.

The above-mentioned result could be an indirect correlation between the degradation of CH3OH and the formation of CO2. Both molecules are believed to be formed in the solid-state primarily from accreted CO on the grain surface. However, observations have revealed CO2 to be an abundant constituent of ices in the dark molecular clouds, whereas CH3OH abundance extends from upper limits to weak detections in several lines of sight. The column densities are listed in Table 4 for comparison. Moreover, laboratory experiments indicate that CO2 can efficiently be formed by energetic processing of CH3OH (e.g. Baratta et al. 2002; Modica & Palumbo 2010; Islam et al. 2014). Hence, CH3OH is most likely formed in the early stages of dark clouds through the hydrogenation of CO ice (Watanabe et al. 2004; Fuchs et al. 2009), and it is partially used up to form CO2.

In addition, we would like to point out that in the mixtures considered in this study, CO2 is not the only species formed after ion irradiation (see Fig. 2). The features due to the other species formed should also be present in the spectra observed along the same lines of sight if cosmic ion irradiation is active in the quiescent regions. In the case of pure CO, carbon dioxide is the most abundant species formed (about 10% with respect to initial CO), while carbon chain oxides (such as C2O, C3O, and C3O2) are formed with an abundance of the order of 0.1% with respect to the initial CO (e.g. Gerakines & Moore 2001; Trottier & Brooks 2004; Loeffler et al. 2005; Palumbo et al. 2008). In the case of CO:N2 mixtures, again carbon dioxide is the most abundant species formed while carbon chain oxides and nitrogen oxides (such as NO, N2O, and NO2) are formed with an abundance of the order of 0.1% (Ioppolo et al. 2009; Sicilia et al. 2012). The IR features due to solid carbon chain oxides and nitrogen oxides would be too weak to be detected in the line of sight to dense molecular clouds with current astronomical instrumentation; however, C2O, C3O, NO, and N2O have been detected in the gas phase, with an abundance with respect to hydrogen of the order of 10-10–10-8 because of their rotational transitions at radio wavelengths (e.g. Matthews et al. 1984; Brown et al. 1985; Ohishi et al. 1991; Ziurys et al. 1994; Halfen et al. 2001; Palumbo et al. 2008). Based on the comparison between the laboratory data and astronomical observations it has been suggested that in dense molecular clouds gas phase carbon chain oxides and nitrogen oxides are formed in the solid phase after cosmic ion irradiation of icy grain mantles and are released to the gas phase after the desorption of grain mantles (Palumbo et al. 2008; Sicilia et al. 2012).

When a sample of H2O deposited on C-grains is considered, both CO and CO2 are formed after ion irradiation with a comparable column density. As discussed by Mennella et al. (2004), the spectral profile and the contribution to the observed column densities make solid CO formed by cosmic-ray irradiation of ice-layered carbon grains a good candidate for the red component (at about 2136 cm-1) of the interstellar CO stretching feature, which is generally attributed to CO mixed with water ice or to CO mixed with CH3OH (e.g. Palumbo & Strazzulla 1992; Cuppen et al. 2011). According to Pontoppidan et al. (2003), the red component accounts for about 10% to 60% of total CO column density. Mennella et al. (2004) have found that the column density of CO formed after cosmic-ion irradiation of ice-layered C-grains can give a contribution of 3–36% to the total observed column density suggesting that an additional feature could contribute to the red component.

After ion irradiation of solid methanol, several absorption features are present in the IR spectra which are assigned to molecular species formed after irradiation. The most abundant are CO, CO2, formaldehyde (H2CO), and methane (CH4) (e.g. Hudson & Moore 2000; Baratta et al. 2002). As shown by Palumbo & Strazzulla (1993), Chiar et al. (1995), and Teixeira et al. (1998), the profile of the CO band produced after ion irradiation of methanol reproduces well the red component of the interstellar CO stretching feature. The most intense IR band of H2CO in the solid phase occurs at about 1720 cm-1 (5.8 μm). In astronomical spectra, this feature is expected to be blended in with the broad 1660 cm-1 (6.0 μm) water band and in fact it has been suggested that formaldehyde could contribute to the C1 component of the interstellar 5–8 μm absorption feature (Boogert et al. 2008) in the line of sight to young stellar objects. In laboratory spectra of irradiated methanol samples, the abundance of H2CO is of the order of 10% with respect to methanol which would make the contribution of formaldehyde to the 6 μm band observed in the line of sight to background sources almost negligible, in agreement with the analysis of the 5–8 μm absorption feature observed towards field stars presented by Knez et al. (2005). The column density of CH4 formed after ion irradiation of methanol is expected to be of the order of 10% with respect to observed methanol (Baratta et al. 2002) and the intensity of its absorption bands would be below the detection limit of current astronomical instrumentation in the line of sight to background sources. Other species such as methyl formate, ethylene glycol, and glycolaldehyde are formed after ion irradiation of methanol-rich samples with an abundance of the order of 0.1% with respect to methanol (e.g. Hudson & Moore 2000; Modica & Palumbo 2010). The absorption features of these species would be too weak to be detected in interstellar ices; however, it is generally accepted that the amount formed in the solid phase contributes to the gas phase abundance after desorption of icy grain mantles (e.g. Öberg et al. 2009; Modica & Palumbo 2010; Chuang et al. 2017).

5. Conclusion

Laboratory spectra of ion irradiated mixtures are used to fit the profile of the solid CO2 bending mode band observed by the Spitzer Space Telescope in the line of sight to quiescent dark cloud regions (i.e. background sources). From a qualitative point of view, good fits are obtained by using either three or four components. From a quantitative point of view, a better result is obtained with four components, i.e. when the spectrum of CO2 formed after ion irradiation of CH3OH is added to the fitting procedure. The results presented support the hypothesis that energetic processing of icy grain mantles can be an efficient means by which CO2 is formed in quiescent dark cloud regions; furthermore, this finding indirectly suggests the presence of CH3OH in icy grain mantles in these cold regions. Owing to the intrinsic weakness of background stars, only a few spectra with a good signal-to-noise ratio have been obtained by the Spitzer Space Telescope. We expect that the James Webb Space Telescope (JWST) will be able to study the profile of the CO2 bands towards a larger sample of background sources. In the near future, the comparison between new observations and laboratory spectra will provide us with a better knowledge of the chemical composition and physical properties of icy grain mantles in quiescent dense molecular clouds.


Acknowledgments

The authors are grateful to D. C. B. Whittet for the IRS Spitzer Space Telescope data and to V. Greco for writing an updated version of the Polyfit software. This research has been partially supported by the European COST Action CM 1401, Our Astro-Chemical History: Short Term Scientific Missions (STSM), the Italian Ministero dell’Istruzione, dell’Università e della Ricerca through the grant Progetti Premiali 2012-iALMA (CUP C52I13000140001), the NRW International Graduate School of Chemistry, Münster, and the LASSIE Initial Training Network, which is supported by the European Community’s Seventh Framework Programme under Grant Agreement Number 238258. S.I. acknowledges the Royal Society for financial support.

References

All Tables

Table 1

New experiments performed in this work.

Table 2

List of the band strength (A) values used.

Table 3

Laboratory spectra used to fit observations.

Table 4

Observed column densities (in units of 1017 cm-2) of CO, H2O, CO2, and CH3OH and AV values for the lines of sight studied here.

Table 5

Expected column density (in units of 1017 cm-2) of CO2 obtained after irradiation of three components – H2O on C-grains +200 keV H+, CO (pure or mixed) +200 keV H+ (16 K), and CO (pure or mixed) +200 keV H+ (40 K) – and column density of CO2 obtained from the fit for some selected sources.

Table 6

Expected column density (in units of 1017 cm-2) of CO2 obtained after irradiation of four components – H2O on C-grains +200 keV H+, CO mixtures +200 keV H+ (16 K), CO mixtures +200 keV H+ (40 K), and CH3OH +200 keV H+ at 16 K – and column density of CO2 obtained from the fit for some selected sources.

All Figures

thumbnail Fig. 1

Schematic depiction of the UHV chamber.

Open with DEXTER
In the text
thumbnail Fig. 2

Selected infrared spectra, in the range 3800–560 cm-1, used in the fitting procedure. From top to bottom the spectra refer to pure CO after bombardment with 200 keV H+ at 16 K, a mixture CO:N2 = 1:1 after bombardment with 200 keV H+ at 16 K, and pure CH3OH after bombardment with 200 keV H+ at 16 K. Spectra are normalized to their maximum optical depth value and are shifted for clarity. The main features formed after bombardment are labelled. The vertical grey arrows indicate the spectral range where most of the features assigned to carbon chain oxides and nitrogen oxides occur. A detailed description of these spectra is given by Palumbo et al. (2008), Sicilia et al. (2012), and Modica & Palumbo (2010).

Open with DEXTER
In the text
thumbnail Fig. 3

Column density of CO2 divided by the initial column density of CO after ion bombardment with 200 keV H+ of a H2O:CO = 8:1 mixture at 17 K. Experimental data are fitted by an exponential curve.

Open with DEXTER
In the text
thumbnail Fig. 4

Column density of CO2 divided by the initial column density of CO after ion bombardment of a CO:CH4:N2 = 1:1:1 mixture at 17 K. Experimental data are fitted by an exponential curve.

Open with DEXTER
In the text
thumbnail Fig. 5

Column density of CO2 divided by the initial column density of CH4 after ion bombardment of a H2O:CH4:NH3 = 1:1:1 mixture at 17 K. Experimental data are fitted by two exponential curves.

Open with DEXTER
In the text
thumbnail Fig. 6

Three-component fit. Laboratory spectra are: H2O on C-grains +200 keV H+, CO (pure or mixed) +200 keV H+ (16 K), and CO (pure or mixed) +200 keV H+ (40 K).

Open with DEXTER
In the text
thumbnail Fig. 7

Comparison between the band profile of the CO2 bending mode in H2O on C-grains +200 keV H+ and in a H2O:CO (8:1) mixture +200 keV H+.

Open with DEXTER
In the text
thumbnail Fig. 8

Four-component fit. Laboratory spectra are: H2O on C-grains +200 keV H+, CO mixtures +200 keV H+ (16 K), and CO mixtures +200 keV H+ (40 K) and CH3OH +200 keV H+ (16 K).

Open with DEXTER
In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.