Seeds Of Life In Space (SOLIS): The Organic Composition Diversity at 300–1000 au Scale in Solar-type Star-forming Regions^*

L1544 is a prestellar core in the Taurus complex, at a distance of 140 pc (Kenyon et al. 1994; Schlafly et al. 2014). This prestellar core has been studied extensively, and its chemical and physical structure is well-known. The gas temperature reaches a minimum of 6 K toward the central 2000 au and a maximum of 13 K at 15,000 au (Crapsi et al. 2007). Caselli et al. (1999) measured large amounts of CO freeze-out within the central few 1000 au, and within the CO-depleted zone, deuterated molecules thrive (Caselli et al. 2002, 2003; Bacmann et al. 2003; Vastel et al. 2006; Crapsi et al. 2007). The ground state transition of ortho-H₂O reveals an inverse P-Cygni profile that can be reproduced with a simple chemical-dynamical model describing the quasi-equilibrium contraction of a Bonnor–Ebert sphere (Caselli et al. 2012; Keto et al. 2015; Quenard et al. 2016). L1544 is hence a prestellar core on the verge of star formation. Thus its study will provide us with crucial information about the chemical processes before the switch-on of the protostar. Several iCOMs are detected toward this core (Bizzocchi et al. 2014; Vastel et al. 2014; Jiménez-Serra et al. 2016). NOEMA observations are centered on the position of the CH₃OH peak (Bizzocchi et al. 2014), where iCOMs are also expected to peak (Vastel et al. 2014; Jiménez-Serra et al. 2016), about 5000 au away from the dust peak, as confirmed by the theoretical work of Vasyunin et al. (2017).

L1521F is another dense core in the Taurus molecular cloud. It was studied by Crapsi et al. (2004), who found a striking similarity with L1544, with the exception of a factor of 2 lower deuterium fraction and a peculiar velocity field. Based on these results, Crapsi et al. (2004) defined L1521F as "a highly evolved starless core." In fact, two years later, a faint infrared source was discovered at its dust peak by the Spitzer Space Telescope (Bourke et al. 2006). This source belongs to the group of VeLLOs (i.e., objects with a luminosity lower than about $0.1\,{L}_{\odot }$) embedded in dense cores (Kauffmann et al. 2005; Dunham et al. 2006). More recently, the central regions of L1521F were studied with ALMA: Tokuda et al. (2014) imaged dust continuum emission and simple molecules (HCO⁺, HCN, CS), finding large central densities and sub-structure with an arc-like morphology. This structure resembles the hydrodynamical simulations of infalling material around protostars carried out by Matsumoto et al. (2015). Tokuda et al. (2016) studied in detail the structure of the L1521F core, from 10s to ∼10,000 au, revealing an inner region with a flat density structure where an unresolved protostellar source is next to a starless high-density core. In summary, L1521F represents the next evolutionary stage of a prestellar core such as L1544.

NGC 1333-IRAS4A is located in the Perseus NGC 1333 region at a distance of (260 ± 26) pc (Hirota et al. 2008; Schlafly et al. 2014). It was the second discovered hot corino (Bottinelli et al. 2004)—namely hot (≥100 K), dense ($\geqslant {10}^{7}$ cm⁻³), and compact (≤100 au in radius) regions enriched of iCOMs (Ceccarelli et al. 2007; Caselli & Ceccarelli 2012). Interferometric observations showed that it is a binary system composed of two Class 0 objects separated by $\sim 1\buildrel{\prime\prime}\over{.} 8$ (423 au): IRAS4-A1 and IRAS4-A2 (e.g., Looney et al. 2000; Santangelo et al. 2015; Tobin et al. 2016). The overall bolometric luminosity of IRAS4A and its total envelope mass are estimated to be $9.1\,{L}_{\odot }$ and $5.6\,{M}_{\odot }$, respectively (Kristensen et al. 2012; Karska et al. 2013; Tobin et al. 2016). While IRAS4-A1 is more than three times brighter at 1.3 mm than its companion, only IRAS4-A2 is associated with a hot corino (Coutens et al. 2015; Taquet et al. 2015; De Simone et al. 2017; López-Sepulcre et al. 2017). In addition, IRAS4-A1 drives a fast collimated jet associated with bright H₂ emission, whereas IRAS4-A2 powers a slower and precessing jet (Santangelo et al. 2015). These findings suggest that the two Class 0 sources are likely in different evolutionary stages, although it is not possible so far to firmly conclude whether this is due to evolution or rather accretion luminosity effects/bursts (De Simone et al. 2017; López-Sepulcre et al. 2017). The compact jets from IRAS4-A1 and IRAS4-A2 create a shock site $\sim 12^{\prime\prime}$ south, where SOLIS observations show emission from iCOMs, for the first time (see Section 3.2).

CepE is located in the second most massive molecular cloud of the Cepheus OB3 association at a distance of 730 pc (Sargent 1977; Few & Cohen 1983). It is an isolated Class 0 source of intermediate mass associated with the infrared source IRAS 23011+6126. Its luminosity is $\sim 100\,{L}_{\odot }$ and its mass is $35\,{M}_{\odot }$ (Lefloch et al. 1996; Crimier et al. 2009). CepE drives an exceptionally powerful outflow, the southern lobe of which is terminated by the Herbig-Haro object HH377 (Lefloch et al. 1996, 2015). Studies of the millimeter CO and the near-infrared H₂ and [Fe ii] lines revealed the presence of several outflows (Eislöffel et al. 1996; Ladd & Hodapp 1997). Continuum and line observations obtained with the IRAM-PdBI (Plateau de Bure Interferometer) at $1^{\prime\prime}$ scale confirm the presence of a binary protostellar system, with each component driving a high-velocity jet detected in lines of H₂, CO, and SiO (J. Ospina-Zamudio et al. 2017, in preparation). The dynamical timescale of these ejections is short—typically 1000 years. The same observations also show evidence for the presence of a hot corino in one of the two components, so that this is a good target to understand the influence of luminosity on the formation and destruction of iCOMs.

NGC 1333-SVS13A is located in the Perseus NGC 1333 region at a distance of 260 ± 26 pc (Hirota et al. 2008; Schlafly et al. 2014), close to the other SOLIS source NGC 1333-IRAS4A. SVS13A is part of a small cluster of protostars, of which the two brightest sources, labelled A and B, are separated by $\sim 15^{\prime\prime}$ (e.g., Chini et al. 1997; Bachiller et al. 1998; Looney et al. 2000; Chen et al. 2009; Tobin et al. 2016). SVS13A is relatively evolved, as it possesses a ratio ${L}_{\mathrm{submm}}/{L}_{\mathrm{bol}}\sim 0.08$, typical of Class I sources. Its luminosity is $\sim 33\,{L}_{\odot }$ (Tobin et al. 2016). SVS13A drives the famous chain of Herbig-Haro (HH) objects 7–11 (Reipurth et al. 1993) and is associated with an extended outflow (>0.07 pc: Lefloch et al. 1998; Codella et al. 1999). For comparison, SVS13B is a Class 0 protostar with ${L}_{\mathrm{bol}}\sim 1\,{L}_{\odot }$ (Tobin et al. 2016; De Simone et al. 2017) that drives a well-collimated SiO jet (Bachiller et al. 1998). Recent observations revealed the presence of a compact (∼50 au), hot (∼200 K), and dense ($\geqslant {10}^{7}$ cm⁻³) region in SVS13A, hinting at the presence of a hot corino region, the first ever detected in a Class I source (Codella et al. 2016; De Simone et al. 2017). More recently, Bianchi et al. (2017) measured a deuterium fraction in SVS13A smaller than in Solar-like prestellar cores and Class 0 sources (their Figure 9), showing, for the first time, that there may be a chemical evolution between Class 0 and I sources. Thus SVS13A is a perfect target to understand the evolutionary effects in the iCOMs abundances—namely formation and destruction routes.

OMC-2 FIR4 belongs to the Orion Molecular Complex, at a distance of 420 pc (Hirota et al. 2007; Schlafly et al. 2014). The total FIR4 mass and luminosity are around $30\,{M}_{\odot }$ (Mezger et al. 1990; Crimier et al. 2009) and less than $1000\,{L}_{\odot }$ (Crimier et al. 2009; Furlan et al. 2014), respectively. Interferometric observations showed that FIR4 contains several protostars and clumps, even though their number is uncertain (Shimajiri et al. 2008; López-Sepulcre et al. 2013; Gonzalez-Garcia et al. 2016). Our new SOLIS maps detect at least eight sources, of which only one has a FIR counterpart so that the others are, likely, very young Class 0 or prestellar objects (R. Neri et al. 2017, in preparation). Therefore, FIR4 is certainly a young protocluster. The nature of the hosted objects is practically unknown, including their mass, temperature, chemical composition, and dynamical status. However, what makes this protocluster particularly interesting and the reason why it is a SOLIS target is that FIR4 is, at present, the best known analogue of the Sun progenitor for two reasons. The first one is that the Sun was born in a cluster of stars and not in an isolated clump (Adams 2010). The second reason is that the solar system experienced a large irradiation of energetic (≥10 MeV) particles during its first phases of formation (Gounelle et al. 2006). A similar dose of energetic particles is present somewhere in or close to FIR4, producing an enhanced degree of gas ionization (Ceccarelli et al. 2014; Fontani et al. 2017). Finally, close to FIR4 lie two additional sources, FIR3 and FIR5, about $25^{\prime\prime}$ north and south, respectively. Not much was known about these two sources before our SOLIS observations. Shimajiri et al. (2015) obtained an unbiased spectral survey at 3 and 0.8 mm toward FIR3, with single-dish telescopes and corresponding angular resolution of about $20^{\prime\prime}$. Based on those observations, Shimajiri et al. claim that the observed line emission is dominated by the shocked gas from the outflow rising from FIR3 and hitting FIR4 rather than a hot core/corino sources.

This famous shocked region is associated with the Class 0 protostar L1157-mm, at a distance of 250 pc (Looney et al. 2007). L1157-mm has a luminosity of $\simeq 3\,{L}_{\odot }$ (Tobin et al. 2010), and drives an episodic and precessing jet (Gueth et al. 1996, 1998; Podio et al. 2016). In the last $\leqslant 2000\,\mathrm{yr}$, the jet accelerated and entrained material, creating a spectacular bipolar outflow and opening cavities along the path. The south and blueshifted lobe of the outflow is formed by two cavities, the respective apexes of which are known as B1 and B2 (e.g., Gueth et al. 1998). The latter is more distant from L1157-mm and, hence, older than B1. The SOLIS target is B1, the age of which is estimated to be less than 2000 years (Podio et al. 2016). This low age and the extent of the shock allow us to probe chemical composition variations along the shock and, consequently, to test the chemistry where the dust icy mantles are recently released in the gas phase via sputtering. In the context of the SOLIS goal, B1 provides us with the possibility to test the different theories of iCOMs formation. Indeed, the different ejection episodes caused several shocks where the jet impacted the cavities walls. In these sites, the abundance of a large number of species is dramatically enhanced: from diatomic, such as SiO and CS, to more complex molecules, such as CH₃CN and CH₃CHO (see, e.g., Bachiller et al. 2001; Benedettini et al. 2013; Codella et al. 2015). Lately, several iCOMs were detected in B1 with abundances as large as or more than in hot corinos (Lefloch et al. 2017a). However, just a few iCOMs (methanol, methyl cyanide, and acetaldehyde) have been imaged with interferometers so far (Codella et al. 2007, 2015; Fontani et al. 2014). These images show a inhomogeneous distribution of the molecules, associated with the different ejection episodes. The comparison of the distribution of acetaldehyde and formamide obtained with the new SOLIS observations provides us with very stringent constraints on the formation route of formamide (Codella et al. 2017).

The comparison between the SOLIS (Figure 2) and ASAI (Figure 4) spectra provides straightforward information on the emission at the 3000–10,000 au scales probed by the ASAI observations and filtered out by the SOLIS ones. For example, the CCS line is very bright in the ASAI spectra, and weak (L1157-B1 SHOCK 1 and 2, and FIR4-E), if not absent (IRAS4Ac and FIR5), in the SOLIS ones, a clear indication that CCS is abundant in the extended molecular clouds to which L1157-B1, OMC-2 FIR4, and NGC 1333-IRAS4A belong. The same can be said of the H₂CCO, c-C₃H₂, and HC₅N lines, which are prominent in the ASAI spectra and weak or absent in the SOLIS ones. All these species are small hydrocarbons or unsaturated C-bearing chains particularly abundant in the skin of the parental clouds (e.g., Spezzano et al. 2017). Two dedicated SOLIS articles discuss the details regarding the cyanopolyynes (Fontani et al. 2017) and c-C₃H₂ (Favre et al. 2017) toward OMC-2 FIR4.

Note that the ASAI spectra are relatively similar in OMC-2 FIR4 and NGC 1333-IRAS4A, with the biggest exception being the HC₅N line, totally absent in NGC 1333-IRAS4A and very bright in OMC-2 FIR4. In Fontani et al. (2017), we argue that this is due to the anomalously large energetic (≥10 MeV) particle irradiation of the OMC-2 FIR4 region, first revealed by Herschel observations (Ceccarelli et al. 2014). On the other hand, the similarity suggests the presence of an embedded hot corino in OMC-2 FIR4.

The shocked region L1157-B1 shows a relatively different line spectrum with respect to both OMC-2 FIR4 and NGC 1333-IRAS4A, with bright lines of formamide (Codella et al. 2017) and SO₂ (S. Feng et al. 2017, in preparation). This difference with respect to OMC-2 FIR4 is less surprising, as the single-dish ASAI spectrum in the latter is likely dominated by a dense and cold gas highly irradiated by energetic particles (Ceccarelli et al. 2014; Fontani et al. 2017); namely it is likely due to the very different environment of OMC-2 FIR4 and L1157-B1. A bit more surprising is the difference between L1157-B1 and NGC 1333-IRAS4, as in both sources the emission presumably (also) originates from species directly or indirectly injected by the grain mantles. This difference is probably due to either a different chemical timescale or a different initial grain mantle composition.

The sources and the regions surrounding them appear particularly different at the 300–1000 au scale. The spectra of Figures 2 and 3 clearly identify either different zones of line emission or different chemical composition across each target source.

All the 80.8–84.3 GHz spectra are dominated by the HC₃N (9-8) line and, to a lesser extent, the SO₂ (${8}_{\mathrm{1,7}}\mbox{--}{8}_{\mathrm{0,8}}$) line (see Table 6). Yet the relative intensity with respect to other lines varies from region to region. In particular, the HC₃N (9-8) to SO₂ (${8}_{\mathrm{1,7}}\mbox{--}{8}_{\mathrm{0,8}}$) intensity ratio varies from about unity in NGC 1333-IRAS4As to 20 in OMC-2 FIR4-E. Such a large difference is unlikely due to excitation effects, as the two lines have rather similar upper level energies, so it probably reflects a real chemical difference. Figure 5 shows the predicted line ratio as a function of the excitation temperature, assuming LTE and optically thin lines. The figure shows that this ratio varies by a factor of about four between 6 and 100 K. Therefore, NGC 1333-IRAS4As appears to be a region enriched in SO₂ compared with OMC-2 FIR4-E or, alternatively, poorer in HC₃N. L1157-B1 lies in between, with a HC₃N (9-8) to SO₂ (${8}_{\mathrm{1,7}}\mbox{--}{8}_{\mathrm{0,8}}$) line ratio of 3–6. This is indeed consistent with previous observations that show an enhancement in shocked regions of SO₂ but not of HC₃N (e.g., Bachiller & Pérez Gutiérrez 1997; Benedettini et al. 2013).

Zoom In Zoom Out Reset image size

Even more marked is the difference in intensity of weaker lines. Interestingly, the CCS (${6}_{7}\mbox{--}{5}_{6}$) line is relatively bright in L1157-B1 SHOCKS 1 and 2, and in OMC-2 FIR4-E, whereas it is undetected in the other sources/regions. On the contrary, the H₂CCO (${4}_{\mathrm{1,3}}\mbox{--}{3}_{\mathrm{1,2}}$) line is bright in L1157-B1 SHOCK2 and OMC-2 FIR4-E, weak in NGC 1333-IRAS4Ac and OMC-2 FIR4-W, and undetected elsewhere. Also, in this case, the largest differences are likely caused by a real difference in chemical composition rather than excitation conditions, as the two lines have, again, similar upper level energies. This is quantitatively shown in Figure 5: the CCS (${6}_{7}\mbox{--}{5}_{6}$) over H₂CCO (${4}_{\mathrm{1,3}}\mbox{--}{3}_{\mathrm{1,2}}$) line ratio is practically constant for an excitation temperature, varying between 6 and 100 K.

More relevant for the present work and the goal of SOLIS, lines from formamide are variably present in the sources/regions. For example, the NH₂CHO (${4}_{\mathrm{1,4}}\mbox{--}{3}_{\mathrm{1,3}}$) line is detected in L1157-B1 SHOCKS 2 and 3, NGC 1333-IRAS4A hot corino and shock, tentatively in OMC-2 FIR4 west and east, and undetected in OMC-2 FIR3 and FIR5 (Figure 3). When considering the ratio between the formamide and CCS lines, which have very similar upper level energies and Einstein coefficients, the chemical differentiation is particularly evident with CCS/NH₂CHO varying from 1.5 to 3 in OMC-2 FIR4-E and L1157-B1 SHOCK 2 to less than unity in the other regions where NH₂CHO is detected. Figure 5 quantifies how different the excitation conditions should be to explain the observed difference in the CCS/NH₂CHO line ratio. For excitation temperatures between 10 and 80 K, the line ratio varies by no more than a factor of two. Therefore, dedicated studies of the various sources/regions, coupled with chemical modeling, as in Codella et al. (2017), are needed to shed light on the processes behind the observed line intensity differences among the various sources.

In summary, the distribution and relative abundances of (complex) organic molecules change significantly among the different sources, and this diversity is already clear at 300–3000 au scales.

The spectra of the two studied regions, the hot corino and the shocked region, are definitively very different (Figure 1). The hot corino shows the typical lines of iCOMs (not reported in Tables 5 and 6 because they are many but not relevant in the context of this article: the entire list will be the subject of a forthcoming publication), such as formamide, dimethyl ether, and methanol, whereas the shocked region possesses only two (relatively) bright lines characteristic of hot corinos, the NH₂CHO (${4}_{\mathrm{1,4}}\mbox{--}{3}_{\mathrm{1,3}}$) and SO₂ (${8}_{\mathrm{1,7}}\mbox{--}{8}_{\mathrm{0,8}}$) lines. The distribution of the SO₂ line intensity is shown in Figure 1. The line clearly maps the contour of the outflowing/shocked gas. Particularly intriguing is the formamide detection toward the shocked region, in a position that is at the border of the region where SO₂ emits. Note that SO₂ traces fairly well the two jets emitted by A1 and A2, observed in SiO by Santangelo et al. (2015) and here unresolved. As in the case of L1157-B1 (Codella et al. 2017; S. Feng et al. 2017, in preparation), the SO₂ versus formamide spatial segregation could be caused by a time effect, in addition to the release in the gas phase of the formamide precursors, namely formaldehyde and amidogen. It is worth mentioning that methanol is, rather surprisingly, not detected toward the shocked region, where formamide is present (Table 6). In addition, the shocked region emits weak lines of H₂CCO and HC₃N (9-8), also detected in the hot corino. A forthcoming article will report on a detailed analysis of this jet plus shocked region (A. López-Sepulcre et al. 2017, in preparation).

The three objects in the NOEMA primary beam, FIR3, FIR4 and FIR5, show clear and remarkable differences. The difference in the cyanopolyynes and c-C₃H₂ are discussed in Fontani et al. (2017) and Favre et al. (2017), respectively. Briefly, cyanopolyynes probe the presence of a source of energetic ($\geqslant 10$ Mev) particles irradiating the eastern part of OMC-2 FIR4 and FIR5. c-C₃H₂ is, instead, a good thermometer of the region, which shows a rather uniform temperature with a possible small increasing gradient toward east, as also found with the low angular resolution temperature map from NH₃ (Friesen et al. 2017). This is also consistent with the fact that the region is irradiated by energetic particles at/from the east.

In addition, the spectra in Figures 2 and 3 indicate the presence of a high lying (E_up = 103 K) line of CH₃OH, typical of hot corinos, only toward FIR4-E. The SO₂ line is bright toward FIR4-W and FIR4-E, and it is detected in FIR3 but not in FIR5 (Table 6). Since SO₂ is often also a hot corino species or, more generally, a warm gas indicator, this would suggest that FIR3 harbors a young protostar, maybe a hot corino, while FIR5 is a cold object. Similarly, one or more hot corinos might be present in the FIR4 region. Toward FIR4, the eastern part (the one irradiated by energetic particles) is the region with the richest spectrum, with all of the lines listed in Table 6 detected. It is likely that one or more hot cores/corinos are embedded, therefore, in the east region of FIR4. In the western part, only H₂CCO and c-C₃H₂ are detected. In general, they are species that are abundant in the exterior of molecular clouds, where residual UV photons produce a (small) fraction of neutral carbon (e.g., Spezzano et al. 2016).

We do not find any evidence that FIR4 north is affected by the outflow from FIR3, as suggested by Shimajiri et al. (2015). An increase of the gas temperature or broadband emission at north are also not detected in the SOLIS observations. Since the Shimajiri et al. observations were carried out with single-dish telescopes with a beam of $\sim 20^{\prime\prime}$, it is possible that the emission that they observed is contaminated by the hot corinos in the east FIR4 region.

Finally, a weak formamide line is present in both FIR4-E and FIR4-W. To confirm its detection, additional SOLIS observations are underway. Finally, incoming higher angular resolution SOLIS observations will definitively establish whether hot corinos or shocks or both are present and, if so, where in the region (R. Neri et al. 2017, in preparation).

L1157-B1 can be separated into three zones when the relative abundance of acetaldehyde and formamide is considered (Codella et al. 2017). The three zones identify three different shocks caused by the episodic ejections and the resulting jet impacting the cavity wall excavated by the outflow from the central star. The three shocks have been created at different times, within a lapse of time of about 1000 yr and with the one in the north being the youngest one (Podio et al. 2016). The difference between the three zones is also evident looking at the CCS, H₂CCO and c-C₃H₂ lines. The CCS and c-C₃H₂ lines are bright in the north and central zones, but absent in the south, whereas H₂CCO is only bright in the central zone. This diversity in the CCS, c-C₃H₂, and H₂CCO line distribution cannot be attributed to sensitivity, but must reflect a difference in the excitation conditions or chemical composition of the three regions. The upper level energies of the CCS, H₂CCO, and c-C₃H₂ (3${}_{\mathrm{1,2}}\mbox{--}{3}_{\mathrm{0,3}}$) transitions are 15, 22, and 16 K, whereas the Einstein coefficients are 2.5, 0.5, and 1.0 $\times {10}^{-5}$ s⁻¹, respectively, so that different excitation conditions seem unlikely. The plots of these line ratios, reported in Figure 5, support this interpretation. Based on the analysis of formamide, the line of which is detected only toward SHOCKs 2 and 3, we suggest that the differentiation of CCS, H₂CCO and c-C₃H₂ is also linked to the evolution of the gas chemical composition on a short timescale (∼1000 years; Codella et al. 2017).