DETECTION OF FORMAMIDE, THE SIMPLEST BUT CRUCIAL AMIDE, IN A SOLAR-TYPE PROTOSTAR

The origin of life on Earth is obviously one of the most important open questions of modern science. Finding the answer implies putting together and solving a plethora of riddles. Many of them have one word in common: chemistry. What was the basic chemical mechanism that led atoms from molecules to life? Was it the first step connected to the metabolism, conversion of energy, or to the genetic passage of information? Or both? Is there a molecule which may have been the key to both of them? A recent study by Saladino et al. (2012) argues that formamide (NH₂CHO) may have been the starting point for the prebiotic synthesis of both metabolic and genetic species: amino acids, nucleic acid bases, acyclonucleosides, sugars, amino sugars, and carboxylic acids. This is because formamide is the simplest possible amide, and contains within its diverse chemistry the functional groups and chemical bonds of central biomolecule and is a versatile solvent and a reactant.

Although the chemistry in the interstellar medium (ISM) is clearly different from that on Earth and that the conditions are obviously not appropriate for life, how complex the interstellar chemistry is and what prebiotic molecules are synthesized in space are long-standing and important questions that have always raised the curiosity of even non-specialized researchers. As an example, the discovery of highly deuterated amino acids in meteoritic material has led to the hypothesis that amino acids are also synthesized in the harsh ISM conditions, in particular during the Sun protostellar phase (Pizzarello & Huang 2005; see also the recent review by Caselli & Ceccarelli 2012). Regardless of whether or not amino acids may have had a role in the terrestrial life appearance, the presence of important prebiotic molecules in space deserves our attention. In this context, whether formamide is synthesized in regions that will give birth to planetary systems like our own is of the greatest interest.

Before this work, formamide was detected toward only two interstellar sources, in the direction of the two galactic brightest millimeter sources: Orion-KL and SgrB2 (Turner 1989; Nummelin et al. 1998; Halfen et al. 2011). Both regions are active sites of massive star formation and present environmental conditions (strong UV emission, intense X-ray irradiation, violent shocks, ...) very different from those existing in regions giving birth to solar-type stars and planetary systems like the solar system. The examples of the molecular deuteration or the abundance of organic molecules make this point very clearly. In Orion-KL and SgrB2, the molecular deuteration is several orders of magnitude lower than in solar-type protostars (Ceccarelli et al. 2007). For example, while doubly deuterated water has been measured to be ∼10⁻³ with respect to the hydrogenated water toward the solar-type protostar IRAS 16293−2422 (Butner et al. 2007; Vastel et al. 2010), it remains undetected toward SgrB2 and Orion-KL. We know how important water and deuterated water are for understanding the origin of terrestrial oceans (e.g., Hartogh et al. 2011). On the same vein, the abundance (normalized to methanol) of several complex organic molecules (COMs) is larger by more than one order of magnitude in solar-type protostars than massive star-forming regions such as SgrB2 and Ori-KL (e.g., Bottinelli et al. 2007; Taquet et al. 2012).

When it comes to the specific case of formamide, various laboratory experiments have shown that, as for other prebiotic molecules, it can be synthesized in UV and energetic particle irradiated ices (e.g., Danger et al. 2011; Jones et al. 2011 and references therein), so that the origin of formamide in Orion-KL and SgrB2 may reflect the importance of these energetic processes. On the contrary, the environments of solar-type forming stars are much less exposed to those energetic processes. Therefore, before the present work, it remained to be proven that formamide can be synthesized in the more quiescent and colder gas surrounding solar-type protostars.

In this Letter, we report the successful search of formamide in one of such sources, IRAS 16293−2422 (hereinafter IRAS16293). Located in the L1689N cloud (d = 120 pc; Loinard et al. 2008), IRAS16293 hosts, in a common colder envelope, two hot corinos, A (south-east) and B (north-west), separated by about 4''. Despite its relatively complex structure at arcsecond scales, IRAS16293 is the brightest and thus most appropriate source to carry out the search for new species in low-mass star-forming regions, as proven by the numerous new detections toward this source (e.g., Ceccarelli et al. 2000).

In addition, formamide has been detected in the comet Hale-Bopp (Bockelee-Morvan et al. 2000) whose chemical composition is suspected to be representative of the chemical composition of the primitive Solar Nebula. Thus, the comparison of the formamide abundance in Hale-Bopp with that of a solar-type protostar can contribute to our understanding of the solar system astrochemical heritage.

The data presented here combine IRAM 30 m spectra from the TIMASSS survey (Caux et al. 2011) with more recent observations, performed in 2012 March, in four selected frequency ranges at 3, 2, and 1 mm, with the new broadband EMIR receivers at the IRAM 30 m telescope. Both sets of observations were centered on the B (northwest) source at α(2000.0) = 16^h32^m226, δ(2000.0) = −24°28'33'. Note, however, that the A and B sources, separated by 4'', are encompassed by the beam of our observations, even at the highest frequencies. All observations were performed in double-beam-switch (DBS) observing mode with a 90'' throw. The pointing and focus were checked every two hours on planets or on continuum radio sources (1741-038 or 1730-130). Both sets of spectra were resampled at the same frequency resolution and averaged using the "stitch" command of the IRAM data reduction package CLASS⁴ to carefully check the coherence of the data in terms of line positions, shapes, and intensities. Comparison of the line intensities in both sets of data shows that the calibration is accurate within 15%. Table 1 summarizes the observed bands and the details of the observations. Several representative spectra are plotted in Figure 1.

Zoom In Zoom Out Reset image size

The new IRAM 30 m observations obtained in 2012 March present a systematically better spectral resolution and an rms in most cases comparable to or better than the TIMASSS spectra (see Table 1). Line identification was checked separately on both sets of data, based on the line frequencies given in the CDMS database (Müller et al. 2005). To limit our search to the lines expected to be the most intense, we have looked for transitions with an Einstein coefficient A_ij ⩾ 3 × 10⁻⁵ s⁻¹ and an upper energy E_up ⩽ 250 K using the CASSIS software.⁵ With these criteria, 115 transitions lie in the (80–280 GHz) TIMASSS survey frequency range and a total of 39 lines were detected at the expected frequencies. In the frequency range covered by the 2012 March observations, we have detected 21 lines among the 37 transitions corresponding to the search criteria.

To verify whether the lines attributed to formamide constitute a reliable identification, we have performed the following tests.

• 1.  
  We have built an LTE model reproducing the fluxes of the detected lines (see Section 4.1) and verified that it does not predict detectable but not detected lines in the observed bands (see also point 2).
• 2.  
  We have classified the formamide lines filling the above-mentioned search criteria into four categories: (1) the detected and isolated lines; (2) the detected but blended lines; (3) the undetected, weak, or severely blended lines; and (4) the "missing" lines, i.e., those for which the predicted line is more than 50% brighter than the observed one. The histograms plotted in Figure 2 show the distribution of the lines in the four categories.
• 3.  
  We have applied the same method to two other important COMs: (1) methyl formate (HCOOCH₃), firmly detected, and (2) glycine (NH₂CH₂COOH), a long searched for and not yet detected COM. The results are also shown in Figure 2.

Zoom In Zoom Out Reset image size

A comparison of these distributions shows that in contrast to glycine, neither methyl formate nor formamide shows the "missing" lines (M). When all the spectra of the TIMASS survey are included, the formamide and the glycine line distributions have a large fraction of undetected lines (N). However, when the noise decreases and the spectral resolution increases, as in the 2012 March observations, both distributions evolve in opposite ways: many undetected lines (N) of glycine become "missing" lines (M) and, in contrast, they become detected (isolated, DI or blended, DB) lines in the case of formamide. With the improved 2012 March data, the formamide line distribution appears similar to that of the methyl formate. We are thus confident that our new observations confirm the detection of formamide in IRAS16293.

4.1. Formamide Abundance

4.2. Chemistry

Several reaction pathways have been previously proposed to explain the abundance of gas-phase formamide in the interstellar medium: formamide might form either in the icy mantles of irradiated dust grains and then be released in the gas phase or formamide could be produced directly in the gas phase. The pioneering work of Hagen et al. (1979) had shown that formamide was formed by vacuum ultraviolet (VUV) irradiation of the ice mixture CO:NH₃:H₂O:CO₂. More recently, formamide was identified by Raunier et al. (2004) as one of the products formed by VUV photolysis of solid isocyanic acid (HNCO) at 10 K. These authors were also able to tentatively assign solid formamide in the ISO-SWS infrared spectra of massive protostellar sources, an identification never confirmed. Very recently, Jones et al. (2011) experimentally produced formamide by electron irradiating ice mixtures CO:NH₃. Energetic processing of icy mantles can thus be at least partially responsible for the presence of formamide in the gas phase. It is, however, very difficult to quantify to what extent (Jones et al. 2011). In the gas phase, to our knowledge there is no available measurements but a possible route is via the exothermic neutral–radical reaction between H₂CO and NH₂⁶:

$$\begin{equation} \rm H_2CO + NH_2 \rightarrow NH_2CHO + H. \end{equation} \tag{ 1 }$$

Indeed, both NH₂ and H₂CO have high abundances in protostellar sources. In fact, the abundance of formaldehyde is ∼10⁻⁷ in the inner region of the envelope of IRAS 16293−2422 (Ceccarelli et al. 2000) while the abundance of NH₂ is ∼10⁻⁹ in the cold outer envelope (Hily-Blant et al. 2010). If we assume that NH₂CHO is removed by the primary molecular ion HCO⁺, whose abundance is ∼10⁻⁸ (Jorgensen et al. 2004), by the reaction

$$\begin{equation} \rm NH_2CHO + HCO^+ \rightarrow CO + NH_2CH_2O^+, \end{equation} \tag{ 2 }$$

then the formamide abundance at steady state is

$$\begin{equation} \rm [NH_2CHO] = \frac{\it k_1(T)}{\it k_2(T)}\frac{[NH_2][H_2CO]}{[HCO^+]}, \end{equation} \tag{ 3 }$$

where k₁(T) and k₂(T) are the rate coefficient of reactions (1) and (2). These rate coefficients are unknown but if the reactions are barrierless, standard values at 100 K are k₁ ∼ 10⁻¹⁰ for a neutral–radical reaction and k₂ ∼ 5 × 10⁻⁹ for a ion–dipole reaction.⁷ We thus obtain a formamide abundance of ∼2 × 10⁻¹⁰, in good agreement with the present determination in the hot corino of IRAS16293. Of course, rate measurements or calculations are necessary to confirm our estimates (especially for reaction (1)) but it is interesting to note that invoking grain-surface chemistry is not necessary to explain the gas-phase abundance of formamide. The situation is different for acetamide (not detected in this work) because there is no obvious exothermic gas-phase formation route. Thus, the high abundance of acetamide in massive star-forming regions (Halfen et al. 2011) and its non-detection in IRAS16293 are consistent with a synthesis in irradiated ices.

To compare the formamide abundance found in IRAS16293 with that in the comet Hale-Bopp, we have used the formamide abundance relative to H₂O. According to the recent measurements by Coutens et al. (2012), the abundance of water in the IRAS16293 hot corino is about 5 × 10⁻⁶ so that the abundance ratio [NH₂CHO]/[H₂O] is 2.5 (−2, +6) × 10⁻⁵. In Hale-Bopp, this ratio is ∼6 times larger (0.01%–0.02%). We have plotted in Figure 3 the abundance of formamide relative to H₂ in the three interstellar sources SgrB2, Orion-KL, and IRAS16293 as well as its abundance relative to H₂O in the three same sources and in the comet Hale-Bopp. First, it is striking that the formamide abundance with respect to H₂ in the three interstellar sources agree within a factor of two. This result suggests that the presence of highly energetic radiative processes in massive star-forming regions does not play a dominant role in the synthesis of formamide. Second, the formamide abundance with respect to water in the three interstellar sources is similar (within a factor of 10) to that in Hale-Bopp, providing new evidence for the similarity between interstellar and cometary chemistry (see the discussion in Bockelee-Morvan et al. 2000 and Caselli & Ceccarelli 2012). Our results thus suggest that the abundances of cometary organic volatiles could reflect not only the grain-surface chemistry, as it is generally assumed, but also the gas-phase chemistry of the interstellar medium.

Zoom In Zoom Out Reset image size

Finally, the singly deuterated isotopologues of formamide (NHDCHO and NH₂CDO) have been searched but not detected (millimeter frequencies will be published elsewhere by Margules et al.), which is consistent with the observed relatively low abundance of the main isotopologue.

We have reported the first detection of formamide, the simplest molecule with a peptide-like bond, in the hot corino of the solar-type protostar IRAS16293. The millimeter data combine spectra from the TIMASS survey and more recent observations performed at the IRAM 30 m telescope. The abundance of formamide with respect to H₂ is ∼1.3 × 10⁻¹⁰, which is similar to that found in the massive star-forming regions SgrB2 and Orion-KL. This result questions the relevance of photolysis processes in the synthesis of formamide and possibly argues in favor of a gas-phase synthesis. Assuming that the neutral–neutral reaction between the abundant NH₂ and H₂CO molecules has no barrier, we predict a gas-phase abundance of ∼2 × 10⁻¹⁰, in good agreement with the observations. The abundance of formamide with respect to H₂O is, within a factor six, similar to the abundance observed in the comet Hale-Bopp. Assuming that the formamide abundance measured in the coma of Hale-Bopp is representative of cometary ices, our results provide a new evidence for a possible direct link between cometary and interstellar chemistry.

We thank our colleague Nadia Balucani for very fruitful discussions on the formamide chemistry and Laurent Margules and Roman Motiyenko for providing us the deuterated formamide line frequencies before publication. Alexandre Ratajczak is aknowledged for his help in the observations. This work has been supported by l' Agence Nationale pour la Recherche (ANR), France (project FORCOMS, contract ANR-08-BLAN-0225).