Detectability of Glycine in Solar-type System Precursors
Izaskun Jimenez-Serra, Leonardo testi, Paola Caselli, Serena Viti
aa r X i v : . [ a s t r o - ph . S R ] A p r Draft version July 3, 2018
Preprint typeset using L A TEX style emulateapj v. 11/26/03
DETECTABILITY OF GLYCINE IN SOLAR-TYPE SYSTEM PRECURSORS
Izaskun Jim´enez-Serra , Leonardo testi , Paola Caselli and Serena Viti Draft version July 3, 2018
ABSTRACTGlycine (NH CH COOH) is the simplest amino acid relevant for life. Its detection in the interstellarmedium is key to understand the formation mechanisms of pre-biotic molecules and their subsequentdelivery onto planetary systems. Glycine has extensively been searched for toward hot molecularcores, although these studies did not yield any firm detection. In contrast to hot cores, low-mass starforming regions, and in particular their earliest stages represented by cold pre-stellar cores, may bebetter suited for the detection of glycine as well as more relevant for the study of pre-biotic chemistryin young Solar System analogs. We present 1D spherically symmetric radiative transfer calculationsof the glycine emission expected to arise from the low-mass pre-stellar core L1544. Water vapour hasrecently been reported toward this core, indicating that a small fraction of the grain mantles in L1544( ∼ ∼ − with respect to water,our calculations reveal that several glycine lines between 67 GHz and 80 GHz have peak intensitieslarger than 10 mK. These results show for the first time that glycine could reach detectable levelsin cold objects such as L1544. This opens up the possibility to detect glycine, and other pre-bioticspecies, at the coldest and earliest stages in the formation of Solar-type systems with near-futureinstrumentation such as the Band 2 receivers of ALMA. Subject headings: astrobiology — astrochemistry — ISM: molecules — stars: formation INTRODUCTION
Glycine (NH CH COOH) is the simplest amino acidand a key constituent of living organisms. It is believedthat its formation may have occurred in the interstellarmedium (ISM) since glycine and other amino acids havebeen found in meteorites (Pizzarello et al. 1991; Ehren-freund et al. 2001; Glavin et al. 2006) and comets (Wild2; Elsila et al. 2009). Laboratory experiments have alsoreported the formation of amino acids, including glycine,via ultraviolet (UV) and ion photolysis of interstellar iceanalogues (Mu˜noz Caro et al. 2002; Holtom et al. 2005),supporting the idea that amino acids may have an inter-stellar origin (Ehrenfreund & Charnley 2000). The de-tection of glycine is thus key to understand the formationof pre-biotic molecules in the ISM and their subsequentdelivery onto planetary systems.Over a decade glycine has extensively been searched forin high-mass star forming regions such as the hot molec-ular cores in SgrB2, Orion KL and W51 e1/e2 (Kuan etal. 2003; Belloche et al. 2008, 2013). Hot cores, with tem-peratures ∼ European Southern Observatory, Karl-Schwarzschild-Str. 2,85748 Garching (Germany); [email protected], [email protected] School of Physics & Astronomy, University of Leeds, LS2 9JT,Leeds (UK) Max-Planck-Institut f¨ur extraterrestrische Physik (MPE),Gießenbachstr., 85741 Garching (Germany); [email protected] Department of Physics & Astronomy, University College Lon-don, Gower Place, WC1E 6BT, London (UK); [email protected]
Several challenges are faced in the search of glycine inhot cores. First, the spectral line density in these ob-jects is high, leading to high levels of line blending andline confusion. Second, the linewidths of the molecu-lar line emission in hot cores are broad (several km s − ),which prevents clear identifications of weak lines fromless abundant species. Finally, hot cores are typically lo-cated at distances ≥ GLYCINE IN LOW-MASS STAR FORMING REGIONS
In contrast with their high-mass counterparts, low-mass star forming regions may be better suited for thedetection of glycine for several reasons. The level of lineconfusion, especially at the earliest stages represented bypre-stellar cores (Caselli et al. 2002a; Crapsi et al. 2005),is low because the measured gas temperatures are ≤
10 K(Crapsi et al. 2007; Pagani et al. 2007) and the num-ber of molecular lines excited at these temperatures issmaller than in hot cores. The molecular emission indark cloud cores also shows linewidths ≤ − (e.g.Caselli et al. 2002b; Crapsi et al. 2007), which allowsaccurate identifications of the observed transitions sincethey suffer less from line blending. COMs such as propy-lene (CH CHCH ), acetaldehyde (CH CHO), dimethylether (CH OCH ) or methyl formate (HCOOCH ) haveindeed been detected in dark cloud cores such as TMC-1and B1 (Marcelino et al. 2007) and the pre-stellar coreL1689B (Bacmann et al. 2012), unexpectedly revealing ahigh chemical complexity in the cold gas of these objects.Previous studies toward hot cores mainly targetedglycine lines with frequencies ≥
130 GHz because theirEinstein A coefficients, A ul , are A ul ≥ − s − , and theirupper level energies are E u ≥
60 K, in agreement withthe temperatures in these objects (Snyder et al. 2005). Jim´enez-Serra et al.However, for pre-stellar cores, the detection of glycineshould be attempted via observations of glycine lineswith low values of E u while having not too small A ul coefficients. In the frequency range between 60 GHz and130 GHz, glycine (conformer I) has several transitionswhose energy levels lie below 30 K and whose A ul coef-ficients are ≥ − s − , only a factor of 10 lower thanthose of the glycine lines observed in hot cores (Kuan etal. 2003; Snyder et al. 2005). In this Letter, we test thedetectability of glycine toward a well-known pre-stellarcore, L1544 (Caselli et al. 1999), by performing simpleradiative transfer calculations of glycine between 60 GHzand 250 GHz, and by making reasonable assumptions ofthe dominant chemical processes in pre-stellar cores andof the abundance of glycine on ices and in the gas phase. RADIATIVE TRANSFER MODELLING OF GLYCINE INTHE L1544 PRE-STELLAR CORE
The Model
For our calculations, we have considered the L1544 pre-stellar core located in the Taurus molecular cloud (dis-tance of 140 pc; Elias 1978). This core has largely beenstudied in the past and its internal physical structure isrelatively well-constrained (Ward-Thompson et al. 1999;Caselli et al. 2002a,b; Keto & Caselli 2010). Figure 1 re-ports the 1D spherically symmetric distribution of theH density, n , and gas temperature, T , derived by Keto& Caselli (2010) for this core, and considered in our cal-culations. We have also included in the model the gasvelocity profile (due to the subsonic phase of the corecontraction) and the linewidth of the emission deducedby Caselli et al. (2012).L1544 represents an excellent candidate for our anal-ysis of the detectability of glycine in pre-stellar coresbecause water vapour has recently been found towardthe central few thousand AU of this core (Caselli et al.2012). The distribution of the gas-phase water abun-dance, χ (H O), is shown in Figure 1 (blue line), where χ (H O) increases by a factor of ≥
100 within the cen-tral 10000 AU (from ∼ × − to ∼ × − ) to decreaseagain to ∼ − . As proposed by Caselli et al. (2012),water vapour in L1544 (with ∼ molecules. The models of Caselliet al. (2012) do not include the gas-phase chemistry ofwater. However, their simple calculations not only re-cover the results from the comprehensive gas-grain mod-els of Hollenbach et al. (2009), but also reproduce theinverse P-Cygni profile observed for water in this core.Since gas-phase chemistry is not the dominant processin the production of water vapour in L1544 (Caselli etal. 2012; Keto et al. 2014), we assume that it is mostlyproduced by comic ray induced FUV photons. We notethat this is consistent with the depletion of CO in L1544(by a factor of 100; Caselli et al. 1999) since a higherCO desorption rate (30 times higher than that used byHasegawa & Herbst 1993) is needed to explain the ob-served C O and C O lines (Keto & Caselli 2010). Thisindicates that CO is still present in the gas phase in the We only refer to glycine conformer I because the ground vibra-tional level of glycine conformer II lies 700 cm − ( ∼ Fig. 1.—
1D spherically symmetric distribution of the H num-ber density, n , and gas temperature, T , assumed in our calculationsfor the L1544 pre-stellar core (Keto & Caselli 2010). The radialdistribution of water vapour, χ (H O), was inferred by Caselli etal. (2012) from Herschel HIFI data (blue line). The glycine abun-dance, χ (Gly), is assumed to follow the distribution of water vapourin L1544, but scaled down by the fraction of solid glycine expectedto be formed on ices (red line; see Sections 3.1 and 3.3). core center as a result of the ice photodesorption by cos-mic ray induced FUV photons (Keto & Caselli 2010).In our calculations, we also consider that glycine isFUV photo-desorbed together with water in L1544. Mul-tilayer simulations of ice mantle formation indeed showthat complex organics (e.g. methanol and formaldehyde)are formed in the outer layers of the mantle, while wa-ter is uniformingly distributed across these layers (Cup-pen et al. 2009; Taquet et al. 2012). Therefore, most ofthe complex organics are expected to be released along-side with water once the mantle outer layers are photo-desorbed . From this, the distribution of gas phaseglycine is assumed to follow the water vapour abundanceprofile deduced in L1544 (see discussion in Section 3.3),but scaled down by the fraction of solid glycine expectedto be formed on ices. This can be calculated as: χ ( Gly )( r ) = χ ( Gly ) m χ ( H O ) m × χ ( H O )( r ) , (1)where χ ( Gly ) m is the abundance of solid glycine on ices, χ ( H O ) m is the water abundance (relative to H ) in themantles ( χ ( H O ) m =7.25 × − ; Whittet & Duley 1991),and χ ( H O )( r ) is the distribution of gas phase water inL1544 (Caselli et al. 2012). We assume a solid glycineabundance of χ ( Gly ) m ∼ − with respect to water onices ( ∼ − with respect to H ; see Section 3.3).As shown in Figure 1, the maximum gas-phase abun-dance of glycine reached in the model (red line) is χ ( Gly ) ∼ × − , which corresponds to column densi-ties ∼ cm − . This glycine abundance is reasonablesince it is a factor of ∼
100 lower than that of amino ace-tonitrile (NH CH CN, a precursor of glycine) measuredin SgrB2(N) (of ∼ × − ; Belloche et al. 2008). Theabundance ratio NH CH COOHNH CH CN ∼ is of the same orderof magnitude as the ratio between acetic acid and methylcyanide in SgrB2(N) ( CH COOHCH CN ∼ ; Belloche et al.2008), as expected if the pairs glycine/amino-acetonitrile The penetrability of UV photons is in the range between100 nm and 200 nm (Mu˜noz Caro & Dartois 2013). etectability of Glycine in Solar-type System Precursors 3
Fig. 2.—
Upper panel:
Simulations of the spectrum of glycine (conformer I) obtained for the frequency range between 60 GHz and250 GHz, considering the physical structure of the L1544 pre-stellar core (Figure 1 and Caselli et al. 2012), a solid glycine abundance of ∼ − on ices (Sections 3.1 and 3.3), and LTE conditions. Horizontal red lines indicate the frequency coverage of ALMA Bands 2, 3, 4and 6. The glycine lines with frequencies 60-80 GHz show peak intensities ≥ Lower panel:
LTE spectrum of glycine predicted for the same frequency range (60 GHz-250 GHz) but forthe AFGL2591 hot molecular core. The hot core’s physical structure derived by Jim´enez-Serra et al. (2012) is used in our calculations.We assume that all glycine within the mantles is released into the gas phase (abundance of ∼ − ; Section 4), and the line spectrum iscorrected by the beam filling factor of the hot core within a single-dish telescope beam (in this case, the IRAM 30 m telescope). The glycinelines at 1 mm are brighter than in the pre-stellar core case. However, line confusion and line blending is a major issue in the detection ofglycine in hot cores. Horizontal red lines are as in the upper panel. and acetic-acid/methyl-cyanide were formed by similarchemical pathways.The spectrum of glycine is obtained toward the corepeak position considering spherical symmetry and LTEconditions. The 1D radiative transfer model follows theprocedure described in Myers et al. (1996) and de Vries &Myers (2005), which was used for pre-stellar cores. Thecolumn density of glycine and the optical depth of everyline are calculated as a function of radius and the ra-diative transfer is performed along the line-of-sight con-sidering the core’s velocity profile and linewidth of theemission at every position. The spectroscopic informa-tion of glycine is extracted from the CDMS (M¨uller etal. 2005).Non-LTE calculations are not possible due to the lackof collisional coefficients of glycine with He or H . As-suming that these coefficients are of the same order asthose derived for methyl formate (HCOOCH , coeffi-cients of ∼ × − cm s − ; Faure et al. 2014), theestimated critical density for the glycine lines is n cr ∼ × cm − , similar to the average H density measuredtoward L1544 (n(H ) ∼ × cm − ; Caselli et al. 2002b).This implies that the glycine emission is probably not farfrom LTE. Results
In Figure 2 (upper panel), we show the simulated spec-trum of glycine obtained between 60 GHz and 250 GHzconsidering the physical and kinematic structure of L1544 (Section 3.1). While the lines at 1 mm and 2 mmshow brightness temperatures ≤ ≥
10 mK. The brighter emission at 3 mm and4 mm is due to the combination of low values of E u ( ≤ g u , for these lines.Since E u progressively increases for higher frequencies,the higher energy levels are more difficult to populatemaking their expected line intensities very low. Theglycine emission is optically thin across the core becausethe maximum optical depths attained for the brightestglycine lines are ≤ − .Since the glycine emission is extended across the core,this emission is expected to fill the beam of the telescopeand the brightness temperature, T B , will be similar tothe main beam temperature, T mb , measured by the tele-scope (T B ∼ T mb , beam-filling factor ∼ for reasonable abun-dances of glycine in the gas-phase, the glycine lines couldreach detectable levels in pre-stellar cores between 60 GHzand 80 GHz. We note that even if glycine did not follow the gas-phase water profile deduced for L1544 (Caselli et al.2012), peak intensities ≥
10 mK would still be measuredbetween 60 GHz and 80 GHz if a constant glycine abun-dance of 1.5 × − were present in the gas phase in this Jim´enez-Serra et al. Table 1. Sample of brightest glycine lines predicted for L1544
Line Transition Frequency a A ul E u g u (MHz) (s − ) (K)1 10 , → , × − , → , × − , → , × − , → , × − , → , × − , → , × − , → , × − , → , × − , → , × − a Extracted from the CDMS (M¨uller et al. 2005). core. This gas-phase abundance is the lowest level thatcould make glycine detectable in cold pre-stellar cores.
Formation and Destruction Mechanisms of Glycine
In its solid form, the amount of glycine that canbe synthesized on ices has been investigated in labo-ratory experiments of ion- and UV-irradiated interstel-lar ice analogs. Mu˜noz Caro et al. (2002) and Bern-stein et al. (2002) reported the formation of glycine onUV-irradiated ices with fractions ∼ ∼ × − -2.5 × − with respect to H . As explained inBernstein et al. (2002), the UV photon flux used in theseexperiments corresponds to the typical interstellar dosewithin dense molecular clouds for cloud ages ∼ yr andvisual extinctions A v ∼ ∼ cm − (Holtom et al. 2005). The interaction ofenergetic electrons in the track of a cosmic-ray impacton a dust grain yields HOCO and NH CH neighbour-ing complexes which subsequently recombine formingglycine (Holtom et al. 2005). The formation of glycinevia the reaction HOCO + NH CH → NH CH COOH re-quires no entrance barrier, and thus, it is feasible at 10 K(Holtom et al. 2005). Therefore, even if glycine werenot photo-desorbed from ices in L1544, column densitiesof ∼ cm − could still be found in the gas phase as-suming a 1% efficiency for the desorption of glycine viacosmic ray impacts (L´eger et al. 1985) and/or exother-mic surface reactions (Garrod et al. 2007). These columndensities are similar to those considered in our modelling(Section 3.2).The destruction rate of glycine by pure gas-phase re-actions at low temperatures (mainly with H , H O + and HCO + ) is expected to decrease the abundance ofthis molecule by a factor of 10 in ∼ × yrs (as de-rived for methyl formate, a precursor of glycine, forwhich its gas phase reaction rates are relatively well-constrained). However, in the cold gas of pre-stellarcores, other processes such as surface diffusion by tun-neling effects of atomic oxygen (Minissale et al. 2014),cosmic rays induced diffusion (or CRID; Reboussin et al.2014), and episodic chemical explosions on ices (Rawl-ings et al. 2013) could also enhance the abundance ofCOMs by factors from a few to ∼ A v ∼ ∼ − could be presentin the cold gas of pre-stellar cores, as considered in ourcalculations. COMPARISON WITH MASSIVE HOT CORES
For comparison, in Figure 2 (lower panel) we also showthe LTE spectrum of glycine calculated for a hot molecu-lar core from 60 GHz to 250 GHz. This is to illustrate thechanges in the predicted spectrum of glycine at temper-atures typical of hot cores. For our calculations, we haveselected the AFGL2591 hot core for which its internalphysical structure is well constrained (van der Tak et al.1999; Jim´enez-Serra et al. 2012). The hot core’s physicalstructure, corrected by its new distance ( ∼ χ ( Gly ) ∼ − ]. We make predictions for asource size ∼ ∼ .Despite the larger abundance of glycine in hot sources,Figure 2 (lower panel) shows that the 3 mm and 4 mmglycine lines are factors ∼ ≥
130 K; and ii) the large beam dilution produced by thesingle-dish telescope. If observed with interferometers(beam sizes ∼ ≤ -10 cm − ). As noted inSection 1, the presence of more abundant COMs in hotcores whose emission has linewidths of some km s − , is amajor issue in the detection/identification of glycine inhot sources due to line blending and line confusion. IMPLICATIONS FOR ALMA BAND 2
The advent of new instrumentation providing higher-sensitivity observations such as the Atacama Large Mil-limeter Array (ALMA), opens up the possibility to detecta large number of COMs of biochemical interest such asglycine in low-mass star forming regions. Our modellingof Section 3 is based on reasonable assumptions of thedominant chemical processes in pre-stellar cores and ofthe glycine abundance on ices and in the gas phase. Thismodelling not only shows that glycine could be detected The derived T B needs to be corrected by the beam filling factordefined as θ s / [ θ s + θ b ], with θ s the source size and θ b the Gaussianbeam size. θ b is given by the diffraction limit (Maret et al. 2011). etectability of Glycine in Solar-type System Precursors 5in these cold objects if abundances as low as ∼ − were present in the gas phase, but more importantly,that this emission, if present, should be searched for inthe frequency range between 60 GHz and 80 GHz wherethe predicted glycine intensities are ≥ , making thisband very well suited for the discovery of glycine andother pre-biotic molecules in Solar-type system precur-sors. In Table 1, we summarize the glycine transitionsthat represent the best targets to detect glycine in coldsources. We note that other glycine lines with E u ≤
20 Kdo exist at frequencies ≤