This site uses cookies. By continuing to use this site you agree to our use of cookies. To find out more, see our Privacy and Cookies policy. Close this notification
NOTICE: Ukraine: Read IOP Publishing's statement.

The American Astronomical Society (AAS), established in 1899 and based in Washington, DC, is the major organization of professional astronomers in North America. Its membership of about 7,000 individuals also includes physicists, mathematicians, geologists, engineers, and others whose research and educational interests lie within the broad spectrum of subjects comprising contemporary astronomy. The mission of the AAS is to enhance and share humanity's scientific understanding of the universe.

The Institute of Physics (IOP) is a leading scientific society promoting physics and bringing physicists together for the benefit of all. It has a worldwide membership of around 50 000 comprising physicists from all sectors, as well as those with an interest in physics. It works to advance physics research, application and education; and engages with policy makers and the public to develop awareness and understanding of physics. Its publishing company, IOP Publishing, is a world leader in professional scientific communications.


Published 2010 September 29 © 2010. The American Astronomical Society. All rights reserved.
, , Citation Jérôme Aléon 2010 ApJ 722 1342



Isotopic fractionation and mixing calculations compared with coupled hydrogen and nitrogen isotopic composition of organic molecules from primitive chondrites, interplanetary dust particles (IDPs), and comets C/1995 O1 (Hale-Bopp) and 81P/Wild2 reveal that meteoritic and cometary organic matter contains three different isotopic components of different origins. (1) A major component of carbonaceous chondrites, IDPs, and comets Hale-Bopp and Wild2 shows correlated H and N isotopic compositions attributable to isotope exchange between an organic matter of solar composition and a reservoir formed by ion–molecule reactions at T < 25 K under conditions where competing reactions are strongly inhibited, possibly in the final evolutionary stages of the presolar cloud core, or more likely in the coldest outer regions of the solar protoplanetary disk. (2) In carbonaceous chondrites, IDPs, and comet Wild2, this component is mixed with a 15N-rich component having identical 15N and D enrichments relative to the protosolar gas. Temperatures > 100 K deduced from the low D/H ratio and an anti-correlation between the abundance of this component and meteoritic age indicate a late origin in the solar protoplanetary disk. N2 self-shielding and the non-thermal nucleosynthesis of 15N upon irradiation are possible but unlikely sources of this component, and a chemical origin is preferred. (3) An interstellar component with highly fractionated hydrogen isotopes and unfractionated nitrogen isotopes is present in ordinary chondrites. A dominantly solar origin of D and 15N excesses in primitive solar system bodies shows that isotopic anomalies do not necessarily fingerprint an interstellar origin and implies that only a very small fraction of volatile interstellar matter survived the events of solar system formation.

Export citation and abstract BibTeX RIS


Nitrogen has two isotopes of a partly primary and a partly secondary origin in stellar nucleosynthesis. All determinations of 14N/15N ratios in interstellar HCN and NH3 in the galaxy yielded 14N/15N ratios on the order of 400 or higher, including giant molecular clouds (Wannier et al. 1981), warm dense clouds (Güsten & Ungerechts 1985; Dahmen et al. 1995), cold prestellar cores, and a class 0 protostar (Gérin et al. 2009; Lis et al. 2010). This ratio is broadly consistent with current knowledge of stellar nucleosynthesis and galactic chemical evolution. The bulk 14N/15N ratio of the initial protosolar nebula (PSN) as determined from NH3 in the atmosphere of Jupiter (14N/15N = 435 ± 57, Owen et al. 2001, 14N/15N = 450 ± 106, Fouchet et al. 2004) from osbornite (TiN) in an early solar system condensate (14N/15N = 424 ± 3; Meibom et al. 2007) and more recently from solar wind N ions implanted in sample targets returned by the Genesis spacecraft (14N/15N = 442 ± 66; Marty et al. 2010) agrees with these interstellar values. The lowest interstellar 14N/15N ratios are at most marginally consistent with the terrestrial atmosphere (14N/15N = 272) but remain significantly higher than those recently found in cometary CN (14N/15N ∼ 140; Arpigny et al. 2003; Manfroid et al. 2009) and HCN (Bockelée-Morvan et al. 2008) and in organic molecules from primitive chondritic meteorites (14N/15N ratios down to ∼50; Bonal et al. 2009; Briani et al. 2009). The origin of these 15N excesses in early solar system materials remains poorly understood. Because the largest 15N excesses are carried out by organic molecules as are the largest D excesses (e.g., Messenger 2000; Aléon et al. 2001, 2003; Busemann et al. 2006; Briani et al. 2009) and because D enrichments up to a factor 104 in organic molecules are widespread in cold molecular clouds due to low-temperature ion–molecule reactions (e.g., Millar 2002), an interstellar origin has repeatedly been proposed for these organic molecules and for both their 15N and D excesses (e.g., Messenger 2000; Floss et al. 2006; Busemann et al. 2006, 2009), despite classical ion–molecule reaction models failing to produce such 15N enhancements (Terzieva & Herbst 2000). To overcome this difficulty, highly specific conditions (Charnley & Rodgers 2002; Aléon & Robert 2004; Rodgers & Charnley 2008a) or specific reaction rates (Rodgers & Charnley 2008b) have been proposed, so that large 15N enrichments can theoretically be expected in prestellar dense cores despite the lack of convincing observations.

Recent astronomical observations have demonstrated that extreme D enrichments are ubiquitous in low-mass star formation environments (Ceccarelli et al. 2007) including prestellar cores (e.g., Caselli et al. 2003; Vastel et al. 2004), class 0 protostars (e.g., Ceccarelli et al. 2001), and the outer midplane of evolved protoplanetary disks (e.g., Ceccarelli et al. 2004; Qi et al. 2008). D excesses alone thus no longer appear very good diagnostics of a prestellar versus protoplanetary disk origin of organic molecules in primitive solar system objects. In this respect, the interstellar origin of meteoritic organic matter has recently been called into question with the discoveries that (1) D-excesses are anti-correlated with the C−H bonding energy (Remusat et al. 2006) and (2) organic radicals in the Orgueil meteorite carry highly exchangeable hydrogen with D/H ratios reaching interstellar values at 1.5 × 10−2 (Gourier et al. 2008). Such a distribution of D excesses at the molecular scale strongly suggests that organics in meteorites initially had a solar D/H ratio and became deuterated lately by isotopic exchange with a D-rich gas in the cold dense parts of the solar protoplanetary disk rather than being inherited from the parent molecular cloud (Remusat et al. 2009). In the latter case, the largest D-excesses should be associated with the strongest C−H bonds (aromatic H), contrary to what is observed. The recent discovery of ultra-carbonaceous antarctic micrometeorites (UCAMMS) of likely cometary origin having widespread very high D/H ratios suggests a similar situation in comets (Duprat et al. 2010). Indeed, silicate inclusions in the D-rich organic matter are dominantly crystalline with a solar O isotopic composition, which indicates that this D-rich organic matter condensed in a region where amorphous interstellar silicates are a minor component, i.e., in a disk rather than in a dense core. As a result, H isotopes cannot be used to support an interstellar origin of 15N excesses in solar system primitive objects.

Furthermore, the systematic imaging at the ∼μm scale of H and N isotopes in meteoritic organic matter and interplanetary dust particles (IDPs) has shown that the largest D excesses and the largest 15N excesses are not spatially correlated (Busemann et al. 2006, 2009; Robert et al. 2006; Briani et al. 2009). It has thus been suggested that the 15N enrichments may result from processes independent from those responsible for the D excesses. Among alternative hypotheses, the self-shielding of N2 in the early solar system, in a similar manner to the self-shielding of CO, may yield the observed 15N excesses. This has not been demonstrated although detailed modeling is currently under study (Lyons et al. 2009). In addition, if direct nucleosynthetic inputs from massive stars, supernovae ejecta, or novae are unlikely, the in situ production by solar or galactic cosmic ray (GCR) spallation or the trapping of GCR in the protosolar core might be worth taking into consideration. In order to investigate the respective role of ion–molecule reactions and other possible processes, I revisit here the systematics of H and N isotopes in organic molecules from primitive chondrites, IDPs, and the long- and short-period comets C/1995 O1 (Hale-Bopp) and 81P/Wild2, respectively, by comparing published coupled H and N isotopic data (Table 1) with a combination of isotopic fractionation and mixing calculations.

Table 1. H and N Isotopic Composition of Meteoritic and Cometary Organic Matter and Fractionations Relative to the PSN

Samples D/H (10−6) σ D/Ha 14N/15N σ 14/15a Refb αD log(αD) α15N log(α15N)
Bulk IOM CCs                  
Orgueil (CI) 306 n 264 n A07 15 1.16 1.59 0.20
Ivuna (CI) 307 n 264 n A07 15 1.16 1.59 0.20
Tagish Lake (ungr. C2) 247 n 253 n A07 12 1.07 1.65 0.22
Bells (anomal. CM) 664 n 192 n A07 32 1.50 2.18 0.34
Kivasvaara (CM) 265 n 271 n A07 13 1.10 1.54 0.19
Murchison (CM) 275 n 272 n A07 13 1.12 1.54 0.19
Mighei (CM) 279 n 272 n A07 13 1.12 1.54 0.19
Murray (CM) 293 n 270 n A07 14 1.15 1.55 0.19
DOM03183 (CM) 254 n 273 n A07 12 1.08 1.54 0.19
ColdBokk (CM) 269 n 273 n A07 13 1.11 1.54 0.19
ALH83100 (CM) 267 n 274 n A07 13 1.10 1.53 0.18
MET01070 (CM) 262 n 274 n A07 12 1.10 1.53 0.18
MET00426 (CR) 628 n 231 n A07 30 1.48 1.81 0.26
QUE99177 (CR) 642 n 229 n A07 31 1.49 1.83 0.26
EET92042 (CR) 620 n 230 n A07 30 1.47 1.82 0.26
GRA95229 (CR) 606 n 236 n A07 29 1.46 1.77 0.25
Al Rais (CR) 560 n 234 n A07 27 1.43 1.79 0.25
GRO95577 (CR) 616 n 221 n A07 29 1.47 1.90 0.28
LEW85332 (CR) 702 n 208 n A07 33 1.52 2.01 0.30
Leoville (CV) 366 n 277 n A07 17 1.24 1.51 0.18
Vigarano (CV) 285 n 278 n A07 14 1.13 1.51 0.18
Efremovka (CV) 266 n 279 n A07 13 1.10 1.50 0.18
Bulk IOM UOC                  
Semarkona (LL3.05) 515 n 265 n A07 25 1.39 1.58 0.20
QUE97008 (L3.05) 651 n 272 n A07 31 1.49 1.54 0.19
MET00526 (L3.05) 857 n 273 n A07 41 1.61 1.53 0.19
MET00452 (H3.5) 570 n 274 n A07 27 1.43 1.53 0.18
Bishunpur (L/LL3.15) 581 n 269 n A07 28 1.44 1.56 0.19
LEW86018 (L3.1) 547 n 262 n A07 26 1.42 1.59 0.20
GRO95502 (L3.2) 666 n 273 n A07 32 1.50 1.53 0.19
GRO95504 (L3.2) 647 n 271 n A07 31 1.49 1.55 0.19
GRO95505 (L3.4) 776 n 282 n A07 37 1.57 1.49 0.17
MET96503 (L3.1) 773 n 275 n A07 37 1.57 1.52 0.18
MET96515 (L3.1) 775 n 274 n A07 37 1.57 1.53 0.18
Krymka (LL3.2) 452 n 275 n A07 22 1.33 1.52 0.18
WSG95300 (H3.3) 1113 n 283 n A07 53 1.72 1.48 0.17
MET00506 (H3.4) 761 n 275 n A07 36 1.56 1.52 0.18
Tieschitz (H/L3.6) 1016 n 270 n A07 48 1.68 1.55 0.19
IOM hotspots                  
Maximad 3162 713 65 11 Bu06 151 2.18 6.46 0.81
Tagish Lake G15–1 556 9 185 3 NM06 26 1.42 2.27 0.36
Tagish Lake G15–2 616 32 192 4 NM06 29 1.47 2.18 0.34
Tagish Lake G15–3 584 52 184 4 NM06 28 1.44 2.27 0.36
Tagish Lake G8–1 635 17 136 1 NM06 30 1.48 3.09 0.49
Tagish Lake G8–2 1413 177 173 3 NM06 67 1.83 2.42 0.38
Tagish Lake G8–3 1076 274 188 4 NM06 51 1.71 2.23 0.35
Tagish Lake G5–1 699 84 179 2 NM06 33 1.52 2.34 0.37
Tagish Lake G5–2 432 80 187 3 NM06 21 1.31 2.24 0.35
Organic clasts                  
Isheyevo PX 18 hotspots 182e 38e 53 2 Br09 9 0.94 7.86 0.90
Isheyevo PX 18 hotspots 182e 38e 46 2 Br09 9 0.94 9.08 0.96
Isheyevo PX 18 hotspots 182e 38e 58 4 Br09 9 0.94 7.23 0.86
Isheyevo PX 18 hotspots 182e 38e 66 8 Br09 9 0.94 6.31 0.80
Isheyevo PX 18 area 1 192 5 239 3 Br p.c. 9 0.96 1.75 0.24
Isheyevo PX 18 area 2 164 4 167 2 Br p.c. 8 0.89 2.51 0.40
Isheyevo PX 18 area 3 182 4 160 2 Br p.c. 9 0.94 2.62 0.42
OM1 253 18 225 9 A01 A03 12 1.08 1.86 0.27
OM2 1510 116 198 8 A01 A03 72 1.86 2.11 0.32
OM3 1880 112 197 3 A01 A03 90 1.95 2.13 0.33
L2054 E1-A hotspot 3410 n.a. 170 n.a. Bu09 162 2.21 2.50 0.40
Wild2 233 u.l. 128 2 D09 11 1.04 3.26 0.51
Wild2 310 26 274 1 D09 15 1.17 1.53 0.18
Hale-Bopp 2300 400 172f 24f M98 B08 110 2.04 2.43 0.39
PSN 21 5 424 3 GG98 M07 1 0 1.00 0.00

Notes. aAbbreviations. n: negligible compared to uncertainty on PSN values; u.l.: upper limit; n.a.: not available. 1σ error. bAbbreviations. A07: Alexander et al. 2007; Bu06: Busemann et al. 2006; NM06: Nakamura-Messenger et al. 2006; Br09: Briani et al. 2009; Br p.c.: G. Briani 2009, private communication; A01: Aléon et al. 2001; A03: Aléon et al. 2003; Bu09: Busemann et al. 2009; D09: DeGregorio et al. 2009; M98: Meier et al. 1998; B08: Bockelée-Morvan et al. 2008; GG98: Geiss & Gloeckler 1998; M07: Meibom et al. 2007. cAbbreviations. CCs: carbonaceous chondrites; UOC: unequilibrated ordinary chondrites; ColdBokk: Cold Bokkeveld; ungr.: ungrouped; anomal.: anomalous; OM1, 2, 3: endmembers defined from the mixing of numerous hotspots. dN isotopes are not available for maximum H hotspots and vice versa, the maxima reported here are from different hotspots and can be considered upper limits for hotspots in carbonaceous chondrite IOMs. eMedian of D/H ratios reported in Br09 (∼homogeneous in the sample). Total range in D/H ratios taken as the error. fWeighted mean of the three estimates reported in B08.

Download table as:  ASCIITypeset images: 1 2


Because in meteorites multiple H and N carriers are present and often mixed at a very fine scale, it is crucial to obtain H and N isotopic composition of pure organic phases that can be subsequently compared. The lack of careful data selection can induce confusing apparent relationships (e.g., Messenger & Walker 1997) since in many objects water is much less D-rich than organic matter, as previously evidenced in carbonaceous chondrites or IDPs (e.g., Aléon et al. 2001). The data used in the present model (Table 1) are coupled H and N isotopes from (1) bulk acid insoluble macromolecular organic matter (IOM) extracted from carbonaceous and ordinary chondrites (Alexander et al. 2007), (2) D-rich and/or 15N-rich subgrains of chondritic IOMs, chondritic clasts, and IDPs, hereafter referred to as "hotspots," (Aléon et al. 2001, 2003; Busemann et al. 2006, 2009; Briani et al. 2009), for which a macromolecular organic nature is evidenced by C/H ratios or Raman microspectroscopy, (3) organic nanoglobules from the Tagish Lake carbonaceous chondrite (Nakamura-Messenger et al. 2006) and from the short-period comet 81P/Wild2 (DeGregorio et al. 2009), and (4) the revised isotopic composition of HCN in the long-period comet C/1995 O1 (Hale-Bopp) (Meier et al. 1998; Bockelée-Morvan et al. 2008). Despite the N isotopic composition of CN being known for 18 comets of the various taxonomic groups (Manfroid et al. 2009), H and N isotopes are available simultaneously for only four of these comets (Jehin et al. 2009). In three of these four comets, H isotopes have been measured in proxies for water (H2O, OH, and atomic H; Jehin et al. 2009). As a result, only the data on Hale-Bopp HCN are of use for the present study. Finally, note that details on the classification of meteorites and IDPs can be found in Krot et al. (2003) and Bradley (2003), respectively.

Fractionation calculations are done following Aléon & Robert (2004) and are only summarized here. In an isotope exchange reaction, the reaction constant is also called the fractionation factor α. At equilibrium, α is a function of the free enthalpy of reaction. At temperatures typical of the interstellar medium (ISM) (T ≤ 100 K), the free enthalpy of reaction is controlled by the difference in zero-point energy between the various isotopologues and is negative (exothermic reactions). It corresponds to the exothermicity of reaction, noted as ΔE in the following. As a result, α can be written as a function of exothermicity:

Equation (1)

where k is the Boltzmann constant. A pre-exponential term depending on the symmetry of the molecules and almost always close to 1 (Terzieva & Herbst 2000) is assumed to be 1 in the following. Cases where it could be equal to 2 are neglected assuming that they would result in specific discrepancies between observations and calculations that can easily be traced back.

With αH and αN, the fractionation factors for H and N isotopes, and ΔEH and ΔEN, the exothermicities of reaction, the relationship between the H and N isotopic fractionation factors can be written as

Equation (2)

which implies that a straight line going through the origin in a (log (αH), log (αN)) diagram corresponds to coupled H and N isotope exchange (or fractionation) reactions.

Most ion–molecule reactions have no activation energy barriers in the forward direction and are rapid. As a result, the equilibrium assumption is not necessarily valid. In particular, the final isotopic fractionation depends on the relative reaction rates of the isotope exchange reaction(s) and of competing reactions. Such competing reactions are for instance dissociative recombination, radiative association, or reactions with other molecular species such as CO (e.g., Millar 2002). Thus, the observed isotopic fractionation will be all the more close to that predicted by the isotope exchange reaction(s) at equilibrium as competing reactions are efficiently suppressed.

In the following, equilibrium fractionation factors are calculated relative to the initial D/H and 14N/15N ratios of the major reservoirs H2 and N2 (or N assuming initial isotopic equilibrium with N2) in the PSN or in the ISM 4.57 Gyr ago. The initial D/H ratio of the PSN (D/H = 2.1 ± 0.5 × 10−5) is taken from measurements of H2 in the atmosphere of Jupiter (Mahaffy et al. 1998) and (3He + D)/H relative abundances in the solar photosphere (Geiss & Gloeckler 1998). The nitrogen isotopic composition of the initial PSN gas is taken from the high precision analysis of TiN from an early solar system condensate, a Ca–Al-rich inclusion from the Isheyevo carbonaceous chondrite (14N/15N = 424 ± 3; Meibom et al. 2007). Despite the most accurate estimate being that of the solar wind samples returned by the Genesis mission (Marty et al. 2010), because the Sun represents more than 99.8 wt% of the solar system, the TiN value is preferred here to avoid artificially large errors on the fractionation factors and obtain errors closer to the uncertainties on the composition of organic matter.

Mass balance mixing calculations are used to define apparent fractionation factors in order to unravel possible mixtures of several components having different levels of isotopic fractionation for H and N. The mixing equation in 15N/14N versus D/H space is

Equation (3)

which is the equation of a hyperbola with D/H0 and 15N/14N0, the asymptotes parallel to the N isotopes and H isotopes coordinates, respectively, and [D/H0 × 15N/14N0 − γ] is the curvature factor. D/H0, 15N/14N0, and γ are constants (given in the Appendix) depending on the D/H and 15N/14N ratios of the endmember components 1 and 2 and on their relative H/N ratios. Because the shape and curvature of the mixing lines depend on the respective H and N concentrations in the two endmembers, it is necessary to distinguish different mixing involving different molecules. Molecules considered here are macromolecular organic matter with a chemical composition comparable to meteoritic IOMs and CHON grains from comet Halley (H/N = 20) (Alexander et al. 2007), ammonia (H/N = 3), and HCN (H/N = 1). Specific mixing cases reported here include isochemical mixing (two endmembers with similar H/N ratios, e.g., CHON + CHON or HCN + HCN) and mixing with strong chemical contrast (i.e., when the difference in the H/N ratios is maximized, such as CHON + HCN which could possibly correspond to addition of small nitrile functionalities onto a large macromolecular structure). Mixing cases involving ammonia are always intermediate between the two. As a result, variations of (H/N)1/(H/N)2 between 20 and 1/20 are considered here, a wide range of possible chemical mixing.

Finally, isotopic fractionation factors are calculated using the 15N/14N ratio and results are discussed in the text using either the 14N/15N ratio or the δ15N notation, in order to facilitate the comparison with astronomical and meteoritic data, respectively. The δ15N notation gives N isotopic ratios as per mil deviations from that of the terrestrial atmospheric N2 (AIR), taken as a reference value [δ15N = (15N/14Nsple15N/14Nair)/(15N/14N)air × 1000; with 15N/14Nair = 3.673 × 10−3, i.e., 14N/15Nair = 272].


3.1. Overall Distribution of H and N Isotopic Fractionations Relative to the PSN

Three groups of isotopic compositions can be distinguished (Figure 1).

Figure 1.

Figure 1. N isotopic fractionation between organic matter and the protosolar gas as a function of H isotopic fractionation. Black dots (group 1) correspond to bulk IOM in carbonaceous chondrites, D-rich hotspots in IDPs, and Hale-Bopp HCN. Gray dots (group 2) correspond to hotspots in carbonaceous chondrites, IDPs, and comet Wild 2. White dots (group 3) correspond to bulk IOM of unequilibrated ordinary chondrites. Arrows correspond to maximal D/H and 15N/14N ratios in (different) hotspots from carbonaceous chondrites (Busemann et al. 2006). The shaded area corresponds to the locus of hotspots from IDPs collected during the meteor shower of comet Grigg-Skjellerup, which contains data from both groups 1 and 2. Also reported are mixing lines between the extreme 15N-rich component in Isheyevo and the extreme D-rich component observed in gas-phase molecules in the ISM. Plain curve: isochemical mixing; dashed curve: Isheyevo HCN + ISM CHON; dotted curve: Isheyevo CHON + ISM HCN. None of the meteoritic and cometary data can be explained by such a mixing. Data shown with 1σ uncertainties.

Standard image High-resolution image

  • 1.  
    Group 1. Bulk IOMs from carbonaceous chondrites (Alexander et al. 2007), the most D-rich hotspots in IDPs (Aléon et al. 2001, 2003; Busemann et al. 2009), including an extreme value recently reported for an IDP collected during the meteor shower of the short-period comet 26P/Grigg-Skjellerup (hereafter GS-IDPs; Busemann et al. 2009), and Hale-Bopp HCN (Bockelée-Morvan et al. 2008) show linearly correlated isotopic fractionation between H and N in log scale. A significant fraction of hotspots in the GS-IDPs follows this fractionation trend with some spread due to analytical uncertainties.
  • 2.  
    Group 2. Nanoglobules in the Tagish Lake ungrouped carbonaceous chondrite (Nakamura-Messenger et al. 2006) and comet Wild2 (DeGregorio et al. 2009), as well as numerous hotspots in GS-IDPs (Busemann et al. 2009) and in carbonaceous chondrite IOMs (Busemann et al. 2006) and the bulk IOMs of Tagish Lake, of the Bells anomalous CM chondrite, and of the Lewis Cliff (LEW) 85332 CR chondrite (Alexander et al. 2007) exhibit significant 15N excesses relative to the above-mentioned correlated H–N isotopic fractionation. These excesses are maximal in organic clasts from the Isheyevo CH/CB chondrite (Briani et al. 2009).
  • 3.  
    Group 3. Bulk IOMs from unequilibrated ordinary chondrites (UOCs; Alexander et al. 2007) show large D excesses not associated with 15N excesses.

None of these groups can be explained by a binary mixture between a 15N-rich, D-poor endmember akin to the extreme Isheyevo hotspots and a D-rich, 15N-poor endmember akin to observed interstellar NH3 and HCN, including both isochemical mixtures and mixtures with a strong chemical contrast (Figure 1). In the following, each group is examined individually.

3.2. Correlated H and N Isotopic Fractionation Due to Ion–Molecule Reactions

The bulk IOMs from most carbonaceous chondrites (CI, CM, the least altered and the least metamorphosed CV, most CR), the most D-rich grains from IDPs, including GS-IDPs, and Hale-Bopp HCN (group 1) exhibit correlated H and N fractionation factors indicating that their isotopic composition is controlled by a fractionation mechanism involving coupled strong D enrichment and moderate 15N enrichment (Figure 2). The fractionation line is adequately fitted using a pair of reaction exothermicities (220 K, 38 K) for ΔEH and ΔEN, respectively. These reaction exothermicities indicate that the fractionation is controlled by reaction (4) between H3+ and HD for hydrogen and by the most fractionating ion–molecule reactions (reactions (5)–(7)) studied theoretically for nitrogen (Terzieva & Herbst 2000):

Equation (4)

Equation (5)

Equation (6)

Equation (7)
Figure 2.

Figure 2. N isotopic fractionation between organic matter and the protosolar gas as a function of H isotopic fractionation among samples of group 1. Gray dots are bulk IOM from carbonaceous chondrites, black dots are D-rich hotspots in IDPs, and white dot is Hale-Bopp HCN. The correlation is well fitted by coupled H,N isotopic fractionation using reaction exothermicities of 220 K and 38 K for H and N fractionation reactions, respectively (plain line). A range of exothermicities still possible for the N fractionation reaction is shown by two dotted lines. Note that the bulk IOM of Tagish Lake (TL), Bells (B), and LEW 85332 (L) depart from this trend. Data shown with 1σ uncertainties. See the text for discussion.

Standard image High-resolution image

Using a more recent estimate of ΔEH for reaction (4) (232 K; Roueff et al. 2007), the fit is obtained for ΔEN = 40 K. The agreement between the data and the considered reactions is surprisingly good given that ion–molecule reactions are considered to be inefficient in producing large 15N enhancements in interstellar conditions because reactions (5) to (7) are by-passed by the cycling between atomic N and N2, which results in a small N abundance and in fine in a small isotopic fractionation (Terzieva & Herbst 2000; Charnley & Rodgers 2002). Similarly, the hydrogen isotope fractionation is usually not explained by reaction (4) alone but other rapid reactions such as dissociative recombination and reactions of H2D+ with neutral molecules, notably CO, need to be taken into account (e.g., Millar 2002). This implies that the fractionation recorded by meteoritic and cometary molecules must have occurred in conditions where rapid competing reactions removing H2D+ and atomic N are efficiently inhibited. For instance, the selective condensation of CO onto grain surfaces relative to N2 has been shown to result in an accumulation of atomic N in the gas relative to N2, which prevents dilution of the equilibrium isotopic fractionation by 15N-poor N2 (Charnley & Rodgers 2002; Rodgers & Charnley 2008a). Detailed chemical models are required here to fully explore the implications of the close-to-equilibrium conditions; however, a very high degree of molecular depletion, notably for CO, appears to be required. H isotopes alone give another constraint: dissociative recombination of H2D+ with electrons must be very small, which implies extremely low fractional electron abundances (x(e)), lower than in most molecular cloud conditions (Millar 2002). The comparison with models (e.g., Roberts et al. 2003) and observations of H2D+ indicates that very low fractional electron abundances are achieved in dense prestellar cores (x(e) ∼ 10−9, Caselli et al. 2003; Vastel et al. 2004) and in the outer midplane of protoplanetary disks (10−10x(e) ≤ 10−9; Ceccarelli et al. 2004). Whether the ΔEN obtained for the fit (38–40 K) is significantly higher than the maximum theoretical ΔEN (36 K) is unclear; if this is the case, it may indicate that the H isotopic fractionation remains slightly below the equilibrium value due to the competing reactions. Altogether, the close-to-equilibrium conditions indicate that the fractionation occurred in an extremely cold and dense, CO-depleted, environment, either in the dense presolar cloud core or in the outer midplane of the solar protoplanetary disk, in agreement with theoretical expectations. However, the distinction between the two cannot be made on the sole basis of the equilibrium fractionation factors, but additional insights can be obtained by comparing the amplitude of the fractionation with the temperatures of reaction.

In a system, where H isotopic fractionation is controlled dominantly by reaction (4), the hydrogen isotopic composition of organic matter can be used to determine the temperatures of reaction. These temperatures are between ∼100 K (bulk CI, CM, CO, CV chondrites) and ∼50 K (D-rich grains in IDPs, Hale-Bopp HCN), a range at which reaction (4) does not alone control the hydrogen isotopic fractionation due to a significant deuteration of CH+3 and C2H+2 directly from HD (e.g., Millar 2002; Roueff et al. 2007). This disagreement suggests that organic molecules do not directly record the isotopic fractionation but rather record the isotopic exchange between an unfractionated solar organic matter with a highly fractionated reservoir, identified to be H2D+ from reaction (4). This is in agreement with the observed anti-correlation between the C−H bonding energy and the level of D-enrichment in organic moieties of chondritic IOM (Remusat et al. 2006). Low amounts of isotopic exchange are observed in bulk IOMs dominated by aromatic and aliphatic hydrogen with high C−H bonding energy, whereas high amounts of exchange are found in the most D-rich subgrains, likely enriched in D-rich organic radicals with highly labile hydrogen (Gourier et al. 2008; Remusat et al. 2009). This indicates that part of the 15N excesses in meteoritic organic matter is similarly due to a late isotopic exchange, coupled with that of hydrogen, between a solar organic matter and a reservoir produced by ion–molecule reactions. A late isotopic exchange requires that similarly to D, the 15N excesses produced by ion–molecule reactions are associated with organic moieties having easily exchangeable N. Unless the bulk D-poor, 15N-poor IOM precursor can be formed with solar-like isotopic compositions in a warm interstellar cloud and acquired its D and 15N excesses by isotope exchange in the presolar dense core, which is unlikely because the aromatic moieties in IOM are smaller than interstellar polycyclic aromatic hydrocarbons (Derenne et al. 2005), the IOM precursor must have formed after the dense core phase and could have acquired its D and 15N excesses in the cold dense regions of the solar protoplanetary disk.

Having a different chemical structure, Hale-Bopp HCN probably did not undergo this preferential isotopic exchange, but more likely provides a direct record of the coupled D-15N fractionation upon transfer of the initial excesses via an ion–molecule reaction network. A possible scenario that could account for the common gas-phase occurrence and deuteration of HCN above 20 K, as observed in the ISM (e.g., Millar et al. 1989; Turner 2001; Roueff et al. 2007), would be that IOM and HCN record isotopic fractionation and isotopic exchange from different layers of a protoplanetary disk (such as the midplane and a warmer molecular layer above the midplane; e.g., Aikawa et al. 2002). Again, a detailed chemical model is required to assess this proposition. Alternatively, different origins would be required for IOM and HCN.

3.3. Addition of a Late 15N-rich Protosolar Component

Numerous samples show evidence for a significant excess of 15N above the ion–molecule reaction fractionation line (hereafter IMR fractionation line), notably μm to sub-μm hotspots in meteoritic IOMs, IDPs, and nanoglobules from Tagish Lake and comet Wild2 (group 2). In some cases, this 15N-rich organic matter is abundant enough to be detected in bulk IOM samples such as in Tagish Lake, Bells, and the LEW 85332 CR chondrite (Figure 2). These 15N-excesses can be explained by mixtures between an extremely 15N-rich component preserved in the PX-18 organic clast from the metal-rich CH–CB chondrite Isheyevo (Briani et al. 2009) with macromolecular components located on the IMR fractionation line (Figure 3). This mixing is not very sensitive to the chemistry of the endmembers and the data are satisfactorily fitted using either isochemical mixing or mixing with a strong chemical contrast. Stepped combustion experiments of Isheyevo (Ivanova et al. 2008), as well as the 15N-rich CH chondrites Allan Hills (ALH) 85085 (Grady & Pillinger 1990) and Acfer 182 (Grady & Pillinger 1993), show that N with a 14N/15N ratio <110 is released above 800 °C, which suggests that this 15N-rich component is associated with thermally resistant organic moieties contrary to the labile 15N excesses attributed to ion–molecule reactions (Section 3.2). In this respect, 15N excesses were found to better survive atmospheric entry heating than D excesses in IDPs (Keller et al. 2004). This may additionally explain the selective survival of 15N-rich hotspots in the Wild2 samples compared to D-rich hotspots (McKeegan et al. 2006) because these samples have undergone a strong thermal stress (up to 2100 K during 0.1 ms; Roskosz et al. 2008) during hypervelocity impact capture in SiO2 aerogel at ∼6.1 km s−1.

Figure 3.

Figure 3. N isotopic fractionation between organic matter and the protosolar gas as a function of H isotopic fractionation among samples of group 2 (gray dots and upper part of the shaded area). Samples of group 1 are shown for reference (black dots and lower part of the shaded area). IMR-fl stands for ion–molecule reaction fractionation line. (a) Mixing between the 15N-rich endmember Isheyevo and a reservoir produced by ion–molecule reactions at 25 K. Plain curve: isochemical mixing; dashed curve: Isheyevo HCN + 25 K CHON; dotted curve: Isheyevo CHON + 25 K HCN. None of the group 2 data can be explained by such a mixing. (b) Same as panel (a) but mixing curves now depict mixing between the Isheyevo endmember and components akin to CV chondrites IOM (Efremovka), CR chondrites IOM (GRA 95229), and the most D-rich IDP hotspot (GS IDP L2054 E1-A), respectively. Data from group 2 are well explained by such mixtures. Data shown with 1σ uncertainties.

Standard image High-resolution image

The origin of the Isheyevo component can possibly be explained by pure fractionation of N isotopes starting from a precursor having an H isotopic composition lower than that of other chondritic IOMs and close to terrestrial values. Self-shielding of N2 in the early solar nebula would be a possible mechanism in this case, despite the fact that preliminary calculations fail to reach the observed enhancement in 15N (Lyons et al. 2009). Addition of cosmogenic 15N produced by flares from the young Sun or by GCRs also seems possible. However the efficiencies of spallation or fusion-evaporation reactions are probably not sufficient to result in 15N enhancements reaching a factor of 3 at the parent-body scale in CH–CB meteorites (notably in Bencubbin and Isheyevo where the bulk 14N/15N ratio is 138 (Prombo & Clayton 1985) and 128 (Ivanova et al. 2008), respectively), although such a scenario remains to be investigated in detail. Alternatively, we note that the calculated N and H isotopic fractionation factors relative to the protosolar gas are identical in this component (∼9). Although this may be fortuitous, it suggests that N-only fractionating processes such as the self-shielding of N2 or admixture of a nucleosynthetic component may not be the best explanations and that a specific chemical reaction could be involved. Extreme N isotopic fractionation by ion–molecule reactions such as that predicted in individual layers of ammonia ice coating interstellar dust grains (fractionation ≥ 10; Rodgers & Charnley 2008a) is unlikely in the absence of associated large D fractionation. A minor contribution of ion–molecule reactions suggests in turn elevated reaction temperatures. A minimal temperature of ∼110 K can be estimated from reaction (4). Taking formation of CH2D+ and C2HD+ into consideration would result in even higher temperatures because of larger exothermicities (∼390 K and ∼550 K, respectively, e.g., Roueff et al. 2007). Such a high temperature rules out fractionation in the dense presolar core (T < 10 K) or in the outer envelope of the class 0 protosun (T < 100 K) but could be achieved either in a protosolar hot corino (T ∼ 100–300 K) or in a warm layer of the solar protoplanetary disk.

The abundance of 15N-rich hotspots in CR chondrites (Busemann et al. 2006) suggests that the abundance of the Isheyevo 15N-rich component in meteorites increases following the sequence CV–CM–CI ≤ Tagish Lake < CR (+ Bells) < CH–CB (Isheyevo), which is roughly opposite to the age of the corresponding meteorite components, since chondrules in CB chondrites are ∼1–2 Myr younger than chondrules in CR chondrites, themselves ∼1–2 Myr younger than chondrules in CV chondrites (Amelin et al. 2002; Bizarro et al. 2004; Connelly et al. 2008; Amelin & Krot 2007; Krot et al. 2005, 2009; Nagashima et al. 2008). The simplest interpretation of this trend is that the Isheyevo 15N-rich reservoir is more and more efficiently produced as the solar protoplanetary disk ages and clears. This tendency is opposite to that expected from N2 self-shielding, which produces a 15N excess in the first 103–104 years of the solar system (Lyons et al. 2009). Such a scenario agrees with the observation of a low abundance of fine-grained matrix, distributed in discrete clasts in the least shocked CB chondrites, which is often attributed to accretion in a region depleted in fine-grained dust in an evolved protoplanetary disk (e.g., Krot et al. 2009). As a result, the hypothesis preferred here is that the Isheyevo component is due to a late chemical or photochemical reaction taking place at a relatively high temperature, possibly in a warm optically thin region during dissipation of the disk.

3.4. Interstellar Organic Matter in Ordinary Chondrites

The isotopic composition of IOM from UOCs (group 3) is best fitted by a mixture of (1) an organic component lying on the IMR fractionation line and akin to the IOM of the CM chondrite Murchison with (2) a 14N-rich, D-rich organic component, isotopically similar to the gas phase molecules observed in dense clouds, prestellar cores, and a class 0 protostar with a highly fractionated D/H ratio and an unfractionated 14N/15N ratio (Figure 4). Again, the mixing is not very sensitive to the chemistry of the endmembers. The interstellar component represents at most ∼1 at% of the IOMs in UOCs, if the interstellar D/H ratio is assumed to be 0.1. If larger ratios are assumed, due for instance to the presence of multiply deuterated molecules, the mixing remains valid but the proportion of interstellar molecules decreases accordingly. The presence of such an interstellar component in UOCs is in good agreement with the large D-excesses observed in UOC hydrated minerals and attributed to silicate hydration by interstellar water (Deloule & Robert 1995; Deloule et al. 1998), whereas the lower D/H ratio of hydrated silicates in carbonaceous chondrites can be attributed to a lower degree of isotope exchange in the solar protoplanetary disk owing to the strength of the O−H bond (Remusat et al. 2006).

Figure 4.

Figure 4. N isotopic fractionation between organic matter and the protosolar gas as a function of H isotopic fractionation among samples of group 3 (white dots, bulk IOM from unequilibrated ordinary chondrites). Samples of group 1 are shown for reference (black dots). IMR-fl stands for ion–molecule reaction fractionation line. Mixing curves depict mixing between a component akin to CM chondrites IOM (Murchison) and observed interstellar molecules. Plain curve: isochemical mixing; dashed curve: Murchison CHON + ISM HCN; dotted curve: Murchison HCN + ISM CHON. IOMs from unequilibrated ordinary chondrites are well explained by a component similar to that in CM chondrites plus a small amount (< 1 atom%) of interstellar molecules. Data shown with 1σ uncertainties.

Standard image High-resolution image

3.5. On the Chronological Significance of 15N Excesses

The 15N excess–chondrule age anti-correlation deduced from the abundance of 15N-rich hotspots in CR chondrites is also visible using bulk isotopic composition (Figure 5). A detailed interpretation of this correlation remains difficult because (1) at least two sources of 15N are present (ion–molecule reactions and the late CH–CB component), (2) the bulk nitrogen isotopic composition can be modified by secondary processes such as thermal metamorphism, spallation by solar energetic particles or GCR after ejection from the parent-body, or impact-induced introduction of foreign N. In this respect, highly heterogeneous meteoritic breccias such as Bells or the ungrouped Kaidun chondrite or meteorites extensively depleted in nitrogen such as the new CB chondrite Frontier Hills (Lauretta et al. 2009) must be excluded. (3) It relies on the concordancy between various chronological systems, which is heavily debated. However, there is increasing agreement that the 26Al and Pb isotopes chronometers used here are concordant (e.g., Connelly et al. 2008; Krot et al. 2009). Note that the observed correlation holds when only average Pb isotopic ages are used (Figure 5).

Figure 5.

Figure 5. Average bulk N isotopic composition of carbonaceous chondrites as a function of average chondrule age in the same chondrite groups. Central dots give the average bulk δ15N vs. average Pb isotope age of chondrules. Total ranges are shown as squares and correspond in Y to the total range of bulk δ15N and in X to the range of 26Al chondrule ages, except for CH–CB chondrites, for which only the Pb isotope age of CB chondrites is known (Krot et al. 2005). In this case, the X range corresponds to the uncertainty on the Pb age. A regression line through central dots and the initial PSN δ15N value is shown as a dotted line. The initial PSN absolute age is assumed to be that of meteoritic Ca–Al-rich inclusions.

Standard image High-resolution image

Despite these potential problems, the general correlation can be extended to the initial isotopic composition of the PSN (Figure 5). In addition, the bulk δ15N values of many differentiated meteorites such as magmatic iron meteorites (e.g., Prombo & Clayton 1993) and monomict ureilites (e.g., Grady et al. 1985; Yamamoto et al. 1998; Rai et al. 2003) broadly agree with their recently determined Hf–W isotopic ages, which indicate accretion within the first 2 Myr of the solar system, and for most of them segregation of a metallic core within this period (Kleine et al. 2005; Schersten et al. 2006; Qin et al. 2008; Lee et al. 2009). Despite N not necessarily being hosted by organic matter in these objects due to a redistribution during differentiation, it seems reasonable to assume that a macromolecular organic material somewhat similar to the IOM of known chondrites was initially the main carrier of N in the chondritic progenitors of these objects. Such an extension to differentiated bodies supports a solar protoplanetary disk origin of both types of 15N-excesses and a general increase with time of the 15N content of organic molecules in the disk. In turn, it suggests that other 15N-rich objects such as 15N-rich IDPs (Floss et al. 2006; Busemann et al. 2009) considered here as a whole group rather than individual objects, or comets could have accreted late in the solar protoplanetary disk, contemporaneously to CR, CH, or CB chondrites, more than 2–3 Myr after the start of solar system formation.

A late accretion of comets agrees with the presence of abundant crystalline silicates in comets (e.g., Crovisier et al. 1997) and with the discovery among the Wild2 samples of high-temperature components usually thought to be formed late in the solar system protoplanetary disk: (1) chondrules (Nakamura et al. 2008), whose formation peaks between 1 and 3 Myr after the first solar system solids (Villeneuve et al. 2009) and (2) a refractory inclusion devoid of 26Al, i.e., melted ⩾2–3 Myr after the first solar system solids (Matzel et al. 2010). Still, the homogeneous 14N/15N ratio (∼140; Manfroid et al. 2009) in many comets accreted either beyond Neptune (Jupiter family comets) or closer to the giant planet regions (Oort cloud comets) may be interpreted as a clue for an inherited interstellar component. However, the abundance of inner solar system dust in comets (e.g., Crovisier et al. 1997; McKeegan et al. 2006; Zolensky et al. 2006) and in IDPs of likely cometary origin (Aléon et al. 2009; Duprat et al. 2010) rather argues for a late protoplanetary disk origin of 15N excesses, unless a total dynamical decoupling exists between silicates and organics (either dust or gas).


This new analysis of coupled H and N isotopic data in organic matter from various classes of meteorites, from IDPs of both asteroidal and cometary origins, and from a short-period and a long-period comet (comets Wild2 and Hale-Bopp, respectively), indicates that the isotopic composition of organic matter in primitive solar system materials can be ascribed to the mixing of at least three components, two of which carry large 15N-excesses and were likely formed in the solar protoplanetary disk and the other being depleted in 15N and inherited from the presolar cloud core. This interpretation implies that the proportion of interstellar organic matter preserved in primitive solar system bodies is extremely small, at most 1% in ordinary chondrites, a group of meteorites paradoxically known to come from a very small number of main belt asteroids (Meibom & Clark 1999). Finally, all samples of likely or verified cometary origin have an H–N isotopic systematics related to that of carbonaceous chondrites rather than to UOCs, similarly to what has been observed for the O isotopic composition of their silicate component (Nakamura et al. 2008; Aléon et al. 2009). This confirms that comets and carbonaceous chondrites belong to the same chemical family, a continuum of primitive solar system objects that have recorded the physico-chemical processes active in the disk around the young Sun.

Numerous discussions with and critical readings of an early draft by Alice Aléon-Toppani, Evelyne Roueff, and Jean Duprat were appreciated. François Robert is warmly thanked for his input in the 2004 paper, on which this work is based. Giacomo Briani is thanked for providing unpublished data on 40 μm × 40 μm areas in the PX-18 clast. Constructive comments from the reviewer were appreciated. This work was supported by a PNP-INSU-CNRS grant.


Mass balance mixing calculations between two components 1 and 2 result in mixing lines, which are hyperbolas in a Ratio 1 versus Ratio 2 space. The mixing equation in 15N/14N versus D/H space is

which is the equation of a hyperbola with D/H0 and 15N/14N0, the asymptotes parallel to the N isotopes and H isotopes coordinates, respectively, and [D/H0 × 15N/14N0 − γ] is the curvature factor

In the specific isochemical mixing case where q = 1, the mixing equation is

which is the equation of a straight line.

Details of the calculations of the mixing equations can be found in Albarède (1995).

Please wait… references are loading.