TOWARD A UNIQUE NITROGEN ISOTOPIC RATIO IN COMETARY ICES

The determination of nitrogen isotopic ratios in solar system objects is of primary importance to achieve a better understanding of the origins of these objects. Measurements of the ¹⁴N/¹⁵N isotopic ratio in different solar system objects and molecules have revealed great diversity. This ratio ranges from 441 ± 5 for the present-day Sun (Marty et al. 2011), considered to be representative of the protosolar nebula, to 50 in some organic materials of chondrite and interplanetary dust particles (Messenger 2000; Bonal et al. 2010). Any object of the solar system (except Jupiter) is actually enriched in ¹⁵N compared to the protosolar nebula.

Different hypotheses have been proposed to explain this enrichment. Chemical models (Rodgers & Charnley 2008) have shown that interstellar chemistry can produce ¹⁵N enrichment for both molecules carrying the nitrile (–CN) functional group and those carrying the amine (–NH) functional group. A recent work suggests that these differences would simply reflect the different interstellar N reservoirs from which N-bearing molecules originate (Hily-Blant et al. 2013). Based on observations of H¹³CN and HC¹⁵N in two prestellar cores, the authors suggested that the molecules carrying the nitrile functional group would be more enriched in ¹⁵N than the molecules carrying the amine functional group. Nevertheless, such a theory still requires further observations and modeling before it can be confirmed.

Comets are interesting targets with which to test this theory because they contain both HCN and NH₃ molecules (leading to CN and NH₂ radicals after photodissociation by solar radiation). So far, the ¹⁴N/¹⁵N ratio has only been measured in comets from HCN and CN (Jehin et al. 2009). It has been measured in about 20 bright comets through optical observations of the CN radical (Arpigny et al. 2003; Manfroid et al. 2009) and for a few comets from millimeter observations of HCN (Bockelée-Morvan et al. 2005, 2008). For both species, the measurements give the same non-terrestrial isotopic composition (¹⁴N/¹⁵N ≃ 150 in comets versus 272 in Earth's atmosphere) and do not depend on the origin of the comets (Jehin et al. 2009).

In this work our objective is to measure the ¹⁴N/¹⁵N isotopic ratio in comets for a radical bearing the amine functional group, NH₂. This radical is assumed to be produced by photodissociation of the NH₃ molecule since: (1) 95% of photodissociated NH₃ molecules in comets produce NH₂ radicals, (2) NH₃ is, by far, the most abundant molecule observed in comets able to produce NH₂, and (3) simultaneous measurements of NH₃ and NH₂ abundances in comets have provided similar values (Kawakita & Mumma 2011). We first conducted laboratory experiments to measure the wavelengths of the most intense ¹⁵NH₂ lines, which were poorly known, with sufficient accuracy. In addition, we processed a large sample of high-resolution cometary spectra obtained during the last decade and coadded them in the corresponding wavelength range. We then obtained an average spectrum with a very high signal-to-noise ratio to search for the faint ¹⁵NH₂ emission lines. We succeeded in detecting seven of them. From this positive detection of ¹⁵NH₂ emission lines, for the first time, it was possible to derive an estimate of the ¹⁴N/¹⁵N isotopic ratio in comets for a species bearing the amine functional group.

Cometary spectra show emission of ¹⁴NH₂ bands around 5700 Å and 6000 Å ((0, 10, 0)–(0, 0, 0) and (0, 9, 0)–(0, 0, 0) bands in linear notation, respectively). Due to the lack of precise line lists with which to identify the cometary ¹⁵NH₂ lines, in the laboratory, we investigated the emission spectrum of ¹⁵NH₂ in the 5550–6250 Å (16,000–18,000 cm⁻¹) spectral range where the most intense lines are expected.

The emission spectrum has been recorded using the Fourier transform (FT) interferometer of the AILES beamline of the synchrotron SOLEIL equipped with a quartz-visible beamsplitter and an avalanche photodiode detector. The experiment was implanted on the AILES beamline as described in Yu et al. (2010) except that in the present case one plane mirror and a single 30 cm focal length lense collected the plasma emission and focused it on the entrance aperture of the FT interferometer. The ¹⁵NH₂ radical was produced in a 13.5 MHz radiofrequency (RF) discharge of about 1 mbar of pure ¹⁵NH₃ continuously flowing through a 100 cm length and 2 cm diameter pyrex cell placed collinear to the RF coil axis. The spectral resolution was set to 0.05 cm⁻¹ (0.017 Å) and the final spectrum is a coaddition of 288 interferograms corresponding to a total of about 2 hr of acquisition time. Together with the weak emission lines of ¹⁵NH₂, we detected numerous intense transitions of ¹⁵N₂ and H₂ in the spectrum. The presence of these spurious lines complicated the spectroscopic analysis of the ¹⁵NH₂ optical transitions but allowed accurate calibration of their frequencies (Bailly & Vervloet 2007; Salumbides et al. 2008; Bailly et al. 2010).

The emission lines measured in this work (both from laboratory and observational spectra) correspond to rovibronic transitions between the two Renner–Teller components of NH₂ which correlate to the ²Π_u electronic ground state in the linear configuration. As for ¹⁴NH₂, the $\tilde{X}^2B_1$ electronic ground state of ¹⁵NH₂ is bent, whereas the $\tilde{A}^2A_1$ excited state is quasilinear (Herzberg 1966). Numerous studies have been devoted to experimental and theoretical understanding of the complex Renner–Teller effect for ¹⁴NH₂, and we refer to Dressler & Ramsay (1959) for the detailed description of the spectroscopic notations. In the following, we adopt the usual description of rotational quantum numbers $N_{K_aK_c}$ for asymmetric top molecules. Due to its unpaired electron each $N_{K_aK_c}$ level is split in two sublevels through spin-rotation interaction identified as F₁ (N = J − 1/2) and F₂ (N = J + 1/2). The vibrational quantum numbers are described as: (ν₁, ν₂, ν₃).

The cometary data are high-resolution spectra obtained at the European Southern Observatory (ESO) using the 8.2 m Kueyen telescope (UT2) of the Very Large Telescope with the Ultraviolet and Visual Echelle Spectrograph (UVES) instrument (Dekker et al. 2000). This instrument is a cross-dispersed echelle spectrograph designed to operate with high efficiency from the atmospheric cut-off at 300 nm to the long-wavelength limit of the CCD detectors (about 1100 nm).

We searched for the ¹⁵NH₂ emission lines in a series of high-resolution spectra collected with UVES on 12 different comets between 2002 and 2011 (resolving power λ/Δλ ≃ 80,000). The reduction procedure is detailed in Manfroid et al. (2009) and Decock et al. (2013). The UVES arc lamp wavelength calibration was corrected using the [O i] night sky emission lines. The BASS2000 solar spectrum was used to remove the absorption features of the cometary dust scattered light. Table 1 presents the detail of the 39 spectra used for our study. Because of the expected faintness of ¹⁵NH₂ emission lines we combined the normalized spectra using weights proportional to the square of the signal-to-noise ratio (column "Ratio" in Table 1). We obtained a single combined spectrum with the best signal-to-noise ratio that could be obtained from all our observational data.

Table 1. Spectra of Comets Used for This Study

Comet	UT Date	MJD	Exp. Time	r	$\dot{r}$	Δ	$\dot{\Delta }$	Ratio
			(s)	(au)	(km s−1)	(au)	(km s−1)
C/2000 WM1 (LINEAR)	2002 Mar 7	52340.3623	1550	1.084	28.258	1.237	0.258	359
C/2000 WM1 (LINEAR)	2002 Mar 7	52340.3808	1550	1.084	28.258	1.237	0.278	386
C/2000 WM1 (LINEAR)	2002 Mar 8	52341.3682	1550	1.100	28.266	1.237	0.132	493
C/2000 WM1 (LINEAR)	2002 Mar 8	52341.3867	1550	1.100	28.266	1.237	0.155	301
C/2002 V1 (NEAT)	2003 Jan 8	52647.0373	2100	1.222	−36.514	0.833	7.871	700
C/2002 V1 (NEAT)	2003 Jan 8	52647.0622	2100	1.221	−36.521	0.833	7.916	723
C/2002 V1 (NEAT)	2003 Jan 10	52649.0312	2100	1.180	−37.105	0.842	8.272	741
C/2002 V1 (NEAT)	2003 Jan 10	52649.0562	1987	1.179	−37.113	0.842	8.313	555
C/2002 X5 (Kudo-Fujikawa)	2003 Feb 19	52689.0132	2000	0.697	43.032	0.865	−5.052	885
C/2002 V1 (NEAT)	2003 Mar 21	52719.9854	600	1.012	39.761	1.626	42.003	467
C/2002 Y1 (Juels-Holvorcem)	2003 May 29	52788.3943	1800	1.142	24.091	1.556	−7.220	682
C/2002 Y1 (Juels-Holvorcem)	2003 May 29	52788.4157	1800	1.142	24.093	1.556	−7.195	600
C/2002 Y1 (Juels-Holvorcem)	2003 May 30	52789.3935	1800	1.156	24.184	1.552	−7.198	508
C/2002 Y1 (Juels-Holvorcem)	2003 May 30	52789.4149	1800	1.156	24.186	1.552	−7.172	517
C/2001 Q4 (NEAT)	2003 Sep 1	52883.2929	4500	3.730	−18.803	3.448	−25.411	724
88P/Howell	2004 May 2	53127.3723	3600	1.385	2.988	1.648	−3.250	382
88P/Howell	2004 May 3	53128.3630	3600	1.387	3.131	1.646	−3.198	350
88P/Howell	2004 May 4	53129.3715	3600	1.388	3.276	1.644	−3.126	163
C/2001 Q4 (NEAT)	2004 May 5	53130.9577	119	0.978	−5.426	0.322	−4.120	269
C/2001 Q4 (NEAT)	2004 May 6	53131.0658	2189	0.977	−5.363	0.322	−3.287	928
C/2002 T7 (LINEAR)	2004 May 6	53131.4214	1080	0.680	15.835	0.607	−65.603	1115
C/2001 Q4 (NEAT)	2004 May 7	53132.0650	2148	0.974	−4.841	0.321	2.228	863
C/2002 T7 (LINEAR)	2004 May 26	53151.9760	2678	0.940	25.576	0.414	54.973	1922
C/2002 T7 (LINEAR)	2004 May 27	53152.0357	1800	0.941	25.585	0.416	55.188	1467
C/2003 K4 (LINEAR)	2004 Nov 20	53329.3438	1499	1.202	14.810	1.510	−28.215	632
73P-C/SW 3	2006 May 27	53882.3666	4799	0.952	−4.168	0.151	12.320	128
8P/Tuttle	2008 Jan 16	54481.0208	3599	1.041	−4.292	0.358	21.631	185
8P/Tuttle	2008 Jan 28	54493.0178	3899	1.027	0.404	0.522	24.711	959
8P/Tuttle	2008 Feb 4	54500.0169	3899	1.034	3.159	0.621	24.155	1033
103P/Hartley 2	2010 Nov 5	55505.3044	2899	1.065	2.533	0.159	7.081	701
103P/Hartley 2	2010 Nov 5	55505.3470	3199	1.065	2.546	0.159	7.194	500
103P/Hartley 2	2010 Nov 10	55510.3036	2899	1.074	4.054	0.181	7.961	716
103P/Hartley 2	2010 Nov 10	55510.3466	3199	1.074	4.067	0.182	8.072	807
103P/Hartley 2	2010 Nov 11	55511.2453	4499	1.076	4.333	0.186	7.954	522
103P/Hartley 2	2010 Nov 11	55511.3048	3599	1.077	4.351	0.186	8.088	506
103P/Hartley 2	2010 Nov 11	55511.3447	2399	1.077	4.363	0.186	8.191	579
C/2009 P1 (Garradd)	2011 Sep 10	55814.0190	4799	2.092	−14.825	1.474	14.622	272
C/2009 P1 (Garradd)	2011 Sep 11	55815.0020	4799	2.084	−14.769	1.482	15.126	330
C/2009 P1 (Garradd)	2011 Sep 12	55816.0102	4799	2.075	−14.710	1.491	15.675	430

Notes. UT Date is for mid-exposure, r is the heliocentric distance, $\dot{r}$ the heliocentric velocity, Δ the geocentric distance, $\dot{\Delta }$ the geocentric velocity, and the ratio is the ratio between the brightest emission line in the range 3899–5695 Å range and the noise.

Download table as:  ASCIITypeset image

We used the combined spectrum to search for ¹⁵NH₂ emission lines. To perform this search we established a list of all the ¹⁴NH₂ emission lines having a significant intensity and searched for their equivalent ¹⁵NH₂ lines for both the (0, 10, 0)–(0, 0, 0) and (0, 9, 0)–(0, 0, 0) bands. Unfortunately the spectral region corresponding to the (0, 9, 0)–(0, 0, 0) band contains numerous bright ¹⁴NH₂ and C₂ emission lines, preventing the detection of faint ¹⁵NH₂ emission lines. In the spectral region corresponding to the (0, 10, 0)–(0, 0, 0) band it was possible to detect seven different faint emission lines corresponding exactly to the ¹⁵NH₂ wavelengths resulting from the rotational analysis of the laboratory spectrum.

Table 2 presents the assignments and the wavelengths of the identified ¹⁵NH₂ lines with their ¹⁴NH₂ counterparts. Figure 1 shows the details of each emission line attributed to ¹⁵NH₂ (two of them, at 5762.995 and 5763.076 Å being unresolved). To identify the lines due to other species we used the cometary line atlas based on spectra obtained on comet 122P/de Vico (Cochran & Cochran 2002).

Zoom In Zoom Out Reset image size

Table 2. ¹⁵NH₂ Emission Lines Identified in the Average Cometary Spectrum

Line Identification	Air Wavelength	14NH2 Counterpart	Ratio
	(Å)	Air Wavelength
		(Å)
110 − 000 (F1 − F1)	5703.582	5693.594	91$^{+50}_{-30}$
111 − 101 (F2 − F2)	5710.959	5700.747	101$^{+60}_{-40}$
111 − 101 (F1 − F1)	5711.275	5700.996	137$^{+50}_{-30}$
110 − 202 (F1 − F1)	5723.810	5713.790	118$^{+60}_{-40}$
110 − 220 (F1 − F1)	5741.556	5731.683	180 ± 80
312 − 422 (F1 − F1)	5762.995	5752.752	135$^{+60}_{-40}$
312 − 422 (F2 − F2)	5763.076	5752.800	135$^{+60}_{-40}$

Notes. The ¹⁴NH₂ wavelengths are computed from the wavenumbers given in Ross et al. (1988). The ratio is the intensity ratio of ¹⁴NH₂ emission lines with the associated ¹⁵NH₂ lines (note identical ratios for the last two lines which appear blended).

Download table as:  ASCIITypeset image

From the detected ¹⁵NH₂ emission lines it is possible to derive a ¹⁴N/¹⁵N ratio for NH₃, the parent molecule of NH₂. In fact, a rigorous calculation of this ratio requires an important work of calculations and significant complementary laboratory experiments. Nevertheless it is possible to get a good estimate of the ¹⁴N/¹⁵N ratio in NH₃ by adopting reasonable assumptions. These assumptions, discussed below, are: (1) a similar photodissociation efficiency for ¹⁴NH₃ and ¹⁵NH₃ to produce, respectively, ¹⁴NH₂ and ¹⁵NH₂ radicals and (2) similar transition probabilities for both ¹⁴NH₂ and ¹⁵NH₂.

For the photodissociation efficiency with respect to the solar radiation it would be necessary to conduct complementary laboratory experiments for measuring the absorption cross section for both ¹⁴NH₃ and ¹⁵NH₃. Some results of laboratory experiments relative to this problem have, nevertheless, already been published (Suto & Lee 1983; Liang et al. 2007). These papers show that the oscillator strength f, defined by 1.13 × 10⁻⁶∫σdν (where σ is the absorption cross section in megabarns (=10⁻¹⁸ cm²) and ν the wavenumber (cm⁻¹)) is only 7.3% greater for ¹⁵NH₃ compared to ¹⁴NH₃ (Liang et al. 2007) in the 165–220 nm range (corresponding to the main region of photodissociation for these molecules). For shorter wavelengths, especially the Lyα region, only ¹⁴NH₃ absorption cross sections are available (Suto & Lee 1983). At the Lyα wavelength the ¹⁴NH₃ cross section is smaller than in the 165–220 nm range (about 10 × 10⁻¹⁸ cm² versus about 20 × 10⁻¹⁸ cm²). Even if the Lyα emission line would correspond to very different absorption cross sections for ¹⁵NH₃ and ¹⁴NH₃ it could not lead to a large overall difference in the photodissociation rate for these two molecules because the solar Lyα flux represents only a few percent of the total solar flux in the 165–220 nm range.⁴ Complementary laboratory data would be welcome to evaluate accurately the difference of photodissociation efficiency for ¹⁴NH₃ and ¹⁵NH₃. It is nevertheless highly probable that no significant difference exists.

Equal transition probabilities are generally adopted for the two isotopic species when measuring isotopic ratios from line intensity ratios. In the case of the ¹⁴NH₂ and ¹⁵NH₂ radicals, this assumption could be invalidated by the Renner–Teller effect. This effect concerns the low-lying rovibronic levels (with K_a ≠ 0) of the electronic excited state with higher rovibronic levels of the electronic ground state. In point of fact, the almost constant isotopic shift between the wavelengths of the selected transitions of Table 2 indicates that no severe perturbations resulting from a strong resonance between two interacting levels occur for the excited rovibronic levels of the subband K_a = 1 (0, 10, 0)–(0, 0, 0) of Table 2.

The intensity ratio between ¹⁴NH₂ and ¹⁵NH₂ emission lines was measured for each detected ¹⁵NH₂ line using the ¹⁴NH₂ corresponding line. In some cases, because of a blend of the ¹⁵NH₂ line with another emission line (due to either ¹⁴NH₂ or C₂) we used an ad hoc fitting of these perturbing lines before measuring the ¹⁵NH₂ line intensity by fitting it with a synthetic line convolved with the instrument response function.

Adopting a Gaussian analytical expression for the line profiles, we have simultaneously adjusted the intensities of the N lines considered in a wavelength range bracketing the line(s) of interest (i.e., for ¹⁴NH₂ or ¹⁵NH₂). This task was performed thanks to a classical Levenberg–Marquardt algorithm (Press et al. 1992). The FWHM, thought to be uniform, has been determined previously in applying the same technique.

In order to maximize the signal-to-noise ratio this analysis has been applied to the combined cometary spectrum. Table 2 presents the ratios measured for each couple of ¹⁴NH₂ and associated ¹⁵NH₂ lines. The uncertainties cited give the deviations of the values for which acceptable fits are obtained using various procedures. From this table it can be seen that, within the error, all the intensity ratios are compatible with each other. In some cases a blend with other very weak and unidentified emission lines cannot be excluded. Such a blend could lead to an underestimation of the ¹⁴N/¹⁵N ratio in one or two lines. From Table 2 we can compute an average ¹⁴NH₂/¹⁵NH₂ ratio of 127. A simple computation of the standard deviation gives σ = 32 but the average of the uncertainties provides a range of values from ∼80 to ∼190. The latter range of values is probably more appropriate because of the difficulty in accurately subtracting the solar continuum for each region of interest. Figure 2 presents the modeling done for the ¹⁵NH₂ lines located at 5710.959 and 5711.275 Å and their ¹⁴NH₂ counterparts.

Zoom In Zoom Out Reset image size

This ratio of 127 obtained with the average spectrum is probably very close to the ¹⁴N/¹⁵N ratio in the NH₃ molecules, with the two reasonable assumptions mentioned above. It is close to the ¹⁴N/¹⁵N ratio measured both in HCN and in CN.

So far, the ¹⁴N/¹⁵N ratio measured in solar system objects presents variations that remain difficult to interpret. The analysis of Genesis solar wind samples (Marty et al. 2011) suggests a ¹⁴N/¹⁵N ratio of 441 ± 5, in agreement with the in situ measurements made in the ammonia of Jupiter's atmosphere (Fouchet et al. 2004) which probably comes from primordial N₂ (Owen et al. 2001). In contrast, with a value of 272 in the atmospheric N₂, Earth's ratio is enriched in ¹⁵N compared to Jupiter and is similar to the bulk of ratios derived from the analysis of comet 81P/wild 2 grains (Mckeegan et al. 2006). Nitrogen isotopic ratios have also been measured in Titan's atmosphere, which is dominated by N₂ molecules. The two existing measurements, found to be 167.7 ± 0.6 and 143 from the Cassini GCMS and INMS data (Niemann et al. 2010; Mandt et al. 2009), respectively, are found to be lower than the Earth's value. Because of the low abundance of primordial Ar observed by Cassini-Huygens, it is generally assumed that N₂ is of secondary origin in this atmosphere and was delivered in a less volatile form, probably NH₃. Different mechanisms have been proposed for the conversion of NH₃ to N₂: photolysis (Atreya et al. 1978), atmospheric shock heating (McKay et al. 1988), endogenic production (Glein et al. 2009), and impacts during the late heavy bombardment (Sekine et al. 2011). Isotopic fractionation may have occurred for nitrogen in Titan's atmosphere, nevertheless the atmospheric model published by Mandt et al. (2009) suggests that the current ¹⁴N/¹⁵N ratio observed in N₂ is close to the value acquired by the primordial ammonia of Titan. Sekine et al. (2011) also predict that ¹⁴N/¹⁵N values in NH₃ in comets and Enceladus' plume would be as high as that of Titan's N₂. A ¹⁴N/¹⁵N ratio in the ammonia of comets significantly lower than the "primordial" solar wind/Jupiter value, and similar to the one measured on Titan, assumed to be representative of its primordial value, favors the hypothesis that comets and Titan were assembled from building blocks sharing a common formation location in the nebula. This idea is supported by the measurement of the D/H ratio in the plumes emitted from Saturn's other moon, Enceladus, by the INMS instrument aboard the Cassini spacecraft, which is found to be similar to the values derived from Oort Cloud comets (Waite et al. 2009; Kavelaars et al. 2011).

The aforementioned measurements suggest that N₂ and NH₃ result from the separation of nitrogen into at least two distinct reservoirs, with distinct ¹⁵N enrichment, which never equilibrated. Ion–molecule chemistry in dense interstellar and/or protostellar material could create ¹⁵N enrichment in the ammonia ice that is produced in these environments (Charnley & Rodgers 2002). Alternatively, the ¹⁵N enrichment observed in cometary ices could result from isotopic fractionation processes that occurred in the late protosolar nebula (Aléon 2010). More observational constraints are needed to test these scenarios such as more accurate and independent measurements of the nitrogen ratio in NH₂ for comets of different origins. Our work demonstrates that such measurements are now possible in the near future.