Tentative detection of cyanoformamide NCCONH 2 in space

The peptide-like molecule cyanoformamide ( NCCONH 2 ) is the cyano ( CN) derivative of formamide ( NH 2 CHO) . It is known to play a role in the synthesis of nucleic acid precursors under prebiotic conditions. In this paper, we present a tentative detection of NCCONH 2 in the interstellar medium with the Atacama Large Millimeter/submillimeter Array ( ALMA) archive data. 10 unblended lines of NCCONH 2 were seen around 3 σ noise levels toward Sagittarius B2( N1E) , a position that is slightly offset from the continuum peak. The column density of NCCONH 2 was estimated to be 2.4 × 10 15 cm − 2 , and the fractional abundance of NCCONH 2 toward Sgr B2( N1E) was 6.9 × 10 − 10 . The abundance ratio between NCCONH 2 and NH 2 CHO is estimated to be ∼ 0.01. We also searched for other peptide-like molecules toward Sgr B2( N1E) . The abundances of NH 2 CHO, CH 3 NCO and CH 3 NHCHO toward Sgr B2( N1E) were about 1/10 of those toward Sgr B2( N1S) , while the abundance of CH 3 CONH 2 was only 1/20 of that toward Sgr B2( N1S) .

The organic species cyanoformamide (NCCONH 2 ) is the cyano (CN) derivative of formamide (NH 2 CHO), a known interstellar molecule with a role in the synthesis of nucleic acid precursors under prebiotic conditions (Winnewisser et al. 2005).Cyanoformamide could form from the reaction between CN and formamide (Winnewisser et al. 2005).As both CN and formamide are observed to be abundant in the ISM, cyanoformamide is expected to be detectable in the ISM (Winnewisser et al. 2005).At room temperature and normal terrestrial pressure, this molecule is rather unstable.The microwave and millimeter-wave spectrum of the gas-phase species was not studied until 2005 (Christiansen 2005;Winnewisser et al. 2005).Recently, Colzi et al. (2021) searched cyanoformamide toward hot core G31.41+0.31but got a negative result.Sgr B2, the giant molecular cloud located in the Galactic central region, is the most massive star-forming region in the Galaxy.It has long been regarded as the best hunting grounds for complex organic molecules (COMs) due to the extraordinary molecular richness (Belloche et al. 2013;Li et al. 2017;Li et al. 2020) (Bonfand et al. 2017). Recently, Li et al. (2021) found evidence for the possible presence of C 2 H 5 CONH 2 , the largest peptide-like molecule, toward Sgr B2(N1E), suggesting that peptide-like molecules are abundant in this region.
In this paper, we present a tentative detection of NCCONH 2 toward Sgr B2(N1E).
Section 2 introduces the observations and data reduction.Section 3 presents the observational results.Section 4 discusses the observing results and possible formation mechanisms of NCCONH 2 .The summary of the work is presented in Section 5.

OBSERVATIONS AND DATA REDUCTION
The data used for this study were acquired from the ALMA Science archive of the ReMoCA survey (Re-exploring Molecular Complexity with ALMA) (Belloche et al. 2019).The ReMoCA survey was conducted with ALMA during Cycle 4 between 2016 and 2017.Detailed description about the observations is presented in Belloche et al. (2019).This is a compete spectral line survey toward Sgr B2(N) covering from 84.1 to 114.4 GHz.Five spectral setups were used in total.The on-source time for each frequency windows range from 47 to 50 minutes.The spectral resolution is 0.488 MHz, which corresponds to a velocity resolution of 1.3-1.7 km s −1 across the observing band.The phase center of the observations is α J2000 = 17 h 47 m 19.87 s , δ J2000 = −28 • 22 ′ 16 ′′ , which locates half way between the two main hot molecular hot cores Sgr B2(N1) and N2.
Our data reduction procedure has been introduced in Li et al. (2021).The data was calibrated using the standard ALMA data calibration pipeline with the Common Astronomy Software Applications package (CASA).CASA version 4.7.0-1 was used for the first spectral setup, while CASA version 4.7.2 was used for the other four spectral setups.The CASA version 5.6.1-8 was used to image the calibrated data.The quasar J1924-2914 was used to calibrate the bandpass for most of the data, while the quasar J1517-2422 was used to calibrate the bandpass for on execution in the second spectral setup.Quasars J1924-2914 or J1733-1304 were used to derive the absolute flux density scale.The quasar J1744-3116 was used to calibrate the phase and amplitude.The TCLEAN deconvolution algorithm in CASA was used to produce the images.Self-calibration is known to introduce artificial features into imaging by including manually chosen clean components if the structure of the target is complicated (i.e., if it is not a simple point source).Sgr B2(N) has complicated substructures (Bonfand et al. 2017;Belloche et al. 2019), therefore, self-calibration will likely introduce artifacts into images, which may affect spectral line identification, especially for the weak lines in our case.We did not perform self-calibration, but still achieved a sufficiently high imaging rms of 0.4 ∼ 1 mJy beam −1 , which is not much higher than that in Belloche et al. (2019).Therefore, we believe that this strategy is suitable for our purpose of identification of weak lines.
Because of high number of spectral lines detected toward hot cores in Sgr B2, the deter-mination of baseline is challenging.As pointed out by Sánchez-Monge et al. (2018), a broad frequency coverage is necessary to ensure the presence of enough line-free frequency intervals to determine the continuum level.They think that a bandwidth of at least 1 GHz is needed.Sánchez-Monge et al. (2018) have simulated spectra that dominated by emission features, and eight methods were used to determine the continuum level of the spectra.We could see from  2018) that all the method over-estimated the rms noise levels by 2% to 113%.We could see from Figure 1(b) in Sánchez-Monge et al. ( 2018) that the continuum level of the spectra could be better determined with several groups of lowest values in the spectra after masking absorption lines.We first compared with spectra toward HII regions to mask absorption lines, then we chose several groups of lowest values to determine the continuum levels.This could properly determine the continuum levels for spectra dominated by emission lines.
The rms noise levels for the spectra window were determined using the median values of channel maps.We first investigated the rms noise levels in regions without either continuum emission or molecular lines.Then we investigated the rms noise level in regions with strong continuum emission, which are significantly larger than rms noise levels in regions without either molecular lines or continuum emission.The median values of rms noise levels were adopted in this paper.

Identification of NCCONH 2
Weeds in Gildas package was used for line identification and spectral modeling.The Jet Population Laboratory (JPL) (Pickett et al. 1998) and the Cologne Database for Molecular Spectroscopy (CDMS) (Müller et al. 2005;Endres et al. 2016) databases were used.The microwave and millimeter-wave spectrum of NCCONH 2 was reported by Christiansen (2005) and Winnewisser et al. (2005).The 3-mm emission of NCCONH 2 is modeled assuming LTE conditions with five parameters: column density, temperature, source size, velocity offset, and linewidth.Each spectral window of each observed setup was modeled separately to account for the varying angular resolution, but with a same set of parameters.
Approximately 120 NCCONH 2 lines were expected to be seen in the observed frequency ranges.Ten unblended transitions, and 21 partially blended transitions of NCCONH 2 were possibly seen toward Sgr B2(N1E) (Figure 1), which is about 1.5 ′′ to the east of the hot core Sgr B2(N1).However, since their signal-to-noise levels are around 3σ, it is not possible to claim secure detection of NCCONH 2 with these data.The large peptide-like molecule, propionamide has been tentatively found to be relatively abundant toward Sgr B2(N1E) (Li et al. 2021).
We also searched for NCCONH 2 toward other regions of Sgr B2(N1).However, serious lineblending prevents detection of this molecule toward other directions.The tentatively detected lines are presented in Figures 2 and 3, and Table 1.Though the remaining ∼ 90 lines that match the considered frequency range and noise level threshold are contaminated by those of other species, the observed results do not contradict with the expected intensities.In Figure 2, the black dashed lines show the 3σ noise levels adopted in Section 2.
Because of low abundance of NCCONH 2 and serious line blending in Sgr B2(N1), all of these unblended lines detected toward Sgr B2(N1E) suffer from line blending toward other directions.Thus we could not get the spatial distribution of NCCONH 2 in Sgr B2(N1).Unfortunately, all of these lines blend seriously with other molecules.The column densities of NCCONH 2 and related molecules are obtained by eye-fitting the spectra in Weeds.The physical size of the emission region is hard to determine, as the morphology of molecules in Sgr B2(N1) does not simply follow a 2D Gaussian profile (Busch et al. 2022).
The emitting size was assumed to be 2.3 ′′ , thus the beam filling factor is ∼1.With a linewidth measurement of 3.0 km s −1 , and an assumption of excitation temperature of 150 K (which is very close to what was obtained for molecules toward Sgr B2(N1S) by Belloche et al. (2019), the column density and centroid velocity are varied to fit the detected transitions.In this way, a column density of 2.4 × 10 15 cm −2 was obtained for cyanoformamide (see Table 2).By adopting the H 2 column density of 3.5×10 24 cm −2 derived with C 18 O (Li et al. 2021), the abundance relative to H 2 was estimated to be 6.9 × 10 −10 for NCCONH 2 .
We also modeled the emission of CH 3 NCO and CH 3 NHCHO toward Sgr B2(N1E) with the same size and excitation temperature (see Table 2).A column density of 2.4×10 16 cm −2 was obtained for CH 3 NCO.A column density of 2.1 × 10 16 cm −2 was obtained for CH 3 NHCHO.
Based on the estimated column density of NCCONH 2 toward Sgr B2(N1E), the abundance ratio between NCCONH 2 and formamide is found to be ∼0.01.This value is consistent with result in Colzi et al. (2021).They found that the abundance ratio of NCCONH 2 to formamide was lower than 0.05 in G31.41+0.31.

Comparison of Sgr B2(N1E) and Sgr B2(N1S)
Sgr B2(N1S) is a position that is about 1 ′′ to the south of Sgr B2(N1) (see Figure 1, also see All the peptide-like molecules detected toward Sgr B2(N1S) were also detected toward Sgr B2(N1E).Table 2 presents the column densities obtained for peptide-like molecules toward Sgr B2(N1E).We compared these results with those of Sgr B2(N1S) presented in Belloche et al. (2019).The abundances of NH 2 CHO, CH 3 NCO and CH 3 NHCHO toward Sgr B2(N1E) were about one tenth of those toward Sgr B2(N1S).The abundance ratio of CH 3 NCO to NH 2 CHO, was 0.086 toward Sgr B2(N1E), which is in agreement with that of Sgr B2(N1S) (Belloche et al. 2019).However, the abundances of CH 3 CONH 2 toward Sgr B2(N1E) was only one twentieth of those toward Sgr B2(N1S).A possible explanation for the low abundance of CH 3 CONH 2 and undetection of NH 2 CONH 2 is that the desorption energy required by CH 3 CONH 2 and NH 2 CONH 2 are higher in comparison with NH 2 CHO, CH 3 NCO and CH 3 NHCHO.According to laboratory studies in Ligterink et al. (2018), the desorption peak of NH 2 CONH 2 is ∼265 K, while the desorption peak of NH 2 CHO is ∼210 K, and the desorption peak of CH 3 CONH 2 is ∼219 K.In this case, both CH 3 CONH 2 and NH 2 CONH 2 may not efficiently desorb into gas phase, which is consistent with results present here.It is noted that the derived rotational temperature of NH 2 CONH 2 toward Sg B2(N1S) seems to be higher than other molecules (see Table 5 in Belloche et al. (2019)), while the rotational temperature of CH 3 CONH 2 is also higher than NH 2 CHO, CH 3 NCO and CH 3 NHCHO (Belloche et al. 2019).We did not search for C 2 H 5 CONH 2 and NCCONH 2 toward Sgr B2(N1S) because of serious line blending from other molecules.

Possible formation mechanism for interstellar cyanoformamide
The formation of NCCONH 2 in the ISM is yet unclear.Winnewisser et al. (2005) proposed that NCCONH 2 could form from the formamide and CN: We ran quantum chemical calculations to further study this reaction.The structures of all the species studied in this work (reactants, intermediates, transition states, and products) were first optimized in the framework of the density functional theory (DFT) employing the M06-2X (Zhao & Truhlar 2008) in conjunction with the 6-311+G(d,p) basis set (Ditchfield, Hehre, & Pople 1971), from which rotational constants, harmonic vibrational frequencies, and zero-point energies(ZPEs) were obtained.High-performance single point energies were also calculated at the M06-2X/aug-cc-pVTZ (Dunning 1989) level using the M06-2X structures.All quantum chemical calculations were run with the Gaussian 16 (Frisch et al. 2016).
NH 2 CHO molecule has several spots that could be attacked by the CN radical: the N, C, and O atoms.Therefore, besides Reaction (1), the following reactions might be also possible.
Thus, we ran quantum mechanical calculations to evaluate these two reactions.
A schematic of the full PES for the reaction between CN and NH 2 CHO can be seen in Figure 6.Reaction (1) may initially proceed via the formation of intermediate I1 (NH 2 C(O)HCN) via transition state TS1.Then the intermediate II is linked to the formation of the products P1 (NH 2 COCN + H) via transition state TS2, as this involves the elimination of an H-atom.
The reaction is exothermic.The barrier energy of this reaction is equal to the energy of transition state minus the energy of the two reactants, which is 9.7 kJ/mol (=1,167 K).We have calculated the rate coefficients of Reaction (1) by the modified Arrhenius equation.The rate coefficients are 2.4 × 10 −13 and 1.94 × 10 −12 cm 3 s −1 at T = 150 and 200 K, respectively.
This means that the exothermic formation reaction (1) is slow in the gas-phase.Reaction ( 2) is via transition state TS3 to form products P2 (NH 2 CN + HCO), and Reaction (3) firstly form the intermediate I2 (NH 2 C(H)OCN), then via transition state TS4 and TS5 to form two products P3 (cis-NCOCHNH + H) and P4 (trans-NCOCHNH + H), respectively.As Figure 6 shows, both reactions (2) and ( 3) are endothermic with high barriers (>55 kJ/mol).Their rates would be very low under the ISM conditions.
Besides, we propose another possible reaction route of NCCONH 2 , where it might be formed upon recombination of two free radicals on the grain surface at low temperature as the following: The radical NH 2 CO can be formed by CN with water molecules of the ice mantle as reported by Rimola et al. (2018).

CONCLUSIONS
In this paper, we present the tentative detection of NCCONH 2 in ISM for the first time.The main results are summarized as follows: 1.Ten unblended transitions, and twenty one partially-blended transition of NCCONH 2 were detected around 3σ noise levels toward Sgr B2(N1E) with the ALMA archive data at 3-mm wavelength.
It is highly desirable to conduct confirmation observations on the existence of NCCONH 2 in space.

Fig. 1 .
Fig. 1.Integrated intensity map of NH2CHO and CH 13 3 CN in Sgr B2(N1).The integrated intensity map of NH2CHO at 103.525 GHz are shown in color scale.The integrated intensity map of CH 13 3 CN at 91.94 GHz is shown in contours, which represent 30% to 90% percent of peak integrated intensity.The white crosses indicate the position of Sgr B2(N1E), Sgr B2(N1S), and the center of Sgr B2(N1).The white ellipse shows the size of the respective synthetic beam.

Fig. 2 .
Fig. 2. Unblended transitions of NCCONH2 toward Sgr B2(N1E).Black lines show continuum-subtracted spectrum observed with the ALMA telescope.The black dashed lines show the 3σ noise levels.The median values of the rms noise levels in the channel maps were used here.The red lines show the modeling results.The cyan lines show the modeling results of other molecules.

Fig. 3 .
Fig. 3. Partially blended transitions of NCCONH2 toward Sgr B2(N1E).Black lines show continuum-subtracted spectrum observed with the ALMA telescope.The black dashed lines show the 3σ noise levels derived byBelloche et al. (2019).The red lines show the modeling results.The cyan lines show the modeling results of other molecules.

Fig. 3 .
Fig. 3. Continued.Note that the NCCONH2 transition at 108650 MHz is affected by 13 CN absorption.

Fig. 4 .
Fig. 4. Transitions of CH3NCO toward Sgr B2(N1E).Black lines show spectrum observed with the ALMA telescope, while the red lines show the modeling results.The black dashed lines show the 3σ noise levels.The cyan lines show the modeling results of other molecules.

Fig. 5 .
Fig. 5. Transitions of CH3NHCHO toward Sgr B2(N1E).The black solid lines show spectrum observed with the ALMA telescope, while the red lines show the modeling results.The black dashed lines show the 3σ noise levels.The cyan lines show the modeling results of other molecules.
. Most O-bearing COMs were first detected toward Sgr B2 (McGuire 2018; McGuire 2022), such as the branched molecule i-C 3 H 7 CN (Belloche et al. 2014) and the chiral molecule CH 3 CHCH 2 O (McGuire et al. 2016).There are two main sites of star formation regions in Sgr B2, namely Sgr B2(N) and Sgr B2(M), both of which host several dense, compact, hot cores that are rich in COMs