Abstract
The photoelectron spectrum of the X1Σ+ → X+ 2Σ+ ionizing transition of hydrogen isocyanide (HNC) is measured for the first time at a fixed photon energy (13 eV). The assignment of the spectrum is supported by wave-packet calculations simulating the photoionization transition spectrum and using ab initio calculations of the potential energy surfaces for the three lowest electronic states of the cation. The photoelectron spectrum allows the retrieval of the fundamental of the CN stretching mode of the cationic ground state and the adiabatic ionization energy of hydrogen isocyanide: IE(HNC) = 12.011 ± 0.010 eV, which is far below that of HCN (IE(HCN) = 13.607 eV). In light of this latter result, the thermodynamics of the HCN+/HNC+ isomers is discussed and a short summary of the values available in the literature is given.
1. Introduction
The hydrogen isocyanide molecule (HNC) is omnipresent in the interstellar medium1–8 with a comparable abundance to that of the hydrogen cyanide (HCN), its more stable isomer.9 These two molecules and their corresponding cations are involved in numerous reactions. For instance, Loison et al. 9 have listed more than 50 reactions of these species in a chemical network describing dark molecular clouds. It is thus crucial to have a good knowledge of the thermochemical data to establish the most important sources and sinks of these species. The HCN+/HNC+ ratio is a puzzling matter in interstellar media. HCN+/HNC+ isomers are produced through various ion-molecule reactions: H+ + HCN, H+ + HNC, CN+ + H2, In dense molecular clouds, knowledge of the specific isomer HCN+/HNC+ is not essential, because both are expected to react rapidly with H2 leading to HCNH+ + H.9,10 For diffuse molecular clouds, however, hydrogen is primarily present as H atoms, and the main destruction pathway for HCN+/HNC+ is electronic dissociative recombination, for which the rate is 30 times larger for HCN+ than HNC+. 11 For the reaction with H+, the ionization energy (IE) is 13.598 eV for H12, 13.607 eV for HCN,13 and around 12 eV for HNC (see Section 3.2). Then, as the rate constant for the H+ + HCN reaction is very high at room temperature,10 close to the Langevin one, the product is likely HNC+ and not HCN+. The CN+ + H2 reaction has been studied by Petrie et al. 14 and Scott et al. 15 They found that it is a fast reaction producing equal amounts of HCN+ and HNC+, the isomers being identified through their different reactivity with CO and CO2. The reaction is supposed to lead only to HCN+ isomer.16,17 There is no data for the reaction but considering the potential energy curves, both isomers should be produced because both H-CN+ and H-NC+ approaches are barrierless. 11,18 It is then important to introduce both isomers in astrochemical networks with relevant rates and branching ratios.
Although the enthalpies of formation of HCN19,20 and HCN+ (see Section 3.4) are well known, the values for HNC and HNC+ are less certain. In particular, although a few theoretical 21–25 and experimental26,27 works are available in the literature, the value of the ionization energy of HNC is still uncertain, as reflected in the range of values reported in the literature (IEcalc. = [11.48 – 12.01] eV and IEexp. = [12.04 – 12.5] eV).
The main drawback of this lack of data is that the two isomers are often described as same species with the same thermodynamic properties. This crude approximation can be a source of error, particularly in reactions like H+ + HCN/HNC → H + HCN+/HNC+, which are either endothermic or exothermic at low temperature (see below).
In this paper, we present the first measurement of photoionizing transitions of HNC in the vicinity of the X1Σ+ → X+ 2Σ+ transition, supported by ab initio calculations.
2. Methods
2.1. Experimental details
The spectrum presented in this paper has been recorded at the DESIRS beamline of the SOLEIL French synchrotron facility 28 with the DELICIOUS III spectrometer. 29 To generate the HNC molecules, we used a flow-tube reactor that has been described in detail elsewhere.30 Briefly, a precursor is diluted in helium and then introduced in the flow tube before mixing with fluorine atoms that abstract H atoms from the precursor to generate hydrogen fluoride (HF).
In the present study, the precursor was methanimine (HN=CH2) which can lead to both isomers, HCN and HNC depending on which hydrogens are removed. It has been synthesized following an improved version of the procedure described in Ref. 31. A tube (30 cm) and two U-traps equipped with stopcocks were fitted on a vacuum line (0.1 mbar). The tube was half-filled with powdered potassium hydroxyde (50 g) and heated up to 90°C. The first trap was immersed in a cold bath (−80°C) and the second one in liquid nitrogen. Aminoacetonitrile (2.8 g, 50 mmol) was slowly vaporized on KOH. High boiling point impurities were trapped in the first trap and methanimine in the second one. At the end of the reaction, the stopcocks of the second U-trap were closed and the compound kept in liquid nitrogen (yield: 0.75 g, 52%, 26 mmol).
HN=CH2 was introduced in the experiment using its vapour pressure at –78°C (dry ice temperature). The flow-tube reactor conditions were optimized to favor two successive H-atom abstractions i.e. the HNC (and HCN) production. Even though the two isomers have the same mass, they have different vibronic structures, including a much lower ionization energy for the HNC isomer (IE(HNC) ≈ 12 eV, see Section 3.2), as compared to HCN (IE(HCN) = 13.607(2) eV13). The ionization energy and corresponding photoelectron energies can thus be used to distinguish the two isomers with the DELICIOUS III spectrometer.
The measurement presented in this paper has been recorded at a fixed photon energy (13 eV) by averaging the signal over 9 hours. The photoelectron spectrum has been obtained from the photoelectron image by using the pBASEX algorithm32 to extract the photoelectron kinetic-energy distribution. The ionization energies of water and oxygen, IE(H2O) = 101766.8(12) cm−1 (= 12.6175(2) eV)33 and IE(O2) = 97348(2) cm−1 (12.0696 eV)34, respectively, present in the background gas inside the spectrometer, was used to calibrate the energy scale. The resulting spectrum has a spectral resolution of 60 meV at a binding energy of 12 eV (1 eV kinetic energy)35 and the accuracy of the calibrated energy scale is about 5 meV.
2.2. Computational details
The potential energy surfaces of HCN+ in the first three electronic states, two of 2A′ and one of 2A″ symmetry, have been calculated here. A correlation consistent F12 double zeta atomic basis set 36, VDZ-F12, was used. Multi-configurational self-consistent field (MCSCF) calculations were performed using 8 and 2 active orbitals of a′ and a″ symmetry, respectively, optimizing the average energy of 4 states: the ground state of neutral, and for the cation, two states of 2A′ symmetry and one 2A″ state. With the resulting molecular orbitals, the ic-MRCI-F12 energies (internally contracted multireference configuration interaction) of the 2A′ and 2A″ states of the cation were calculated using the MOLPRO suite of programs.37
The calculations are done at geometries defined in Jacobi coordinates with r, the CN internuclear vector, R, the vector joining the CN center of mass to H, and cos γ = r · R/rR, where r and R are the modulus of the corresponding Jacobi vectors, see Figure 1. The zero energy is set at the minimum well of the X1Σ+ ground state of HCN (at re = 1.158 Å, Re = 1.697 Å, and γ = 180°), and the absolute energies of the HCN+ potential energy surfaces were determined from the ab initio calculations described above.
Fig. 1.
Representation of Jacobi coordinates (r, R, and γ). The H, C, and N atoms are represented in light grey, dark grey, and blue, respectively.
As discussed below, the lowest electronic states of HCN+ and HNC+ are subject to Σ – Π vibronic couplings due to conical intersections. In order to properly describe the conical intersections, a diabatic representation is considered, formed by three states with well-defined projection of the electronic orbital angular momentum, Λ = 0, ± 1. In this 3×3 subspace, the diabatic states are coupled through the Ve potential term. The Hamiltonian matrix is completely analogous to those used before for treating Σ – Π vibronic effects, 38
| (1) |
where E1 = E–1, and V1(r, R, γ) ∝ sin γ, as γ → 0° or 180°. The adiabatic eigenvalues of this electronic matrix are
and can be associated with the ab initio points of the two 2A′ and the 2A″ surfaces described above. Thus the parameters E0, E1, and V1 can be directly obtained from the ab initio calculations as previously done for the OHF system39.
With the three-dimensional set of ab initio points, the diabatic potential energy surfaces, E0 and E1, and the coupling, V1, of HCN+ are interpolated using 3D cubic splines, with the DB3INK/DB3VAL subroutines based on the method of de Boor40 and distributed by GAMS.41
The photoelectron spectrum of HNC isomer is calculated using a wave-packet method similarly to that applied to calculate the photodissociation of HCN42,43 and photodetachment of OHF–. 39 In the present approach, the transition dipole moment is neglected, and the initial wave-packet is the rovibrational wave function of the neutral HNC isomer. Here we consider three independent transitions, in which the initial wave-packet is placed on one of the 3 diabatic states of the cation. The wave-packet is then propagated within the coupled diabatic representation, using the MADWAVE3 code.44,45 The spectra are calculated by doing a Fourier transform of the autocorrelation function. We have propagated the wave-packets for about 105 iterations, so that doing the Fourier transform integration within a given time window, the resulting simulated spectra are broadened by a homogeneous width of approximately 0.5 meV, much narrower than the experimental resolution (see Section 2.1).
3. Results and discussion
3.1. Topography of the HCN(+)/HNC(+) isomerization
The minimum energy path calculated at each angle γ (optimizing r and R for each value of γ) is presented in Fig. 2. In Table 1, the optimized values of r and R for γ = 180° and 0° are reported. The values obtained for the 11A′ surface are in good agreement with experimental works.46,47
Fig. 2.
Minimum energy path for the HCN/HNC and HCN+/HNC+ isomerization in the ground state of the neutral and the lowest electronic states of the cation as a function of γ, with r and R parameters optimized for each value of γ. (See Figure 1 for geometry conventions.)
Table 1. Optimized geometries of the HCN(+)/HNC(+) neutral and cationic isomers (calculated in this work unless specified). All the values are in Å.
All the states displayed in Fig. 2 exhibit two minima at collinear geometries (γ = 180° and 0°), corresponding to the HCN(+) and HNC(+) isomers, respectively.
For HCN+, three states are very close in energy at linear configuration. Two of them are degenerate, forming the 2Π electronic ground state of HCN+ at γ = 180° and splitting in one 2A′ and one 2A″ components in the Cs symmetry group when the system bends. The third state is slightly higher in energy at γ = 180° and corresponds to the first excited state (2Σ+) of HCN+. Like one of the components of the ground state, the 2Σ+ state becomes a 2A′ state when 180° > γ > 0°. The lowest 2A′ component gives rise to the X+ 2Σ+ ground electronic state of HNC+ and the upper one, combined with the 2A″ component, lead to the 2Π excited states of HNC+ at γ = 0°.
Because they have the same symmetry, the two 2A′ components stemming from the two lowest electronic states (2Σ+ and 2Π states) of HCN+ interact and thus repel each other. This situation is typical near a conical intersection between the two 2A′ states, giving rise to important Σ – Π vibronic couplings, as reported for this system by Köppel et al. 48 The conical intersection is clearly seen in the bottom panel of Fig. 3, at a r distance slightly shorter than the equilibrium distance of the well of the neutral HCN (re = 1.158 Å). At this linear configuration, the equilibrium re distances of the 2Σ and 2Π states shift to lower and longer re distances, respectively. The large change of the r value upon ionization towards the 2Π state and the Σ – Π vibronic couplings lead to the vibrational progression observed in the HCN photoelectron spectrum.13,48–51
Fig. 3.
Potential energy curves of the three lowest electronic states of HCN+/HNC+ system for R = 1.697 Å and γ = 180° (HCN+ isomer, bottom panel), and for R = 1.531 Å and γ = 0° (HNC+ isomer, top panel). The degeneracy of the blue lines (full or dashed) and the red circles corresponds to the 2Π state. The vertical dotted lines indicate the re equilibrium distances of the neutral HNC (re = 1.166 Å) and HCN (re = 1.158 Å) isomers.
The situation for the HNC+ isomer, at γ = 0°, is rather different. In this case the 2Σ(= 11A′) state shifts to lower energy, resulting in a shift of the conical intersection towards considerably larger r values (top panel of Fig. 3). As a consequence, the Σ and Π states have a rather different energy and they do not strongly interact at re = 1.166 Å, in the Frank-Condon region. In addition, the 2Σ state for the HNC+ isomer has an equilibrium distance very close to that of neutral HNC. From that, one can expect to observe a reduced vibrational progression in the HNC photoelectron spectrum with respect to the one of HCN.
3.2. Photoelectron spectrum of the HNC isomer
In Fig. 4, the experimental photoelectron spectrum of the X1Σ+ → X+ 2Σ+ ionizing transition of HNC is depicted (upper panel) and compared with our calculated spectrum (lower panel). To account for the experimental resolution and the rotational structures of the neutral and the cationic states (not considered in our calculation), the calculated spectrum displayed in the lower panel (grey line) has been convoluted with a Gaussian line shape (full width at half maximum (FWHM)= 80 meV). The resulting spectrum is also presented in the lower panel (black line). Note that the calculated spectrum has been shifted up in energy by 0.371 eV in order to have the origin band of the X1Σ+ → X+ 2Σ+ transition at the same energy as in the experimental spectrum. The adiabatic ionization energy of HNC is thus the only fitted parameter in our calculated spectrum.
Fig. 4.
Photoelectron spectra of HNC: experimental spectrum recorded at 13 eV (upper panel) and calculated spectrum between 11 and 15 eV (lower panel). In the lower panel, the grey line is the calculated spectrum and the black line its convolution with a gaussian line shape (FWHM = 80 meV).
The experimental spectrum consists of one main band located at 12.011 eV corresponding to the origin band of the X1Σ+ → X+ 2Σ+ transition and a weaker band at higher energy (12.29 eV) involving the ν = 1 excited state of the cationic ν3 vibrational mode (CN stretching mode, see below). The transition towards the 2Π state of HNC+ should appear at approximately the same photon energy as the one of HCN+ as expected from the calculations. However, the PES of Fig. 4 has been recorded at 13 eV, below the IE of HCN (13.6 eV) to avoid a saturation of the spectrum by the HCN signal. Thus we cannot provide a measurement of the ionizing transition towards the first excited states of the HNC+ cation.
The different values available in the literature concerning the vibrational wavenumbers of HNC+ for the ν1 (NH stretch), ν2 (HNC bend), and ν3 (CN stretch) modes, are compared with our results in Table 2. The overall agreement is satisfactory.
Table 2. Vibrational wavenumbers of the HNC+ electronic ground state.
| Mode | Wavenumber / cm−1 | Method | Ref. |
|---|---|---|---|
| ν1 | 3470 | calc. | this work |
| 3404 | calc. | 25 | |
| 3464 | calc. | 24 | |
| 3365.0 | exp.a | 52 | |
| ν2 | 605 | calc. | this work |
| 523 | calc. | 25 | |
| 618 | calc. | 24 | |
| 577.7 or 584.5b | exp.a | 52 | |
| ν3 | 2260 | calc. | this work |
| 2163 | calc. | 25 | |
| 2212 | calc. | 24 | |
| 2260 | exp. | this workc | |
| 2195.2 | exp.a | 52 |
Measured in neon matrix.
Observed but not assigned to ν2 by the authors. The NIST recommends 577.7 cm−1.53
±80 cm−1.
From our calculations, one should expect to observe in the photoelectron spectrum transitions involving all the vibrational modes of the cation (ν1, ν2, and ν3), see lower panel of Fig. 4. The calculated positions of the bands with respect to the origin band are 0.15 eV (1210 cm−1), 0.28 eV (2260 cm−1), and 0.43 eV (3470 cm−1), respectively, see Table 2. Nevertheless in the present work, the signal-to-noise ratio does not allow an unambiguous assignment of structures corresponding to ν1- and ν2-active transitions. Only the fundamental of the ν3 vibrational mode can be extracted from our spectrum: (= 0.28(1) eV). This value is in very good agreement with our calculations (0.28 eV), see Table 2.
In Table 3, the IE value of HNC obtained in this work is compared with previous experimental works. For completeness, the two most accurate values for HCN are also reported. The only experimental values for HNC found in the literature are the one deduced by Hansel et al. from the HNC+ + Xe charge transfer reaction 27 and the recombination energy of HNC+ measured by Bieri and Jonsson.26 Although the uncertainties do not overlap, the agreement is satisfactory.
Table 3. Experimental values of the first ionization energy of the hydrogen cyanide isomers.
Deduced from the HNC+ + Xe charge transfer reaction.
Estimation, value of the recombination energy of HNC+.
On the side of the theory, several values have been obtained using different approaches. 21–24 The value of Koch et al. calculated using the Moller-Plesset perturbation theory (12.01 eV) presents the best agreement with our measurements. 23
3.3. Photoelectron angular distribution of HNC
The normalized photoelectron angular distribution is described by the well-known parameter, β, through the expression:
| (2) |
where θ is the polar angle of ejection with respect to the ε polarization vector of the light (see Fig. 5). The β values have been extracted from the photoelectron image shown in Figure 5 for the transitions to the ν = 0 (origin band) and ν3 = 1 (ν3 = 1 band) levels of cationic ground state, and correspond to 1.14 and 1.2, respectively. In the absence of theoretical calculations of the photoionization process, which are not within the scope of the present article, we can only compare these values to those found in HCN. In HNC, the highest occupied molecular orbital (HOMO) corresponds to the 5σ orbital, and the (HOMO-1) corresponds to the 1π orbital, while in HCN, the ordering of these two orbitals is reversed. Holland et al. 55 have reported the β parameters for ionization from these two orbitals close to threshold. For the 1π orbital, the origin and ν3 = 1 bands gave β values around 0.0 and 0.25, respectively. Unfortunately, the photoelectron bands for ionization from the 5σ orbital were overlapped by signal from ionization to higher vibrational levels from the 1π band. Nevertheless, the β value extracted for these overlapped bands near threshold is ≈ 0.4, somewhat higher than for ionization from the 1π orbital of HCN. This observation is consistent with the present one for HNC, where ionization from the 5σ orbital is more anisotropic and gives a relatively large value of β.
Fig. 5.
Photoelectron image of HNC recorded at 13 eV photon energy. The left side shows the raw image, while the right side shows the output of the pBasex algorithm. The positive anisotropy discussed in the text is clear for the transitions to both the origin and the ν3 = 1 levels of the cation. The white arrow indicates the polarization vector of the synchrotron radiation.
3.4. Thermodynamics of the HCN/HNC isomers
The enthalpy of formation of HCN is well known and the preferred value in the literature is 135.1(84) kJ.mol−1. 19,20 Using the precise value of the ionization energy of HCN reported by Fridh and Åsbrink,13 one can deduce the enthalpy of formation of HCN+ (ΔfH°(HCN+) = 1448.0(84) kJ.mol−1). These two values are compatible with the ones of the Active Thermochemical Tables, 56,57 see Table 4.
Table 4. Enthalpies of formation of the hydrogen cyanide isomers, their cations, and the CN radical at 300 K.
| ΔfH° / kJ.mol−1 | Ref. | |
|---|---|---|
| HCN | 135.1(84) | 19 and 20 |
| 129.283(91) | 56 | |
| HCN+ | 1448.0(84) | 19, 20, and 13 |
| 1442.06(20) | 56 | |
| HNC | 178(10) | 63, 19, and 20 |
| 197(12) | 64 | |
| 189(3) | this work and 27,61 | |
| 184(3) | this work and 27,60 | |
| 188(4) | 27 | |
| 192.40(38) | 56 | |
| HNC+ | 1348(3) | 27,56 |
| 1342(3) | 27,60 | |
| 1349(4) | 27 | |
| 1352.9(13) | 56 | |
| CN | 439.7(48) | 61,62 |
| 434.5(20) | 60 | |
| 440.0(1) | 56 | |
Although the enthalpy of formation and the IE of HCN are well known, there is relatively little direct experimental data for HNC. Hansel et al. measured the endothermicity of the HNC+ + CO → HCO+ + CN reaction (31 kJ.mol−1). 27 From this value, they derived the enthalpy of formation of HNC+, using literature data. In their paper, the source of the data from the literature was not unambiguously given. Thus, we propose here to use other values from the literature combined with their result. Using ΔfH(CO) = −110.5(10) kJ.mol−1, 20 ΔfH(HCO+) = ΔfH(HCO) + IE(HCO) = 42(1)+ 786.377(6) kJ.mol−1, 58,59 and ΔfH(CN) (434.5(20) kJ.mol−1 60 or 439.7(48) kJ.mol−1 61,62), we obtain ΔfH(HNC+) = 1342(3) kJ.mol−1 or 1348(3) kJ.mol−1. Note that Hansel et al. obtained 1349(4) kJ.mol−1 close to the latter value.
Using our IE(HNC) = 12.01(1) eV and the two values of ΔfH(HNC+) derived above, we obtain ΔfH(HNC) = 184(3) kJ.mol−1 or 189(3) kJ.mol−1, respectively. The latter value is in good agreement with the one from the Active Thermochemical Tables with respect to the error bars, 56 see Table 4.
In addition, there are two direct measurements of the energy difference between HNC and HCN. First, Maki et al. performed a HCN/HNC equilibrium measurement at high temperature using spectroscopic HNC absorption.63 Their measurement gave an energy difference (ΔEneutral iso.) between HNC and HCN of 3600(400) cm−1 = 43.1(48) kJ.mol−1. Using this value and the enthalpy of formation of HCN, one can deduce the enthalpy of formation of HNC (ΔfH0 = 178(10) kJ.mol−1). One year later, Pau and Hehre measured the energy difference between HNC and HCN (ΔEneutral iso. = 61.9(83) kJ.mol−1) using pulsed ion cyclotron double resonance spectroscopy64 which leads to an enthalpy of formation of HNC of 197(12) kJ.mol−1. It is difficult to explain the differences between the value of Maki et al. and the one of Pau and Hehre, but the results obtained by Pau and Hehre are suported by theoretical calculations (ΔEneutral iso. = 59.4 kJ.mol−1). 65 In addition, one could argue that at high temperature HCN and HNC have a non-negligible reactivity that may affect the HCN/HNC equilibrium; thus, the measurement of Maki et al. may be biased. Using the enthalpies of formation of HCN and HNC from the the Active Thermochemical Tables, 56 we obtain ΔEneutral iso. = 63.117 kJ.mol−1, which is in better agreement with the value of Pau and Hehre.
From the discussion above, the values proposed by the Active Thermochemical Tables, 56 seem to be very accurate and always confirmed by experimental measurement. Note that from the enthalpy of formation of HNC and HNC+ proposed by these tables, one can derive an IE value for HNC of 12.028(14) eV, in very good agreement with our present measurement (12.01(1) eV).
4. Concluding remarks
Using a flow-tube reactor and a suitable precursor (HN=CH2), the hydrogen isocyanide isomer (HNC) has been efficiently produced in the gas phase, and its photoionization has been studied for the first time at a fixed photon energy (13 eV). The recorded photoelectron spectrum is in good agreement with the photoionizing transition spectrum predicted by wave-packet method based on ic-MRCI ab initio calculation of the potential energy surfaces for the HCN/HNC system. The differences in the photoelectron spectra of the HCN and HNC isomers are explained by the different energy landscapes of their respective Franck-Condon regions. For HCN, a conical intersection between two of the three lowest potential energy surfaces of the cation occurs close to the equilibrium geometry of the HCN ground state, giving rise to important vibronic couplings in the ground state of the cation. This affects the vibrational progressions in the HCN photoelectron spectrum in the vicinity of the first ionization energy. 13,48–51 On the contrary, for HNC no such vibrational progression is expected since the conical intersection is located far from the HNC equilibrium geometry and at higher energy. The experimental spectrum, recorded for the X1Σ+ → X+ 2Σ+ transition, exhibits mainly the origin band and a vibrational band corresponding to excitation of the cationic CN stretching mode (ν3), for which the fundamental is measured at .
The calculations of the present work predict another photoionization transition, X1Σ+ → A+ 2Π, involving the first excited electronic state of HNC+ and resulting in an extended vibrational progression between 13.7 and 15 eV. It would be very interesting to design a gas-phase source enable to produce efficiently the HNC isomer only. This would allow to confirm our prediction and to investigate the vibronic structure of the first excited electronic state of the cation.
The β anisotropy parameter for the origin band and the ν3 = 1 band is extracted from the photoelectron angular distribution of HNC with values around 1.14 and 1.2 larger than what was measured for HCN. This finding is consistent with the different HOMO involved in the photoionization process.
Finally, the measurement of the first adiabatic ionization energy of HNC (IE(HNC) = 12.01(1) eV) can be used in combination with the formation enthalpy of HNC+ found in the literature to derive the formation enthalpy of neutral HNC: ΔfH(HNC) = 184(3) kJ.mol−1 or 189(3) kJ.mol−1, depending on which value of the literature is employed for the enthalpy of formation of CN. It appears that the calculated thermodynamic data available in the Active Thermochemical Tables 56 are all in agreement with the experimental values for the HCN(+)/HNC(+) isomers. Their very small error bars make them the preferred values for the thermodynamics of these species.
Acknowledgments
This work was performed on the DESIRS beamline under Proposal No. 20160376. We acknowledge SOLEIL for provision of synchrotron radiation facilities and the DESIRS beamline team for their assistance. This work received financial support from the French Agence Nationale de la Recherche (ANR) under Grant No. ANR-12-BS08-0020-02 (project SYNCHROKIN) and the CNRS program "Physique et Chimie du Milieu Interstellaire" (PCMI) cofunded by the Centre National d'Etudes Spatiales (CNES). A. A. and O. R. acknowledge support by the Ministerio de Economía e Innovación under grants No. FIS2014-52172-C2 and FIS2017-83473-C2 and from the European Research Council under the European Union’s Seventh Framework Programme (FP/2007-2013) / ERC Grant Agreement n. 610256 (NANOCOSMOS). Work by was supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, Division of Chemical Sciences, Geosciences, and Biosciences respectively under contract No. DE-AC02-06CH11357. J.-C.G. thanks the Centre National d’Etudes Spatiales (CNES) for financial support.
Notes and references
- 1.Liszt H, Lucas R. Astronomy & Astrophysics. 2001;370:576–585. [Google Scholar]
- 2.Turner BE, Pirogov L, Minh YC. The Astrophysical Journal. 1997;483:235–261. [Google Scholar]
- 3.Irvine W, Schloerb F. The Astrophysical Journal. 1984;282:516–521. doi: 10.1086/162229. [DOI] [PubMed] [Google Scholar]
- 4.Pratap P, Dickens JE, Snell RL, Miralles MP, Bergin EA, Irvine WM, Schloerb FP. The Astrophysical Journal. 1997;486:862–885. doi: 10.1086/304553. [DOI] [PubMed] [Google Scholar]
- 5.Hirota T, Yamamoto S, Mikami H, Ohishi M. The Astrophysical Journal. 1998;503:717–728. [Google Scholar]
- 6.Jørgensen JK, Schöier FL, van Dishoeck EF. Astronomy & Astrophysics. 2004;416:603–622. [Google Scholar]
- 7.Hily-Blant P, Walmsley M, Pineau des Forêts G, Flower D. Astronomy & Astrophysics. 2010;513:A41. [Google Scholar]
- 8.Godard B, Falgarone E, Gerin M, Hily-Blant P, De Luca M. Astronomy & Astrophysics. 2010;520:A20. [Google Scholar]
- 9.Loison J-C, Wakelam V, Hickson KM. Monthly Notices of the Royal Astronomical Society. 2014;443:398. doi: 10.1093/mnras/stv2866. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Anicich VG. JPL-Publication-03-19. 2003. [Google Scholar]
- 11.Talbi D, Padellec AL, Mitchell JBA. Journal of Physics B: Atomic, Molecular and Optical Physics. 2000;33:3631. [Google Scholar]
- 12.Garcia JD, Mack JE. J Opt Soc Am. 1965;55:654–685. [Google Scholar]
- 13.Fridh C, Åsbrink L. Journal of Electron Spectroscopy and Related Phenomena. 1975;7:119–138. [Google Scholar]
- 14.Petrie S, Freeman CG, McEwan MJ, Ferguson EE. Monthly Notices of the Royal Astronomical Society. 1991;248:272–275. [Google Scholar]
- 15.Scott GB, Fairley DA, Freeman CG, McEwan MJ, Spanel P, Smith D. The Journal of Chemical Physics. 1997;106:3982–3987. [Google Scholar]
- 16.Scott GBI, Fairley DA, Freeman CG, McEwan MJ, Anicich VG. The Journal of Physical Chemistry A. 1999;103:1073–1077. [Google Scholar]
- 17.Talbi D. Chemical Physics Letters. 1999;312:291–298. [Google Scholar]
- 18.Sheehan C, Padellec AL, Lennard WN, Talbi D, Mitchell JBA. Journal of Physics B: Atomic, Molecular and Optical Physics. 1999;32:3347. [Google Scholar]
- 19.Chase MJ. NIST-JANAF THERMOCHEMICAL TABLES, Monograph 9, Fourth Edition, Journal of Physical and Chemical Reference Data Monographs. 1998. [Google Scholar]
- 20.Baulch DL, Bowman CT, Cobos CJ, Cox RA, Just T, Kerr JA, Pilling MJ, Stocker D, Troe J, Tsang W, Walker RW, et al. Journal of Physical and Chemical Reference Data. 2005;34 year. [Google Scholar]
- 21.von Niessen W, Cederbaum LS, Domcke W, Diercksen GHF. Molecular Physics. 1976;32:1057–1061. [Google Scholar]
- 22.Murrell JN, Derzi AA. Journal of the Chemical Society, Faraday Transactions 2: Molecular and Chemical Physics. 1980;76:319–323. [Google Scholar]
- 23.Koch W, Frenking G, Schwarz H. Naturwissenschaften. 1984;71:473–474. [Google Scholar]
- 24.Peterson KA, Mayrhofer RC, Woods RC. The Journal of Chemical Physics. 1990;93:4946–4953. [Google Scholar]
- 25.Kraemer WP, Jensen P, Roos BO, Bunker PR. Journal of Molecular Spectroscopy. 1992;153:240–254. doi: 10.1006/jmsp.1997.7452. [DOI] [PubMed] [Google Scholar]
- 26.Bieri G, Jonsson B-Ö. Chemical Physics Letters. 1978;56:446–449. [Google Scholar]
- 27.Hansel A, Scheiring C, Glantschnig M, Lindinger W, Ferguson EE. The Journal of Chemical Physics. 1998;109:1748–1750. [Google Scholar]
- 28.Nahon L, de Oliveira N, Garcia GA, Gil J-F, Pilette B, Marcouillé O, Lagarde B, Polack F. Journal of Synchrotron Radiation. 2012;19:508–520. doi: 10.1107/S0909049512010588. [DOI] [PubMed] [Google Scholar]
- 29.Garcia GA, de Miranda BK, Tia M, Daly S, Nahon L. Review of Scientific Instruments. 2013;84:053112. doi: 10.1063/1.4807751. [DOI] [PubMed] [Google Scholar]
- 30.Garcia GA, Tang X, Gil J-F, Nahon L, Ward M, Batut S, Fittschen C, Taatjes CA, Osborn DL, Loison J-C. The Journal of Chemical Physics. 2015;142 doi: 10.1063/1.4918634. 164201. [DOI] [PubMed] [Google Scholar]
- 31.Guillemin J-C, Denis J-M. Tetrahedron. 1988;44:4431–4446. [Google Scholar]
- 32.Garcia GA, Nahon L, Powis I. Review of Scientific Instruments. 2004;75:4989–4996. [Google Scholar]
- 33.Merkt F, Signorell R, Palm H, Osterwalder A, Sommavilla M. Molecular Physics. 1998;95:1045–1054. [Google Scholar]
- 34.Tonkyn RG, Winniczek JW, White MG. Chemical Physics Letters. 1989;164:137–142. [Google Scholar]
- 35.Tang X, Garcia GA, Gil J-F, Nahon L. Review of Scientific Instruments. 2015;86:123108. doi: 10.1063/1.4937624. [DOI] [PubMed] [Google Scholar]
- 36.Peterson KA, Adler TB, Werner H-J. The Journal of Chemical Physics. 2008;128:84102. doi: 10.1063/1.2831537. [DOI] [PubMed] [Google Scholar]
- 37.Werner H-J, Knowles PJ, Knizia G, Manby FR, Schütz M, et al. MOLPRO, version 2012.1, a package of ab initio programs. 2012 [Google Scholar]
- 38.Köppel H, Domcke W, Cederbaum LS. Multimode Molecular Dynamics Beyond the Born-Oppenheimer Approximation. Vol. 57. Wiley-Blackwell; 1984. p. 59. [Google Scholar]
- 39.Gómez-Carrasco S, Aguado A, Paniagua M, Roncero O. The Journal of Chemical Physics. 2006;125 doi: 10.1063/1.2363988. 164321. [DOI] [PubMed] [Google Scholar]
- 40.de Boor C. A practical guide to splines. Springer-Verlag; Nex York: 1978. [Google Scholar]
- 41.Boisvert RF, Howe SE, Kahaner DK. Transactions on Mathematical Software. 1985;11:313–355. [Google Scholar]
- 42.Chenel A, Roncero O, Aguado A, Agúndez M, Cernicharo J. The Journal of Chemical Physics. 2016;144:144306. doi: 10.1063/1.4945389. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Aguado A, Roncero O, Zanchet A, Agúndez M, Cernicharo J. The Astrophysical Journal. 2017;838:33. doi: 10.3847/1538-4357/aa63ee. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.González-Lezana T, Aguado A, Paniagua M, Roncero O. The Journal of Chemical Physics. 2005;123:194309. doi: 10.1063/1.2118567. [DOI] [PubMed] [Google Scholar]
- 45.Zanchet A, Roncero O, González-Lezana T, Rodríguez-López A, Aguado A, Sanz-Sanz C, Gómez-Carrasco S. The Journal of Physical Chemistry A. 2009;113:14488–14501. doi: 10.1021/jp9038946. [DOI] [PubMed] [Google Scholar]
- 46.Harmony MD, Laurie VW, Kuczkowski RL, Schwendeman RH, Ramsay DA, Lovas FJ, Lafferty WJ, Maki AG. Journal of Physical and Chemical Reference Data. 1979;8:619–722. [Google Scholar]
- 47.Maki AG, Mellau G, Klee S, Winnewisser M, Quapp W. Journal of Molecular Spectroscopy. 2000;202:67–82. doi: 10.1006/jmsp.1999.7794. [DOI] [PubMed] [Google Scholar]
- 48.Köppel H, Cederbaum LS, Domcke W, Niessen WV. Chemical Physics. 1979;37:303–317. [Google Scholar]
- 49.Turner DW. Philosophical Transactions of the Royal Society of London A: Mathematical, Physical and Engineering Sciences. 1970;268:7–31. [Google Scholar]
- 50.Hollas JM, Sutherley TA. Molecular Physics. 1972;24:1123–1131. [Google Scholar]
- 51.Frost DC, Lee ST, McDowell CA. Chemical Physics Letters. 1973;23:472–475. [Google Scholar]
- 52.Forney D, Thompson WE, Jacox ME. The Journal of Chemical Physics. 1992;97:1664–1674. [Google Scholar]
- 53.Linstrom PJ, Mallard WG. NIST Chemistry WebBook, NIST Standard Reference Database Number 69. National Institute of Standards and Technology; 2010. [Google Scholar]
- 54.Eland JHD, Field T, Baltzer P, Hirst DM. Chemical Physics. 1998;229:149–163. [Google Scholar]
- 55.Holland DMP, Parr AC, Dehmer JL. Journal of Physics B: Atomic and Molecular Physics. 1984;17:1343. [Google Scholar]
- 56.Ruscic B, Bross DH. Active Thermochemical Tables (ATcT) values based on ver. 1.122d of the Thermochemical Network. 2018 [Google Scholar]
- 57.Nguyen TL, Baraban JH, Ruscic B, Stanton JF. The Journal of Physical Chemistry A. 2015;119:10929–10934. doi: 10.1021/acs.jpca.5b08406. [DOI] [PubMed] [Google Scholar]
- 58.Chuang M, Foltz MF, Moore CB. The Journal of Chemical Physics. 1987;87:3855–3864. [Google Scholar]
- 59.Mayer E, Grant ER. The Journal of Chemical Physics. 1995;103:10513–10519. [Google Scholar]
- 60.Huang Y, Barts SA, Halpern JB. The Journal of Physical Chemistry. 1992;96:425–428. [Google Scholar]
- 61.Costes M, Naulin C, Dorthe G. Astronomy & Astrophysics. 1990;232:270–276. [Google Scholar]
- 62.Cox JD, Wagman DD, Medvedev VA. CODATA key values for thermodynamics. Hemisphere Pub Corp.,; 1984. [Google Scholar]
- 63.Maki AG, Sams RL. The Journal of Chemical Physics. 1981;75:4178–4182. [Google Scholar]
- 64.Pau CF, Hehre WJ. The Journal of Physical Chemistry. 1982;86:321–322. [Google Scholar]
- 65.DePrince AE, III, Mazziotti DA. The Journal of Physical Chemistry B. 2008;112:16158–16162. doi: 10.1021/jp805752f. [DOI] [PubMed] [Google Scholar]





