Technical Comments

Comment on “Synthesis and characterization of the pentazolate anion cyclo-N5 in (N5)6(H3O)3(NH4)4Cl”

See allHide authors and affiliations

Science  16 Mar 2018:
Vol. 359, Issue 6381, eaao3672
DOI: 10.1126/science.aao3672

Abstract

Zhang et al. (Reports, 27 January 2017, p. 374) reported synthesis of a cyclo-N5 ion putatively stabilized in a solid-state salt by hydrogen bonding from surrounding counterions. We performed theoretical calculations suggesting that HN5 would be favored over the anion in the reported pentazolate salt via proton transfer.

Zhang et al. (1) reported a cyclo-N5 ion stabilized in the (N5)6(H3O)3(NH4)4Cl salt via hydrogen bonding from the adjacent NH4+ and H3O+ ions. The structure was determined by single-crystal x-ray diffraction (XRD) analysis. However, the ellipsoid of the O2 atom in the asymmetry unit plot of the pentazolate salt was much larger than those of the remaining atoms in the reported pentazolate salt, indicating that there should actually be less electron density than assigned in the model. The O2 atom could also be a nitrogen atom. Therefore, we suggest that the authors should provide solid evidence to confirm the O2 atom in the pentazolate salt. Moreover, the phase purity of the pentazolate salt and the stability of its crystal structure at room temperature could not be determined on the basis of the physical evidence afforded at present. In particular, the cyclo-N5 ion in (N5)6(H3O)3(NH4)4Cl possesses C2v symmetry, and thus the analysis of the infrared spectrum of the cyclo-N5 ion based on D5h symmetry is questionable. In addition, the out-of-plane mode peak of the H3O+ ion cannot be observed from the infrared spectrum of the pentazolate salt. On the basis of the nature of the cyclo-N5 ion and H2O, the two expected competitive reaction pathways areEmbedded Image Thus, in the reported pentazolate salt, the question becomes whether the purported cyclo-N5 is actually protonated. Recently, Steele et al. (2) reported that the crystal of pentazole HN5 discovered using first-principles evolutionary search at high pressures is thermodynamically stable, so HN5 may also exist in the pentazolate salt. Because the position of the H atom is difficult to confirm by single-crystal XRD methods, we performed theoretical calculations for all the relevant species as an aid to ascertaining the H atom position in the reported pentazolate salt.

All computations were performed with the Gaussian 09 D.01 program (3). Density functional theory calculations were carried out using the B3LYP (4, 5), wB97X-D (6), and M06-2X (7) methods in combination with the 6-311++G(d,p) (8) basis set and the very large aug-cc-pVQZ (9, 10) basis set for H, O, and N atoms, respectively. For all the species explored here, the initial geometries were further optimized and were followed via frequency analysis to confirm minimum energy structures. The vibrationally stable cyclo-N5 ion possesses D5h symmetry at different levels, in distinct contrast to that of the cyclo-N5 ion (C2v symmetry) in the reported pentazolate salt. However, the experimental C2v symmetry of the cyclo-N5 ion coincides with that of HN5 (C2v symmetry) obtained at different levels. This indicates the possible formation of HN5 during the reaction process.

To illustrate that HN5 can exist in the reported pentazolate salt, we obtained the molecular electrostatic potentials V(r) and average local ionization energies Ī(r) on the molecular surfaces encompassing cyclo-N5 and H2O by means of the Multiwfn program (11). Figure 1A presents the distribution of the local minima values, VS,min(r) and ĪS,min(r), on the van der Waals surface of cyclo-N5 and H2O, which are the electrophilic reactive positions. VS,min(r) and ĪS,min(r) are all associated with the N/O atoms in cyclo-N5 and H2O, except for the partial VS,min(r) of cyclo-N5 located between the two adjacent nitrogen atoms. In comprehensively considering the distribution of VS,min(r) and ĪS,min(r), the strong reactive sites are those associated with the N/O atoms in cyclo-N5 and H2O, respectively, which can be attacked by the electrophilic reagents (e.g., Mn+ and H+).

Fig. 1 Computational results.

(A) Distribution of electrostatic potential minima (left) and average local ionization energy minima (right) on the van der Waals surfaces of cyclo-N5 (top) and H2O (bottom) at the B3LYP/6-311++G(d,p) level. (B) The calculated ΔVS,min(r) [VS,min(r) of H2O – VS,min(r) of cyclo-N5] (red) and ΔĪS,min(r) [ĪS,min(r) of H2O – ĪS,min(r) of cyclo-N5] (blue) at different levels.

As is well known, the addition product, H3O+, is stable in corresponding systems. Zhang et al. (12) reported a cobalt-based metal complex containing the cyclo-N5 ligand, in which the Co2+ and N5 ions are linked together though a Co–N bond. Burke and colleagues (13), using 15N nuclear magnetic resonance spectroscopy, concluded that HN5 was also produced and held in solution as a N5 ion with Zn2+. Steele et al. (2) reported a thermodynamically stable crystal of pentazole HN5 discovered at pressures above 50 GPa using first-principles methods. All of these results, obtained experimentally and theoretically, indicate that our predictive electrophilic reactivities of cyclo-N5 and H2O are reasonable. Further, in looking at the VS,min(r) and ĪS,min(r) values associated with the N/O atoms for cyclo-N5 and H2O, it is notable that the VS,min(r) and ĪS,min(r) values of cyclo-N5 were much lower than those found for H2O at different levels. The computed ΔVS,min(r) [VS,min(r) of H2O – VS,min(r) of cyclo-N5] and ΔĪS,min(r) [ĪS,min(r) of H2O – ĪS,min(r) of cyclo-N5] were also obtained. As shown in Fig. 1B, the ΔVS,min(r) and ΔĪS,min(r) between H2O and cyclo-N5 were more than 84 and 122 kcal mol−1, respectively. The large difference values of VS,min(r) and ĪS,min(r) indicate that the electrophilic reactivity of cyclo-N5 was much higher than that found for H2O, indicating that HN5 was more favored in the reaction mixture of H+, cyclo-N5, and H2O. The topological parameters at the bond critical points (BCPs) of the N-H and O-H bonds for HN5 and H3O+ at different levels are listed in Table 1. The electronic density at the BCPs demonstrates that the N-H bond of HN5 was also stronger than the O-H bond of H3O+. Thus, HN5 can exist in the reported pentazolate salt.

Table 1 Topological parameters at bond critical points of N-H and O-H bonds for HN5 and H3O+ by atoms-in-molecules analysis based on the aug-cc-pVQZ basis set.

Values are atomic units.

View this table:

Deeper insight into possible proton transfer (PT) from H3O+ to cyclo-N5 emerged from theoretical calculations performed for the ion pair N5⋅⋅⋅H3O+ (Fig. 2A), which were taken from the crystal structure of the reported pentazolate salt. The optimized stable structure of the ion pair N5⋅⋅⋅H3O+ was also obtained, as shown in Fig. 2. We obtained new species in which the proton was bonded to the nitrogen atom of cyclo-N5, namely the hydrogen-bonding dimer HN5⋅⋅⋅H2O (Fig. 2B). In addition, the figure presents a continuous decrease in the energy for the species studied, which occurs as a result of the PT from H3O+ to cyclo-N5 and the formation of O⋅⋅⋅H-N hydrogen bonds and N-H bonds with cyclo-N5. The above analysis shows a barrierless PT mechanism during optimization. Furthermore, the calculations at the wB97X-D/aug-cc-pVQZ level show that the deprotonation of H3O+ requires an energy of 166.62 kcal mol−1, whereas the protonation of cyclo-N5 can provide an energy of 319.77 kcal mol−1. As a result, besides enough energy for the PT process, there can be extra energy for release, so it is easy for PT to occur in the ion pair studied. The above analysis further indicates that the HN5 can exist in the reported pentazolate salt via PT.

Fig. 2 Geometry optimization.

(A) Ion pair N5⋅⋅⋅H3O+ obtained from the crystal structure of the pentazolate salt. (B) A barrierless proton transfer during optimization leads to the final optimized structure at the wB97X-D/aug-cc-pVQZ level.

References and Notes

Acknowledgments: Supported by Natural Science Foundation of the Education Commission of Anhui Province of China grant KJ2017A347.
View Abstract

Navigate This Article