Prediction of Superhalogen-Stabilized Noble Gas Compounds - The

Aug 27, 2014 - The discovery of HArF has generated renewed interest in the chemistry of noble gases, particularly their hydrides. Though many weak ...
0 downloads 0 Views 1MB Size
Letter pubs.acs.org/JPCL

Prediction of Superhalogen-Stabilized Noble Gas Compounds Devleena Samanta* Department of Chemistry, Stanford University, Stanford, California 94305, United States S Supporting Information *

ABSTRACT: The discovery of HArF has generated renewed interest in the chemistry of noble gases, particularly their hydrides. Though many weak complexes of noble gases bound by van der Waals interactions are known, the number of halogenated noble gas compounds, HNgX (Ng = noble gas; X = halogen), where the noble gas atom is chemically bound, is limited. These molecules are metastable, and their specialty is that there is substantial ionic bonding between the noble gas atom and the halogen atom. In this Letter, it is shown using density functional theory and second-order Møller−Plesset perturbation theory that by replacing the halogen atoms by superhalogens (Y), whose electron affinities are much larger than those of halogens, more ionic bonds between Ng and Y can be attained. Moreover, the superhalogen-containing noble gas hydrides, HNgY, are more stable compared to their halogenated counterparts. SECTION: Molecular Structure, Quantum Chemistry, and General Theory he noble gases (Ng’s), which have completely filled s and p shells, were initially thought to be inert due to their reluctance to participate in chemical bonding. In 1933, Pauling1 speculated that the heavier Ng’s, Xe and Kr, may form stable compounds. Bartlett2 confirmed this prediction in 1962 by characterizing Xe+[PtF6]−, the first compound of xenon. His synthesis was guided by the rationale that PtF6 is strong enough to oxidize oxygen (IP = 12.07 eV).3 Therefore, he reasoned that it should also be able to oxidize xenon, which has a similar ionization potential (IP = 12.13 eV).4 This revolutionary work not only showed that Ng’s can react with appropriate ligands but also instigated the search for other stable Ng compounds. Over the years, several compounds of Xe as well as Kr were found.5−7 However, molecules of the lighter congeners, Ar, Ne, and He, remained elusive until 2000. The discovery of HArF, the only known neutral argon molecule, was another breakthrough in Ng chemistry.8 Following Pettersson’s pioneering work9 on Ng hydrides, in 2000, Räsänen and co-workers8 successfully prepared this compound by photolysis of HF in an argon matrix at 7.5 K. Their discovery was followed by a series of theoretical and experimental studies on the stabilities and predictions of various Ng hydrides.7,10−12 As the Ar atom is sandwiched between H and F, this type of a compound is called an “insertion compound”. This is a charge-transfer molecule with a partial positive charge on the HAr moiety and a partial negative charge on F.8 Some other experimentally confirmed insertion compounds are HKrF,13 HKrCl,9 HXeNC,14 and so on. An interesting observation is that most Ng compounds contain one or more highly electronegative moieties such as F, CN, and Cl. The success of these particular ligands can be attributed to their high electron affinities (EAs).15−17 It was shown by Gutsev and Boldyrev18 that it is possible to create “superhalogens”, which have EAs even larger than that of

T

© XXXX American Chemical Society

chlorine (EA = 3.62 eV),17 the element with the highest EA. The conventional superhalogen is given by the general formula MXn+1, where n is the maximal valence of a central metal atom, M, surrounded by n + 1 halogen atoms, X. Once the extra electron is attached, it can delocalize over n + 1 halogen atoms, as opposed to one halogen atom. Consequently, the EA of the superhalogen is higher than that of the constituent halogen.18 Over the past 30 years, theoretical as well as experimental studies have shown that the presence of metals or halogens is not mandatory for a cluster to exhibit superhalogen properties.19,20 Examples include BF4 (EA = 6.86 eV),21 which follows the MXn+1 scheme but has no metal, and BO2 (EA = 4.46 eV),22 which has neither a metal atom nor a halogen atom. With the concept of superhalogens in mind, it is clear why PtF6 succeeds in binding with Xe. The EA of PtF6 is 7.00 eV,23 giving it the capability to remove an electron from the filled valence shell of Xe. The present study examines whether superhalogens, denoted by Y, can also stabilize Kr and Ar using density functional theory and second-order Møller−Plesset perturbation theory. Though an appropriate superhalogen can potentially stabilize Ar and Kr in the form Ng+Y−, the EA required would be enormous, and it would be difficult to find a suitable ligand. On the other hand, HArF and HKrF have been experimentally confirmed.8,13 Hence, it is logical to propose that substitution of F with a superhalogen may lead to more stable Ng compounds. The replacement of the F atom by a superhalogen in HArF and HKrF is expected to increase the charge transfer, thereby improving the ionic character of the HNg−Y bond. Moreover, the fragmentation energy corresponding to the Received: July 7, 2014 Accepted: August 27, 2014

3151

dx.doi.org/10.1021/jz501404h | J. Phys. Chem. Lett. 2014, 5, 3151−3156

The Journal of Physical Chemistry Letters

Letter

linear with a longer Ng−F bond compared to the H−Ng bond. The bond lengths shown in the figures are in good agreement with those from prior benchmark CCSD(T) studies.8,26 The H−Ng−O moiety in HNgBO2 is pseudolinear (∼177°) in both HArBO2 and HKrBO2. The Ng−O−B angle is about 116° in HArBO2 and about 117° in HKrBO2. In HArBF4 and HKrBF4, the H−Ng−F angles (where F is the fluorine atom closest to Ng) are 178 and 177°, respectively. The BF4 moiety has one B−F bond, which is slightly elongated (1.46 and 1.47 Å for Ng = Ar and Kr, respectively) compared to the B−F bond length in free BF4− (1.41 Å), whereas one B−F bond is slightly contracted (1.36 Å for both Ng = Ar and Kr). It should be noted that free BF4 (neutral) is a weakly bound complex of the form BF3·F in which all B−F bonds of the BF3 moiety are 1.32 Å.27 The distance between the remaining F atom and the B atom is 2.74 Å. Therefore, the structure of the BF4 moiety in HNgBF4 lies between that of BF3·F and BF4−, closer to the anion. This structure is an indication that there may be significant ionic bonding between Ng and BF4. In general, HNgY molecules are known to have a significantly covalent H−Ng bond and a more ionic Ng−Y bond.8,9,13 The relatively low partial charges on H and Ng and relatively higher partial charges on HNg and Y, computed via natural bond orbital (NBO) analysis, confirm these earlier observations. The NBO charges as well as the Mulliken charges on the HNg moiety calculated at different levels of theory are given in Table 1. These two charge calculation methods were chosen based on literature.19,21 The partial charge on HNg increases and gradually approaches unity as the EA of the ligand Y increases. The charge distribution between the atoms within the BO2 and BF4 moieties in HNgBO2 and HNgBF4 is similar to the partial atomic charges in free BO2− and BF4−. This further corroborates the ionic nature of the Ng−Y bond. The charge-transfer nature of these molecules induces significant dipole moments in the compounds, as can be seen in Table S2 (Supporting Information). The dipole moments increase as Y changes from F to BO2 to BF4. For the insertion compounds synthesized so far, the H−Ng vibrations are particularly intense.8,13 Therefore, the effect of changing Y on the H−Ng stretching frequencies was closely monitored (Table S3, Supporting Information). The H−Ng frequencies in HNgF and HNgBO2 are quite close to one another, while that in HNgBF4 is much larger. The frequencies of HArF and HKrF predicted by B3LYP/6-311++G(2d,2p) are closest to the experimental values. The discrepancy between the theoretically predicted and experimentally determined frequencies can be understood by realizing that the theoretical calculations correspond to single isolated molecules, as opposed to molecules in a solid matrix environment.28 It has also been shown that while matrix effects exist, anharmonic corrections to frequencies are also important to obtain better agreement between theory and experiment.28 To test the stability of these molecules, the energies associated with different fragmentation pathways were calculated. Figure 2 illustrates two possible generic fragmentation paths that might reasonably occur

energy difference of HNgY and HY + Ng, which is expected to be the lowest-energy path, will be lowered compared to the fragmentation energy of HNgX. To test the above hypotheses, a systematic study of HNgY (where Ng = Ar and Kr, and Y = F, BO2, and BF4) molecules was performed. Note, the EAs of F, BO2, and BF4 are 3.4,15 4.46,22 and 6.86 eV,21 respectively (theoretically calculated values are given in Table S1, Supporting Information). The two superhalogens were chosen as they are small in size (to enable faster computations), and their corresponding hydrides, metaboric acid (HBO2), and fluoroboric acid (HBF4) are well characterized. Therefore, synthesis of these HNgY compounds might be feasible using the appropriate Ng and HY starting materials. Insertion compounds were the focus of this study as opposed to association compounds (where the Ng is at one end of the molecule bound by van der Waals interactions as in Ar···BeO or Ar···CuX),24,25 as the former are true molecules of Ng’s, in which the inert gas atom is chemically bound, and the electrons of the Ng are involved in bonding. Alternately, a systematic study of MNgF could also have been performed, where M is an alkali metal atom. However, considering that the ionization potential of alkali metals is very low, the metal atom would have near-unity positive charge, and the fluorine atom would have a near-unity negative charge. There would not be any necessity for the outer electrons of Ng to be involved in bonding. As a result, MNgF might not form. Instead, an association compound, Ng···MF, might result. It should be mentioned here that no compounds of the form MNgF have been reported to date. Moreover, test calculations performed on LiArF resulted in a linear structure with two imaginary frequencies. Upon removal of the imaginary frequencies, the Li atom spontaneously moved to the F end of the molecule over the course of optimization. The MP2/6-311++G(2d,2p)-optimized geometries of HArY and HKrY are presented in Figure 1. HArF and HKrF are both

HNgY → HY + Ng ΔE1 = E(HY) + E(Ng) − E(HNgY)

Figure 1. Optimized geometries for (a) HArF, (b) HKrF, (c) HArBO2, (d) HKrBO2, (e) HArBF4, and (f) HKrBF4; values in parentheses denote bond lengths (Å) computed at the CCSD(T)/augcc-pV5Z level for HArF and at the MP2/aug-cc-pVTZ level for HKrF.8,26

(1)

HNgY → H + Ng + Y ΔE2 = E(H) + E(Ng) + E(Y) − E(HNgY) 3152

(2)

dx.doi.org/10.1021/jz501404h | J. Phys. Chem. Lett. 2014, 5, 3151−3156

The Journal of Physical Chemistry Letters

Letter

Table 1. NBO and Mulliken Charges (e) on the HNg Moiety Calculated at Different Levels of Theory B3LYP 6-311++G(2d,2p)

aug-cc-pVTZ

M06

MP2

6-311++G(2d,2p)

6-311++G(2d,2p)

molecule

NBO

Mulliken

NBO

Mulliken

NBO

Mulliken

NBO

Mulliken

HArF HArBO2 HArBF4 HKrF HKrBO2 HKrBF4

0.710 0.761 0.920 0.727 0.775 0.907

0.591 0.675 0.827 0.560 0.607 0.767

0.719 0.769 0.926 0.723 0.783 0.910

0.659 0.723 0.870 0.661 0.721 0.852

0.743 0.799 0.939 0.762 0.818 0.927

0.633 0.713 0.868 0.592 0.654 0.801

0.860 0.918 0.985 0.845 0.903 0.976

0.750 0.826 0.912 0.703 0.753 0.869

for the molecules studied in this work, and hence, the BSSE at other levels of theory were ignored. The fragmentation along path 1 was expected to be highly exothermic (HNgY compounds characterized so far are metastable), while fragmentation along path 3 was expected to be highly endothermic (separation of molecules into ions requires substantial energy).7 Fragmentation along path 2 should preferably be endothermic for stability of the molecule. It should be noted that all insertion molecules experimentally confirmed have endothermic dissociation along path 2.7 We note from Tables 2 and 3 that this trend is followed for all of the molecules studied with the exception of HArBO2. The fragmentation energy corresponding to pathway 2 is exothermic, signifying that HArBO2 is not stable versus H + Ar + BO2. Irrespective of the sign of the fragmentation energy, these metastable compounds will be protected against dissociation, provided that there is a substantial energy barrier between the intact molecules and their fragments. Therefore, the energy barrier along the bending mode was calculated for HArF, HArBO2, and HArBF4 at the B3LYP/6-311++G(2d,2p) level. The B3LYP/6-311++G(2d,2p) level of theory was chosen as it best reproduces experimental H−Ng frequencies. It also closely matches dipole moments of HArF and HKrF calculated at the CCSD(T)/aug-cc-pV5Z and QCISD/6-311+G(2df,2pd) levels, respectively.30 The energy barrier was computed by carrying out partial optimizations at a fixed HArY angle and then scanning through different angles (Figure 3). The barrier heights were calculated to be 1.30, 0.82, and 0.10 eV, respectively, for X = F, BO2, and BF4. The energy barriers for the krypton-containing compounds showed similar trends (Figure S1, Supporting Information). The barrier heights were

Figure 2. Energetics of fragmentation of HNgY.

We do not know a priori which path is more exothermic, but what matters to the stability of HNgY is the energetic barrier (Eact) between the HNgY species and its fragments. As will be shown, HNgY is metastable as its fragmentation into HY + Ng is exothermic. However, there is a substantial energy barrier that prevents this dissociation. For completeness, we also study the ionization pathway HNgY → HNg + + Y − ΔE3 = E(HNg +) + E(Y −) − E(HNgY)

(3)

For HNgBF4, additional dissociation pathways were studied. The fragmentation energies of these molecules, including zeropoint energy (ZPE) corrections, are given in Tables 2 and 3. The basis set superposition error (BSSE) is usually insignificant for these types of insertion compounds relative to the magnitude of their fragmentation energies.29 Calculations performed at the B3LYP/6-311++G(2d,2p) confirmed this

Table 2. Fragmentation energies of HArY (Y = F, BO2, and BF4) along Different Pathways Including ZPE Corrections Fragmentation Energy, ΔE= Eproduct − Ereactant (eV) B3LYP molecule HArF

HArBO2

HArBF4

dissociation channel HF + Ar H + Ar + F HAr+ + F− HBO2 + Ar H + Ar + BO2 HAr+ + BO2− HBF4 + Ar BF3 + HF + Ar H + Ar + BF4 H + Ar + F + BF3 HAr+ + BF4−

M06

6-311++G(2d,2p)

aug-cc-pVTZ

without ZPE

with BSSE

with ZPE

−5.53 0.50 6.71 −5.25 −0.19 5.15 −4.43 −4.31 1.83 1.72 4.50

−5.54 0.49 6.66 −5.26 −0.21 5.13 −4.52 −4.38 1.80 1.65 4.47

−5.52 0.26 6.62 −5.18 −0.46 5.08 −4.43 −4.34 1.56 1.44 4.42 3153

−5.48 0.30 6.55 −5.16 −0.44 5.07 −4.38 −4.31 1.57 1.47 4.40

MP2

6-311++G(2d,2p) −5.67 0.23 6.48 −5.27 −0.35 4.98 −4.39 −4.25 1.61 1.64 4.44

−6.06 −0.21 6.19 −5.52 −0.18 4.75 −4.50 −4.37 1.43 1.49 4.37

CCSD(T) aug-cc-pVTZ −5.81 0.23 6.41 −5.40 −0.02 4.93 −4.27 −4.10 1.87 1.94 4.52

dx.doi.org/10.1021/jz501404h | J. Phys. Chem. Lett. 2014, 5, 3151−3156

The Journal of Physical Chemistry Letters

Letter

Table 3. Fragmentation Energies of HKrY (Y = F, BO2, and BF4) along Different Pathways Including ZPE Corrections Fragmentation Energy, ΔE= Eproduct − Ereactant (eV) B3LYP molecule HKrF

HKrBO2

HKrBF4

dissociation channel HF + Kr H + Kr + F HKr+ + F− HBO2 + Kr H + Kr + BO2 HKr+ + BO2− HBF4 + Kr BF3 + HF + Kr H + Kr + BF4 H + Kr + F + BF3 HKr+ + BF4−

M06

6-311++G(2d,2p)

aug-cc-pVTZ

without ZPE

with BSSE

with ZPE

−4.78 1.26 6.92 −4.58 0.48 5.28 −3.84 −3.72 2.42 2.31 4.55

−4.80 1.23 6.86 −4.60 0.45 5.25 −3.89 −3.80 2.38 2.23 4.51

−4.76 1.02 6.83 −4.51 0.21 5.20 −3.82 −3.73 2.17 2.05 4.48

1.52, 1.03, and 0.26 eV, respectively. We make two important observations here. First, the barrier heights decrease with increasing size of the superhalogen. As increased EA often entails increased size of the ligand, this poses a limitation to what superhalogen could be used to oxidize a Ng. For example, we can imagine that a very large EA would be required to form a stable compound of He. However, if the size of the ligand is large, it may not favorably bind to He at all. Second, the barrier heights for the HKrY compounds are consistently higher than the corresponding HArY compounds. This is in agreement with the fact that lighter Ng’s are more difficult to stabilize. For a complete picture of the stability of these molecules, energy barriers corresponding to the three-body fragmentation as depicted in pathway 2 need to be computed. Earlier studies have shown that multireference methods such as multireference configuration interaction (MRCI) and complete active space perturbation theory (CASPT2) are essential for accurate description of this dissociation pathway.31 However, these methods become prohibitively expensive with increasing size of the system. To add further complications, different multireference methods yield different results.31 Therefore, these calculations were not attempted. The motivation of this work was to discern the general trends in the stability of HNgY compounds as the EA of Y increases and not to calculate the absolute value of the fragmentation energies of the systems with extreme accuracy. The expected accuracy of these calculations in the regime of density functional theory is within 0.2−0.3 eV, which is acceptable for the purposes of this study. It should be emphasized that though the magnitudes obtained for charge transfer, dipole moments, and ΔE may differ slightly with varying levels of theory, the general trends observed are consistent at all levels of theory. To conclude, this Letter predicts the metastability of hydrides of Ng’s (HNgY) stabilized by superhalogens for Ng = Ar and Kr. Test calculations performed on HArY and HKrY show that as the EA of the ligand, Y, increases, the ionic character of the Ng−Y bond as well as the dipole moment of the molecule increases. These molecules are metastable as they exhibit exothermic fragmentation along the HY + Ng dissociation pathway. The corresponding fragmentation energy is lowered (that is, the molecule is more stabilized) with an increase in EA of the ligand Y. However, the energy barrier to fragmentation decreases as the size of the ligand increases.

−4.68 1.10 6.80 −4.46 0.26 5.21 −3.76 −3.70 2.18 2.08 4.46

MP2

6-311++G(2d,2p) −4.74 1.16 6.77 −4.45 0.48 5.15 −3.63 −3.49 2.37 2.40 4.55

−5.19 0.67 6.50 −4.74 0.60 4.97 −3.87 −3.73 2.07 2.13 4.45

CCSD(T) aug-cc-pVTZ −4.88 1.16 6.75 −4.56 0.82 5.18 −3.59 −3.42 2.55 2.62 4.61

Although this Letter focuses on BO2 and BF4 as potential ligands, the concept of using superhalogens to stabilize Ng hydrides is general and could be extended to other candidates. An important implication of these results is that if the right superhalogen could be identified, chemical compounds of smaller Ng’s such as He and Ne could also be formed. The author hopes that this Letter might motivate experimentalists to attempt to synthesize such compounds.



COMPUTATIONAL METHODS First-principles calculations were performed using Gaussian 0932 available via the corn and barley clusters at Stanford University. The equilibrium structures of neutral and anionic F, BO2, and BF4 were first determined. To establish the superhalogen nature of BO2 and BF4, their adiabatic electron detachment energies (ADEs) were calculated by taking the energy difference between the optimized geometry of the anion and the optimized geometry of the structurally similar corresponding neutral.19 These values are given in Table S1 (Supporting Information). For comparison, literature values are also provided. The geometries of the ground-state HNgY clusters (where Ng = Ar and Kr, and Y = F, BO2, and BF4) were then optimized along with frequency calculations. Absence of any negative (imaginary) frequencies suggested that the structures obtained corresponded to true minima on the potential energy surface. To confirm the charge-transfer nature of these molecules, dipole moments and the charges on the HNg and Y moieties were calculated using NBO analysis. The stability of these clusters was studied by computing the dissociation energies along different pathways. A relatively high level of theory was required to ensure the reliability of the data for these small clusters. Computations were first performed using the B3LYP hybrid functional and the 6-311++G(2d,2p) as well as the aug-cc-pVTZ basis sets. B3LYP was chosen for relative ease of computation, whereas the choice of basis sets was guided by literature on earlier studies of insertion compounds.7,8 Both basis sets gave comparable results, though aug-cc-pVTZ was more computationally expensive. Therefore, using the B3LYP/6-311++G(2d,2p)-optimized structures as initial geometries, further calculations were executed at the MP2/6-311++G(2d,2p) and M06/6-311++G(2d,2p) levels. Apart from B3LYP and MP2, CCSD(T) is also a standard method used to treat these systems.8 However, because 3154

dx.doi.org/10.1021/jz501404h | J. Phys. Chem. Lett. 2014, 5, 3151−3156

The Journal of Physical Chemistry Letters



Letter

ASSOCIATED CONTENT

S Supporting Information *

Adiabatic detachment energies of F, BO2, and BF4, dipole moments of HNgY calculated at different levels of theory, theoretically calculated H−Ng frequencies for HNgY, and potential energy diagram of fragmentation of HKrY into HY + Kr. This material is available free of charge via the Internet at http://pubs.acs.org.



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The author would like to thank Professor Richard N. Zare and Professor Puru Jena for valuable discussions and the Center for Molecular Analysis and Design, Stanford University, for funding.



REFERENCES

(1) Pauling, L. The Formulas of Antimonic Acid and the Antimonates. J. Am. Chem. Soc. 1933, 55, 1895−1900. (2) Bartlett, N. Xenon Hexafluoroplatinate (V) Xe+[PtF6]−. Proc. Chem. Soc. 1962, 218. (3) Tonkyn, R. G.; Winniczek, J. W.; White, M. G. Rotationally Resolved Photoionization of O2+ near Threshold. Chem. Phys. Lett. 1989, 164, 137. (4) Dehmer, P. M.; Dehmer, J. L. Photoelectron Spectrum of Xe2 Van der Waals Molecule. J. Chem. Phys. 1977, 67, 1774−1775. (5) Grochala, W. Atypical Compounds of Gases, Which Have Been Called ‘Noble’. Chem. Soc. Rev. 2007, 36, 1632−1655. (6) Christe, K. O. A Renaissance in Noble Gas Chemistry. Angew. Chem., Int. Ed. 2001, 40, 1419−1421. (7) Khriachtchev, L.; Räsänen, M.; Gerber, R. B. Noble-Gas Hydrides: New Chemistry at Low Temperatures. Acc. Chem. Res. 2009, 42, 183−191. (8) Khriachtchev, L.; Pettersson, M.; Runeberg, N.; Lundell, J.; Räsänen, M. A Stable Argon Compound. Nature 2000, 406, 874−876. (9) Pettersson, M.; Lundell, J.; Räsänen, M. Neutral Rare-Gas Containing Charge-Transfer Molecules in Solid Matrices. I. HXeCl, HXeBr, HXeI, and HKrCl in Kr and Xe. J. Chem. Phys. 1995, 102, 6423−6431. (10) Lundell, J.; Khriachtchev, L.; Pettersson, M.; Räsänen, M. Formation and Characterization of Neutral Krypton and Xenon Hydrides in Low-Temperature Matrices. Low Temp. Phys. 2000, 26, 680−690. (11) Khriachtchev, L.; Isokoski, K.; Cohen, A.; Räsänen, M.; Gerber, R. B. A Small Neutral Molecule with Two Noble-Gas Atoms: HXeOXeH. J. Am. Chem. Soc. 2008, 130, 6114−6118. (12) Lignell, A.; Khriachtchev, L.; Lundell, J.; Tanskanen, H.; Räsänen, M. On Theoretical Predictions of Noble-Gas Hydrides. J. Chem. Phys. 2006, 125, 184514. (13) Pettersson, M.; Khriachtchev, L.; Lignell, A.; Räsänen, M.; Bihary, Z.; Gerber, R. B. HKrF in Solid Krypton. J. Chem. Phys. 2002, 116, 2508−2515. (14) Pettersson, M.; Lundell, J.; Khriachtchev, L.; Räsänen, M. Neutral Rare-Gas Containing Charge-Transfer Molecules in Solid Matrices. III. HXeCN, HXeNC, and HKrCN in Kr and Xe. J. Chem. Phys. 1998, 109, 618−625. (15) Blondel, C.; Delsart, C.; Goldfarb, F. Electron Spectrometry at the μeV Level and the Electron Affinities of Si and F. J. Phys. B: At., Mol. Opt. Phys. 2001, 34, L281−L288.

Figure 3. Potential energy diagram of fragmentation of HArY into HY + Ar.

CCSD(T) optimizations are difficult to perform for five or more atoms, single-point energies were calculated at the CCSD(T)/aug-cc-pVTZ level using the MP2/6-311++G(2d,2p)-optimized structures. The geometry of molecules does not significantly change with method; consequently, the error in calculating the CCSD(T) energy without optimization is not expected to be significant. For similar reasons, larger basis sets such as aug-cc-pVQZ or aug-cc-pV5Z were not used. 3155

dx.doi.org/10.1021/jz501404h | J. Phys. Chem. Lett. 2014, 5, 3151−3156

The Journal of Physical Chemistry Letters

Letter

(16) Bradforth, S. E.; Kim, E. H.; Arnold, D. W.; Neumark, D. M. Photoelectron Spectroscopy of CN−, NCO− and NCS−. J. Chem. Phys. 1993, 98, 800−810. (17) Hotop, H.; Lineberger, W. C. Binding Energies in Atomic Negative Ions. 2. J. Phys. Chem. Ref. Data 1985, 14, 731−750. (18) Gutsev, G. L.; Boldyrev, A. I. DVM-X-α Calculations on the Ionization-Potentials of MX(k+1) Complex Anions and the ElectronAffinities of MXk+1 Superhalogens. Chem. Phys. 1981, 56, 277. (19) Samanta, D.; Wu, M. M.; Jena, P. Unique Spectroscopic Signature of Nearly Degenerate Isomers of Au(CN)3 Anion. J. Phys. Chem. Lett. 2001, 2, 3027−3031. (20) Smuczynska, S.; Skurski, P. Halogenoids as Ligands in Superhalogen Anions. Inorg. Chem. 2009, 48, 10231−10238. (21) Paduani, C.; Wu, M. M.; Willis, M.; Jena, P. Theoretical Study of the Stability and Electronic Structure of Al(BH4)n=1→4 and Al(BF4)n=1→4 and Their Hyperhalogen Behavior. J. Phys. Chem. A 2011, 115, 10237−10243. (22) Zhai, H. J.; Wang, L. M.; Li, S. D.; Wang, L. S. Vibrationally Resolved Photoelectron Spectroscopy of BO− and BO2−: A Joint Experimental and Theoretical Study. J. Phys. Chem. A 2007, 111, 1030−1035. (23) Korobov, M. V.; Kuznetsov, S. V.; Sidorov, L. N.; Shipachev, V. A.; Mitkin, V. N. Gas Phase Negative Ions of Platinum Metal Fluorides. 2. Electron Affinity of Platinum Metal Hexafluorides. Int. J. Mass Spectrom. Ion Phys. 1989, 87, 13. (24) Thompson, C. A.; Andrews, L. Noble Gas Complexes with BeO: Infrared Spectra of Ng-BeO (Ng = Ar, Kr, Xe). J. Am. Chem. Soc. 1994, 116, 423−424. (25) Evans, C.; Gerry, M. C. Noble Gas−Metal Chemical Bonding? The Microwave Spectra, Structures, and Hyperfine Constants of Ar− CuX (X = F, Cl, Br). J. Chem. Phys. 2000, 112, 9363−9375. (26) Chen, Y. L.; Hu, W. P. Rate Constant Calculation for HArF → Ar + HF and HKrF → Kr + HF Reactions by Dual-Level Variational Transition State Theory With Quantized Reactant State Tunneling. J. Phys. Chem. A 2004, 108, 4449−4454. (27) Gutsev, G. L.; Jena, P.; Bartlett, R. J. Structure and Stability of BF3 ∗ F and AlF3 ∗ F Superhalogens. Chem. Phys. Lett. 1998, 292, 289−294. (28) Bihary, Z.; Chaban, G. M.; Gerber, R. B. Vibrational Spectroscopy and Matrix-Site Geometries of HArF, HKrF, HXeCl, and HXeI in Rare-Gas Solids. J. Chem. Phys.. 2002, 116, 5521−5529. (29) Bertolus, M.; Major, M.; Brenner, V. Assessment of Density Functional Theory for Bonds Formed between Rare Gases and OpenShell Atoms: A Computational Study of Small Molecules Containing He, Ar, Kr and Xe. Phys. Chem. Chem. Phys. 2012, 14, 553−561. (30) Yen, S. Y.; Mou, C. H.; Hu, W. P. Strong Hydrogen Bonding between Neutral Noble-Gas Molecules (HNgF, Ng = Ar, Kr, and Xe) and Hydrogen Fluoride: A Theoretical Study. Chem. Phys. Lett. 2004, 383, 606−611. (31) Gerber, R. B. Formation of Novel Rare-Gas Molecules in LowTemperature Matrices. Annu. Rev. Phys. Chem. 2004, 55, 55−78. (32) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, M. J.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, Ö .; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09, revision D.01; Gaussian, Inc.: Wallingford, CT, 2009.

3156

dx.doi.org/10.1021/jz501404h | J. Phys. Chem. Lett. 2014, 5, 3151−3156