A Theoretical Investigation on FRgLF

Subscriber access provided by University of Massachusetts Amherst Libraries

A: New Tools and Methods in Experiment and Theory

Compounds with Rare Gas-Selenium/Tellurium Bonds: A Theoretical Investigation on FRgLF and FRgLF (L= Se and Te, n=1, 3 and 5) n

n-1+

Zhuo Zhe Li, and An Yong Li J. Phys. Chem. A, Just Accepted Manuscript • DOI: 10.1021/acs.jpca.7b12834 • Publication Date (Web): 31 May 2018 Downloaded from http://pubs.acs.org on May 31, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Compounds with Rare Gas−Selenium/Tellurium Bonds: A Theoretical Investigation on FRgLFn and FRgLFn−1+ (Rg = Kr ~ Rn, L= Se and Te, n = 1, 3 and 5) Zhuo Zhe Li, An Yong Li*

School of Chemistry and Chemical Engineering, Southwest University, Chongqing, 400715, P.R.China

Abstract

A new type of interesting insertion compounds FRgLFn (Rg = Kr ~ Rn, L = Se and Te, n = 1, 3 and 5) and ionic FRgLFn−1+ obtained through the insertion of a rare gas atom into the selenium fluorides and tellurium fluorides have been explored theoretically using MP2, CCSD(T), and PBE0 calculations. These predicted species were examined to present the optimized geometries, vibrational modes, molecular properties, thermodynamic and kinetic stabilities and bond nature. The optimized structures are without imaginary frequencies and metastable. In neutral FRgLFn, F−Rg bonds should be of ionic character with large dissociation energy ranging from 150 kcal mol−1 ~ 200 kcal mol−1 that could be best described by F−(RgLFn)+. Rg−L bonds have some covalent character with lower interaction energies within the range of 25 ~ 40 kcal mol−1. In FRgL+ and FRgLF2+, the bonding nature of the F−Rg and Rg−L bonds are somewhat similar to that of the neutral compounds. In FRgLF4+, the F−Rg bond could be of partial covalent type but the Rg−L bond could be considered as an ionic bond.

1

ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 37

Introduction

Rare gas (Rg) elements were no longer perceived as chemically inert since Bartlett

1

firstly synthesized “Xe+PtF6−” in 1962. That could be called a

milestone. In the subsequent decades, new Rg−compounds and bonding motifs are continually being reported, and the Rg chemistry was expanded theoretically and experimentally 2−41. So far, small discovered and predicted Rg molecules or ions can be roughly classified into two types: the “donor–acceptor type” complexes

2−21

and the “inserted type” complexes

22−41

. The saturated electronic

Rg can act as Lewis base to bind with the electron deficient Lewis acid compounds very well through Lewis acid−base interactions. Theoretical studies have contributed immensely to the discovery of different Lewis acid−base Rg 2−3

compounds. The representative species include RgBeO RgSX (X= Be, B+, C2+, N3+, O4+)

4

, RgBeS

5

, RgBeNH/NCN/NBO

RgBeCO3/SO4 7, RgBeHPO42−/ CrO42− 8) and BeH3BeR triangular B3+, square B42+

and its derivatives,

10

9

6

,

etc. Recently, the

, pentagonal B53+ and hexagonal ions B64+

11

have

also been considered to be good Lewis acid to form the BnRgn(n−2)+ compounds with strong covalent Rg−B bonds 12. Furthermore, many efforts were carried to investigate the nature of Rg−M bonds in the rare gas−transition metal complexes RgMX (M represents transition metal element, typically Cu, Ag or Au, X represents halogen element). Over the past two decades, Gerry and coworkers have measured the microwave spectra of ArAgX (X = F, Cl, Br) 13, ArCuX (X = F, Cl, Br) 14, RgAuCl (Rg = Ar, Kr) 15, ArAuX (X = F, Br) 16, KrCuX (X = F, Cl) 17, KrAuF, KrAgF, KrAgBr

18

, XeAuF

19

and XeCuF/Cl

20

, and determined their

geometrical parameters. Lovallo and Klobukowski have calculated the bond 2


Page 3 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


lengths and binding energies of Rg–MX (Rg = Ar, Kr, Xe; M = Cu, Ag, Au; X = F, Cl) at the MP2 level of theory 21.

Since Räsänen and co−workers synthesized a series of inserted HRgY using photodissociation of HY in a Rg matrix and found the ionic−covalent characteristic structure with an ionic bond Rg+Y− and a covalent bond (H−Rg)+, such as HArF

22

, HKrF/Cl and HXeY (Y = OH, SH, OBr, Cl, Br, I, CN)

23−28

,

more efforts towards the novel inserted Rg compounds were currently being explored. These inserted Rg compounds have large electronegativity of halogen elements to stablize the whole Rg compounds. The typical halogen element is fluorine. A few new types of insertion XRgY (X = electronegative atom or group，Y = H, metal, electropositive atom or group) were discovered and studied theoretically and experimentally, for instance, FRgO− Xe)

30

, F−(RgO)n (Rg = He, Ar, Kr, Xe)

29

, FRgBO (Rg = Ar, Kr,

31

, FRgBN− (Rg = Ar, Kr, Xe)

32

,

FRgCCH (Rg = Kr and Xe) 33, FRgCN (Rg = Kr and Xe) 34 and FRgBNR (Rg = Ar, Kr, and Xe; R = H, CH3, CCH, CHCH2, F, and OH)

35

. These compounds

can be considered as a result of Rg atom embedded in the molecules or ions of fluorides XY and usually have considerable binding energies.

In recent decades, Rg inserted fluorides have been received extensive attention. In 2005, Cohen et al focused on the metastable compounds including FArSiF3 and FArCCH 36, and opens a way to a field of organic−argon chemistry. Then Yockel et al

37

theoretically predicted the stability of FKrC/Si/GeF3, and

firstly suggested the existence of the third−row main group forming chemical Rg bonds. Grandinetti 38 and co−workers explored the metastable species of neutral 3



Page 4 of 37

FXeGeFn (n = 1, 3) compounds with Xe–Ge bonds and show thermochemical stability with respect to dissociation into F + Xe + GeFn (n = 1, 3) and kinetic stability with respect to dissociation into Xe + GeFn+1. Chattaraj

39

et al carried

out the structure and stability of FRgSn/PbF3 and FRgSn/PbF (Rg = Kr – Rn) compounds to explore the novel Rg−Sn and Rg−Pb bonds. The natural population analysis reveals that FRgSn/PbF3 and FRgSn/PbF can be best represented as F−(RgSn/PbF3)+ and F−(RgSn/PbF)+. And the Rg–Sn/Pb bonds are covalent in nature. Ghanty 40−41 predicted theoretically a series of netural FRgY compounds with Rg = Kr and Xe, Y= N, P, As, Sb and Bi. The predicted FXeY molecules are found to be thermodynamically stable with respect to all dissociation channels except for the 2−body channel obtaining the global minimum products Xe and FY. The FKrY molecules are thermodynamically unstable with respect to the dissociations into 2FKr + 4Y and 2F + Kr + 4Y according to the calculations by MP2 and CCSD(T) methods. However, all these FRgY molecules are kinetically stable with high S−T barrier heights that can prevent the metastable FRgY molecules from dissociation into the global minima products (Rg + FY).

In the spirit of the previous work of these researchers, we have focused to investigate the interaction of Rg−Se/Te between the heavy rare gas (Kr, Xe, Rn) and the VIA main group element L (selenium and tellurium). This is the first time to show the bonding ability between Te and Rg. The Rg−Se bond has just been reported once by Stefano Borocci 42 since 2008. That idea follows the first found Rg anion FRgO−

29

and the anion analogue FRgS−

43

. The predicted

FRgSe− are similar to FRgO/S− and the FKr/XeSe− are the first predicted rare 4


Page 5 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


gas−selenium metastable species. In this article, Rg atoms are artificially “inserted” in the selenium fluorides and tellurium fluorides, including, LF2, LF4, and LF6 (L= Se and Te) to form the inserted Rg compounds FRgLFn (L= Se and Te, n = 1, 3 and 5). Beyond that, Rg atoms are also inserted in the ionic LF+, LF3+, and LF5+ to form the ionic FRgLFn−1+. This work is more detailed on the bonding ability analysis between Rg and Se/Te than the previous reports. In the present work, we have systematically carried out theoretical calculations using DFT, MP2 and CCSD(T) methods to investigate the structures, energy, charge distributions, characteristic vibrational frequencies and the energy decomposition analysis of the total compounds.

Computational methodology

All the equilibrium structures, harmonic vibrational frequencies were computed as the characterization of the stationary states by the density functional theory PBE0

44−45

and Møller−Plesset theory (MP2)

46

in combination with

def2−nZVPPD (n=T and Q) 47−48 basis set designed by Weigend and Ahlrichs. Moreover, FRgLF were also carried out by CCSD(T) 49 /def2−TZVP 50 level to ensure that the obtained minima are not an artifact of the MP2 level. For Xe, Te and Rn atoms, a quasi−relativistic pseudopotential is used with 28, 28 and 60 core electrons, respectively. In order to analyze the stability of the neutral FRgLFn, five possible dissociation channels were designed: (1) FRgLFn → F + Rg + LFn, (2) FRgLFn →F− + RgLFn+, (3) FRgLFn →FRgLFn−1+ + F−, (4) FRgLFn → RgF2 + LFn−1 and (5) FRgLFn → Rg + LFn+1. The ZPE corrected dissociation energy (D0), dissociation enthalpy (∆H) and Gibbs free energy change (∆G) for these dissociation channels were computed at the MP2/def2−TZVPPD level. 5



Page 6 of 37

Additionally, to improve the accuracy of the dissociation energy ∆E, the single point

energy

calculations

of

the

whole

work

were

performed

by

CCSD(T)/def2TZVPPD //MP2/def2−TZVPPD level. All the calculations have been carried out using the Gaussian09 program 51.

The natural atomic charges and Wiberg bond indices (w) have been calculated by NBO5.0 program

52−53

. The energy decomposition analysis (EDA)

were performed with the ETS−NOCV (extended transition state method with the natural orbitals for chemical valence theory) charge and energy decomposition scheme at the revPBE−D3/TZ2P level using the ADF (2013.01) 54−55 program. All the above analyzes were based on configurations optimized at the MP2/def2−TZVPPD level.

Results and Discussion

Structure and stability

The optimized equilibrium geometries and detailed structural parameters of FRgLFn and FRgLFn−1+ (n = 1, 3 and 5, L = Se and Te) using MP2/def2−TZVPPD method are depicted in Fig. 1. Those parameters obtained at the PBE0/def2− TZVPPD, PBE0/def2−QZVPPD and CCSD(T)/def2−TZVP levels are listed in supporting materials Tables S1~ S3, respectively. As shown in Fig. 1, FRgL+ has obvious linear C∞v symmetry. FRgLF, FRgLF3 and FRgLF2+ exhibit Cs symmetry. Unlike FRgSnF3, FRgPbF3 39 and FXeGeF3 38 with a stable C3v structure, the C3v FRgLF3 structures for L = Se/Te are unstable that have two large imaginary frequencies (nearly −100cm−1 for Se−series and −62cm−1 for Te−series) and should be saddle points. The monovalent Tellurides FRgTeF4+ of Rg−Xe/Rn 6


Page 7 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


exhibit two stable structures with C2v and C3v symmetries respectively; the C3v structures are shown in Fig. S1; the C2v FRgTeF4+ is more stable than the C3v FRgTeF4+. FKrTeF4+ and FRgSeF4+ (Rg = Kr, Xe and Rn) have only the structure of C2v symmetry. In the following sections we will discuss the C2v symmetric FRgLF4+. Moreover, FRgLF5 has a perfect C4v symmetry. To facilitate discussion below, in FRgLF and FRgLF2+, the F atoms bonded with L are named F′; in FRgLF3 and FRgLF4+, the F atoms bonded with L are named F' and F'' respectively for the ones in or out of the symmetry plane; in FRgLF5, the F atom bonded with L and in the symmetry axis is called F' and the others are called F''.

FRgLF

Cs

FRgLF3

Cs

FRgLF5 C4v

FRgL+ C∞v

FRgLF2+

Cs

FRgLF4+ C2v

Fig. 1 Optimized geometrical parameters (Å) of FRgLFn (n = 1, 3 and 5, L = Se and Te) and FRgLFn−1+ calculated by MP2/def2−TZVPPD level. Se−compounds values are given out of parentheses and Te−compounds values are given in parentheses.

7



Page 8 of 37

For the neutral FRgSeF, FRgSeF3 and FRgSeF5, F−Rg bond lengths have been found to be in the range 2.01 ~ 2.03Å for Kr, 2.06 ~ 2.10Å for Xe and 2.11 ~ 2.16Å for Rn, respectively. In general, these bond lengths are longer than the sum of the covalent radii of F and Kr/Xe/Rn (the referenced covalent radii

56−57

are

1.81Å for F−Kr, 1.95Å for F−Xe and 2.06Å for F−Rn, respectively). For FRgTeFn, F−Rg bonds are a bit longer than those in FRgSeFn (except for FKrTeF), and certainly they are also longer than the referenced covalent radii. Note that both the distances of F−Kr and F−Xe bonds in FKrLF and FKrLF3 are longer than in FRgCCH (1.96Å for F−Kr and 2.07Å for F−Xe), FRgCN (1.93Å for F−Kr and 2.04Å for F−Xe) and FRgNC (1.89Å for F−Kr and 1.99Å for F−Xe). By comparison, the calculated F−Xe bonds in FXeLF5 are similar to the bond length values of the corresponding moiety in FXeCCH. In addition, the F−Rg bond lengths in FRgSeF calculated by MP2/def2−TZVPPD method are a bit longer comparing with the CCSD(T)/def2−TZVP method, whereas those in FKr/XeTeF seem to be shorter. The F−Rg distances for the ionic FRgLFn−1+ are visibly smaller than those in the neutral FRgLFn. For Se−compounds, those values are 1.84 ~ 1.97Å for Kr, 1.90 ~ 1.98Å for Xe and 1.98 ~ 2.05Å for Rn, respectively. For a given Rg atom in Te−compounds, the calculated F−Rg bond lengths are also longer than those in the Se−compounds. However, these F−Rg bonds are close to the sum of the covalent radii of F and Rg atoms, meaning that the F−Rg bonds in ionic FRgLFn−1+ could be covalent bonds. To make a comparison with the previous reports

42

, the

calculated F−Kr/Xe bonds in our research are obviously shorter than the reported

8


Page 9 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


bond lengths in FRgSe− (2.345Å and 2.338Å for F−Kr and F−Xe at MP2/aug−cc−pVDZ, respectively).

The Rg−Se distances of FRgSeFn are about 2.51 ~ 2.59Å, 2.61 ~ 2.67Å and 2.68 ~ 2.74Å for Kr, Xe and Rn, respectively. Those exceed the covalent bond lengths (the referenced covalent bond lengths are 2.33, 2.47 and 2.58Å for Kr−Se, Xe−Se, and Rn−Se bonds, respectively). These obtained Rg−Se bonds are also shorter than the reported values of FRgSe− (2.48 and 2.59Å for Kr−Se and Xe−Se at MP2/aug−cc−pVDZ) 42. Moreover, the calculated Rg−Te bonds are beyond the referenced covalent bond lengths between Rg and Te (2.53, 2.67 and 2.87Å for Kr−Te, Xe−Te, and Rn−Te, respectively). In ionic FRgLFn−1+ and for a given L, Rg−L bond lengths are significantly longer than those in the neutral FRgLFn and are certainly longer than the corresponding referenced covalent bond lengths. In addition, the L−F' bond lengths are 1.6~1.8Å and 1.8~2.0Å in Se−series and Te−series, respectively. In FRgLF3, FRgLF5 and FRgLF4+, L−F' are shorter than L−F'' indicating that L−F' should be slightly more stable than L−F''. All the bond angles in FRgLFn and FRgLFn−1+ (n = 1, 3 and 5, L = Se and Te) calculated by MP2/def2−TZVPPD are listed in Table S4. The F–Rg–L moieties in the FRgL+, FRgLF4+ and FRgLF5 compounds are perfectly linear where the F–Rg–L angles are already 180 degrees. In FRgLF and FRgLF3, the F–Rg–L angles are 1.4~2.0°, 1.8~3.0° less than 180.0 degrees, respectively. In FRgLF2+ the F–Rg–L angle is even smaller. The Rg–L–F arrangement is predicted above 90° for the most species except for the Rg–L–F'' in FRgLF3 (that is 4~7° below 90°). Generally, for given Rg and n, FRgTeF1/3 and FRgTeF2+ have less F–Rg–L and Rg–L–F angles than FRgSeF1/3 and FRgSeF2+, respectively. However, for 9



Page 10 of 37

FRgSeF4+ and FRgTeF4+, the results may be reverse (except that Rg–L–F'' in FKrTeF4+ is smaller than that in FKrSeF4+).

Table 1 ZPE corrected 0K dissociation energy (D0, kcal mol−1) by the MP2/def2−TZVPPD,

the

dissociation

energy

(∆E,

kcal

mol−1)

at

the

CCSD(T)/def2−TZVPPD// MP2/def2−TZVPPD level, dissociation enthalpy (∆H, kcal mol−1) and Gibbs free energy change (∆G, kcal mol−1) by the MP2/def2−TZVPPD level for different dissociation channels of FRgSeFn (n = 1, 3 and 5).

FKrSeF FXeSeF FRnSeF FKrSeF3 FXeSeF3 FRnSeF3 FKrSeF5 FXeSeF5 FRnSeF5

(1)

∆E D0 ∆H ∆G

−3.23 2.53 3.44 −11.42

20.86 26.80 27.69 12.77

32.33 39.22 40.06 25.23

−6.88 −0.11 0.70 −16.06

16.00 21.78 22.63 5.49

27.26 33.88 34.66 17.62

−4.27 4.89 5.61 −11.17

20.47 27.93 28.68 11.57

32.42 40.73 41.40 24.45

(2)

∆E D0 ∆H ∆G

151.19 149.37 149.81 141.96

161.00 159.04 159.47 151.56

164.74 163.25 163.60 155.80

108.90 107.83 108.29 99.65

129.01 126.69 127.16 118.40

138.60 136.75 137.15 128.63

152.39 152.83 153.17 145.23

169.55 167.71 168.01 160.18

175.71 173.97 174.20 166.61

(3)

∆E D0 ∆H ∆G

257.72 263.67 264.42 257.58

247.52 254.53 255.29 248.30

242.99 249.92 250.67 243.65

207.77 208.17 208.87 200.26

200.14 200.60 201.32 192.48

197.25 197.35 198.06 189.21

228.20 229.69 230.66 220.11

213.35 213.82 214.97 203.43

208.97 209.13 210.25 198.72

∆E D0 ∆H ∆G

84.31 94.25 94.49 89.13

70.53 82.59 82.85 77.35

67.87 79.99 80.27 74.71

34.89 37.21 37.19 27.41

19.90 23.16 23.24 12.99

17.02 20.24 20.33 10.03

10.16 12.35 12.52 0.93

−2.97 −0.55 −0.29 −12.30

−5.15 −2.76 −2.51 −14.52

∆E D0 ∆H ∆G

−87.55 −87.82 −87.74 −94.87

−63.46 −63.55 −63.49 −70.68

−51.99 −51.13 −51.13 −58.22

−99.94 −98.96 −99.27 −105.60

−77.06 −77.07 −77.34 −84.05

−65.80 −64.98 −65.31 −71.92

−102.26 −97.98 −98.56 −103.47

−77.52 −74.94 −75.49 −80.74

−65.57 −62.13 −62.77 −67.86

(4)

(5)

(1) ~ (5) represents the dissociation process of FRgSeFn → F + Rg + SeFn, FRgSeFn → F− + RgSeFn+, FRgSeFn → FRgSeFn−1+ + F−, FRgSeFn → RgF2 + SeFn−1 and FRgSeFn → Rg + SeFn+1, respectively. 10


Page 11 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 2 ZPE corrected 0K dissociation energy (D0, kcal mol−1) by the MP2/def2−TZVPPD,

the

dissociation

energy

(∆E,

kcal

mol−1)

at

the

CCSD(T)/def2−TZVPPD// MP2/def2−TZVPPD level, dissociation enthalpy (∆H, kcal mol−1) and Gibbs free energy change (∆G, kcal mol−1) by the MP2/def2−TZVPPD level for different dissociation channels of FRgTeFn (n = 1, 3 and 5).

FKrTeF FXeTeF FRnTeF FKrTeF3 FXeTeF3 FRnTeF3 FKrTeF5 FXeTeF5

(1)

(2)

(3)

(4)

(5)

∆E D0 ∆H ∆G ∆E D0 ∆H ∆G ∆E D0 ∆H ∆G ∆E D0 ∆H ∆G ∆E D0 ∆H ∆G

FRnTeF5

−2.39 3.93 4.75 −9.90

19.06 27.17 27.96 13.27

31.05 39.33 40.07 25.45

−3.49 3.77 4.61 −13.24

17.28 25.50 26.32 8.31

29.12 37.19 37.95 20.02

−2.55 6.04 6.72 −9.87

20.86 29.75 30.47 12.72

34.25 42.87 43.52 26.69

137.51 135.22 135.65 127.76

148.50 146.97 147.38 139.48

153.78 152.38 152.76 144.94

106.12 104.56 105.01 96.52

124.02 122.64 123.06 114.61

133.41 131.88 132.24 123.99

150.19 150.74 150.95 143.70

165.44 165.00 165.26 157.73

172.29 171.60 171.83 165.22

242.23 244.75 245.49 238.62

235.75 239.92 240.58 233.92

232.13 236.28 236.93 230.28

211.53 211.19 211.99 202.83

207.51 207.14 207.75 199.19

204.85 204.53 205.13 196.58

237.71 236.58 237.47 227.80

224.25 223.67 224.68 213.34

220.43 219.51 220.51 209.96

84.62 92.62 92.74 87.71

69.11 79.93 80.07 74.92

66.05 77.07 77.23 72.00

53.62 55.98 55.83 46.67

37.44 41.78 41.66 32.25

34.23 38.45 38.34 28.87

17.51 17.37 17.30 7.13

3.98 5.15 5.17 −6.25

2.32 3.25 3.27 −7.37

−90.67 −90.69 −90.63 −97.29

−69.23 −67.45 −67.42 −74.12

−57.24 −55.28 −55.31 −61.93

−105.90 −104.91 −105.13 −111.68

−85.13 −83.19 −83.41 −90.13

−73.29 −71.49 −71.78 −78.42

−105.55 −101.32 −101.83 −105.85

−82.13 −77.61 −78.08 −83.26

−68.74 −64.49 −65.03 −69.29

(1) ~ (5) represents the dissociation process of FRgTeFn → F + Rg + TeFn, FRgTeFn → F− + RgTeFn+, FRgTeFn → FRgTeFn−1+ + F−, FRgTeFn → RgF2 + TeFn−1 and FRgTeFn → Rg + TeFn+1, respectively.

To investigate the stability of FRgLFn and FRgLFn−1+, five possible dissociation channels are considered. Table 1 and 2 list the ZPE corrected 11



Page 12 of 37

dissociation energy (D0), dissociation enthalpy (∆H) and free energy change (∆G) (at gas phase, 298.15 K and 1atm) for FRgSeFn and FRgTeFn that were obtained at the MP2/def2−TZVPPD method, respectively. Furthermore, to improve the accuracy of energy, the dissociation energy (∆E) were also calculated at the CCSD(T)/def2−TZVPPD // MP2/def2−TZVPPD level that were put in the first row of Table 1 and 2. The dissociation path (1) results in three parts, F, Rg and LFn. It is easily seen that the ∆E for FKrLFn are small negative values and grow steadily from Kr to Rn. For FRgLF, the corresponding ∆E values for these channels are 3 ~ 5 kcal mol−1 or 1 ~ 2 kcal mol−1 larger than FRgSeF3 and FRgTeF3, respectively. FRgLF3 seem to have the lowest ∆E values. In the case of FRgLF5, the dissociation energy (∆E = −4.27 ~ 32.42 kcal mol−1 for FRgSeF5 and ∆E = −2.55 ~ 34.25 kcal mol−1 for FRgTeF5) are 0.5 ~5 kcal mol−1 larger than FRgLF3 with the same Rg and L. Here the regularity of D0 is consistent with that of ∆E, however, D0 calculated by MP2 method overestimates approximately 5 ~ 9 kcal mol−1 with respect to ∆E that was by CCSD(T) level for the whole Se and Te systems. Note that the corresponding ∆E and D0 values for these 3B dissociation channels in the FRgSeFn cases are slightly lower than those in the FRgTeFn cases, whereas FXe/RnSeF have larger ∆E than FXe/RnTeF. Except for FKrLF that those dissociation processes are shown to be endothermic and spontaneous, other compounds including FXe/RnLF, FRgLF3 and FRgLF5, the three−body (3B) dissociation processes are endothermic but non−spontaneous. Comparisons with the previously reported FKrCN, FKrNC (calculated at CCSD(T)/def2−TZVPPD level)

and

FKrCCH

(calculated

at

CCSD(T)/aug−cc−PVTZ−PP//MP2/aug−cc−PVTZ−PP level), the dissociation 12


Page 13 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


energy values of FKrLFn are slightly more negative but those values of FXeLFn are notably more positive. The dissociation pathway (2) is an ionic channel, resulting in F− and accompanying the breakage of the F−Rg bond, and has a significantly high D0 and ∆E values. The calculated dissociation results increase from Kr to Rn, but decrease from FRgLF5, FRgLF to FRgLF3. The ∆E of FRgLF are shown to be 20 ~ 43 kcal mol−1 larger than FRgLF3. For FRgLF5 compounds, FRgSeF5 are 1 ~ 11 kcal mol−1 higher than FRgSeF. However, those of FRgTeF5 are even more that the difference between FRgTeF5 and FRgTeF is 12 ~ 20 kcal mol−1. For FKrSeF5 and FKrTeF5, ∆E values are 0.4 ~ 0.5 kcal mol−1 lower than D0. But beyond that, ∆E are 0.5 ~ 3 kcal mol−1 higher than D0. In addition, all these ionic dissociation ways are endothermic and non−spontaneous meaning that FRgLFn are thermodynamic stable through this path. The path (3) is also an ionic dissociation channel, breaking L−F bond and resulting in F−. It is even highly endothermic and non−spontaneous and has significantly larger energy change than the path (2).

The two−body dissociation channels (4) of FRgLFn produce RgF2 and L or LF2. Here we can see that all the results including D0, ∆E, ∆H and ∆G values decreases dramatically from Kr to Rn and from FRgLF, FRgLF3 to FRgLF5, respectively. For FXeTeF5 and FRnTeF5, the dissociation energy are several kcal mol−1. However, for FXeSeF5 and FRnSeF5, these dissociation energy are almost small negative values. For FRgLFn (n = 1 and 3) and FKrLF5, these two−body dissociation channels are endergonic and non−spontaneous in nature. However, 13



Page 14 of 37

for FXe/RnSeF5 the dissociation processes are exothermal and spontaneous. Those for FXeTeF5 and FRnTeF5 are also spontaneous but endothermic.

The two−body dissociation path (5), which FRgLFn decompose into Rg and LFn+1 are exergonic and spontaneous for all compounds. Here the dissociation energy absolute values decrease from Kr to Rn. FRgLF have the highest dissociation energy values that are −51 ~ −88 kcal mol−1 and −55 ~ −91 kcal mol−1 for FRgSeF and FRgTeF, respectively. The ∆E for FRgLF3 are 12 ~14 kcal mol−1 and 15 ~ 16 kcal mol−1 lower than those for FRgLF, respectively. Moreover, the ∆E for FKr/XeSeF5 are 0.5 ~ 2.5 kcal mol−1 lower than those for FKr/XeSeF, respectively. However, for FRnSeF5 and FRgTeF5, the results show opposite laws. All these D0, ∆E, ∆H and ∆G of FRgTeFn are several kcal mol−1 smaller than those of FRgSeFn. And in comparison with dissociation processes of FRgCN, FRgNC and FRgCCH producing into Rg + R (R represents CN, NC and CCH), FRgLFn have much more negative dissociation energy with respect to dissociation into Rg + LFn.

Nature of bonding

Table 3 Natural atomic charges q (e) of Rg, Se and the F atom attached to Rg, Wiberg Bond Indices w and ionic component ion% of the bonds in Se−compounds calculated at the MP2/def2−TZVPPD level

14


Page 15 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


qF

qRg

qSe

wF−Rg

wRg−Se

FKrSeF

−0.742

0.638

0.680

0.23

0.75

93.5%

24.6%

FXeSeF

−0.776

0.874

0.488

0.25

0.84

92.6%

67.3%

FRnSeF

−0.795

0.972

0.414

0.24

0.84

92.3%

9.9%

FKrSeF3

−0.689

0.669

1.890

0.27

0.66

91.4%

23.2%

FXeSeF3

−0.736

0.936

1.709

0.30

0.74

89.1%

5.3%

FRnSeF3

−0.759

1.038

1.646

0.29

0.73

90.1%

15.3%

FKrSeF5

−0.658

0.712

2.704

0.30

0.53

88.7%

34.3%

FXeSeF5

−0.708

0.982

2.583

0.34

0.58

62.4%

35.5%

FRnSeF5

−0.730

1.086

2.543

0.33

0.56

87.3%

6.6%

FKrSe+

−0.453

0.845

0.609

0.52

0.61

37.8%

76.0%

FXeSe+

−0.606

1.114

0.492

0.48

0.71

58.2%

67.7%

FRnSe+

−0.653

1.211

0.442

0.44

0.75

63.4%

51.4%

FKrSeF2+

−0.432

0.808

1.662

0.51

0.53

77.9%

31.4%

FXeSeF2+

−0.571

1.104

1.520

0.52

0.55

54.7%

49.7%

FRnSeF2+

−0.614

1.192

1.481

0.50

0.53

83.8%

48.2%

FKrSeF4+

−0.300

0.988

2.462

0.73

0.28

18.9%

91.3%

FXeSeF4+

−0.458

1.259

2.399

0.71

0.23

41.1%

92.7%

FRnSeF4+

−0.512

1.323

2.393

0.67

0.23

46.9%

92.5%

ion%F−Rg ion%Rg−Se

To explore the bonding nature of FRgLFn and FRgLFn−1+, the charge distribution, Wiberg Bond Indices (w) and the ionic component of the chemical bonds based on the natural resonance theory (NRT) using MP2/ def2−TZVPPD 15



Page 16 of 37

level were performed. The results of the Se−compounds are listed in Table 3 and Table S5 and those of the Te−compounds are shown in Table S6. For neutral FRgLF, FRgLF3 and FRgLF5, the F atom attached to Rg contains a more negative charge than the F attached to L. Rg atoms (qRg) possess positive charges that increase from 0.6 to 1.1e in FRgSeFn and from 0.5 to 1.0 e in FRgTeFn, respectively, as summarized in Table 3 and S6. L possesses a positive charge that decreases slightly from Kr to Rn but increases with the number of the F atoms. For a given number of the F atoms, Rg possesses a more positive charge in Se−compounds than in Te−compounds, whereas Te has a larger positive charge than Se. Thus electron density of both the Rg and Se/Te atoms is transferred into the F atoms.

Following the charge distribution, for fixed Rg atoms, wF−Rg increase from FRgLF, FRgLF3 to FRgLF5. However, wF−Rg are less than 0.5 (0.2~0.3) and the ionic components calculated using NRT are larger than 60% that further confirm the ionic character of the F−Rg bond. So these systems could be represented as F− (RgLFn)+. Note that the F−Xe bonds are slightly stronger than F−Kr/Rn bonds with 0.01~0.05 more w values. In contrast, wRg−L decrease slightly from n = 1, 3 to 5. In FRgLF, wRg−L values are within the range from 0.7 to 0.9. Those are 0.6 ~ 0.8 and 0.5 ~0.7 in FRgLF3 and FRgLF5, respectively. Thus the Rg−L bonds have some covalent character with considerable w and small ionic components. These results are similar or consistent with the previous reports about FRgSn/PbF3

and

FRgSn/PbF.

In

comparision

with

Te−compounds, 16


Page 17 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Se−compounds have slightly stronger F−Rg bonds and weaker Rg−L bonds (except that wRg−Te in FRgTeF are smaller than wRg−Se in FRgSeF). The charge distributions of ionic FRgLFn−1+ are much different from the neutral FRgLFn. In FRgLFn−1+, F attached to Rg possesses −0.4 ~ −0.7e that is 0.2~0.3e less than the neutral FRgLFn. However, Rg contains 0.8~1.3e positive charge that is 0.15~0.3e larger than FRgLFn. It seems that the bonding properties of FRgLFn and FRgLFn−1+ are different from each other. For FRgL+ and FRgLF2+, wRg−L are close to 0.5 (about 0.5 ~ 0.7) and are a little larger than wF−Rg. The ionic component of F−Rg bond is greater than Rg−L bond (except for FKr/XeSe+), thus the F−Rg bond may have more ionic character and the Rg−L bond may have more covalent character. So these two ionic systems FRgL+ and FRgLF2+ are somewhat similar to the neutral compounds. For FRgLF4+, the cases are reverse. Here wF−Rg are remarkably large (nearly 0.6 to 0.7) and wRg−L are small (about 0.2~0.3). Table 3 shows that F−Rg bond has a relatively small ionic component (0.9). Thus the F−Rg bond in the ionic systems FRgLF4+ might have covalent character whereas the Rg−L bond can be considered as an ionic bond. It is noted that for FRgLFn−1+ with the increasing n, the bonding ability between F and Rg enhances but the Rg−L bonding ability weakens.

17



Page 18 of 37

Table 4 Energy decomposition (kcal mol−1) characterizing the F−Rg and Rg−Se bonds of FRgSeFn (n = 1, 3 and 5) calculated at the revPBE−D3/TZ2P//MP2/def2−TZVPPD level.

Systems

Fragments

∆Epauli

∆Εelstat

∆Εorb

∆Εdis

∆Εint

FKrSeF

F−···KrSeF+

124.76

−184.00

−107.16

−0.13

−166.53

FXeSeF

F−···XeSeF+

145.47

−209.66

−103.48

−0.11

−167.78

FRnSeF

F−···RnSeF+

149.32

−219.48

−98.93

−0.10

−169.19

FKrSeF3

F−···KrSeF3+

119.97

−176.84

−111.34

−0.15

−168.35

FXeSeF3

F−···XeSeF3+

153.87

−213.73

−112.81

−0.12

−172.78

FRnSeF3

F−···RnSeF3+

158.55

−225.70

−108.18

−0.11

−175.44

FKrSeF5

F−···KrSeF5+

124.55

−185.35

−120.77

−0.16

−181.72

FXeSeF5

F−···XeSeF5+

161.50

−225.41

−122.82

−0.13

−186.85

FRnSeF5

F−···RnSeF5+

166.29

−238.52

−117.71

−0.11

−190.05

FKrSeF

FKr···SeF

125.87

−45.11

−114.20

−0.75

−34.19

FXeSeF

FXe···SeF

183.22

−81.57

−138.62

−0.77

−37.74

FRnSeF

FRn···SeF

178.79

−73.62

−140.70

−0.84

−36.37

FKrSeF3

FKr···SeF3

97.58

−29.42

−91.97

−1.77

−25.57

FXeSeF3

FXe···SeF3

140.91

−45.46

−121.57

−1.78

−27.90

FRnSeF3

FRn···SeF3

153.21

−51.70

−128.09

−1.93

−28.50

FKrSeF5

FKr···SeF5

82.59

−21.37

−87.97

−0.88

−27.62

FXeSeF5

FXe···SeF5

121.07

−32.80

−116.20

−2.55

−30.49

FRnSeF5

FRn···SeF5

130.86

−37.36

−122.74

−2.75

−31.99 18


Page 19 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The energy decomposition analysis (EDA) was performed with the ETS−NOCV (extended transition state method with the natural orbitals for chemical valence theory) charge and energy decomposition scheme at the revPBE−D3/TZ2P//MP2/def2−TZVPPD level. The interaction energy ∆Eint is decomposed into four components: the destabilizing Pauli repulsion ∆EPauli accounting for the repulsive Pauli interaction between occupied orbitals of the fragments, the electrostatic energy ∆Eelstat corresponding to the classical electrostatic interaction between the fragments as they are brought to their positions in the final molecule, the stabilizing orbital interaction ∆Eorb representing the interactions between the occupied MOs of one fragment and the unoccupied MOs of the other fragment, the dispersion energy ∆Edisp representing the dispersion energy correction toward the total attraction energy. To give a clear demonstration of bonding nature, the results of FRgLFn and FRgLFn−1+ will be discussed separately. Those for FRgSeFn and FRgSeFn−1+ are shown in Table 4 and 5, respectively, and those for FRgTeFn and FRgTeFn−1+ reveal similar properties are listed in Table S7 and S8, respectively. The first part in Table 4 listed that FRgLFn are dissociated into F− and (RgLFn)+ on the basis of the NBO analysis. As expected from the ionic character of F−Rg bonds, ∆Eelstat contributes dominantly, ranging from 60 to 70% towards the total attractive interaction. The next important contribution comes from ∆Eorb ranging from 30% ~ 40% in the whole neutral systems. The dispersion ∆Edis (

A Theoretical Investigation on FRgLF

Recommend Documents