Selenium(IV) fluoride and oxofluoride anions

Article abstract of DOI:10.1016/j.jfluchem.2010.04.009

The [((C₆H₅)₃P)₂N]⁺, [(C₆H₅)₄P]⁺ and [N(CH ₃)₄]⁺ salts of SeF₅^-, SeF₆^2- and SeO

Full text of DOI:10.1016/j.jfluchem.2010.04.009

Journal of Fluorine Chemistry 131 (2010) 791–799
Contents lists available at ScienceDirect
Journal of Fluorine Chemistry
journal homepage: www.elsevier.com/locate/fluor
Selenium(IV) fluoride and oxofluoride anions
b
a,
a
b
Karl O. Christe a, , David A. Dixon , Ralf Haiges , Mathias Hopfinger , Virgil E. Jackson ,
*
*
Thomas M. Klapo¨tke c, , Burkhard Krumm , Matthias Scherr
c
c
*
a Loker Research Institute and Department of Chemistry, University of Southern California, Los Angeles, CA 90089-1661, USA
b Department of Chemistry, University of Alabama, Tuscaloosa, AL 35487-0336, USA
c Department of Chemistry, Ludwig-Maximilian University of Munich, Butenandtstr. 5-13(D), D-81377 Munich, Germany
A R T I C L E I N F O
A B S T R A C T
2ꢀ
ꢀ
The [((C6H5)3P)2N]+, [(C6H5)4P]+ and [N(CH3)4]+ salts of SeF5ꢀ, SeF6 and SeOF3 and CsSeO2F were
prepared and characterized. Crystal structures were obtained for [((C6H5)3P)2N][ꢀSeF5] and
[((C6H5)3P)2N][SeOF3] CH2Cl2. In contrast to oxygen-bridged dimeric TeOF3ꢀ, the SeOF3 anion in
[((C6H5)3P)2N][SeOF3] CH2Cl2 is monomeric and represents the first experimentally well determined
molecular structure of a monomeric trifluoro-chalcogenite anion. Similarly, [((C6H5)3P)2N][SeF5]
represents the first example of a structure containing a well-isolated undistorted SeF5ꢀ anion. The NMR
and the vibrational spectra and their assignments were re-examined and corrected by comparison with
highꢀ-level theoretical calculatiꢀons. Whereas the previously published normal coordinate analysis of
SeF5 is correct, that for SeOF3 needs major revision.
Article history:
Received 13 March 2010
Received in revised form 15 April 2010
Accepted 20 April 2010
Available online 28 April 2010
Keywords:
Selenium fluorides
Selenium oxofluorides
Multinuclear NMR spectroscopy
X-ray crystallography
Theoretical calculations
ß 2010 Elsevier B.V. All rights reserved.
Dedicated to Prof. Dr. Henry Selig on the
occasion of his ACS Fluorine Award.
1. Introduction
of selenium tetrafluoride with one and two equivalents of silver
fluoride, respectively, to form the corresponding silver salts. The
The properties of the binary selenium(IV)–fluorine compounds
subsequent reactions with equimolar amounts of [PNP]Cl resulted
in the formation of insoluble AgCl and the corresponding PNP-salts.
The single crystals of [PNP][SeOF3]ꢁCH2Cl2 were obtained by the
hydrolysis of [PNP][SeF5] in CH2Cl2 solution with small amounts of
adventitious water.
The selenium fluorides A[SeF5], A2[SeF6] and A[SeOF3] (A = tet-
ramethyl ammonium, TMA+, tetraphenyl phosphonium, TPP+, and
PNP+) were prepared (at USC) by reaction of stoichiometric
amounts of SeF4 and SeOF2, respectively, with the corresponding
poly-bifluorides A[HF2ꢁx HF] in acetonitrile solution. The SeF4 was
prepared in high yield and purity by the reaction of SeO2 with ClF3
in HF solution at room temperature, and the SeOF2 by con-
proportionation of SeO2 and SeF4 at room temperature. The
Cs[SeO2F] salt was prepared from CsF and H2SeO3 in aqueous
solution by a literature method [16].
ꢀ
2ꢀ
SeF4 [1–5], SeF5 [6,7], and SeF6 [8,9] have been studied for a
number of years. The oxygen containing selenium(IV) fluorides
SeOF2 [10–14], SeO2Fꢀ [15–19], and SeOF3ꢀ [20–22] have also been
ꢀ
studied. A calculated structure of SOF3 [23] and the molecular
ꢀ
structure of the TeOF3 dimer, [Te2O6F2]2ꢀ, [24,25] have been
ꢀ
reported, but the selenium analogue SeOF3 had only been
ꢀ
characterized by vibrational spectroscopy. Furthermore, for SeF5
,
, only the crystal structure of a tetrameric anion is known [6],
which was strongly distorted from the ideal C4v geometry
predicted [26–28] for the free anion. In this paper, we present a
reinvestigation of the above selenium(IV) compounds and also the
ꢀ
crystal structures of the monomeric SeF5 anion and the
ꢀ
trifluoroselenite anion, SeOF3
.
2. Results and discussion
2.2. NMR spectra
2.1. Synthesis
All compounds were characterized by 19F and 77Se NMR
spectroscopy (Table 1). At room temperature, the 19F and 77Se NMR
resonances were generally broad and it was not possible to observe
The selenium fluorides [PNP][SeF5] and [PNP]2[SeF6],
(PNP+ = [((C6H5)3P)2N]+), were synthesized (at LMU) by reaction
ꢀ
any F–F couplings. For SeF5 in either DMSO or CH3CN solution,
* Corresponding authors.
two 19F resonances were observed at about 18 and 57 ppm with an
area ratio of 4:1. These resonances can be attributed to the four
E-mail addresses: kchriste@usc.edu (K.O. Christe), haiges@usc.edu (R. Haiges),
tmk@cup.uni-muenchen.de (T.M. Klapo¨tke).
0022-1139/$ – see front matter ß 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.jfluchem.2010.04.009

792
K.O. Christe et al. / Journal of Fluorine Chemistry 131 (2010) 791–799
Table 1
Experimentally observed and computed 19F and 77Se NMR chemical shifts.
Molecule
d
(
19F)/ppma
d (
77Se)/ppmb
Expt.c
Calc.
Expt.c
1083i
Calc.d,e,f
SeF4 (CH3F, ꢀ140 8Cb, ꢀ135 8Cc)
37.7, tg
12.1, tg
ax 86d, 83e, 76f, 55h
eq 56d, 49e, 49f, 16h
1042d, 1159e, 1162f, 1100h, 1153j
SeF4 (CH3F, 20 8C)b
SeF4 (CD2Cl2, 0 8C)
SeF4 (CH3CN, 25 8C)
SeF4 (neat, 25 8C)
28.7
32.1
1120
20.2
18.4, 24.9k, 27.5l
1093
1114, 1092m
[TPP][SeF5] (DMSO, 25 8C)
57.7 br (1 F)
19.6 br (4 F)
eq 79d, 56e, 67f
ax 29d, 10e, 20f
970, dk
962
880d, 1001e, 1053f
[TPP][SeF5] (CH3CN, 25 8C)
56.2 br (1 F)
17.2 br (4 F)
[TMA][SeF5] (CH3CN)
[TMA]2[SeF6] (CH3CN, 25 8C)
[PNP]2[SeF6] (CD2Cl2, 0 8C)
SeOF2 (CH3CN, 25 8C)
SeOF2 (CD2Cl2, 0 8C)
SeOF2 (neat)
14.4n
43.9 br
41.5 br
34.6
0d, ꢀ32e, ꢀ12f
1255
945d, 1088e, 1211f
1253
69d, 70e, 59f
1398 to
1397 tp
1392, 1378i,m,r
1267d, 1370e, 1303f, 1402l
42.8
37.0 br, 38.6q
[TMA][SeOF3] (CH3CN, 25 8C)
19.0 br
ax 44d, 21e, 31f
eq 43d, 28e, 37f
1358
1129d, 1261e, 1230f
[TPP][SeOF3] (CH3CN, 25 8C)
CsSeO2F (D2O, 25 8C)
16.1
1385
1314
Not observed
45d, 24e, 40f
1376d, 1511e, 1383f
a
F chemical shifts relative to CFCl3. Absolute F chemical shift =156.1 ppm at the GIAO B3LYP/AhlrichsVTZP level; absolute F chemical shift = 135.3 ppm at the GIAO BLYP/
TZ2P level with the ADF code.
b
Se chemical shifts relative to Se(CH3)2. Absolute Se chemical shift = 1698.8 ppm at the GIAO B3LYP/AhlrichsVTZP level; absolute Se chemical shift = 1616.2 ppm at the
GIAO BLYP/TZ2P level with the ADF code.
c
Data from this study, unless noted otherwise.
d
Data from this study, MPW1PW91/aug-cc-pVDZ.
e
Data from this study, GIAO B3LYP/AhlrichsVTZP.
f
Data from this study, BLYP-ZORA/TZ2P.
g
Data from Ref. [3].
h
Data from [46] using GIAO-MP2/6-311 + G(3df,3pd)//MP2/6-311 + G(3df,3pd).
i
Data from Ref. [5].
j
Data from [47] using PBE0/TZVP(P)//PBE0/TZVP.
k
Data from C. Lau, J. Passmore, J. Fluorine Chem. 6 (1975) 77–81.
l
Data from G.A. Olah, M. Nojima, I. Kerekes, J. Am. Chem. Soc. 46 (1974) 925–927.
m
Data from R. Birchall, J. Gillespie, S.L. Vekris, Can. J. Chem. 43 (1965) 1672–1679.
n
Data from Ref. [6].
o
J(SeF) 786 Hz.
p
J(SeF) 854 Hz.
q
Data from [k].
r
J(SeF) 837 Hz from [k].
equatorial and one axial fluorine atoms of the anion possessing a
C4v structure, which was also found in the crystal structure (see
below). 77Se satellites could only be observed for the axial fluorine
signal at 57.7 ppm in DMSO (1J(SeFax) = 1170 Hz). This assignment
is further substantiated by the observed broad doublet at 970 ppm
in the 77Se NMR spectrum of this sample, which exhibits the same
1J(SeFax) coupling constant. The failure to observe the coupling to
the four equatorial fluorines is attributed to the broadness of the
signals and the small value expected for this coupling. In the
structurally related C4v molecule, ClF5, 1J35Cl19Fax was found to be
192 Hz, whereas 1J35Cl19Feq was ꢂ20 Hz [29]. It is interesting to
note, that the 77Se NMR spectrum of the [SeF5]ꢀ anion in CH3CN
solution showed only a broad singlet at 962 ppm, indicating a more
fluxional structure than in DMSO solution. The only previously
reported value for the NMR spectra of SeF5ꢀ was a broad signal at
14.4 ppm in the 19F spectrum, recorded for [TMA]+[SeF5]ꢀ in
CH3CN at an unspecified temperature [6]. It appears that this signal
may have been due to the four equatorial fluorines with the axial
fluorine signal not having been observed or reported.
hexamethylpiperidinium) in CH3CN solution at an unspecified
temperature. This 19F signal was shifted to higher values upon
addition of SeF5 and to lower values upon addition of Fꢀ, which
ꢀ
was interpreted in terms of a rapid intermolecular fluorine
exchange between SeF5ꢀ, SeF6 and Fꢀ; however, it should be
2ꢀ
noted that tꢀhis signal is very close to that reported by these authors
[8] for SeF5 and, therefore, might have been due to the latter. In
our current study, broad resonances were observed at about
43 ppm in the 19F and at about 1254 ppm in the 77Se spectra of
2ꢀ
SeF6
in either CH3CN or CH2Cl2 solution. Whereas the 19F
resonances of SeF62ꢀ (43 ppm) and exchange-averaged equatorial/
axial SeF5ꢀ (26 ppm) are similar and, therefore, do not allow a cleaꢀr
distinction between SeF5ꢀ and SeF62ꢀ, our 77Se resonances of SeF5
2ꢀ
(970 ppm) and SeF6 (1254 ppm) differ by a very large amount
and imply that the spectra observed by us for SeF62ꢀ are not largely
due to a mixture of rapidly exchanging SeF5ꢀ and Fꢀ. A substantial
difference in the 77Se resonances of SeF5ꢀ (970 ppm) and SeF62ꢀ is
consistent with the calculations discussed below.
In Table 1, the experimentally observed shifts are also
compared to values calculated at the density functional theory
level using GIAO [30] to deal with the gauge problem with different
exchange-correlation functionals and basis sets: MPW1PW91/
aug-cc-pVDZ [31–34], B3LYP/AhlrichsVTZP [35–37], BLYP/TZ2P
For the [SeF6]2ꢀ anion, our values are also somewhat at odds
with the previous literature report [8] which listed no 77Se signal
and a single resonance without 77Se satellites at 13.4 ppm for the
19F NMR spectrum of the [PIP+]2[SeF6]2ꢀ salt (PIP+ = 1,13,3,5,5-

K.O. Christe et al. / Journal of Fluorine Chemistry 131 (2010) 791–799
793
Table 3
[38,39], and at the ZORA-BLYP/TZ2P level to include the effects of
Summary of crystal data and refinement results.
relativity [40–45]. We also compare to previous calculated results
at the MP2/6-311 + G(3df,3pd)//MP2/6-311 + G(3df,3pd) [46] and
PBE0/TZVP(P)//ꢀPBE0/TZVP [47] levels of theory. In general, the
calculated 19F shifts reflect the observed trends reasonably well.
The best results for SeF4 are those at the MP2 level [46]. The effects
of relativity at the ZORA level with the BLYP functional are small for
the 19F chemical shifts. In general the DFT values are up to 30–
50 ppm too large as compared to experiment.
The calculated 77Se chemical shifts are in qualitative agreement
with experiment. For SeF4 and SeF5ꢀ, the calculated chemical shifts
bracket the experimental values. As noted previously [47,48]
calculations of 77Se shifts are quite sensitive to small changes in the
geometry, solvent used, temperature and computational methods
and, therefore, show larger variations than the calculated 19F shifts.
Property
[PNP][SeOF3]
36H30F3NOP2Se
[PNP][SeF5]
Empirical formula
Formula mass
C
C36H30F5NP2Se
712.51
690.53
Temperature [K]
Crystal size [mm]
Space group
100
298
0.42 ꢃ 0.28 ꢃ 0.27
0.24 ꢃ 0.16 ꢃ 0.11
C2/c
¯
P1
˚
a [A]
11.7382(11)
11.8451(10)
14.1545(14)
66.790(9)
70.321(9)
71.023(8)
1660.5(3)
2
16.1712(12)
13.7199(10)
16.0646(12)
90
˚
b [A]
˚
c [A]
a
b
g
[8]
[8]
[8]
112.961 (1)
90
3
˚
V [A ]
3281.8(4)
4
Z
rcalc [g cmꢀ3
[mmꢀ1
]
1.3811
1.442
m
]
1.273
1.297
The effect of the relativistic correction for the Se chemical shifts
Reflections collected
Reflections unique
11647
9937
ꢀ
can be to either increase or decrease the chemical shift. For SeOF2
,
9652 (Rint = 0.0187)
0.0429, 0.0855
0.0527, 0.0963
1.138
3687 [Rint = 0.0475]
0.0395, 0.0922
0.0612, 0.0961
1.009
the calculated chemical shifts are greater than experiment,
R1, wR2 (2s data)
R1, wR2 (all data)
GOOF on F2
ꢀ
whereas for SeF62ꢀ, SeOF2, and SeOF3 the calculated values are
smaller than experiment. The observed experimental trend that
the chemical shift of SeF5ꢀ is less than that of SeF62ꢀ is confirmed
by the calculations.
For the 19F chemical shifts, the only exception for the calculated
chemical shifts being larger than experiment is for SeF62–where the
trend is in the opposite direction and the calculated 19F chemical
shiftsaremore negativethanexperiment. Thisisnosurprisebecause
the free valence electron pair of Se is partially sterically active [8,49]
requiring very high-level computations for reliable results [50ꢀ–53].
The structures of the isoelectronic molecules KrF6 and BrF6 are
over night at LMU. When taken out of the solution, the crystals
rapidly decomposed upon contact with air at room temperature.
Therefore, a crystal was mounted on the goniometer head under an
inert atmosphere at 253 K and immediately transferred into the
cold nitrogen stream of the diffractometer. The resulting structure
ꢀ
was that of the PNP-salt of the trifluoroselenite anion, SeOF3
,
containing one molecule of disordered dichloromethane as a
solvate, [PNP][SeOF3]ꢁCH2Cl2. The formation of this anion can be
explained by the reaction of [PNP][SeF5] with traces of adventitious
water.
ꢀ
octahedral but the structures of XeF6 and IF6 are C3v with the
octahedral structure slightly higher in energy. We attempted to
reproduce the previous calculations [49] for SeF62ꢀ at the B3LYP/6-
311 + G(2df) level but were unable to obtain a C3v structure lower
than the Oh structure. We note that the previous calculations
predicted the C3v structure to be only 0.4 kcal/mol below the Oh
structure. We also note that using the same approach the previous
authors [49] predicted that KrF6, BrF6ꢀ, and AsF63ꢀ are all octahedral
eventhoughthe lonepair would beexpectedtobelargerinthe latter
than in SeF62ꢀ. Our B3LYP/aVTZ-PP calculations as well as previous
[27,54] calculations resulted in a sterically inactive valence electron
pair on Se and an Oh geometry. In addition, starting from the C3v
structure, we optimized the geometry at the CCSD(T)/aug-cc-pVTZ
level and found an Oh structure. In order to further study this ꢀissue,
we used the same approach as we did for XeF6 as well as IF6 and
calculated the energy difference between the Oh and C3v structures
for SeF62ꢀ at the CCSD(T) level as a function of the basis set. We were
able to optimize the Oh structure at the CCSD(T) level but had to use
estimated structures for the C3v conformer. We used the Hartree–
Fock structure, the reported DFT optimized structure and the
experimental crystal structure. The energy differences are strongly
basis set dependent as shown in Table 2.
[PNP][SeOF3]ꢁCH2Cl2 crystallizes in the triclinic crystal system,
ꢀ
¯
space group P1 with Z = 2 (Table 3). The SeOF3 anion (Fig. 1)
adopts a pseudo-trigonal bipyramidal structure with two fluorine
atoms in the axial positions and the three equatorial positions
being occupied by an oxygen, a fluorine and a sterically active free
2.3. Crystal structures
Colorless crystals of [PNP][SeOF3]ꢁCH2Cl2 were obtained by
storing a concentrated solution of [PNP][SeF5] in CH2Cl2 at ꢀ32 8C
Table 2
2ꢀ
Oh–C3v energy differences for SeF6 at the CCSD(T) level in kcal/mol.
Method
aug-cc-pVDZ
aug-cc-pVTZ
aug-cc-pVQZ
DFTa
Expb
HFc
13.1
14.8
23.0
12.2
10.6
11.1
12.4
10.1
9.6
Fig. 1. Molecular structure of the SeOF3ꢀ anion in [PNP][SeOF3]. Thermal ellipsoids
˚
are shown at the 50% probability level. Selected bond lengths [A] and angles [8]:
a
˚
Se1–F1 1.9020(10), Se1–F2 1.906(3), Se1–O1 1.609(3), Se1–F3 1.7668(10), F2–Se1–
F1 174.79(10), F1–Se1–F3 90.70(11), F2–Se1–F3 87.52(11), F1–Se1–O1 91.80(12),
F2–Se1–O1 93.39(12), F3–Se1–O1 104.32(17).
Geometry parameters for the C3v structure: r(Se–F) = 1.907 and 2.051 A.
b
c
˚
r(Se–F) = 1.840 and 2.03 A.
˚
r(Se–F) = 1.719 and 2.167 A.

794
K.O. Christe et al. / Journal of Fluorine Chemistry 131 (2010) 791–799
Table 4
ꢀ
55
Comparison of the experimental and calculated (CCSD(T) and DFT) structures of SeOF3 with those of isoelectronic BrOF3 and of closely related ClOF356.a.
ꢀ
Parameter
X–Fax
SeOF3
CCSD(T)/aug-cc-pVQZ
1.901
B3LYP/aug-cc-pVQZ
1.923
BrOF3
ClOF3
1.713(3)
1.906(3)
182.0(3)
183.9(4)
1.902(1)
X–Feq
X–O
1.767(1)
1.609(3)
1.763
1.599
1.781
1.595
1.725(3)
1.569(4)
169.8(2)
1.603(4)
1.405(3)
F
F
ax–X–Fax
eq–X–O
174.8(1)
167.9
166.3
170.5(41)
108.9(9)
104.3(2)
104.1
104.2
103.3(2)
Feq–X–Fax
87.5(1)
90.7(1)
86.8
95.8
86.1
96.4
86.3(2)
87.9(12)
O–X–Fax
91.8(1)
93.4(1)
94.3(2)
94.7(20)
94.7(20)
a
˚
Bond lengths in A, angles in degree.
valence electron pair. This structure is analogous to those
previously found for isoelectronic BrOF3 [55] and the related
ClOF3 molecule [56] (Table 4). The structure of SeOF3 is in
Slightly pink single crystals of [PNP][SeF5] were grown from
CH3CN solution by slow evacuation of the solvent in vacuo at
ꢀ20 8C at USC. The compound crystallizes in the monoclinic space
group C2/c with Z = 4 (Table 3). In contrast to the previously
reported [6] crysꢀtal structure of [N(CH3)4][SeF5], which contains
ꢀ
excellent agreement with our theoretical calculations (Table 4),
showing that the oxygen and the equatorial fluorine are not
suffering from disorder. Although theoretical calculations were
carried out at different levels of theory and with different basis
sets, only those obtained at the highest levels (CCSD(T)) with the
best basis sets are given in Table 4. As expected, the results from
the coupled cluster calculations are superior to those using density
functional theory.
tetrameric [SeF5
]
4
units in which the individual anions are
strongly distorted from the ideal C4v symmetry, [PNP][SeF5]
ꢀ
contains well-isolated SeF5 anions with almost perfect square
˚
pyramidal symmetry (Fig. 2). The axial Se–F bond length of 1.68 A
˚
is significantly shorter than the equatorial ones of 1.81 A, in
excellent accord with Christe’s Rule III for hypervalent species
containing a sterically active free valence electron pair [62]. This
rule states that ‘‘the free valence electron pairs on the central
atoms seek high s-character, i.e., spn hybridization. If the number of
ligands (including the free electron pairs) is larger than four, then
so many F ligands form linear semi-ionic 3 center ꢀ4 electron (3c–
4e) bonds [57,58] as are required to allow the free electron pairs to
form an spn hybrid with the remaining F ligands. These semi-ionic
3c–4e bonds are significantly weaker and longer than the mainly
covalent spn hybrid bonds.’’ The increased polarity of the four
Consistent with the experimental structures of BrOF3 [55] and
ClOF3 [56] and the calculated structure of the analogous sꢀulfur
ꢀ
compound, SOF3 [23], the axial Se–F bond lengths in SeOF3 are
significantly longer than the equatorial one and, as expected,
longer than those in SOF3ꢀ, ClOF3 and BrOF3. The bond lengthening
in SeOF3ꢀ, when compared to the neutral BrOF3 molecule, is due to
the formal negative charge in SeOF3ꢀ which increases the ionicity
of the bonds and makes them longer. The fact that the bond length
of the axial Se–F bonds is much larger than that of the equatorial
one suggests that the negative charges in SeOF3ꢀ are larger on the
two axial fluorine atoms (Mulliken charge q = ꢀ0.70e (B3LYP/aug-
cc-pVTZ)) than on the equatorial fluorine atom (q = ꢀ0.59e).
ꢀ
equatorial Se–F bonds in SeF5 is also reflected by the Mulliken
charges of ꢀ0.60 and ꢀ0.81 calculated by us at the B3LYP/aug-cc-
pVTZ level for the axial and the four equatorial bonds, respectively.
Because the free valence electron pair of selenium is more diffuse
and space-demanding than the bonding electron pairs, the Fax–Se–
ꢀ
Therefore, the bonding in SeOF3 is best described by a model
assuming a semi-ionic 3c–4e bond [57,58] for the two axial fluorine
ligands and predominantly covalent bonding for the equatorial
Feq angles are slightly compressed from the ideal value of 90–848.
fluorine and oxygen ligands, as discussed below in more detail for
The [SeF5]ꢀ anions are located in channels, formed by the cations,
with the axial F ligands pointing up and down in an alternating
pattern. They are very well separated because of the rather large
ꢀ
SeF5
. The Se–O distance is significantly shorter than the
equatorial Se–F distance implying that the Se–O bond has high
double bond character the nature of which has recently been
analyzed [59] using the Bader’s AIM method [60]. By analogy with
BrOF3 and ClOF3, the axial fluorine atoms are bent away from the
Se55O double bond into the sector between the Se–F single bond
and the free lone electron pair of Se. This demonstrates that in the
axial direction the steric repulsion from the Se55O double bond is
larger than the repulsion from either the lone pair on Se or the Se–F
single bond. The angles in the equatorial plane [56], however, show
that in the equatorial direction the repulsion by the lone pair is
largest, followed by the Se55O double bond, with the Se–F single
bond being smallest consistent with VSEPR arguments. These
directional repulsion effects by lone valence electron pairs and
p
bonds in trigonal bipyramidal molecules have been discussed by us
ꢀ
in detail previously [61]. A discussion of trends within the SOF3
,
SeOF3ꢀ, and TeOF3 series is not warranted at this point because
for SOꢀF3ꢀ only a low-level calculated structure is available, and for
TeOF3 the experimental structure contains oxygen-bridged
dimeric anions [24] which are also present in the complex salt
[C5H6N]2[TeF5][Te2O2F6]0.5 [25]. Therefore, the crystal structure of
ꢀ
Fig. 2. Molecular structure of the SeF5ꢀ anion in [PNP][SeF5]. Thermal ellipsoids are
˚
shown at the 50% probability level. Selected bond lengths [A] and angles [8]: Se–F1
ꢀ
SeOF3 represents the first experimentally well-determined
1.681(2), Se–F2 1.809(2), Se–F3 1.812(2), F1–Se–F2 84.46(8), F1–Se–F3 84.13(7),
F2–Se–F3 90.31(11), F2–Se–F2A 168.92(16), F3–Se–F3A 168.25(13).
molecular structure of a monomeric trifluoro-chalcogenite anion.

K.O. Christe et al. / Journal of Fluorine Chemistry 131 (2010) 791–799
795
Table 5
Computational and experimental geometries for SeF5 (bond lengths in A, angles in degree).
ꢀ
˚
Parameter
Point group
Se–Fax
MPW1PW91/aug-cc-PVDZ
CCSD(T)/aug-cc-pVQZ
Exptl. [PNP][SeF5]
1.681(2)
Exptl.a [TMA][SeF5]
C4v
C4v
1.752
1.715
1.707(3)
1.717(3)
Se–Feq
1.863
1.836
1.809(2)
1.812(2)
1.827(3)
1.869(3)
Fax–Se–Feq2
84.7
84.2
84.46(8)
84.13(7)
83.1(1)
84.3(1)
F
F
F
ax–Se–Feq3
84.7
84.2
eq2–Se–Feq2A
eq3–Se–Feq3A
169.4
169.4
168.4
168.4
168.92(16)
168.25(13)
166.2(2)
168.6(2)
a
Data from Reference [6].
the Se–F bonds in the Se(+IV) anion series, SeO2Fꢀ (1.851 A),
˚
ꢀ
ꢀ
˚
˚
SeOF3 (unique Se–F = 1.763 A), SeF5 (unique Se–F = 1.715 A),
shows a pronounced shortening from SeO2Fꢀ to SeF5ꢀ. This is due
to the fact that the electron withdrawing effect of a doubly bonded
oxygen is substantially lower than that of two fluorine ligands,
thereby causing the unique Se–F bond in SeF5ꢀ to be the least polar
and hence the shortest one.
2.4. Vibrational spectra
The previously reported [7] vibrational spectra of CsSeF5 and
their assignments were examined for their correctness by
comparison with those calculated at the MPW1PW91/aug-cc-
PVDZ, B3LYP/aug-cc-PVTZ and CCSD(T)/aug-cc-PVDZ levels
(Table 6). Whereas the observed stretching mode frequencies
are slightly lower than the calculated ones, the observed
deformation mode frequencies are higher than the calculated
ones. The calculated stretches are higher than experiment due to
the lack of anharmonic corrections and the bends are predicted to
be too low as is typically found for these types of compounds. The
biggest differences are ꢄ40 cmꢀ1 between theory and experiment
Fig. 3. Structure of the SeO2Fꢀ anion, calculated at the CCSD(T)/aug-cc-pVQZ level of
˚
theory. Selected bond lengths [A] and angles [8]: Se–F 1.851, Se–O 1.632, F–Se–O
100.3, O–Se–O 110.9.
[PNP]+ counter-ions, and the closest Se–F distances between
˚
neighboring anions are 6.4 A. Table 5 shows a comparison of the
experimental structure [6] of tetrameric SeF5ꢀ in [TMA][SeF5] and
ꢀ
our structure of monomeric SeF5 in [PNP][SeF5] with values
calculated for the free gaseous anion at different levels of theory.
The agreement between the two experimental structures is quite
good in spite of the association in the TMA salt, and the fit with the
coupled cluster CCSD(T) calculations is excellent.
Since we could not obtain a well ordered crystal structure for
the SeO2Fꢀ anion, we have calculated its structure at the CCSD(T)/
aug-cc-pVQZ level of theory (Fig. 3). In this anion, Se has only a
coordination number of 4, so it is not hypervalent and no semi-
ionic 3c–4e bonding contributions need to be invoked. Compared
for the n7 stretch and the n5 and n8 bends. The excellent agreement
between theory and experiment confirms our assignments.
Vibrational data of alkali trifluoroselenites have been known for
many years [20–22], and a complete normal coordinate analysis
had been carried out previously [20]. However, a comparison with
our calculated spectra (Table 7) reveals that in the previous
assignments, the antisymmetric axial SeF2 stretching frequency
had been misassigned to a band having [26,27] too high a
frequency, resulting in a value for the axial stretching force
˚
to the calculated structure for neutral SeO2 (rSe5O = 1.611 A), the
ꢀ
˚
Se55O bond distance (1.632 A) in SeO2F has increased only
constant which is much too high. In addition, n4, n5, n8, and n9 were
˚
slightly, but the Se–F bond length (1.851 A) is unusually long.
assigned incorrectly, and some of the mode descriptions were
misleading. By analogy to the closely related pseudo-trigonal
bipyramidal SF4 molecule [63], the axial and equatorial scissoring
Therefore, the Se–F bond is highly ionic and carries most of the
negative charge. Inspection of the CCSD(T)/aug-cc-pVQZ values for
Table 6
ꢀ
Observed and calculated vibrational frequencies (cmꢀ1) of SeF5 and their assignments in point group C4v .
.
Assignment IR,
RA activ
Approx mode
descript
CsSeF5 [7] obsd
freq, rel intens
Calcd freq, (IR) [RA] intensa
Calcd freq,
Calcd freq, (IR) intens
CCSD(T)/aug-cc-pVDZb
MPW1PW91/aug-cc-pVDZ
(IR) [RA] intensa
B3LYP/aug-cc-pVTZ
RA
IR
A1 (RA, IR) n1
n2
n3
nSeF
nsym SeF
in phase
4
666 [10]
515 [7.5]
332 [3.2]
460 [7.0]
236 [0.6]
282 [2.6]
480 sh
665 vs
520 sh
335 s
–
674 (84) [13]
524 (2.6) [14]
311 (31) [2.3]
469 (0) [14]
660 (83) [13]
504 (2.5) [14]
289 (31) [3.1]
443 (0) [13]
674 (84)
527 (0.01)
305 (48)
475(0)
axial
dumbrella
nsym SeF
dasym SeF
dsym SeF
nasym SeF
B1 (RA, ꢀ) n4
out of phase
4
n5
–
183 (0) [0.1]
259 (0) [2.1]
527 (881) [1.3]
348 (0.2) [1.6]
188 (0.3) [0.4]
181 (0) [0.2]
258 (0) [2.1]
492 (874) [1.3]
350 (0.4) [1.4]
185 (1.4) [0.3]
191 (0)
out of plane
in plane
4
B2 (RA, ꢀ) n6
–
265 (0)
4
E (RA, IR) n7
475 vs, br
398 mw
n.o.
526 (393)
359 (4.9)
194 (0.1)
4
n8
n9
dF SeF
0
399 [1.9]
202 [0.7]
4
dasym SeF
in plane
4
a
IR intensities in km/mol, RA intensities in A4 amuꢀ1
IR intensities at the CCSD(T)/cc-pVDZ level.
.
˚
b

796
K.O. Christe et al. / Journal of Fluorine Chemistry 131 (2010) 791–799
Table 7
ꢀ
Observed and calculated vibrational frequencies (cmꢀ1) of SeOF3 and their assignments in point group Cs.
Assignment
Approx mode descript
CsSeOF3 [20]
obsd RA freq,
rel intens
Calc freq,
Calc freq,
Calc freq,
Calc freq CCSD(T)/
aug-cc-pVDZ
(IR) [RA] intensa
MPW1PW91/
aug-cc-pVDZ
(IR) [RA] intensa
B3LYP/aug-cc-pVTZ
(IR) [RA] intensa
MP2/aug-cc-pVDZ
A0 n1
n2
n3
n4
n5
nSeO
nSeF
nsym SeF
dscissOSeF
dsym SeF
dsym SeF
958 [10]
560 [10]
420 [8,br]
350 [2]
965 (119) [26]
584 (146) [9.0]
440 (5.9) [12]
326 (28) [2.9]
198 (28) [3]
971 (119) [28]
564 (145) [8.6]
414 (6.2) [12]
326 (26) [3.1]
187 (11) [0.2]
162 (6.2) [0.3]
1006 (138) [27]
570 (159) [13]
435 (5.4) [16]
330 (29.7) [3.8]
195 (9.4) [0.2]
166 (5.2) [0.8]
934
581
440
324
189
162
eq
2 ax
þ
dsym SeF
eq
2axðumbrellaÞ
230 [0+]
197 [1]
2 ax out of FꢀSeEꢀF plane
n6
ꢀ
dscissOSeF
191 (11) [0.2]
2 ax
eqðBerry mechanismÞ
A00 n7
n8
n9
nasym SeF
420, br
375 [1]
303 [2]
465 (457) [0.3]
338 (3.5) [2.0]
277 (0.2) [1.4]
431 (436) [0.3]
331 (21) [2.0]
273 (0.9) [1.4]
460 (487) [0.5]
343 (2.8) [2.4]
282 (0.4) [1.4]
455
334
271
2 ax
dwagOSeF
eq
pOSeF
eq out of plane
a
IR intensities in km/mol, RA intensities in A4 amuꢀ1
.
˚
Table 8
Observed and calculated vibrational frequencies (cmꢀ1) of SeO2Fꢀ and their assignments in point group Cs.
Assignment
Approx mode
descript
CsSeO2Fa obsd
Et4NSeO2Fb
Calcd freq, (IR) [RA]
intensc B3LYP/aug-cc-pVTZ
Calcd freq,
CCSD(T)/
RA freq, rel intens
obsd (IR) [RA]
aug-cc-pVDZ
A0 n1
n2
n3
n4
A00 n5
n6
nsym SeO
nSeF
nsciss SeO
nsym FSeO
nasym SeO
dasym FSeO
879 [10]
429 [2.0]
402 [2.0]
320 [1.8]
857 [4.4]
281 [2.1]
910 [10]
452 [1]
389 [2]
297 [1]
864 [1]
266 [1]
896 (65) [37]
443 (182) [5.0]
379 (38) [3.0]
273 (3.5) [3.6]
900 (267) [16]
243 (5.5) [2.9]
856
464
366
270
869
240
2
2
2
2
2
a
Data from this work.
b
data from Ref. [16].
IR intensities in km/mol, RA intensities in A4 amuꢀ1
c
˚
.
modes are strongly coupled, resulting in a symmetric (
antisymmetric ( 6) combination of the corresponding symmetry
coordinates. Therefore, ( 4) is best described as an umbrella type
deformation, while ( 6) corresponds to the mode which becomes
n
4) and
binding energy of Fꢀ to A to form AFꢀ) of the three neutral
compounds and they are 75.8, 68.1, and 64.1 kcal/mol for SeF4,
SeO2, and SeOF2, respectively. The fluoride affinities provide an
estimate of the Lewis acidity of a compound [66] and show that
these selenium compounds are moderate Lewis acids.
n
n
n
the reaction coordinate for the axial–equatorial fluorine ligand
exchange by the Berry mechanism [64]. Again, excellent agree-
ment is found between theory and experiment with the largest
difference being ꢄ40 cmꢀ1 for the n5 and n8 bending modes.
The vibrational spectra of the SeO2Fꢀ anion have previously
been reported and analyzed in detail [16]. Since its four lower
frequency modes have relatively similar frequencies and intensi-
ties, we have calculated its vibrational spectra at the CCSD(T)/aug-
cc-pVDZ and B3LYP/aug-cc-pVTZ levels of theory (Table 8) to verify
the previously proposed assignments [16]. It was found that our
current Raman spectrum for CsSeO2F was in good agreement with
that previously reported and that the previously proposed
assignments are also correct. The largest deviations between the
calculated and observed frequencies are attributed in part to anion
association and solid state effects, because the deviations become
smaller with increasing cation size and for solutions of the salts in
either CH3CN or (CH3)2SO [16]. We note that the SeO2 stretching
frequencies are calculated to be closer to each other than was
experimentally observed and that their order should be reversed
with the SeO2 antisymmetric stretching frequency being higher
than the symmetric one.
3. Experimental
3.1. General procedures
At LMU, all manipulation of air- and moisture-sensitive
materials were performed under an inert atmosphere of dry argon
using flame-dried glass vessels or oven-dried plastic equipment
and Schlenk techniques [67]. The selenium fluorides were handled
in PFA-vessels (perfluoroalkoxy-copolymer) due to the extreme
sensitivity towards glass. For all NMR measurements, 4 mm PFA-
tubes were used, which were placed into standard 5 mm NMR
glass tubes. Selenium tetrafluoride (Galaxy Chemicals), silver
fluoride (ABCR) and bis(triphenylphosphoranylidene)ammonium
chloride ([PNP]Cl, Aldrich) were used as received. The salts
[PNP][SeF5] and [PNP]2[SeF6] were prepared as previously reported
[68]. The solvents dichloromethane and acetonitrile were dried by
standard methods, and freshly distilled prior to use. NMR spectra
were recorded in CD2Cl2 at 0 8C on a JEOL Eclipse 400 instrument,
and chemical shifts are with respect to CFCl3 (19F, 376.1 MHz) and
Me2Se (77Se, 76.3 MHz).
2.5. Calculated heats of formation
At USC, all reactions were carried out in Teflon-FEP ampules or
stainless steel cylinders that were closed by stainless steel valves
and had been passivated with ClF3 prior to use. Volatile materials
were handled in stainless steel/Teflon-FEP or grease-free Pyrex-
glass vacuum lines [69]. Nonvolatile materials were handled in the
dry argon atmosphere of a glove box. Raman spectra were recorded
at ꢀ80 8C in the range 4000–80 cmꢀ1 on a Bruker Equinox 55 FT-RA
spectrophotometer using a Nd-YAG laser at 1064 nm with power
levels less than 100 mW. Teflon-FEP tubes with stainless steel
Our composite approach [65] to the prediction of the
thermodynamic properties of molecules is based on CCSD(T)
calculations extrapolated to the complete basis set limit plus the
inclusion of additional corrections. This approach was used to
calculate the heats of formation of SeF5ꢀ, SeO2Fꢀ, and SeOF3 as
well as those of SeF4, SeO2, and SeOF2 as given in Table 9. These
values can be used to predict the fluoride affinities (negative of the
ꢀ

K.O. Christe et al. / Journal of Fluorine Chemistry 131 (2010) 791–799
797
Table 9
behind off-white solids. Single crystals of [((C6H5)3P)2N][SeF5]
were grown from CH3CN solution by slow evaporation of the
solvent in vacuo.
Calculated heats of formation in kcal/mol.
Molecule
DHf (0 K)
DHf (298 K)
SeF4ꢀ
SeF5
ꢀ197.3
ꢀ331.6
ꢀ126.7
ꢀ249.2
ꢀ34.7
ꢀ202.9
ꢀ338.2
ꢀ131.1
ꢀ254.7
ꢀ38.1
3.3. X-ray crystallography
SeOF2ꢀ
SeOF3
SeO2
SeO2Fꢀ
At LMU, an Oxford Xcalibur diffractometer with a CCD area
detector was employed for the data collection using Mo K
a
ꢀ161.2
ꢀ165.7
radiation. The structure was solved using direct methods (SIR97)
[74] and refined by full-matrix least squares on F2 (SHELXL) [75].
All non-hydrogen atoms were refined anisotropically. Satisfactory
atomic positions of the solvent molecules in [PNP][SeOF3] could
not be determined reliably. Therefore, the disordered solvent
valves were used as sample containers. NMR spectra were
recorded at 470.51 MHz (19F) and 95.36 MHz (77Se) on a Bruker
AMX 500 spectrometer using neat compounds or solutions of the
compounds in acetonitrile or DMSO in sealed 4 mm Teflon-FEP
sample tubes. Neat CFCl3 (0.00 ppm) and (CH3)2Se (0.00 ppm) were
used as external references for 19F and 77Se, respectively. The
starting materials SeO2, [P(C6H5)4]Cl, [((C6H5)3P)2N]Cl (all from
Aldrich), and ClF3 (Matheson Co.) were used without further
purification. The HF (Matheson Co.) was dried by storage over BiF5
(Ozark Mahoning) [70]. Solvents were dried by standard
methods and freshly distilled prior to use. [P(C6H5)4][HF2]
and [((C6H5)3P)2N][HF2] were prepared from [P(C6H5)4]N3
and [((C6H5)3P)2N]N3, respectively, and HF. [P(C6H5)4]N3 and
[((C6H5)3P)2N]N3 were prepared from [P(C6H5)4]Cl and
[((C6H5)3P)2N]Cl, respectively, and NaN3 by ion-exchange [71].
Literature methods were used for the preparation of anhydrous
[N(CH3)4]F [72], H2SeO3 [73], and Cs[SeO2F] [16].
molecules were treated as a diffuse contribution using the program
3
˚
SQUEEZE [76,77] SQUEEZE calculated 338.0 A of void space per unit
cell and 86 electrons; 2 molecules of dichloromethane require 84
electrons per unit cell.
At USC, the single crystal X-ray diffraction data of
[((C6H5)3P)2N][SeF5] were collected on a Bruker 3-circle platform
diffractometer, equipped with a SMART CCD (APEX) detector with
the
x-axis fixed at 54.748, and using Mo Ka radiation (Graphite
monochromator) from a fine-focus tube. The diffractometer was
equipped with an LT-3 apparatus for low-temperature data
collection using controlled liquid nitrogen boil off. Cell constants
were determined from 60 ten-second frames.
A complete
hemisphere of data was scanned on omega (0.38) with a run time
of 10-s per frame at a detector resolution of 512 ꢃ 512 pixels using
the SMART software package [78]. A total of 1271 frames were
collected in three sets and a final set of 50 frames, identical to the
first 50 frames, was also collected to determine any crystal decay.
The frames were then processed on a PC, running Windows 2000
software, by using the SAINT software package [79] to give the hkl
files corrected for Lp/decay. The absorption correction was
performed using the SADABS program [80]. The structure was
solved by the direct method using the SHELX-90 program and
refined by the least squares method on F2, SHELXL-97 incorporated
in SHELXTL Suite 6.12 for Windows NT/2000 [75]. All non-
hydrogen atoms were refined anisotropically. For the anisotropic
displacement parameters, the Ueq is defined as one third of the
trace of the orthogonalized Uij tensor. ORTEP drawings were
prepared using the ORTEP-3 for Windows V1.076 program [81].
Further details of the crystal structure investigations reported in
this paper may be obtained from the Cambridge Crystallographic
Data Centre (CCDC), 12 Union Road, Cambridge CB2 1EZ (UK) (fax:
+44 1223 336 033 or email: deposit@ccdc cam ac uk) by quoting
the depository numbers CCDC 686702 and 641947.
3.2. Preparation and analytical data of the fluoroselenites
[PNP][SeF5]: Silver fluoride (1.77 mmol) was added at room
temperature to a solution of SeF4 (1.77 mmol) in CH3CN (4 mL).
The resulting solution was stirred for 2 h and then [PNP]Cl
(1.77 mmol) was added. After additional stirring for 30 min, the
pale yellow solution was decanted from a gray precipitate and all
volatile material was pumped off, yielding a colorless solid
[PNP]2[SeF6]: Silver fluoride (3.55 mmol) was added at room
temperature to a solution of SeF4 (1.77 mmol) in CH3CN (6 mL).
The resulting solution was stirred for 2 h and then [PNP]Cl
(3.55 mmol) was added. After additional stirring for 30 min, the
pale yellow solution was decanted from a gray precipitate and all
volatile material was pumped off, yielding a colorless solid. 19F
NMR
d d 1253 ppm
41.5 (br) ppm; 77Se NMR
[PNP][SeOF3]: This compound was obtained as a by-product in
the preparation of [PNP][SeF5]. The reaction of SeOF2, which was
present in our sample of SeF4, with AgF in CH2Cl2 furnished
[PNP][SeOF3] as colorless crystals suitable for X-ray analysis
SeF4: A sample of SeO2 (100 mmol) was loaded into a 75 mL
stainless steel cylinder. The cylinder was evacuated, cooled to
ꢀ196 8C, and HF (180 mmol) and ClF3 (140 mmol) were condensed
in. The reaction mixture was allowed to warm to ambient
temperature. After 12 h, all volatile material was pumped off
through traps at ꢀ45 and ꢀ196 8C. The ꢀ45 8C trap contained
95 mmol of pale yellow SeF4.
SeOF2: A sample of SeO2 (30 mmol) was loaded into a 15 mL
stainless steel cylinder. The cylinder was connected to the steel
line, evacuated and SeF4 (25 mmol) condensed in at ꢀ196 8C. The
reaction mixture was warmed to ambient temperature. After 10 h,
the SeOF2 was pumped off, collected in a trap at ꢀ196 8C, and
22 mmol of SeOF2 was obtained as pale yellow liquid.
3.4. Computational details
Geometries were initially optimized at the density functional
theory level with the B3LYP exchange-correlation functional
[35,36] and the standard aug-cc-pVnZ, with n = D and T, basis
sets were used for O and F [33]. A small core relativistic effective
core potential (RECP) was used for Se, which subsumes the (1s2,
2s2, 2p6,) orbital space into the 10-electron core, and a 24 electron
space (3s2, 3p6, 4s2, 3d10 and 4p4) with the electrons handled
explicitly [34]. We denote this combination of basis sets as aug-cc-
pVnZ. Only the spherical component subset (e.g., 5-term d
functions, 7-term f functions, etc.) of the Cartesian polarization
functions were used. Frequencies including the IR and Raman
intensities were calculated with the B3LYP functional as well as
with the MPW1PW91 exchange-correlation functional. The B3LYP
geometries were used for the starting point for optimizations at the
CCSD(T) level with the aug-cc-pVnZ basis sets for n = D, T, and in
some cases Q. Frequencies were calculated at the CCSD(T)/aug-cc-
pVDZ level for some molecules. Only the 4s and 4p electrons on Se
A+[SeF5]ꢀ, (A+)2[SeF6]2ꢀ and A+[SeOF3]ꢀ (A = [N(CH3)4], [P(C6H5)4],
[((C6H5)3P)2N]): Stoichiometric amounts of SeF4 or SeOF2, respec-
tively, were condensed onto frozen solutions of A+[HF2]ꢀ
(0.30 mmol) in 1.5 mL CH3CN at ꢀ196 8C. The reaction mixtures
were stirred at ambient temperature. After 30 min, all volatile
material was pumped off from the pale orange solutions, leaving

798
K.O. Christe et al. / Journal of Fluorine Chemistry 131 (2010) 791–799
[5] R. Damerius, P. Huppmann, D. Lentz, K. Seppelt, J. Chem. Soc., Dalton Trans. (1984)
2821–2826.
[6] A.R. Mahjoub, D. Leopold, K. Seppelt, Z. Anorg. Allg. Chem. 618 (1992) 83–88.
[7] K.O. Christe, E.C. Curtis, C.J. Schack, D. Pilipovich, Inorg. Chem. 11 (1972) 1679–
1682.
[8] A.R. Mahjoub, X. Zhang, K. Seppelt, Chem. Eur. J. 1 (1995) 261–265.
[9] K.H. Moock, R.T. Boere, J. Fluorine Chem. 68 (1994) 175–179.
[10] I.C. Bowater, R.D. Brown, F.R. Burden, J. Mol. Spectrosc. 23 (1967) 272–279.
[11] R. Paetzold, K. Aurich, Z. Anorg. Allg. Chem. 315 (1962) 72–78.
[12] J.A. Rolfe, L.A. Woodward, Trans. Faraday Soc. 51 (1955) 778–780.
[13] E.B.R. Prideaux, C.B. Cox, J. Chem. Soc. (1928) 739–745.
[14] E.B.R. Prideaux, C.B. Cox, J. Chem. Soc. (1927) 928–929.
[15] C. Feldmann, M. Jansen, Chem. Ber. 127 (1994) 2173–2176.
[16] J. Milne, Inorg. Chem. 17 (1978) 3592–3595.
and the 2s and 2p electrons in O and F were correlated in the
valence electron correlation calculations. The ZORA NMR calcula-
tions [40–45] were done, as described above, with the ADF code
[39] at the B3LYP/aVTZ-PP geometries. NMR calculations were also
done with the B3LYP and BLYP [36,38] functionals with the aug-cc-
pVDZ and Ahlrichs TZ2P [37] basis sets.
The heat of formation calculations were done at the CCSD(T)
level. For the open shell atomic calculations, we used the R/
UCCSD(T) (restricted method for the starting Hartree–Fock
wavefunction and then relaxed the spin restriction in the coupled
cluster portion of the calculation) approach [82]. The CCSD(T)/aug-
cc-pVnZ valence energies were extrapolated to the complete basis
set (CBS) limit by using a mixed exponential/Gaussian function of
the form [83]:
[17] E.J. Baran, J. Fluorine Chem. 10 (1977) 255–259.
[18] R.J. Gillespie, P. Spekkens, J.B. Milne, D. Moffett, J. Fluorine Chem. 7 (1976) 43–54.
[19] R. Paetzold, K. Aurich, Z. Anorg. Allg. Chem. 335 (1965) 281–288.
[20] J. Milne, P. Lahaie, Inorg. Chem. 22 (1983) 2425–2428.
[21] R. Paetzold, K. Aurich, Z. Anorg. Allg. Chem. 348 (1966) 94–106.
[22] R. Paetzold, K. Aurich, Z. Chem. 6 (1966) 152–153.
[23] A. Kornath, D. Kadzimirsz, R. Ludwig, Inorg. Chem. 38 (1999) 3066–3069.
[24] U. Kessler, M. Jansen, Z. Anorg. Allg. Chem. 627 (2001) 1782–1786.
[25] T.M. Klapo¨tke, B. Krumm, P. Mayer, I. Schwab, Acta Crystallogr. E 61 (2005)
o2984–o2986.
[26] L. Wang, Int. J. Mass Spectrom. 264 (2007) 84–91.
[27] M. Atanasov, D. Reinen, Inorg. Chem. 43 (2004) 1998–2012.
[28] Q. Li, W. Xu, Y. Xie, H.F. Schaefer III, J. Phys. Chem. A 103 (1999) 7496–7505.
[29] M. Alexandre, P. Rigny, Can. J. Chem. 52 (1974) 3676–3681.
[30] K. Wolinski, J.F. Hinton, P. Pulay, J. Am. Chem. Soc. 112 (1990) 8251–8260.
[31] C. Adamo, V. Barone, J. Chem. Phys. 108 (1998) 664–675.
[32] J.P. Perdew, Y. Wang, Phys. Rev. B 45 (1991) 13244.
[33] (a) T.H. Dunning Jr., J. Chem. Phys. 90 (1989) 1007–1023;
(b) R.A. Kendall, T.H. Dunning Jr., R.J. Harrison, J. Chem. Phys. 96 (1992) 6796–
6806;
2
EðnÞ ¼ ECBS þ Beꢀðnꢀ1Þ þ Ceꢀðnꢀ1Þ
(1)
with n = 2 (aug-cc-pVDZ), 3 (aug-cc-pVTZ), 4 (aug-cc-pVQZ).
Core-valence corrections, ECV, were obtained as the difference
D
between frozen-core and all-electrons correlated calculations at
the CCSD(T)/cc-pwCVTZ-PP level [84]. A scalar relativistic correc-
tion,
DESR, due to the F and O atoms was evaluated from the
expectation values for the two dominant terms in the Breit–Pauli
Hamiltonian (the mass-velocity and one-electron Darwin (MVD)
corrections) [85] from configuration interaction singles and
doubles (CISD) calculations with a VTZ basis set at the CCSD(T)/
aug-cc-pVTZ geometry. Any ‘‘double counting’’ of the relativistic
effect on the Se when applying a MVD correction to an energy,
which already includes most of the relativistic effects via the RECP,
is small. A second relativistic correction is due to atomic spin orbit
effects and the values are 0.22 kcal/mol for O, 0.39 for F and
2.70 kcal/mol for Se were taken from the excitation energies
(c) D.E. Woon, T.H. Dunning Jr., J. Chem. Phys. 98 (1993) 1358–1371;
(d) T.H. Dunning Jr., K.A. Peterson, A.K. Wilson, J. Chem. Phys. 114 (2001) 9244–
9253;
(e) A.K. Wilson, D.E. Woon, K.A. Peterson, T.H. Dunning Jr., J. Chem. Phys 110
(1999) 7667–7676.
[34] (a) K.A. Peterson, J. Chem. Phys. 119 (2003) 11099–11112;
(b) K.A. Peterson, D. Figgen, E. Goll, H. Stoll, M. Dolg, J. Chem. Phys. 119 (2003)
11113–11123.
[35] A.D. Becke, J. Chem. Phys. 98 (1993) 5648–5652.
compiled by Moore [86]. By combining our computed
given by the following expression:
S
D0 values
[36] C. Lee, C.W. Yang, R.G. Parr, Phys. Rev. B 37 (1988) 785–789.
[37] A. Scha¨fer, H. Horn, R. Ahlrichs, J. Chem. Phys. 97 (1992) 2571–2577.
[38] A.D. Becke, Phys. Rev. A 38 (1988) 3098–3100.
[39] (a) ADF 2004.01, ADF Users Guide, SCM, Theoretical Chemistry, Vrije Universiteit,
Amsterdam, http://www.scm.com.
(b) G. te Velde, F.M. Bickelhaupt, E.J. Baerends, C. Fonseca Guerra, J.A. van
Gisbergen, J.G. Snijders, T. Ziegler, J. Comput. Chem. 22 (2001) 931–967.
[40] G. Schreckenbach, T. Ziegler, J. Phys. Chem. 99 (1995) 606–611.
[41] G. Schreckenbach, T. Ziegler, Int. J. Quantum Chem. 61 (1997) 899–918.
[42] S.K. Wolff, T. Ziegler, J. Chem. Phys. 109 (1998) 895–905.
[43] S.K. Wolff, T. Ziegler, E. van Lenthe, E.J. Baerends, J. Chem. Phys. 110 (1999) 7689–
7698.
[44] E. van Lenthe, E.J. Baerends, J.G. Snijders, J. Chem. Phys. 99 (1993) 4597–4610.
[45] J. Autschbach, T. Ziegler, in: M. Kaupp, M. Buhl, V.G. Malkin (Eds.), Calculation of
NMR and EPR Parameters: Theory and Application, Wiley-VCH & Co., Weinheim,
2004, pp. 249–264.
[46] H. Poleschner, K. Seppelt, Chem. Eur. J. 10 (2004) 6565–6574.
[47] T. Maaninen, H.M. Tuononen, K. Kosunen, R. Oilunkaniemi, J. Hitola, R. Laitinen, T.
Chivers, Z. Anorg. Allg. Chem. 630 (2004) 1947–1954.
[48] G. Schreckenbach, Y. Ruiz-Morales, T. Ziegler, J. Chem. Phys. 104 (1996) 8605–8612.
[49] J. Pilme, E.A. Robinson, R.J. Gillespie, Inorg. Chem. 45 (2006) 6198–6204.
[50] D.A. Dixon, W.A. de Jong, K.A. Peterson, K.O. Christe, G.J. Schrobilgen, J. Am. Chem.
Soc. 127 (2005) 8627–8634.
X
D0
¼
D
EelecðCBSÞ ꢀ
D
EZPE
þ
D
ECV
þ
D
ESR
þ
D
ESO
(2)
with the known [87,88] heats of formation at 0 K for the
elements,
kcal=mol, and
D
Hf0ðOÞ ¼ 58:99 kcal=mol,
D
Hf0ðFÞ ¼ 18:47 ꢅ 0:07
D
Hf0ðSeÞ ¼ 54:11 kcal=mol, we can derive
D
Hf0
values for the molecules under study. Heats of formation at
298 K were obtained by following the procedures outlined by
Curtiss et al. [89].
All CCSD(T) calculations were performed with the MOLPRO
program system [90] on the Dell Intel cluster at UA or on a Penguin
AMD cluster at UA. The DFT and nonrelativistic NMR chemical shift
calculations were done with the Gaussian program system [91]
The ZORA NMR calculations were done with the ADF program
system [39].
Acknowledgements
[51] D.A. Dixon, D.J. Grant, K.O. Christe, K.A. Peterson, Inorg. Chem. 47 (2008) 5485–
5494.
[52] D.A. Dixon, T.-H. Wang, D.J. Grant, K.A. Peterson, K.O. Christe, Inorg. Chem. 46
(2007) 10016–10021.
[53] M. Kaupp, Ch. van Wuellen, R. Franke, F. Schmitz, W. Kutzelnigg, J. Am. Chem. Soc.
118 (1996) 11939–11950.
[54] M. Klobukowski, J. Comput. Chem. 14 (1993) 1234–1239.
[55] A. Ellern, J.A. Boatz, K.O. Christe, T. Drews, K. Seppelt, Z. Anorg. Allg. Chem. 628
(2002) 1991–1999.
[56] H. Oberhammer, K.O. Christe, Inorg. Chem. 21 (1982) 273–275.
[57] G.C. Pimentel, J. Chem. Phys. 19 (1951) 446–448.
Financial support from the University of Munich, the Fonds der
Chemischen Industrie, and the Deutsche Forschungsgemeinschaft
(KL636/10-1) is gratefully acknowledged. The work at USC was
financially supported by the National Science Foundation, the Air
Force Office of Scientific Research, the Office of Naval Research, and
the Defense Threat Reduction Agency. This research was sup-
ported, in part, by the U.S. Department of Energy, Office of Basis
Energy Research, Chemical Sciences, in the geosciences program.
[58] R.J. Hach, R.E. Rundle, J. Am. Chem. Soc. 73 (1951) 4321–4324.
[59] I. Love, J. Phys. Chem. A 113 (2009) 2640–2646.
[60] R.F.W. Bader, Atoms in Molecules—A Quantum Theory, Oxford University Press,
Oxford, 1990.
References
[61] K.O. Christe, H. Oberhammer, Inorg. Chem. 20 (1981) 296–297.
[62] K.O. Christe, XXIVth International Congress of Pure and Applied Chemistry, vol. 4,
1974, 115–141.
[63] K.O. Christe, W. Sawodny, P. Pulay, J. Mol. Struct. 21 (1974) 158–164.
[64] R.S. Berry, J. Chem. Phys. 32 (1960) 933–938.
[1] W. Nakanishi, S. Hayashi, J. Phys. Chem. A 103 (1999) 6074–6081.
[2] R. Kniep, L. Korte, R. Kryschi, W. Poll, Angew. Chem., Int. Ed. 23 (1984) 388–389.
[3] K. Seppelt, Z. Anorg. Allg. Chem. 416 (1975) 12–18.
[4] I.C. Bowater, R.D. Brown, F.R. Burden, J. Mol. Spectrosc. 28 (1968) 454–460.

K.O. Christe et al. / Journal of Fluorine Chemistry 131 (2010) 791–799
799
[65] (a) D. Feller, D.A. Dixon, J. Phys. Chem. A 104 (2000) 3048–3056;
(b) D. Feller, D.A. Dixon, J. Chem. Phys. 115 (2001) 3484–3496;
(c) D.A. Dixon, D. Feller, K.A. Peterson, J. Chem. Phys. 115 (2001) 2576–2581;
(d) D. Feller, D.A. Dixon, J. Phys. Chem. A 107 (2003) 9641–9651;
(e) D.A. Dixon, D. Feller, K.O. Christe, W.W. Wilson, A. Vij, V. Vij, H.D.B. Jenkins,
R.M. Olson, M.S. Gordon, J. Am. Chem. Soc. 126 (2004) 834–843;
(f) D.A. Dixon, M. Gutowski, J. Phys. Chem. A 109 (2005) 5129–5135;
(g) L. Pollack, T.L. Windus, W.A. de Jong, D.A. Dixon, J. Phys. Chem. A 109 (2005)
6934–6938;
[80] SADABS, Software for the CCD Detector System, Bruker AXS Madison, WI, 2001.
[81] L.J. Farrugia, J. Appl. Cryst. 30 (1997) 565.
[82] (a) P.J. Knowles, C. Hampel, H.-J. Werner, J. Chem. Phys. 99 (1993) 5219–5227;
(b) M.J.O. Deegan, P.J. Knowles, J. Chem. Phys. Lett. 227 (1994) 321–326.
[83] K.A. Peterson, D.E. Woon, T.H. Dunning Jr., J. Chem. Phys. 100 (1994) 7410–7415.
[84] (a) K.A. Peterson, T.H. Dunning Jr., J. Chem. Phys. 117 (2002) 10548–10560;
(b) D.E. Woon, T.H. Dunning Jr., J. Chem. Phys. 98 (1993) 1358–1371.
[85] E.R. Davidson, Y. Ishikawa, G.L. Malli, Chem. Phys. Lett. 84 (1981) 226–227.
[86] C.E. Moore, Atomic energy levels as derived from the analysis of optical spectra,
Vol. 1, H to V, U.S. National Bureau of Standards Circular 467, U.S. Department of
Commerce, National Technical Information Service, COM-72-50282, Washington,
DC, 1949.
[87] M.W. Chase, Jr., NIST-JANAF Tables (4th edition), J. Phys. Chem. Ref. Data, Mono. 9,
Suppl. 1 (1998).
[88] D.D. Wagman, W.H. Evans, V.B. Parker, R.H. Schumm, I. Halow, S.M. Bailey, K.L.
Churney, R.L. Nuttall, J. Phys. Chem. Ref. Data 11 (1982) (Suppl. 2).
[89] L.A. Curtiss, K. Raghavachari, P.C. Redfern, J.A. Pople, J. Chem. Phys. 106 (1997)
1063–1079.
[90] H.-J. Werner, P.J. Knowles, R. Lindh, F.R. Manby, M. Schu¨ tz, P. Celani, T. Korona, A.
Mitrushenkov, G. Rauhut, T.B. Adler, R.D. Amos, A. Bernhardsson, A. Berning, D.L.
Cooper, M.J.O. Deegan, A.J. Dobbyn, F. Eckert, E. Goll, C. Hampel, G. Hetzer, T.
Hrenar, G. Knizia, C. Ko¨ppl, Y. Liu, A.W. Lloyd, R.A. Mata, A.J. May, S.J. McNicholas,
W. Meyer, M.E. Mura, A. Nicklaß, P. Palmieri, K. Pflu¨ger, R. Pitzer, M. Reiher, U.
Schumann, H. Stoll, A.J. Stone, R. Tarroni, T. Thorsteinsson, M. Wang, A. Wolf,
MOLPRO, Version 2009. A Package of ab initio Programs, University of Cardiff/
Universita¨t Stuttgart, Wales, UK/Stuttgart, Germany, 2009.
[91] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman,
J.A. Montgomery Jr., T. Vreven, K.N. Kudin, J.C. Burant, J.M. Millam, S.S. Iyengar, J.
Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G.A. Petersson, H.
Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T.
Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J.E. Knox, H.P. Hratchian, J.B.
Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R.E. Stratmann, O. Yazyev,
A.J. Austin, R. Cammi, C. Pomelli, J.W. Ochterski, P.Y. Ayala, K. Morokuma, G.A.
Voth, P. Salvador, J.J. Dannenberg, V.G. Zakrzewski, S. Dapprich, A.D. Daniels, M.C.
Strain, O. Farkas, D.K. Malick, A.D. Rabuck, K. Raghavachari, J.B. Foresman, J.V.
Ortiz, Q. Cui, A.G. Baboul, S. Clifford, J. Cioslowski, B.B. Stefanov, G. Liu, A.
Liashenko, P. Piskorz, I. Komaromi, R.L. Martin, D.J. Fox, T. Keith, M.A. Al-Laham,
C.Y. Peng, A. Nanayakkara, M. Challacombe, P.M.W. Gill, B. Johnson, W. Chen,
M.W. Wong, C. Gonzalez, J.A. Pople, Gaussian 03. Revision E.01, Gaussian, Inc.,
Wallingford, CT, 2004.
(h) D. Feller, K.A. Peterson, D.A. Dixon, J. Chem. Phys. 129 (2008) 204105 (32 pp.);
(i) B. Ruscic, A.F. Wagner, L.B. Harding, R.L. Asher, D. Feller, D.A. Dixon, K.A.
Peterson, Y. Song, X. Qian, C. Ng, J. Liu, W. Chen, D.W. Schwenke, J. Phys. Chem. A
106 (2002) 2727–2747.
[66] K.O. Christe, D.A. Dixon, D. McLemore, W.W. Wilson, J. Sheehy, J.A. Boatz, J.
Fluorine Chem. 101 (2000) 151–153; Chem. Eng. News, pp. 48–49, March 3, 2003.
[67] D.F. Shriver, M.A. Drezdzon, The Manipulation of Air Sensitive Compounds, Wiley,
New York, 1986.
[68] T.M. Klapo¨tke, B. Krumm, M. Scherr, R. Haiges, K.O. Christe, Angew. Chem., Int. Ed.
46 (2007) 4686–8690.
[69] K.O. Christe, W.W. Wilson, C.J. Schack, R.D. Wilson, Inorg. Synth. 24 (1986) 39–48.
[70] K.O. Christe, W.W. Wilson, C.J. Schack, J. Fluorine Chem. 11 (1978) 71–85.
[71] R. Haiges, T. Schroer, M. Yousufuddin, K.O. Christe, Z. Anorg. Allg. Chem. 631
(2005) 2691–2695.
[72] K.O. Christe, W.W. Wilson, R.D. Wilson, R. Bau, J. Feng, J. Am. Chem. Soc. 112
(1990) 7619–7625.
[73] A. Rosenheim, L. Krause, Z. Anorg. Allg. Chem. 118 (1921) 177–192.
[74] A. Altomare, G. Cascarano, C. Giacovazzo, A. Guagliardi, A.G.G. Moliterni, M.C.
Burla, G. Polidori, M. Camalli, R. Spagna, SIR97: A New Program for Solving and
Refining Crystal Structures, Istituto de Cristallografia, Bari, Italy, 1997 http://
www.ic.cnr.it/sir97.php.
[75] (a) G.M. Sheldrick, SHELXS-90, Program for the Solution of Crystal Structure,
University of Go¨ttingen, Germany, 1990, SHELXL-97, Program for the Refinement
of Crystal Structure, University of Go¨ttingen, Germany, 1997.;
(b) SHELXTL 6.12 for Windows NT/2000, Program library for Structure Solution
and Molecular Graphics, Bruker AXS, Madison, WI, 2000.
[76] A.L. Spek, SQUEEZE software, University of Utrecht, The Netherlands, 2001.
[77] P. Van der Sluis, A.L. Spek, Acta Crystallogr. A 46 (1990) 194–201.
[78] SMART V 5.625, Software for the CCD Detector System, Bruker AXS Madison, WI,
2001.
[79] SAINT V 6.22, Software forthe CCDDetector System, BrukerAXSMadison, WI, 2001.