XeOF3-, an example of an AX3YE2 valence shell electron pair repulsion arrangement; Syntheses and structural characterizations of [M][XeOF3] (M = Cs, N(CH3) 4)

Article abstract of DOI:10.1021/ja103730c

The XeOF₃^- anion has been synthesized as its Cs ⁺ and N(CH₃)₄⁺ salts and structurally characterized in the solid state by low-temperature Raman spectroscopy and quantum-chemical calculations. Vibrational frequency assignments for [Cs][XeOF₃] and [N(CH₃) ₄][XeOF₃] were aided by ¹⁸O enrichment. The calculated anion geometry is based on a square planar AX₃YE ₂ valence-shell electron-pair repulsion arrangement with the longest Xe-F bond trans to the oxygen atom. The F-Xe-F angle is bent away from the oxygen atom to accommodate the greater spatial requirement of the oxygen double bond domain. The experimental vibrational frequencies and trends in their isotopic shifts are reproduced by the calculated gas-phase frequencies at several levels of theory. The XeOF₃^- anion of the Cs ⁺ salt is fluorine-bridged in the solid state, whereas the anion of the N(CH₃)₄⁺ salt has been shown to best approximate the gas-phase anion. Although [Cs][XeOF₃] and [N(CH ₃)₄][XeOF₃] are shock-sensitive explosives, the decomposition pathways for the anions have been inferred from their decomposition products at 20°C. The latter consist of XeF₂, [Cs][XeO₂F₃], and [N(CH₃)₄][F]. Enthalpies and Gibbs free energies of reaction obtained from Born-Fajans-Haber thermochemical cycles support the proposed decomposition pathways and show that both disproportionation to XeF₂, [Cs][XeO₂F₃], and CsF and reduction to XeF₂, CsF, and O₂ are favorable for [Cs][XeOF₃], while only reduction to XeF₂ accompanied by [N(CH₃)₄][F] and O₂ formation are favorable for [N(CH₃)₄][XeOF₃]. In all cases, the decomposition pathways are dominated by the lattice enthalpies of the products.

Full text of DOI:10.1021/ja103730c

Published on Web 07/15/2010
-
XeOF3 , an Example of an AX3YE2 Valence Shell Electron Pair
Repulsion Arrangement; Syntheses and Structural
Characterizations of [M][XeOF3] (M ) Cs, N(CH3)4)
David S. Brock, He´le`ne P. A. Mercier, and Gary J. Schrobilgen*
Department of Chemistry, McMaster UniVersity, Hamilton, Ontario L8S 4M1, Canada
Received May 1, 2010; E-mail: schrobil@mcmaster.ca
Abstract: The XeOF3 anion has been synthesized as its Cs+ and N(CH3)4 salts and structurally
characterized in the solid state by low-temperature Raman spectroscopy and quantum-chemical calculations.
Vibrational frequency assignments for [Cs][XeOF3] and [N(CH3)4][XeOF3] were aided by 18O enrichment.
The calculated anion geometry is based on a square planar AX3YE2 valence-shell electron-pair repulsion
arrangement with the longest Xe-F bond trans to the oxygen atom. The F-Xe-F angle is bent away from
the oxygen atom to accommodate the greater spatial requirement of the oxygen double bond domain. The
experimental vibrational frequencies and trends in their isotopic shifts are reproduced by the calculated
gas-phase frequencies at several levels of theory. The XeOF3- anion of the Cs+ salt is fluorine-bridged in
-
+
+
the solid state, whereas the anion of the N(CH3)4 salt has been shown to best approximate the gas-
phase anion. Although [Cs][XeOF3] and [N(CH3)4][XeOF3] are shock-sensitive explosives, the decomposition
pathways for the anions have been inferred from their decomposition products at 20 °C. The latter consist
of XeF2, [Cs][XeO2F3], and [N(CH3)4][F]. Enthalpies and Gibbs free energies of reaction obtained from
Born-Fajans-Haber thermochemical cycles support the proposed decomposition pathways and show that
both disproportionation to XeF2, [Cs][XeO2F3], and CsF and reduction to XeF2, CsF, and O2 are favorable
for [Cs][XeOF3], while only reduction to XeF2 accompanied by [N(CH3)4][F] and O2 formation are favorable
for [N(CH3)4][XeOF3]. In all cases, the decomposition pathways are dominated by the lattice enthalpies of
the products.
copy.9 The synthesis (eqs 1 and 2) utilized anhydrous HF (aHF)
Introduction
as the solvent medium, which was complicated by the fact that
Although the xenon(IV) fluoride species XeF3 ,1-3 XeF4,4-6
+
8
both F- and XeOF2 form HF solvates. Removal of HF under
and XeF5 7 have been well characterized spectroscopically and
by single-crystal X-ray diffraction, the absence of a facile
synthetic route to high-purity XeOF2 had prevented extensive
exploration of its fluoride ion donor-acceptor properties. The
-
dynamic vacuum at -78 °C produced a mixture of XeOF2 and
[Cs][F(HF)n].
HF
8
CsF + mHF
8 [Cs][F(HF)m]
(1)
recent synthesis of pure XeOF2 has provided an opportunity
-78 °C
to extend the oxide fluoride chemistry of Xe(IV) and to study
the fluoride ion acceptor properties of XeOF2 with the view to
[Cs][F(HF)m] + XeOF2 dynamic vac8 [Cs][XeOF3] + mHF (2)
-
synthesize and characterize salts of the XeOF3 anion.
0 °C
A prior publication has reported the synthesis of [Cs][XeOF3]
and its characterization by low-temperature Raman spectros-
Slow warming of the product mixture to room temperature under
dynamic vacuum resulted in further removal of HF. The authors
concluded that, upon solvent removal, the reaction proceeded
to the formation of [Cs][XeOF3], along with traces of
[Cs][XeO2F3]; the latter was attributed to a disproportionation
(eq 3).9 The present work will show that a mixture of XeF2,
XeOF2, [Cs][XeF5], and [Cs][XeO3F] was actually formed in
the earlier reported work and that the XeOF3- anion has eluded
synthesis until the present work.
(1) Gillespie, R. J.; Schrobilgen, G. J.; Landa, B. J. Chem. Soc. D 1971,
1543–1544.
(2) Boldrini, P.; Gillespie, R. J.; Ireland, P. R.; Schrobilgen, G. J. Inorg.
Chem. 1974, 13, 1690–1694.
(3) McKee, D. E.; Zalkin, A.; Bartlett, N. Inorg. Chem. 1973, 12, 1713–
1717.
(4) Claassen, H. H.; Chernick, C. L.; Malm, J. G. J. Am. Chem. Soc. 1963,
85, 1927–1928.
(5) Schumacher, G. A.; Schrobilgen, G. J. Inorg. Chem. 1984, 23, 2923–
2929.
2[Cs][XeOF3] HF 8 [Cs][XeO2F3] + XeF2 + CsF (3)
(6) Levy, H. A.; Burns, J. H.; Agron, P. A. Science 1963, 139, 1208–
1209.
0 °C
(7) Christe, K. O.; Curtis, E. C.; Dixon, D. A.; Mercier, H. P.; Sanders,
J. C. P.; Schrobilgen, G. J. J. Am. Chem. Soc. 1991, 113, 3351–3361.
(8) Brock, D. S.; Bilir, V.; Mercier, H. P. A.; Schrobilgen, G. J. J. Am.
Chem. Soc. 2007, 129, 3598–3611.
(9) Gillespie, R. J.; Schrobilgen, G. J. J. Chem. Soc., Chem. Commun.
1977, 595–597.
9
10.1021/ja103730c  2010 American Chemical Society
J. AM. CHEM. SOC. 2010, 132, 10935–10943 10935

A R T I C L E S
Brock et al.
The XeOF3- anion is of special interest because, to the best
of the authors’ knowledge, it represents the only example of an
AX3YE2 valence-shell electron-pair repulsion (VSEPR)10 ar-
rangement in which the valence electron lone-pair domains and
a double-bond domain occupy positions where they are ap-
proximately 90° to one another (Structure I).
further pumping, vigorous mixing, and slow warming to -10
°C to remove additional bound HF, the original pale yellow
color of the sample intensified. The Raman spectrum of the
resulting solid revealed a mixture of XeOF2 and [Cs][XeOF3]
(and presumably [Cs][F(HF)n], where n , m).
The previously reported synthesis of “[Cs][XeOF3]”, which
was carried out at 0 °C,9 closely resembles the present synthesis,
which was carried out at -10 °C. However, the Raman spectra
of their respective products are significantly different. The
Raman spectrum of the previously reported product mixture9
has now been reassigned to XeF2, XeOF2, [Cs][XeF5], and
[Cs][XeO3F] (Table S1 in the Supporting Information). It is
likely that none of the reactions leading to the latter products
involve [Cs][XeOF3] as an intermediate because none of its
vibrational bands were observed in the Raman spectra of the
product mixtures based on the present Raman assignments for
[Cs][XeOF3] (cf. Tables 1 and S1). The neutral molecules,
XeOF2 and XeF2, likely form upon removal of HF from
XeOF2 ·nHF (eq 5),8 followed by either xenon(IV) reduction
to xenon(II) (eq 6)8 or disproportionation (eq 7).8,9 The previous
report also noted that the reaction of XeF4 and H2O in HF
solvent at -60 °C, which was used to synthesize XeOF2, was
incomplete and therefore contained unreacted XeF4 and H2O.9
Xenon tetrafluoride, which has been shown to react with CsF
to form [Cs][XeF5],7 presumably reacts with [Cs][F(HF)n]
according to eq 8.11 Water and [Cs][F(HF)n] would form
[Cs][XeO3F] by either hydrolysis of XeO2F2 (eq 9),12 followed
by fluoride ion addition13 (eq 10), or by fluoride ion addition
to XeO2F214 (eq 11), followed by hydrolysis (eq 12). Equations
11 and 12 represent the most likely route to [Cs][XeO3F]
because [Cs][XeO2F3] was often observed in the Raman spectra
The present paper details the syntheses of the Cs+ and
N(CH3)4+ salts of the XeOF3- anion and their characterization
by Raman spectroscopy. Quantum-chemical calculations and
18O enrichment have been employed to assign the Raman spectra
-
of the Xe16/18OF3 anion and to aid in the assessment of its
chemical bonding.
Results and Discussion
Syntheses of [M][XeOF3 ] (M ) N(CH4), Cs). (a) [M]-
[XeOF3] in CH3CN Solvent. Reactions and the purities of all
products were routinely monitored by recording the low-
temperature Raman spectra (-150 °C) of the natural abundance
and 18O-enriched (98.6 atom %) XeOF3 salts.
-
The XeOF3 anion was obtained as the N(CH3)4 and Cs+
salts by the low-temperature reaction of XeOF2 with
[N(CH3)4][F](CsF) in dry CH3CN solvent according to eq 4,
where M ) N(CH3)4, Cs.
-
+
9
-
of product mixtures but XeO3 was not. The hydrolysis of XeF5
in the presence of unreacted H2O, which competes with Xe(VI)
species in eqs 9 and 12, does not occur and is consistent with
the higher oxophilicities of Xe(VI) species (see Reactivities of
XeOF2 + [M][F] CH CN8 [M][XeOF3]
(4)
3
-45 °C
-
XeOF3 Salts).
Tetramethylammonium fluoride or CsF was added in ca. 3-5%
excess to circumvent possible XeOF2 contamination of the
product. The [N(CH3)4][XeOF3] salt precipitated from solution
as a very pale yellow, amorphous powder and was isolated by
removal of the solvent under dynamic vacuum at -45 to -42
°C over a period of several hours. The synthesis of [Cs][XeOF3]
was complicated by the low solubilities of CsF and [Cs][XeOF3]
in CH3CN solvent. Upon warming, XeOF2 reacted in varying
degrees with CsF, forming insoluble [Cs][XeOF3] in admixture
with unreacted CsF. Unreacted XeOF2, which imparted a yellow
color to the supernatant, remained in solution and was removed
by use of a cannula as previously described.8 The residual
solvent was removed under dynamic vacuum at -45 to -42
°C over a period of several hours, leaving behind a pale yellow
powder. Reaction of [M][F] with XeOF2 in 2:1 molar ratios
yielded only [M][XeOF3] and did not result in the formation of
[M]2[XeOF4]. Conversely, reaction of the fluoride salts with
XeOF2 in 1:2 molar ratios did not result in the formation of
XeOF2 · nHF f XeOF2 + nHFv
XeOF2 f XeF2 + ½O2
(5)
(6)
(7)
(8)
(9)
XeOF2 f ½XeF2 + ½XeO2F2
XeF4 + [Cs][F(HF)n] f [Cs][XeF5] + nHFv
XeO2F2 + H2O f XeO3 + 2HF
XeO3 + [Cs][F(HF)n] f [Cs][XeO3F] + nHFv (10)
XeO2F2 + [Cs][F(HF)n] f [Cs][XeO2F3] + nHFv
[Cs][XeO2F3] + H2O f [Cs][XeO3F] + 2HF
(11)
(12)
(c) An Alternative Synthesis of [Cs][XeOF3]. High-purity
[Cs][XeOF3] was obtained when HF solvent was removed from
a stoichiometric mixture of CsF and XeOF2 at -78 °C, followed
by warming the reaction mixture to -45 °C under dynamic
vacuum. Solid XeOF2, which forms an insoluble HF solvate,8
appeared to liquify and became intense yellow in color prior to
formation of a dry yellow powder under dynamic vacuum. The
Raman spectrum of this intermediate mixture showed only
XeOF2. It is presumed that [Cs][F(HF)m] was also present but
could not be observed in the Raman spectrum because the bands
were too broad and weak. Extraction of residual HF with an
-
F2OXe---F---XeOF2 salts, yielding only [M][XeOF3] and
unreacted XeOF2.
(b) Attempts To Replicate the Prior Synthesis of “[Cs]-
[XeOF3]” in aHF. A 1:1 molar ratio of CsF and XeOF2 was
prepared in aHF and dried under dynamic vacuum at -78 °C.
The product was shown by Raman spectroscopy to be
XeOF2 ·nHF, but presumably also contained [Cs][F(HF)m]. With
(10) Gillespie, R. J.; Hargittai, I. The VSEPR Model of Molecular Geometry;
Allyn and Bacon: Boston, MA, 1991; pp 154-155.
9
10936 J. AM. CHEM. SOC. VOL. 132, NO. 31, 2010

XeOF3-, an Example of a VSEPR Arrangement
A R T I C L E S
XeOF3 + H2O CH3CN8 [XeO2F-] + 2HF
(15)
(16)
aliquot of CH3CN at -20 °C resulted in high-purity
[Cs][XeOF3]. The present method avoids decomposition arising
from higher reaction temperatures and heterogeneous reaction
conditions that result from the synthetic procedure outlined in
the section above by using CH3CN to partially or completely
solubilize XeOF2 and CsF.
-
-45 °C
[XeO2F-] CH3CN8 ½XeO3F- + ½Xe + ½O2 + ½F-
-45 °C
-
Reactivities of XeOF3 Salts. The insolubilities and/or reac-
tivities of the N(CH3)4+ and Cs+ salts in CH3CN, SO2, SO2ClF,
HF, XeOF4, ONF, and O2NF have prevented crystal growth and
NMR characterization.
Attempts to dissolve [N(CH3)4][XeOF3] in SO2 at -78 °C
resulted in sample detonation, and it is proposed that XeOF3
Hydrolytic and Thermal Stabilities of [M][XeOF3] (M )
N(CH4), Cs). Dry [N(CH3)4][XeOF3] is shock sensitive at low
temperatures but does not exhibit shock sensitivity under CH3CN
upon standing at room temperature for 10 min. Dry
[N(CH3)4][XeOF3] is kinetically stable at -78 °C for indefinite
periods of time but begins to slowly decompose upon warming
from 0 to 10 °C. The low-temperature (-150 °C) Raman spectra
of samples briefly warmed to 25 °C and then quenched at -150
-
oxidizes SO2 to SO3 according to eq 17.
neat
15
°C show the growth of the νs(XeF2) band of XeF2 at 498 cm-1
[N(CH3)4][XeOF3] + SO2
8 XeF2 + SO3 +
-78 °C
and the ν(CH3) bands of [N(CH3)4][F] at 2957 and 3038
16-18
cm-1
.
The [Cs][XeOF3] salt is also stable at -78 °C for
[N(CH3)4][F] (17)
indefinite periods of time but slowly decomposes at room
temperature to XeF2, XeO2F3-, O2, and CsF. Although
[Cs][XeOF3] appears to be less shock sensitive than the
An attempt to dissolve [Cs][XeOF3] in XeOF4 at ca. 20 °C
resulted in oxygen/fluorine metathesis to give XeF4, XeO2F2,
and [Cs][F(XeOF4)m]20 according to eq 18.
+
N(CH3)4 salt, detonations have occurred when attempting to
transfer the finely powdered salt in a drybox.
Two decomposition pathways have been inferred for XeOF3
on the basis of these observations. The major decomposition
pathway is through reduction to Xe(II) (eq 13), and the minor
decomposition pathway is through disproportionation (eq 14)
to Xe(II) and Xe(VI) (see Thermochemistry).
[Cs][XeOF3] + (m + 1)XeOF4 neat 8 XeF4 + XeO2F2 +
-
20 °C
[Cs][F(XeOF4)m] (18)
The syntheses of the NO+ and NO2 salts of XeOF3 were
attempted. Contact of ONF with XeOF2 at -78 °C resulted in
an explosion in which XeOF2 likely oxidizes ONF to O2NF
(eq 19).
+
-
[M][XeOF3] f XeF2 + ½O2 + [M][F]
(13)
(14)
neat
[M][XeOF3] f ½XeF2 + ½[M][XeO2F3] + ½[M][F]
ONF + XeOF2
8 O2NF + XeF2
(19)
-78 °C
While [Cs][XeOF3] and [N(CH3)4][XeOF3] can be obtained
as pure salts when exact stoichiometric ratios of XeF4 and H2O
are used for the synthesis of XeOF2, both salts hydrolyze when
H2O is used in excess. Characterization of the hydrolysis
products by Raman spectroscopy showed several bands in the
frequency range 720-785 cm-1. The bands appearing between
720 and 765 cm-1 correspond to the ν(XeO) regions for XeOF2,8
F2OXeNCCH3,8 XeOF2 ·nHF,8 and XeOF3-, while the bands
appearing between 745 and 785 cm-1 correspond to the νs(XeO3)
regions for XeO3F-13 and XeO3.19 The findings indicate that
the decomposition pathway resulting from hydrolysis involves
disproportionation to Xe(VI)-containing species that likely
occurs according to eqs 15 and 16. Although the exact nature
of the decomposition product(s) could not be established, it is
likely to be a compound(s) having the general formulation(s)
[Cs]x[(XeO3)yFx(XeOF2)z].
The white residue that remained and coated the fractured FEP
reaction vessel had an odor reminiscent of XeF2. Addition of
O2NF to XeOF2 at -78 °C resulted in an immediate color
change from yellow to white upon contact with liquid O2NF.
The Raman spectrum of the product mixture showed modes
corresponding to XeF4, N2O5 (the spectrum corresponded to the
21
-
ionic solid-state formulation [NO2][NO3] ), and XeF5 anion
bands. The Raman spectrum (Table S2 in the Supporting
Information) indicates that XeOF2 transfers oxygen to O2NF to
give XeF4 and N2O5 (eq 20) and a subsequent reaction of O2NF
-
with XeF4 gives the XeF5 anion (eq 21).
neat
2O2NF + XeOF2
XeF4 + O2NF
8 XeF4 + N2O5
(20)
(21)
-78 °C
neat
8 [NO2][XeF5]
-78 °C
(11) To confirm this assumption, an intimate mixture of XeF4 and
[Cs][F(HF)n] was created by dissolving an equimolar mixture of XeF4
and CsF in aHF. The solution was quenched at-196 °C, and HF was
removed under dynamic vacuum at-78 °C. Warming this XeF4/
[Cs][F(HF)n] mixture to 0 °C under dynamic vacuum yielded a small
amount of [Cs][XeF5] in admixture with XeF4, which was verified by
Raman spectroscopy at-150 °C.
Raman Spectroscopy. The low-temperature Raman spectra
of [N(CH3)4][Xe16/18OF3] and [Cs][Xe16/18OF3] are shown in
Figures 1 and 2, respectively. The observed and calculated
frequencies and their assignments are listed in Tables 1 and 2
and in Tables S3 and S4 of the Supporting Information. In the
absence of a crystal structure, the energy-minimized geometry
(12) Huston, J. L. J. Phys. Chem. 1967, 71, 3339–3341.
(13) LaBonville, P.; Ferraro, J. R.; Spittler, T. M. J. Chem. Phys. 1971,
55, 631–640.
(14) Christe, K. O.; Wilson, W. W. Inorg. Chem. 1988, 27, 3763–3768.
(15) Agron, P. A.; Begun, G. M.; Levy, H. A.; Mason, A. A.; Jones, C. G.;
Smith, D. F. Science 1963, 139, 842–844.
-
(20) There is precedent for polynuclear anions of the type F(XeOF4)m
with the synthesis and structural characterization of the (XeOF4)3F-
anion as its Cs+ salt: Holloway, J. H.; Kaucˇicˇ, V.; Martin-Rovet, D.;
Russell, D. R.; Schrobilgen, G. J.; Selig, H. Inorg. Chem. 1985, 24,
678–683.
(16) Berg, R. W. Spectrochim. Acta 1978, 34A, 655–659.
(17) Kabisch, G.; Klose, M. J. Raman Spectrosc. 1978, 7, 311–315.
(18) Kabisch, G. J. Raman Spectrosc. 1980, 9, 279–285.
(19) Smith, D. F. J. Am. Chem. Soc. 1963, 85, 816–817.
(21) Wilson, W. W.; Christe, K. O Inorg. Chem. 1987, 26, 1631–1633.
9
J. AM. CHEM. SOC. VOL. 132, NO. 31, 2010 10937

A R T I C L E S
Brock et al.
Figure 3. Calculated geometries [B3LYP/aug-cc-pVTZ(-PP)] for (a)
-
XeOF3 and (b) XeOF2.
with the previous literature.16-18 The XeOF3 anion of
-
[N(CH3)4][XeOF3] is in overall good agreement with the
calculated gas-phase anion. As predicted from the calculated
frequencies and Raman intensities, the highest frequency and
most intense XeOF3- band occurs at 730.1 cm-1 and is assigned
to ν(XeO). This mode displays a large low-frequency shift
(-36.1 cm-1) upon 18O substitution that is in very good
agreement with the calculated values (-33.5 to -42.8 cm-1).
The asymmetric νas(XeF2a) mode is expected to be weak to very
weak in the Raman spectrum and was not observed. The bands
at 457.3/471.3 and 447.6 cm-1 are of medium intensity and are
assigned to the in-plane stretching modes νs(XeF2a) + ν(XeFb)
and νs(XeF2a) - ν(XeFb), respectively. Their frequencies are
essentially unshifted upon 18O-enrichment, which is expected
for modes that do not involve oxygen atom displacements. The
band at 332.0 cm-1 is assigned to the out-of-plane deformation
mode, δ(XeOF3), and appears as a weak band, in agreement
with the calculated Raman intensity, and shows no 18O
dependence although the calculations predict a small isotopic
shift (-2.8 to -3.8 cm-1). The split band at 268.7, 271.5 cm-1
is sensitive to isotopic substitution and is assigned to the in-
plane deformation mode, Frock(XeOF2a). The experimental 18O
isotopic shift for this mode, -9.3, -9.6 cm-1, is in good
agreement with the calculated values (-7.5 to -9.3 cm-1). The
weak bands at 189.5, 193.9 cm-1 are insensitive to 18O
substitution and are assigned to the in-plane XeF2 bending mode.
The band at 156.5 cm-1 is assigned to the out-of-plane δ(XeOFb)
- δ(XeF2a) mode and shows a small isotopic shift (-3.6 cm-1),
in agreement with the calculated values (-1.4 to -1.9 cm-1).
The lowest frequency band at 137.8 cm-1 is also weak,
displaying no isotopic shift, in accordance with the small
calculated values (-0.1 to -1.2 cm-1), and is assigned to the
in-plane deformation mode, Frock(OXeF2a) + δ(FbXeFa).
The low-frequency shifts of the stretching modes relative to
their counterparts in XeOF2 and F2OXeNCCH38 are consistent
with anion formation and increased Xe-O and Xe-F bond
polarities in the anion. These low-frequency shifts are repro-
duced by the quantum-chemical calculations.
Figure 1. Raman spectra of natural abundance (lower trace) and 97.8%
18O-enriched (upper trace) [N(CH3)4][XeOF3] recorded at -150 °C using
1064-nm excitation. Symbols denote FEP sample tube lines (*), instrumental
artifact (†), and minor hydrolysis product(s) (‡).
Figure 2. Raman spectra of natural abundance (lower trace) and 97.8%
18O-enriched (upper trace) Cs[XeOF3] recorded at -150 °C using 1064-
nm excitation. Symbols denote FEP sample tube lines (*), instrumental
artifact (†), and minor hydrolysis product(s) (‡).
of the XeOF3- anion was calculated at several levels of theory
(Tables 2, S3, and S4) using the aug-cc-pVDZ(-PP) and aug-
cc-pVTZ(-PP) basis sets. The calculated vibrational frequencies,
intensities, and 16/18O-isotopic shifts were used to assign the
Raman spectra of Xe16OF3 and Xe18OF3 in their Cs+ and
-
-
+
N(CH3)4 salts. Regardless of the level of theory or basis set
used, the frequency, intensity, and isotopic shift trends are
similar and consistent. The frequency ranges cited in the ensuing
discussion refer to isotopic shifts and are the values obtained
for the entire range of theory levels, showing good agreement
with experiment regardless of the level of theory used. The
experimental and calculated vibrational frequencies for mono-
meric XeOF2 and their assignments (Table S5 in the Supporting
(b) [Cs][XeOF3]. The Raman spectrum of the Cs+ salt is very
+
similar to that of the N(CH3)4 salt, although the bands are
shifted to higher frequencies, which likely results from signifi-
cant cation-anion contacts. The bands that are the most affected
are shifted by as much as 30 cm-1 and correspond to modes
that involve oxygen atom displacements, i.e., ν(XeO) and
Information) have been previously discussed8 and were also used
-
to aid in the vibrational assignments of XeOF3
.
F
rock(XeOF2a), indicating that the Cs+ cation likely coordinates
The Xe16/18OF3- anion possesses C2V symmetry, which results
in nine fundamental vibrational modes that span the irreducible
representations 4A1 + 3B1 + 2B2 (the xz-plane is the molecular
plane and the Xe-O bond lies along the z-axis) which are both
Raman and infrared active.
most strongly to the oxygen atom. These findings are in
agreement with the NBO analysis (see Computational Results),
which assigns most of the negative charge to the oxygen atom.
Weak bands between 223 and 276 cm-1 in the Cs+ salt are
absent in the N(CH3)4+ salt. These bands are in the appropriate
region and are of appropriate intensity for stretching modes
associated with fluorine bridges,20 suggesting an oligomeric or
(a) [N(CH3)4][XeOF3]. The frequencies associated with the
N(CH3)4+ cation have been assigned on the basis of comparisons
9
10938 J. AM. CHEM. SOC. VOL. 132, NO. 31, 2010

XeOF3-, an Example of a VSEPR Arrangement
Table 1. Experimental Raman Frequenciesa and Intensitiesb for XeOF3 in Cs[XeOF3] and [N(CH3)4][XeOF3]
A R T I C L E S
-
a Frequencies are given in cm-1 b Values in parentheses denote relative Raman intensities. c Raman spectra were recorded in FEP sample tubes at
.
-150 °C using 1064-nm excitation. d The abbreviations denote shoulder (sh) and not observed (n.o.). e Weak bands at 454.7(4) and 763.1(sh) in the 16O
f
+
spectrum and at 454.5(7) and 724.8(sh) in the 18O spectrum are attributed to a hydrolysis product(s). The N(CH3)4 cation modes were observed at
ν8(E), 377(2); ν19(T2), 460(2); ν3(A1), 761(13); ν18(T2), 951(10); ν7(E), 1176(2), 1187(1); ν17(T2), 1287(1); ν16(T2), 1405(2), 1408(2); ν2(A1), ν6(E),
1462(12), 1479(1), 1487(10); ν(CH3) and combination bands, 2800(2), 2881sh, 2908(3), 2938(3), 2957(12), 2978sh, 2999sh, 3038(18) cm-1 g Weak
.
bands at 756.2(8), 749.4(5), 744.7(3), 721.3(5) 710.0(6), 702.1sh, 488.8(1), 481.9(3), and 478.2(6) in the 18O spectrum are attributed to hydrolysis
product(s). The band at 472.0(16) also partially overlaps with a hydrolysis product band. h The abbreviations denote symmetric (s), asymmetric (as),
stretch (ν), bend (δ), rock (Frock), in-plane bend (i.p.), and out-of-plane bend (o.o.p.). The in-plane and out-of-plane mode descriptions are relative to the
XeOF2aFb plane, i.e., the anion lies in the xz-plane with the Xe-O bond collinear with the z-axis (see Figure 3).
chain structure for [Cs][XeOF3]. Such long contacts most likely
result in coupling between XeOF3- units which not only would
account for the splittings of the vibrational bands but also may
result in symmetry reduction and observation of the νas(XeF2a)
stretching mode at 498.9, 509.7, and 512.8 cm-1 which was
the use of a larger basis set tends to give shorter bond lengths
and smaller O-Xe-Fa bond angles. With the exception of the
BP86 method, where the bonds are significantly elongated, all
bond lengths fall within relatively narrow ranges (Xe-O,
1.821-1.859 [1.854-1.890] Å; Xe-Fa, 2.032-2.064 [2.061-
2.096] Å; Xe-Fb, 2.129-2.195 [2.131-2.205] Å). As predicted
by the VSEPR rules,10 the Xe-Fa bonds are shorter than the
Xe-Fb bond. In general, all bond angles are also within a narrow
range, with the exception of the MP2 value, which gives a larger
O-Xe-Fa angle than other levels of theory. The fluorine atoms,
Fa, are displaced away from the oxygen atom, with an O-Xe-Fa
angle of 92.8-93.7 [93.1-94.2]° and an Fa-Xe-Fa angle of
172.5-174.4 [171.7-173.9]°. When compared with XeOF2, the
+
not observed in the N(CH3)4 salt.
Computational Results. Because a crystal structure for a
XeOF3- salt is not available, a series of quantum-chemical
calculations at different levels of theory with different basis sets
were carried out (Table 2 and 3 and Tables S3, S4, and S6 in the
Supporting Information). All calculations resulted in stationary
points with all frequencies real. The calculations established that
trends in the vibrational frequencies persisted at all levels of theory
and for all basis sets used. The calculated frequencies and their
trends were then compared with the observed trends, allowing
reliable assignments of the experimental Raman frequencies to be
made. The parent compound, XeOF2,8 was used as a benchmark
for the calculations (Table S5). Values for the aug-cc-pVTZ(-PP)
basis set are reported along with those of the aug-cc-pVDZ(-PP)
basis set. The latter are given in square brackets.
-
net -1 charge of XeOF3 results in longer Xe-Fa and Xe-O
bonds and a Fa-Xe-Fa angle that is closer to linearity than in
XeOF2 (164.8-170.8°).
(b) Natural Bond Orbital (NBO) Analyses. The NBO22-25
analyses were carried out for all optimized gas-phase geometries
(22) Reed, A. E.; Weinstock, R. B.; Weinhold, F. J. Chem. Phys. 1985,
83, 735–746.
(a) Geometries. The geometry of the XeOF3- anion optimized
at C2V symmetry, yielding a planar structure that is in accord
with that predicted by the VSEPR model of molecular geom-
etry10 (see Raman Spectroscopy, above). As noted for XeOF2,8
(23) Reed, A. E.; Curtiss, L. A.; Weinhold, F. Chem. ReV. 1998, 88, 899–
926.
(24) Glendening, E. D.; Reed, A. E.; Carpenter, J. E.; Weinhold, F. NBO
Version 3.1; Gaussian Inc.: Pittsburgh, PA, 1990.
9
J. AM. CHEM. SOC. VOL. 132, NO. 31, 2010 10939

A R T I C L E S
Table 2. Selected Calculated Vibrational Frequenciesa and Infrared and Raman Intensitiesb for the Xe16/18OF3 Anionc
Brock et al.
-
MP2
PBE1PBE
B3LYP
assignmentd
-
Xe16OF3
867.9 (25) [241]
493.6 (<1) [283]
460.7 (27) [152]
379.2 (20) [34]
286.8 (<1) [54]
250.6 (4) [<1]
171.2 (<1) [2]
132.6 (<1) [<0.1]
150.5 (<1) [2]
771.2 (59) [160]
491.2 (<0.1) [340]
460.5 (34) [54]
372.3 (10) [131]
269.4 (<1) [48]
259.9 (3) [<1]
183.6 (<1) [1]
136.6 (<1) [<1]
115.6 (1) [1]
729.6 (62) [152]
467.3 (<0.1) [319]
432.7 (37) [66]
356.9 (13) [106]
256.3 (<1) [47]
246.9 (4) [<1]
177.2 (<1) [2]
127.5 (<1) [<1]
119.8 (1) [1]
-
ν(XeO)
νas(XeF2a)
νs(XeF2a) + ν(XeFb)
νs(XeF2a) - ν(XeFb)
δ(XeOF3) o.o.p.
F
rock(OXeF2a) i.p.
δ(XeF2a) i.p.
δ(OXeFb) - δ(XeF2a) o.o.p.
rock(OXeF2a) + δ(FbXeFa) i.p.
F
Xe18OF3
825.1 (22) [224]
494.8 (<1) [284]
461.0 (27) [151]
379.3 (20) [33]
283.0 (<1) [52]
242.6 (4) [<1]
171.2 (<1) [2]
131.1 (<1) [0]
149.3 (<1) [2]
733.4 (53) [51]
492.5 (<0.1) [341]
460.6 (10) [129]
372.3 (10) [129]
266.6 (<1) [46]
250.8 (3) [<1]
183.6 (<1) [1]
134.7 (<1) [<1]
115.0 (1) [1]
694.0 (54) [145]
468.5 (<0.1) [319]
432.8 (38) [65]
356.8 (13) [104]
253.4 (<1) [46]
238.4 (4) [<1]
177.2 (<1) [2]
125.8 (<1) [<1]
119.2 (1) [1]
ν(XeO)
νas(XeF2a)
νs(XeF2a) + ν(XeFb)
νs(XeF2a) - ν(XeFb)
δ(XeOF3) o.o.p.
F
rock(OXeF2a) i.p.
δ(XeF2a) i.p.
δ(OXeFb) - δ(XeF2a) o.o.p.
F
rock(OXeF2a) + δ(FbXeFa) i.p.
a Frequencies are given in cm-1 b Values in parentheses denote calculated Raman intensities (Å4 u-1). Values in square brackets denote calculated
.
infrared intensities (km mol-1). c The aug-cc-pVTZ(-PP) basis set was used. d The abbreviations denote symmetric (s), asymmetric (as), stretch (ν), bend
(δ), rock (Frock), in-plane bend (i.p.), and out-of-plane bend (o.o.p.). The in-plane and out-of-plane mode descriptions are relative to the XeOF2aFb plane
(see Figure 3 and footnote h of Table 1.
a
-
Table 3. Calculated Geometrical Parameters for the XeOF3 Anion
MP2
SVWN5
BP86
PBE1PBE
B3LYP
B3PW91
MPW1PW91
Bond Lengths (Å)
2.106
1.881
2.188
Xe-Fa
Xe-O
Xe-Fb
2.060
1.821
2.129
2.064
1.859
2.143
2.032
1.830
2.173
2.064
1.852
2.195
2.048
1.840
2.178
2.034
1.831
2.175
Bond Angles (deg)
O-Xe-Fa
Fa-Xe-Fb
Fa-Xe-Fa
94.8
85.2
170.3
93.7
86.3
172.6
93.7
86.3
172.5
92.9
87.1
174.3
93.1
86.9
173.9
93.0
87.0
173.9
92.8
87.2
174.4
a The aug-cc-pVTZ(-PP) basis set was used.
-
of XeOF3 and XeOF2 for comparison and are summarized in
Tables S7 and S8 of the Supporting Information. The results at
all levels of theory are similar, and therefore only the MP2
values are referred to in the ensuing discussion.
Xe-Fa and Xe-O/Xe-Fb bond order ratios, 3.43 and 2.65,
respectively, and Xe/O/Fa/Fb valencies (1.509/0.703/0.213/0.257)
are in overall agreement with a composite of these resonance
contributions.
The natural population analysis (NPA) charges of 2.11 and
2.13 for Xe in XeOF3- and XeOF2, respectively, show that the
charge on xenon remains essentially unchanged upon anion
formation and is approximately half of the formal charge that
is given by a purely ionic model, indicating that the anion bonds
-
are polar covalent. Upon formation of XeOF3 from XeOF2,
the net negative charge of the anion is dispersed among the
more electronegative atoms, yielding natural charges for O
(-1.038), Fa (-0.762 each), and Fb (-0.654) that are higher
than the corresponding values in XeOF2. The charge on O is
about half of its formal oxidation number, while the F charges
are approximately two-thirds of their formal oxidation number,
indicating that the bonds are more ionic in the anion than in
neutral XeOF2. Among the plausible valence bond Structures
I-IV and VI-IX for XeOF3- (see Scheme S1 in the Supporting
Information), the calculated charges are best represented by
Structure V, which is an average of Structures II-IV, with a
somewhat larger contribution from Structure II. The Xe-O/
Upon anion formation, donation of electron density by the
fluoride ion into the xenon valence shell results in substantial
decreases in the Xe-O and Xe-F bond orders and Xe valency
and increases in the O and F valencies relative to those of
XeOF2. Thus, the Xe-O and Xe-F bonds are somewhat more
(25) Glendening, E. D.; Badenhoop, J. K.; Reed, A. E.; Carpenter, J. E.;
Bohmann, C. M.; Morales, C. M.; Weinhold, F. NBO Version 5.0;
Theoretical Chemistry Institute, University of Wisconsin: Madison,
WI, 2001.
-
ionic in XeOF3 when compared with those in XeOF2.
9
10940 J. AM. CHEM. SOC. VOL. 132, NO. 31, 2010

XeOF3-, an Example of a VSEPR Arrangement
Table 4. Estimated Volumes, Lattice Enthalpies, and Entropies for N(CH3)4 and Cs+ Salts of F-, XeOF3-, and XeO2F3
A R T I C L E S
+
-
a The lattice enthalpies (∆H°L) were calculated as described in ref 28. b The standard entropies were calculated as described in ref 30. c The formula
unit volumes, Vm, for [N(CH3)4][F] and CsF were obtained from their crystallographic unit cells at -163 (ref 32) and 25 °C (ref 33), respectively. d The
values for Vm[M][XeOF3] were estimated by Vm(F2OXeNCCH3; 0.1324 nm3, ref 8) - Vm(CH3CN; 0.0616 nm3, ref 34) + Vm([M][F]). e The Vm values
for [M][XeOF3] were estimated by Vm([M][F]) + [Vm(XeF2; 0.0622 nm3, ref 35) + Vm(XeO2F2; 0.0842 nm3, ref 36)]/2. f The Vm values for [M][XeO2F3]
were estimated by Vm([M][F]) + Vm(XeO2F2; 0.0842 nm3, ref 36).
Table 5. Values of ∆H°, ∆S°, and ∆G° Calculated for the Decomposition Reactions of [M][XeOF3] (X ) N(CH3)4, Cs)a
∆H° (kJ mol-1
)
∆S° (J mol-1 K-1
)
∆G° (kJ mol-1
)
1
[N(CH3)4][XeOF3](s) f XeF2(s) + /2O2(g) + [N(CH3)4][F](s)
-58.2
-59.6
121.3
18.4
118.1
15.2
-94.3
-0.6
-94.7
-1.1
1
[N(CH3)4][XeOF3](s) f /2XeF2(s)
+
4.9
3.5
1/2[N(CH3)4][XeO2F3](s) + /2[N(CH3)4][F](s)
1
1
[Cs][XeOF3](s) f XeF2(s) + /2O2(g) + [Cs][F](s)
-146.3
-35.6
-145.1
-38.4
121.3
18.4
118.1
15.2
-182.4
-41.1
-184.3
-42.9
1
1
1
[Cs][XeOF3](s) f /2XeF2(s) + /2[Cs][XeO2F3](s) + /2[Cs][F](s)
a The first column under the headings ∆H°, ∆S°, and ∆G° was obtained from Vm([M][XeOF3]) ) Vm(F2OXeNCCH3) - Vm(CH3CN) + Vm([M][F]),
whereas the second column was obtained from Vm([M][XeOF3]) ) Vm([M][F]) + [Vm(XeF2) + Vm(XeO2F2)]/2.
Thermochemistry. To account for the different decomposition
routes for [N(CH3)4][XeOF3] and [Cs][XeOF3], quantum-
chemical calculations and established semiempirical methods26-30
were used in conjunction with known thermodynamic quantities
to estimate ∆H°, ∆S°, and ∆G° for eqs 13 and 14.
The standard enthalpies for the decomposition reactions were
determined by analyzing their Born-Fajans-Haber cycles. The
enthalpy changes for the gas-phase reduction (eq 22) and
disproportionation (eq 23) reactions were calculated using the
MP2 method.
generalized by Jenkins et al.28,29 in eq 24, where R is the gas
constant (8.314 J K-1 mol-1), I is the ionicity of the salt, and
the constants R, ꢀ, and p depend on the nature of the salt.
R
∆H°L ) 2I
+ ꢀ + pRT
(24)
3 V
(
)
√
m
For the salts under investigation, which are singly charged and
nonlinear, the following values were used: I ) 1, R ) 117.3
nm kJ mol-1, ꢀ ) 51.9 kJ mol-1, and p ) 2. In this formalism,
∆H°L is the lattice enthalpy, defined as the energy required to
break the crystal lattice, and therefore has a positive value. This
approach is generally accurate to ∼4% for salts with ∆H°L <
-
XeOF3
f XeF2(g) + ½O2(g) + F-
(22)
(g)
(g)
5000 kJ mol-1 29 and is particularly useful because the formula
,
∆H° ) +52.6kJ mol mol-1
MP2/aug-cc-pVTZ(-PP)
unit volume (Vm) of an unknown salt can be estimated with
reasonable accuracy using several methods.29,37 The net en-
thalpies of decomposition (eqs 25 and 26) calculated for the
-
-
XeOF3
f ½XeF2(g) + ½XeO2F3
+ ½F-
(g)
(g)
(g)
(23)
reductions (∆H°red
) and disproportionations (∆H°dis) of
∆H° ) +56.3kJ mol mol-1
MP2/aug-cc-pVTZ(-PP)
[M][XeOF3 ] (M ) N(CH3)4, Cs) are summarized in Table 5.
∆H°red ) ∆H°L([M][XeOF3]) - ∆H°(eq 22) -
∆H°L([M][F]) - ∆H°(sub XeF2) (25)
The experimental value for the enthalpy of sublimation (∆Hsub
)
for XeF2 (55.71 kJ mol-1) was used.31 The lattice enthalpies of
[M][F], [M][XeOF3], and [M][XeO2F3] (Table 4) were estimated
by use of the volume-based method of Bartlett et al.,26,27 as
∆H°dis ) ∆H°L([M][XeOF3]) - ∆H°(eq 23) -
[
]
½∆H° ([M][F]) - ½∆H° ( M XeO F ) - ∆H°(sub XeF )
[
]
L
L
2
3
2
(26)
(26) Bartlett, N.; Yeh, S.; Kourtakis, K.; Mallouk, T. J. Fluorine Chem.
1984, 26, 97–116.
(27) Shen, C.; Hagiwara, R.; Mallouk, T.; Bartlett, N. In Inorganic Fluorine
Chemistry, Toward the 21st Century; Thrasher, J. S., Strauss, S. H.,
Eds.; ACS Symposium Series 555; American Chemical Society:
Washington, DC, 1994; Chapter 2, pp 26-39.
A method for estimating the absolute standard entropy of a
salt from its unit volume has been reported by Jenkins and
Glasser (eq 27), where k ) 1360 J K-1 mol-1 (nm-3 formula
30
unit-1) and c ) 15 J K-1 mol-1
.
(28) Jenkins, H. D. B.; Tudela, D.; Glasser, L. Inorg. Chem. 2002, 41,
2364–2367.
(29) Jenkins, H. D. B.; Roobottom, H. K.; Passmore, J.; Glasser, L. Inorg.
Chem. 1999, 38, 3609–3620.
S° ) kVm + c
(27)
(30) Jenkins, H. D. B.; Glasser, L. Inorg. Chem. 2003, 42, 8702–8708.
(31) Osborne, D. W.; Flotow, H. E.; Malm, J. G. J. Chem. Phys. 1972, 57,
4670–4675.
Entropies for the [M][F], [M][XeOF3], and [M][XeO2F3] salts
under consideration are provided in Table 4. When coupled with
9
J. AM. CHEM. SOC. VOL. 132, NO. 31, 2010 10941

A R T I C L E S
Brock et al.
the experimental standard entropies of O2(g) (206 J mol-1 K-1)38
and XeF2(s) (115.09 J mol-1 K-1),31 this method allows ∆S°
(eqs 28 and 29) and ∆G° (eq 30) to be calculated for the
reactions of interest. The ∆S°and ∆G° values obtained for these
reactions are summarized in Table 5. Estimates of the unit
volume of XeOF2 using either the difference method
(Vm(F2OXeNCCH3) - Vm(CH3CN) ) 0.0708 nm3) or an average
of the unit volumes of XeF2 and XeO2F2 (0.0731 nm3) differ
by only 0.0023 nm3 and lead to essentially the same ∆H° and
∆G° values.
salts, [M][XeOF3 ] (M ) N(CH3)4, Cs), in high purity. Both
salts are kinetically stable at -78 °C but slowly decompose at
10-25 °C. Their decomposition pathways, inferred from their
decomposition products, are supported by their thermochemical
cycles. The latter show that the proposed disproportionation and
reduction pathways are mainly driven by lattice energy contri-
butions, with entropy playing a significant role in the pathways
that lead to reduction of Xe(IV) to Xe(II) and O2 evolution.
The thermochemical cycles also reveal that disproportionation
of Xe(IV) to Xe(II) and Xe(VI) is favored for [Cs][XeOF3] but
not for [N(CH3)4][XeOF3], in accordance with experiment.
Comparison of the solid-state vibrational spectrum of
∆S°red ) ½S°(O2) + S°([M][F]) + S°(XeF2) -
S°([M][XeOF3]) (28)
-
[N(CH3)4][XeOF3] with that of the calculated gas-phase XeOF3
anion indicates little interaction between the anion and cation.
∆S° ) ½S°([M][F]) + ½S°(XeF2) + ½S°([M][XeO2F3]) -
S°([M][XeOF3]) (29)
+
-
The Raman spectra of the Cs+ and N(CH3)4 salts of XeOF3
also show that significant interactions occur between the Cs+
-
cation and the oxygen atom of the XeOF3 anion and that the
∆G° ) ∆H° - T∆S°
(30)
anions interact with one another by means of fluorine bridges.
The XeOF3- anion is presently the only example of an AX3YE2
VSEPR arrangement known in which a double-bond domain
subtends angles of ca. 90° with two valence electron lone-pair
domains. Comparisons of the experimental frequencies of
[Cs][XeOF3] with those previously reported for [Cs][XeOF3]9
reveal that the product obtained in an earlier study was a mixture
of XeF2, XeOF2, [Cs][XeF5], and [Cs][XeO3F]. Thus, the present
work represents the first bona fide synthesis and characterization
Based on the aforementioned thermochemical calculations,
both the gas-phase reduction and disproportionation pathways
are endothermic, and the decompositions of the [M][XeOF3]
salts are largely driven by lattice enthalpies. This is illustrated
by the greater ∆H° and ∆G° values for the Cs+ salt relative to
+
those of the N(CH3)4 salt which result from the smaller size
of Cs+ and greater lattice enthalpies of CsF and [Cs][XeO2F3].
The entropy term is also a major contributor in the reduction
pathways because O2 gas is evolved. The liberation of O2 greatly
increases the entropy of the reaction (119.7 J mol-1 K-1) relative
to the small entropy gain in the disproportionation pathways
(16.8 J mol-1 K-1). This results in the contribution of an
additional -30.7 kJ mol-1 to the Gibbs free energies for the
reduction pathways which, combined with the lattice enthalpies,
renders the reduction pathways significantly more favorable than
the disproportionation pathways for both salts.
-
of the XeOF3 anion.
Experimental Section
Caution. The xenon(IV) oxide fluoride species, XeOF2 and
[M][XeOF3], are highly energetic, shock-sensitiVe materials and
are only stable at the low temperatures described in the experi-
mental procedures that outline their syntheses. Both XeOF2 and
XeOF3- salts can detonate at low temperatures when mechanically
or thermally shocked. Thus, adequate protectiVe apparel and
working behind adequate shielding are crucial for the safe
manipulation of these materials. In the case of XeOF2, detonations
may occur upon freezing or further cooling of its CH3CN solutions
to -196 °C. It is strongly recommended that only small scale (<100
Although the disproportionation of [Cs][XeOF3] is less
favorable than reduction, the former is spontaneous. In contrast,
∆H° and ∆G° of the corresponding disproportionation reaction
for [N(CH3)4][XeOF3] are close to zero. The larger size of the
+
-
N(CH3)4 cation lowers the lattice enthalpies of [N(CH3)4][F]
mg) syntheses of XeOF2 and XeOF3 salts be undertaken.
and [N(CH3)4][XeO2F3] such that they no longer overcome the
gas-phase disproportionation enthalpy. The thermochemical
calculations are in accordance with the observed decomposition
products ([N(CH3)4][F] and XeF2) for [N(CH3)4][XeOF3], which
can only decompose by the reduction pathway (eq 13). In
contrast, [Cs][XeOF3], which can decompose by either the
reduction (eq 13) or disproportionation (eq 14) pathway, resulted
in the formation of predominantly XeF2 and a small amount of
[Cs][XeO2F3].
Apparatus and Materials. (a) General. All manipulations
involving air-sensitive materials were carried out under strictly
anhydrous conditions as previously described.39 Reaction vessels/
1
Raman sample tubes were fabricated from /4-in. o.d. FEP tubing
and outfitted with Kel-F valves. All reaction vessels and sample
tubes were rigorously dried under dynamic vacuum prior to
passivation with 1 atm of F2 gas.
Acetonitrile (Caledon, HPLC grade)40 was purified by the
literature method and transferred under static vacuum on a glass
vacuum line. Sulfur dioxide was distilled into a thick-wall glass
storage vessel containing P4O10 and dried, with frequent agitation,
for ca. 1 week prior to use. Dry SO2 was then distilled directly
from the storage vessel into the reaction vessel. Anhydrous HF
(Harshaw Chemicals Co.) was purified by the literature method.41
The high-purity syntheses and transfer methods for ONF and O2NF
are described in the Supporting Information. Xenon oxide difluo-
Conclusion
The fluoride ion acceptor properties of XeOF2 have been
demonstrated by the high-yield syntheses of the endothermic
(32) Christe, K. O.; Wilson, W. W.; Wilson, R. D.; Bau, R.; Feng, J. J. Am.
Chem. Soc. 1990, 112, 7619–7625.
5
ride8 and XeOF4 were prepared and purified according to the
literature methods. Cesium fluoride (CsF, ICN-KCK Laboratories
Inc., 99.9%) was dried by fusion in a platinum crucible, followed
by immediate transfer of the melt to a drybox port which was
immediately evacuated. Upon transferring to a nitrogen atmosphere
(33) Cortona, P. Phys. ReV. B: Condens. Matter 1992, 46, 2008–2014.
(34) Torrie, B. H.; Powell, B. M. Mol. Phys. 1992, 75, 613–622.
(35) Lehmann, J. F. Ph.D. Thesis, McMaster University, Hamilton, ON,
2004.
(36) Peterson, S. W.; Willett, R. D.; Huston, J. L. J. Chem. Phys. 1973,
59, 453–459.
(37) Jenkins, H. D. B.; Glasser, L.; Klapo¨tke, T. M.; Crawford, M.-J.;
Bhasin, K. K.; Lee, J.; Schrobilgen, G. J.; Sunderlin, L. S.; Liebman,
J. F. Inorg. Chem. 2004, 43, 6238–6248.
(39) Casteel, W. J., Jr.; Dixon, D. A.; Mercier, H. P. A.; Schrobilgen, G. J.
Inorg. Chem. 1996, 35, 4310–4322.
(40) Winfield, J. M. J. Fluorine Chem. 1984, 25, 91–98.
(41) Emara, A. A. A.; Schrobilgen, G. J. Inorg. Chem. 1992, 31, 1323–
1332.
(38) Chase, M. W., Jr. NIST JANAF Thermochemical Tables; American
Institute of Physics: New York, 1998.
9
10942 J. AM. CHEM. SOC. VOL. 132, NO. 31, 2010

XeOF3-, an Example of a VSEPR Arrangement
A R T I C L E S
drybox, the sample was ground to a fine powder and stored in a
PFA container inside the drybox until used. The naked fluoride
ion source, [N(CH3)4][F], was prepared according to the literature
method32 and was stored in an FEP tube in the drybox until used.
Both H2O (Caledon, HPLC grade) and H218O (MSD Isotopes, 97.8
atom % 18O) were used without further purification to prepare 2.00
M H216/18O solutions in previously dried and purified CH3CN40
which were used in the synthesis of XeOF2.8 High-purity Ar
(99.998%, Air Liquide) or N2 (obtained from liquid N2 boil-off
and dried by passage through a column of dry 3 Å molecular sieves)
gases were used for backfilling vessels.
a Raman spectrum of the white powder was recorded, revealing it
to be a mixture XeF4, XeO2F2, and [Cs][F(XeOF4)m].20
Raman Spectroscopy. Raman spectra were recorded on a Bruker
RFS 100 FT-Raman spectrometer at -150 °C using 1064 nm
excitation, 300 mW laser power, and 1 cm-1 resolution as previously
described.42
Computational Details. Geometries were fully optimized using
MP2 and DFT (SVWN5, BP86, PBE1PBE, B3LYP, B3PW91,
MPW1PW91) methods, and aug-cc-pVTZ and aug-cc-pVDZ basis
sets were used for all atoms. Pseudopotentials were used for xenon
(aug-cc-pVTZ-PP and aug-cc-pVDZ-PP), and their uses in com-
bination with aug-cc-pVTZ and aug-cc-pVDZ are indicated by aug-
cc-pVTZ(-PP) and aug-cc-pVDZ(-PP).43 Fundamental vibrations
were calculated, and NBO analyses22-26 were performed for the
optimized local minima at all levels. Quantum-chemical calculations
were carried out using the program Gaussian 0344 for geometry
optimizations, vibrational frequencies, and their intensities. The
program GaussView45 was used to visualize the vibrational
displacements that form the basis of the vibrational mode descrip-
tions given in Table 2 and Tables S3-S5 of the Supporting
Information.
(b) Syntheses of [M][Xe16/18OF3] (M ) N(CH3)4, Cs). In typical
syntheses, a 1/4-in. o.d. FEP reactor containing 0.306 mmol
(N(CH3)4+ salt) or 0.166 mmol (Cs+ salt) of Xe16/18OF2 was cooled
to -196 °C, and sufficient CH3CN was condensed onto the walls
of the reactor so that upon warming to -42 °C the solvent thawed,
contacting, without detonation, and dissolving Xe16/18OF2. Alter-
natively, CH3CN was syringed into the reactor in a manner identical
to that used for the synthesis of XeOF2.8 Note: In the eVent that
the amount of CH3CN added is insufficient to dissolVe XeOF2, the
former method requires refreezing the sample at -196 °C, which
often leads to sample detonation.8 The latter method is far less
likely to lead to detonation and is preferred because CH3CN is added
at a higher temperature (-78 °C). The solution was quickly frozen
at -78 °C and introduced into a drybox through a cold port, and
0.319 mmol of [N(CH3)4][F] (0.171 mmol of CsF) was added while
maintaining the reactor below -78 °C. The cold reactor was
removed from the drybox and warmed to -42 °C, and the contents
were thoroughly mixed, whereupon a pale yellow solid precipitated
from solution. The solvent was removed under dynamic vacuum
while maintaining the sample between -45 and -42 °C. The
completeness of solvent removal was monitored by Raman spec-
troscopy and required several hours for the last traces of CH3CN
Acknowledgment. We thank the Natural Sciences and Engi-
neering Research Council of Canada for support in the form of a
Discovery Grant (G.J.S.), the Ontario Graduate Scholarship in
Science and Technology and the McMaster Internal Prestige
“Ontario Graduate Fellowships” Programs for support (D.S.B.), and
the computational resources provided by SHARCNet (Shared
Hierarchical Academic Research Computing Network; www.
sharcnet.ca).
Supporting Information Available: Reassigned Raman spec-
trum of products formerly attributed to [Cs][XeOF3] (Table S1);
Raman spectrum of the products formed in the reaction of
XeOF2 with O2NF (Table S2); calculated vibrational frequencies
to be removed, resulting in pale yellow Cs+ and N(CH3)4 salts.
+
(c) Attempted Syntheses of [M′][XeOF3] (M′ ) NO, NO2).
Approximately 0.2 mL of ONF or O2NF was distilled into an
evacuated sample tube containing XeOF2 (43.3 mg for the ONF
reaction; 49.7 mg for the O2NF reaction) and frozen onto the vessel
walls at -196 °C. The ONF or O2NF was melted onto XeOF2 at
-78 °C. In the case of the ONF reaction, an immediate explosion
ensued upon contact with XeOF2. In the case of the O2NF reaction,
the solid immediately changed from yellow to white upon contact
with liquid O2NF.
(d) Reactivities of [M][XeOF3] with SO2 and XeOF4. In an
effort to solubilize [N(CH3)4][XeOF3], ca. 0.2 mL of SO2 was
distilled into an evacuated sample tube containing 72.1 mg of
[N(CH3)4][XeOF3] by freezing the condensate onto the walls of
the reaction vessel at -196 °C. Upon warming the sample to -78
°C, contact of SO2 vapor with the [N(CH3)4][XeOF3] resulted in
an immediate explosion. In a second attempt, ca. 0.2 mL of XeOF4
was frozen onto the upper walls of an evacuated 4-mm o.d. FEP
sample tube containing 11.9 mg (0.0353 mmol) of [Cs][XeOF3] at
-196 °C. The XeOF4 was melted onto [Cs][XeOF3] at ca. 20 °C,
and upon contact the solid rapidly turned from yellow to white.
The XeOF4 was removed under dynamic vacuum at ca. 20 °C, and
-
for XeOF3 (Tables S3 and S4) and XeOF2 (Table S5);
calculated geometrical parameters for XeOF3- (Tables S6); NBO
-
valencies, bond orders, and NPA charges for XeOF3 (Table
S7) and XeOF2 (Table S8); calculated geometrical parameters
for XeOF2 (Table S8); valence bond structures I-IV and
VI-IX; detailed O2NF and ONF syntheses; and complete ref
44. This material is available free of charge via the Internet at
http://pubs.acs.org.
JA103730C
(42) Gerken, M.; Dixon, D. A.; Schrobilgen, G. J. Inorg. Chem. 2000, 39,
4244–4255.
(43) Basis sets were obtained from the Extensible Computational Chemistry
Environment Basis set Database, version 2/25/04, as developed and
distributed by the Molecular Science Computing Facility, Environ-
mental and Molecular Science Laboratory, which is part of the Pacific
Northwest Laboratory, P.O. Box 999, Richland, WA 99352.
(44) Frisch, M. J.; et al. Gaussian 03, revision D.01; Gaussian, Inc.:
Wallingford, CT, 2004.
(45) GaussView, release 3.0; Gaussian Inc.: Pittsburgh, PA, 2003.
9
J. AM. CHEM. SOC. VOL. 132, NO. 31, 2010 10943