Xe3OF3+, a precursor to a noble-gas nitrate; Syntheses and structural characterizations of FXeONO2, XeF2·HNO3, and XeF2·N 2O4

Published on Web 09/15/2010

+

Xe₃OF₃, a Precursor to a Noble-Gas Nitrate; Syntheses and

Structural Characterizations of FXeONO₂, XeF₂·HNO₃, and

XeF₂·N₂O₄

Matthew D. Moran, David S. Brock, He´le`ne P. A. Mercier, and Gary J. Schrobilgen*

Department of Chemistry, McMaster UniVersity, Hamilton, Ontario L8S 4M1, Canada

Received June 26, 2010; E-mail: schrobil@mcmaster.ca

Abstract: Xenon ﬂuoride nitrate has been synthesized by reaction of NO₂F with [FXeOXeFXeF][AsF₆] at

-50 °C. It was characterized in SO₂ClF and CH₃CN solutions by low-temperature ¹⁴N, ¹⁹F, and ¹²⁹Xe NMR

spectroscopy and in the solid state by low-temperature Raman spectroscopy (-160 °C) and single-crystal

X-ray diffraction (-173 °C). The reactions were carried out using natural abundance and ¹⁸O-enriched

[FXeOXeFXeF][AsF₆] and ¹⁵NO₂F to aid in the vibrational assignments of FXeONO₂and to establish the

likely reaction pathway. Raman spectroscopy showed that FXe¹⁶ON(¹⁶O¹⁸O) was formed, along with XeF₂

and [NO₂][AsF₆], when an excess of N¹⁶O₂F reacted with [FXe¹⁸OXeFXeF][AsF₆]. A reaction mechanism

consistent with these ﬁndings is discussed. The crystal structure consists of well-separated FXeONO₂

molecules which display no signiﬁcant intermolecular interactions, providing geometric parameters that

are in good agreement with the gas-phase values determined from quantum-chemical calculations.

Decomposition of solid FXeONO₂is proposed to occur by three reaction pathways to give XeF₂, Xe, O₂,

N₂O₅, N₂O₄, and NO₂F. Attempts to synthesize FXeONO₂and Xe(ONO₂)₂by reaction of XeF₂with HNO₃

in SO₂ClF solution below -30 °C led to XeF₂·HNO₃. The structure of XeF₂·HNO₃includes a hydrogen

bond between HNO₃and a ﬂuorine atom of XeF₂, as well as an interaction between the xenon atom and

an oxygen atom of HNO₃, leading to a crystal lattice comprised of layered sheets. A molecular addition

compound between XeF₂and N₂O₄crystallized from liquid N₂O₄below 0 °C. The crystal structure of

XeF₂·N₂O₄displayed weak interactions between the xenon atom of XeF₂and the oxygen atoms of N₂O₄.

Quantum-chemical calculations have been used to assign the vibrational spectra of FXeONO₂, XeF₂·HNO₃,

and XeF₂·N₂O₄and to better understand the nature of the interactions of HNO₃and N₂O₄with XeF₂. The

synthesis of [XeONO₂][AsF₆] was attempted by the reaction of FXeONO₂with excess liquid AsF₅between

-78 and -50 °C, but resulted in slow formation of [NO₂][AsF₆], Xe, and O₂. Thermodynamic calculations

show that the pathways to [XeONO₂][AsF₆] formation and decomposition are exothermic and spontaneous

under standard conditions and at -78 °C.

10

1. Introduction

force in these reactions. Alternative syntheses of the OSeF₅

and OTeF₅^11,12derivatives also exist which do not involve HF

displacement.

The greatest variety of polyatomic ligand groups bonded to

xenon occurs for xenon in the +2 oxidation state, and those

bonded through oxygen are the most prevalent. Neutral mono-

and disubstituted oxygen-bonded derivatives having the formu-

lations FXeL and/or XeL₂have been reported where L )

OC(O)CF₃,¹OP(O)F₂,²OSO₂CF₃,³OSO₂F,^4,5OClO₃,⁵OSeF₅,⁶

XeF₂+ nHL f F_2-nXeL_n+ nHF

(1)

Two prior studies have reported the formation of xenon(II)

nitrates by HF displacement. The earlier study reports the

reactions of XeF₂with anhydrous HNO₃containing 20% NO₂

by weight at 20 °C.¹³Red-brown solids were obtained that

rapidly decomposed at 23 °C, forming an intense, transient blue

color. It is likely that the blue color arose from N₂O₃.¹⁴The

formulations FXeONO₂and Xe(ONO₂)₂were suggested, but

no structural characterizations were provided. When these

OTeF₅,^7,8and OIOF₄.⁹With the exception of OP(O)F₂and

2

OIOF₄,⁹their syntheses have been accomplished by reaction

of the parent acid, HL, with XeF₂according to eq 1 (n ) 1, 2),

leading to HF formation, a signiﬁcant thermodynamic driving

(1) Musher, J. I. J. Am. Chem. Soc. 1968, 90, 7371–7372.

(2) Eisenberg, M.; DesMarteau, D. D. Inorg. Chem. 1972, 11, 1901–1904.

(3) Naumann, D.; Tyrra, W.; Gnann, R.; Pfolk, D. J. Chem. Soc., Chem.

Commun. 1994, 2651–2653.

(4) Bartlett, N.; Wechsberg, M.; Jones, G. R.; Burbank, R. D. Inorg. Chem.

1972, 11, 1124–1127.

(10) Seppelt, K.; Lentz, D. Inorg. Synth. 1986, 24, 27–31.

(11) Schumacher, G. A.; Schrobilgen, G. J. Inorg. Chem. 1984, 23, 2923–

(5) Wechsberg, M.; Bulliner, P. A.; Sladky, F. O.; Mews, R.; Bartlett, N.

Inorg. Chem. 1972, 11, 3063–3070.

2929.

(12) Jacob, E.; Lentz, D.; Seppelt, K. Z. Anorg. Allg. Chem. 1981, 472,

(6) Seppelt, K.; Nothe, D. Inorg. Chem. 1973, 12, 2727–2730.

(7) Sladky, F. O. Monatsh. Chem. 1970, 101, 1559–1570.

(8) Sladky, F. O. Monatsh. Chem. 1970, 101, 1571–1577.

(9) Syvret, R. G.; Schrobilgen, G. J. Inorg. Chem. 1989, 28, 1564–1573.

7–25.

(13) Eisenberg, M.; DesMarteau, D. D. Inorg. Nucl. Chem. Lett. 1970, 6,

29–34.

(14) Beattie, I. R. Prog. Inorg. Chem. 1963, 5, 1–26.

9

10.1021/ja105618w  2010 American Chemical Society

J. AM. CHEM. SOC. 2010, 132, 13823–13839 13823

A R T I C L E S

Moran et al.

Scheme 1. Proposed Pathway for the Reaction of FXe¹⁸OXeFXeF⁺with NO₂F

2. Results and Discussion

reactions were carried out at -20 °C, the red-brown solids were

not observed. In a subsequent study, FXeONO₂was reported

to have been generated by the reaction of XeF₂with HNO₃in

CH₂Cl₂at -30 °C, and it was, in turn, reacted in situ with

various alkenes to give 1,2-disubstituted ﬂuoro-nitrato alkanes

(1 ) F, 2 ) ONO₂).¹⁵The proposed intermediate xenon

compound was not characterized in solution, nor was its isolation

attempted.

2.1. Synthesis and Decomposition of FXeONO₂. Deep red-

orange to magenta [FXeOXeFXeF][AsF₆]²⁰was allowed to react

as a suspension with liquid NO₂F at -50 °C. A suspension of

white solid formed over a 5 h period. Removal of excess NO₂F

under vacuum at -110 °C yielded a white, powdered mixture

of FXeONO₂, XeF₂, and [NO₂][AsF₆]. The reaction also

proceeded to completion at -78 °C but required reaction times

of 2-3 weeks. The latter method proved useful in establishing

the proposed reaction pathway (Scheme 1). The solid changed

in color from deep magenta to yellow-orange after 24-48 h,

and ﬁnally to white. The yellow-orange color and resulting

Raman spectrum of the intermediate reaction mixture showed

the presence of a transient xenon oxide ﬂuoride that is proposed

to be O(XeF)₂in Scheme 1. The latter compound will be the

subject of a future publication. The proposed intermediates,

The absence of a well-characterized xenon(II) nitrate is

surprising because the nitrate anion meets the general criteria

that are normally associated with a ligand that is suitable for

stabilization of Xe in its +2 oxidation state: (1) the least

electronegative atom of the ligand (nitrogen) is in its highest

-

oxidation state, (2) NO₃is the conjugate base of a strong

monoprotic acid, and correspondingly, (3) the electronegativity

of the -ONO₂ligand group is high. Empirical correlations based

1

on the H chemical shift difference between the R- and ꢀ-¹H

FXeON(O)F⁺and FXeFNO₂, were shown to be stable species

+

resonances of CH₃CH₂X derivatives have been previously used

to assign group electronegativities¹⁶based on the halogen

electronegativities of Huggins.¹⁷This approach has been

employed here using Pauling electronegativities¹⁸to give a

group electronegativity for ONO₂of 3.95,¹⁶which is close to

those of ﬂuorine (3.98)¹⁸and OTeF₅(3.87).¹⁹

+

with all frequencies real, where the FXeFNO₂isomer is ca.

80-90 kJ mol^-1more stable than FXeON(O)F⁺(Table S1 and

Figure S1, Supporting Information). Moreover, FXeON(O)F⁺

is also plausible in view of the existence of the structurally

related XN(O)F⁺cations (X ) N₃,²¹Cl,²²CF₃,²²and F²³). The

reaction pathway was also supported by an ¹⁸O-enrichment and

Raman spectroscopic study which showed that only

FXe¹⁶ON(¹⁶O¹⁸O) was formed when [FXe¹⁸OXeFXeF][AsF₆]

was used as the starting material. Failure to observe FXe¹⁸ONO₂

indicated that no oxygen isotope scrambling had occurred among

the bridging ¹⁶O and terminal ¹⁸O atoms (see Raman Spectros-

copy). A weak Raman band was observed at 1362 cm^-1in the

¹⁸O-enriched spectrum of the FXe¹⁶ON(¹⁶O¹⁸O)/[NO₂][AsF₆]

product mixture. The band is attributed to the mixed ¹⁸ON¹⁶O⁺

cation which results from a minor exchange pathway (Scheme

2; see Raman Spectroscopy).

Recent work has shown that XeF₂and [H₃O][AsF₆] react in

anhydrous HF to form the only known oxide ﬂuoride cation of

Xe(II), [FXeOXeFXeF][AsF₆].²⁰The present paper details the

synthesis of FXeONO₂from the FXeOXeFXeF⁺cation, provid-

ing an application of this interesting cation to the synthesis of

a novel xenon compound. The decomposition of FXeONO₂, the

attempted synthesis of XeONO₂⁺, and the syntheses and struc-

tural characterizations of the molecular adduct, XeF₂·HNO₃,

and the molecular addition compound, XeF₂·N₂O₄, are also

described.

The polar-covalent compounds, FXeONO₂and XeF₂, were

separated from [NO₂][AsF₆] and the decomposition products,

N₂O₄and N₂O₅, by rapidly extracting FXeONO₂/XeF₂into

SO₂ClF at -30 °C, followed by decanting the supernatant from

(15) Zeﬁrov, N. S.; Gakh, A. A.; Zhdankin, V. V.; Stang, P. J. J. Org.

Chem. 1991, 56, 1416–1418.

(16) Dailey, B. P.; Shoolery, J. N. J. Am. Chem. Soc. 1955, 77, 3977–

3981.

(17) Huggins, M. L. J. Am. Chem. Soc. 1953, 75, 4123–4126.

(18) Allred, A. L. J. Inorg. Nucl. Chem. 1961, 17, 215–221.

(19) Birchall, T.; Myers, R. D.; de Waard, H.; Schrobilgen, G. J. Inorg.

Chem. 1982, 21, 1068–1073.

(21) Wilson, W. W.; Haiges, R.; Boatz, J. A.; Christe, K. O. Angew. Chem.,

Int. Ed. 2007, 46, 3023–3027.

(20) Gerken, M.; Moran, M. D.; Mercier, H. P. A.; Pointner, B. E.;

Schrobilgen, G. J.; Hoge, B.; Christe, K. O.; Boatz, J. A. J. Am. Chem.

Soc. 2009, 131, 13474–13489.

(22) Minkwitz, R.; Bernstein, D.; Preut, H.; Sartori, P. Inorg. Chem. 1991,

30, 2157–2161.

(23) Vij, A.; Zhang, X.; Christe, K. O. Inorg. Chem. 2001, 40, 416–417.

9

13824 J. AM. CHEM. SOC. VOL. 132, NO. 39, 2010

Xe₃OF₃⁺, Precursor to a Noble-Gas Nitrate

A R T I C L E S

Scheme 2. Proposed Pathway for the Formation of the ¹⁸ON¹⁶O⁺

and 3) was found to be spontaneous under standard conditions

(∆G° ) -75.7 kJ mol^-1) and at -78 °C (∆G_195.15) -71.0 kJ

mol^-1) for the decomposition of 1 mol of FXeONO₂to XeF₂,

N₂O₆, and Xe. A further -22.5 kJ mol^-1(∆G_195.15) -17.9 kJ

mol^-1) is released when 0.5 mol of N₂O₆decomposes to N₂O₅

and O₂, giving a net ∆G° ) -98.2 kJ mol^-1(∆G_195.15) -88.9

kJ mol^-1) for the decomposition of 1 mol of FXeONO₂. The

remaining proposed pathways, which involve the nonsponta-

Cation

neous generation of FXeNO₂(eq 4, ∆G° ) 26.5 kJ mol^-1

,

∆G_195.15) 35.7 kJ mol^-1) or FXeONO (eq 5, ∆G° ) 37.9 kJ

mol^-1, ∆G_195.15) 46.5 kJ mol^-1) as intermediates, are more

than compensated for by their spontaneous decompositions to

XeF₂, Xe, and N₂O₄(FXeNO₂[one-half of eq 6], ∆G° ) -139.2

kJ mol^-1, ∆G_195.15) -136.3 kJ mol^-1; FXeONO [one-half of

solid [NO₂][AsF₆], N₂O₄, and N₂O₅. A mixture of FXeONO₂

and XeF₂was then precipitated from the supernatant at -78

°C (see Experimental Section). Prolonged and repeated extrac-

tions resulted in dissolution of signiﬁcant amounts of

[NO₂][AsF₆] and its coextraction with XeF₂and FXeONO₂(see

Experimental Section).

eq 9], ∆G° ) -150.6 kJ mol^-1, ∆G_195.15) -147.1 kJ mol^-1

)

,

The ¹⁹F NMR spectrum of a SO₂ClF solution containing

extracted FXeONO₂and XeF₂was monitored at 0 °C and

showed that FXeONO₂was 50% decomposed after 6.5 h when

the decomposition was quenched at -40 °C. The only ﬂuorine-

or to NO₂F and Xe (FXeNO₂[eq 8], ∆G° ) -217.6 kJ mol^-1

∆G_195.15) -207.4 kJ mol^-1; FXeONO [eq 10], ∆G° ) -229.0

kJ mol^-1, ∆G_195.15) -218.2 kJ mol^-1).

2.2. Reactions of XeF₂with HNO₃. Attempts were made to

prepare FXeONO₂and Xe(ONO₂)₂by reaction of XeF₂with

HNO₃. Xenon diﬂuoride was allowed to react with 2 equiv of

containing decomposition product observed was XeF₂(-177.6

1

¹⁹F-¹²⁹Xe

ppm, J

) 5611 Hz). Low-temperature Raman spectra

of FXeONO₂in admixture with XeF₂and [NO₂][AsF₆] (Figure

S2, Supporting Information) revealed that FXeONO₂had

partially decomposed to N₂O₅and presumably XeF₂at -50 °C

after 5 h, whereas FXeONO₂had partially decomposed to N₂O₄

and XeF₂·N₂O₄at -78 °C after 7 days. The former decomposi-

tion pathway, leading to N₂O₅, likely proceeds through the

unstable intermediate, O₂NOONO₂(eq 2), which is known to

rapidly decompose at -78 °C according to eq 3,^24,25and by

analogy with the decomposition of FXeOSO₂F to XeF₂,

FO₂SOOSO₂F, and Xe.⁵

100% HNO₃in SO₂ClF solution. NMR spectroscopy at -30

1

¹⁹F-¹²⁹Xe

°C in SO₂ClF solution revealed XeF₂(-1684 ppm, J

) 5630 Hz) and Xe gas (-5374 ppm) in the ¹²⁹Xe spectrum and

only XeF₂(-183.4 ppm, ¹J

_Xe) 5633 Hz) and HF (-180.2

¹⁹F-¹²⁹

ppm, ∆ν _/₂) 550 Hz) in the ¹⁹F spectrum. Although the gas-phase

syntheses (Table S2, Supporting Information) of FXeONO₂(eq

11, ∆G° ) -17.6 kJ mol^-1, ∆G_195.15) -20.2 kJ mol^-1) and

Xe(ONO₂)₂(eq 12, ∆G° ) -17.5 kJ mol^-1, ∆G_195.15) -20.8 kJ

mol^-1) are near equilibrium, their rapid decompositions (FXeONO₂,

eqs 2 and 3 or eqs 4-10; Xe(ONO₂)₂, eq 13 followed by eq 3)

may preclude their observation by NMR spectroscopy.

1

2FXeONO₂f XeF₂+ [O₂NOONO₂] + Xe

(2)

(3)

XeF₂+ HNO₃f FXeONO₂+ HF

FXeONO₂+ HNO₃f Xe(ONO₂)₂+ HF

Xe(ONO₂)₂f O₂NOONO₂+ Xe

(11)

(12)

(13)

1

[O₂NOONO₂] f N₂O₅+ /₂O₂

The latter decomposition pathway, leading to N₂O₄and

XeF₂·N₂O₄, may occur through the unstable N-nitrito or

O-nitrito intermediates, FXeNO₂(eq 4) and FXeONO (eq 5),

which would most likely decompose according to eqs 6-8 and

eqs 9 and 10, respectively.

Although FXeONO₂and/or Xe(ONO₂)₂could account for the

decomposition products observed by NMR spectroscopy, in either

case, the degree of reaction must have been small because neither

decomposition product, N₂O₄or N₂O₅, was observed in the Raman

spectrum. Rather, XeF₂·HNO₃crystallized between -40 and -60

°C from SO₂ClF solution and was characterized in the solid state

by Raman spectroscopy and by single-crystal X-ray diffraction (see

X-ray Crystallography and Raman Spectroscopy). These ﬁndings

indicate that if Xe(ONO₂)₂forms, it is inherently unstable toward

decomposition. The gas-phase thermochemical calculations (Table

S2, Supporting Information) show a large negative Gibbs free

energy for the spontaneous decomposition of Xe(ONO₂)₂to Xe

and the unstable intermediate, N₂O₆(eq 13, ∆G° ) -151.4 kJ

mol^-1, ∆G_195.15) -141.4 kJ mol^-1).

2.3. Reactions of XeF₂and [XeF][AsF₆] with N₂O₅. Attempts

to react XeF₂or [XeF][AsF₆] with N₂O₅in SO₂ClF to give

FXeONO₂and/or Xe(ONO₂)₂were monitored by Raman

spectroscopy. A 1.5:1 mixture of XeF₂and N₂O₅that was

warmed stepwise from -40 to 10 °C showed essentially no

reaction, with only a small amount of N₂O₅decomposition to

N₂O₄and subsequent formation of XeF₂·N₂O₄(vide infra).

However, at -78 °C, equimolar amounts of [XeF][AsF₆] and

N₂O₅formed a transient orange, clumpy mixture which was

consistent with [Xe₃OF₃][AsF₆] formation (eq 14). Gradual

warming of the reaction mixture above -30 °C to completely

1

FXeONO₂f [FXeNO₂] + /₂O₂

(4)

1

FXeONO₂f [FXeONO] + /₂O₂

(5)

(6)

2[FXeNO₂] f XeF₂+ Xe + N₂O₄

XeF₂+ N₂O₄f XeF₂· N₂O₄

[FXeNO₂] f Xe + NO₂F

(7)

(8)

2[FXeONO] f XeF₂+ Xe + N₂O₄

[FXeONO] f Xe + NO₂F

(9)

(10)

The gas-phase thermochemical calculations using the MP2

method (Table S2, Supporting Information) support the proposed

decomposition pathways. The decomposition pathway for

FXeONO₂that leads to XeF₂, N₂O₅(via N₂O₆), and Xe (eqs 2

(24) Khadzhi-Ogly, M. R.; Yagodovskaya, T. V.; Nekrasov, L. I. Zh. Fiz.

Khim. 1981, 55, 3124–3127.

(25) Khadzhi-Ogly, M. R.; Yagodovskaya, T. V.; Nekrasov, L. I. Zh. Fiz.

Khim. 1982, 56, 1807–1809.

9

J. AM. CHEM. SOC. VOL. 132, NO. 39, 2010 13825

A R T I C L E S

Moran et al.

Table 1. NMR Parameters of FXeONO₂and Related Species^a

2.5. Reaction of FXeONO₂with AsF₅. In an attempt to form

a salt of the XeONO₂⁺cation, a mixture of FXeONO₂and XeF₂

was allowed to react with excess liquid AsF₅at -78 °C. The

reaction was monitored by low-temperature Raman spectroscopy

over a period of 24 h and showed that AsF₅reacted with XeF₂

to yield [XeF][AsF₆], but FXeONO₂did not react. Warming

the reaction mixture to -50 °C for 12 h with periodic quenching

and monitoring of the spectrum by Raman spectroscopy at -160

°C showed [XeF][AsF₆] and [NO₂][AsF₆] were the only

products. The ﬁndings are consistent with the formation of an

temp,

solvent °C

δ(¹⁹F), δ(¹²⁹Xe), δ(¹⁴N),

1

19 129

solute

species

ppm

J

_Xe, Hz

F-

FXeONO₂SO₂ClF -70 FXeONO₂-131.9 -1974

5456

5603

XeF₂

-177.3 -1864

b

FXeONO₂SO₂ClF -50 FXeONO₂

-65.8

FXeONO₂CH₃CN -40 FXeONO₂-135.1 -1870

XeF₂-179.2 -1775

5503

5697

c

XeF₂

30 FXeONO₂-130.1 -1989

5408

5642

N₂O₄

+

unstable XeONO₂salt (eq 19) that rapidly decomposed

according to eq 20.

XeF₂

NO₂F

Xe

-179.4 -1841

397.8

-5243

FXeONO_2(s)+ AsF_5(l)f [XeONO₂][AsF₆]_(s)

(19)

^aUnless otherwise noted, the reaction products result from Scheme 1.

^bAttempts to acquire a ¹⁵N NMR spectrum on ¹⁵N-enriched (98+%)

FXeONO₂(SO₂ClF solvent, -70 °C) failed using relaxation delays as

long as 180 s. ^cThe observed species arise from eqs 16-18.

1

[XeONO₂][AsF₆]_(s)f [NO₂][AsF₆]_(s)+ Xe_(g)+ /₂O_2(g)

(20)

dissolve N₂O₅resulted in a white product. Low-temperature

Raman spectroscopy showed a mixture of XeF₂, [NO₂][AsF₆],

and unreacted N₂O₅and the absence of FXeONO₂. The

[Xe₃OF₃][AsF₆] salt, which is unstable above -30 °C,²⁰likely

decomposed (eq 15) to form XeF₂, O₂, Xe, and [XeF][AsF₆].

The latter salt is presumed to re-enter the cycle (eqs 14 and 15)

until completely consumed.

Thermochemical calculations for the gas-phase reactions

(Table S2, Supporting Information) show there is a considerable

barrier to ﬂuoride ion abstraction from FXeONO₂by AsF₅to

form the XeONO₂⁺and AsF₆^-(eq 19 in the gas phase, ∆H° )

359.0 kJ mol^-1, ∆G° ) 365.0 kJ mol^-1; ∆H_195.15) 358.7 kJ

mol^-1, ∆G_195.15) 362.9 kJ mol^-1), but this barrier is consider-

ably less than that for the gas-phase abstraction of ﬂuoride ion

from XeF₂by AsF₅(∆H° ) 493.2 kJ mol^-1, ∆G° ) 499.9 kJ

mol^-1; ∆H_195.15) 493.4 kJ mol^-1, ∆G_195.15) 497.7 kJ mol^-1).

Volume-based thermodynamics (VBT) calculations have been

used to arrive at the thermodynamic properties of these systems

in the solid state (see Solid-State Thermochemistry in the

Supporting Information) and show that ﬂuoride ion abstraction

is essentially mitigated by the lattice enthalpy of [XeONO₂]-

[AsF₆]. Once formed, the gas-phase XeONO₂⁺cation has a large

negative Gibbs free energy that accompanies its spontaneous

decomposition to Xe, O₂, and NO₂⁺(eq 21, ∆H° ) -142.7 kJ

mol^-1, ∆G° ) -189.5 kJ mol^-1; ∆H_195.15) -143.0, ∆G_195.15

) -173.4 kJ mol^-1).

3[XeF][AsF₆] + N₂O₅f [FXeOXeFXeF][AsF₆] +

2[NO₂][AsF₆] (14)

1

[FXeOXeFXeF][AsF₆] f XeF₂+ /₂O₂+ Xe +

[XeF][AsF₆] (15)

2.4. Reaction of XeF₂with N₂O₄. Dissolution of XeF₂in liquid

N₂O₄at 0-30 °C resulted in NO₂F formation (eq 16, δ(¹⁹F) )

397.8 ppm at 30 °C). A small, steady-state concentration of

FXeONO₂(Table 1) was also observed which apparently arose

as a result of the self-ionization of N₂O₄(eq 17) and the reaction

-

of XeF₂with NO₃(eq 18).

+

1

+

XeONO₂f Xe_(g)+ /₂O_2(g)+ NO₂

(21)

(g)

N₂O₄+ XeF₂f 2NO₂F + Xe

(16)

(17)

(18)

The thermochemical parameters were also determined for eq

20 by application of VBT (see Solid State Thermochemistry in

the Supporting Information). The appropriate thermochemical

cycle (eq 22) using the gas-phase standard enthalpy, ∆H°(21)

(Table S2 in the Supporting Information), and Gibbs free energy,

∆G°(21), given by eqs 23 and 24 shows that the solid-state

decomposition of [XeONO₂][AsF₆] (eq 20) is signiﬁcantly more

-

N₂O₄U NO⁺+ NO₃

-

XeF₂+ NO⁺+ NO₃f FXeONO₂+ NOF

The self-ionization of N₂O₄has been inferred from electrical

conductivity measurements (2.36 × 10^-13Ω^-1cm^-1at 17 °C)²⁶

and by measurement of the self-ionization constant of N₂O₄in

sulfolane at 30 °C (K_{N O}) 7.1 × 10^-8mol L^-1).²⁷The small

+

exothermic than the gas-phase decomposition of the XeONO₂

cation (eq 21).

2

4

-

steady-state concentration of the NO₃anion (eq 17) and the

instability of FXeONO₂(eqs 2-10) at room temperature (vide

supra) account for the low product concentrations (see NMR

Spectroscopy).

∆H°(20) ) ∆H°(21) - ∆H°_L([NO₂][AsF₆]) +

∆H°_L([XeONO₂][AsF₆]) ) -188 kJ mol^-1(22)

1

Dissolution of XeF₂in liquid N₂O₄at 35 °C also resulted in

the formation of XeF₂·N₂O₄, which was observed by Raman

spectroscopy of the frozen mixture at -160 °C (see Raman

Spectroscopy). The molecular addition compound, XeF₂·N₂O₄,

crystallized from a solution of XeF₂in N₂O₄at -3 °C and was

characterized by single-crystal X-ray diffraction (see X-ray

Crystallography).

∆S°(20) ) S°([NO₂][AsF₆]_(s)) + S°(Xe_(g)) + /₂S°(O_2(g)) -

S°([XeONO₂][AsF₆]_(s)) ) 207 J mol^-1K^-1(23)

∆G°(20) ) ∆H°(20) - T∆S°(20) ) -250 kJ mol^-1

(24)

The gas-phase reaction (eq 25) has also been considered and

is shown to be considerably more endothermic (∆H° ) 522.0

(26) Bradley, R. S. Trans. Faraday Soc. 1956, 52, 1255–1259.

(27) Wartel, M.; Boughriet, A.; Fischer, J. C. Anal. Chem. Acta 1979, 110,

211–217.

kJ mol^-1, ∆G° ) 527.4 kJ mol^-1; ∆H_195.15) 521.8 kJ mol^-1

,

∆G_195.15) 525.5 kJ mol^-1) than the gas-phase counterpart of

9

13826 J. AM. CHEM. SOC. VOL. 132, NO. 39, 2010

Xe₃OF₃⁺, Precursor to a Noble-Gas Nitrate

A R T I C L E S

Table 2. Crystallographic Data for FXeONO₂, XeF₂·HNO₃, and

XeF₂·N₂O₄

eq 19, and therefore it has not been considered as a likely

alternative reaction pathway.

chem formula

space group

a (Å)

b (Å)

c (Å)

NO₃FXe

P2₁/c

4.6663(4)

8.7995(7)

9.4153(8)

90

90.325(5)

90

386.60(6)

4

212.30

3.648

-173

8.80

0.0417

0.0807

N₂O₄F₂Xe

P1

HNO₃F₂Xe

Pnma

17.3543(7)

5.6539(2)

4.7658(2)

90

467.62(5)

4

232.30

3.300

j

-

FXeONO₂+ AsF₅f XeF⁺+ AsF₅ONO₂

(25)

4.5822(3)

5.0597(3)

6.2761(5)

79.170(7)

88.454(5)

81.083(5)

141.19(2)

1

261.30

3.073

-173

6.10

2.6. Solution Characterization of FXeONO₂by ^14/15N, ¹⁹F,

and ¹²⁹Xe NMR Spectroscopy. The ¹⁹F and ¹²⁹Xe NMR spectra

of FXeONO₂and XeF₂mixtures were recorded in SO₂ClF at

-70 °C and in CH₃CN at -40 °C, whereas the ¹⁴N NMR

spectra were recorded in SO₂ClF at -50 °C. The ¹⁹F and

¹²⁹Xe NMR spectra of XeF₂, FXeONO₂, and NO₂F were

obtained in N₂O₄solution at 30 °C (see Reaction of XeF₂

with N₂O₄). The NMR parameters for the three solvent

systems are provided in Table 1.

R (deg)

ꢀ (deg)

γ (deg)

V (Å³)

Z (molecules/unit cell)

mol wt (g mol^-1

)

F_calcd(g cm^-3

T (°C)

)

-160

7.32

0.0140

0.0317

µ (mm^-1

)

a

R₁

0.0385

0.0742

The ¹⁹F NMR spectrum of FXeO¹⁵NO₂(98+% ¹⁵N enrich-

ment) in SO₂ClF consisted of a singlet (δ(¹⁹F) ) -131.9 ppm)

wR₂

b

^aR₁) ∑||F_o| - |F_c||/∑|F_o| for I > 2σ(I). wR₂is deﬁned as {∑[w(F_o

b

2

with accompanying ¹²⁹Xe (I ) ¹/₂, 26.44%) satellites (¹J

¹⁹F-¹²⁹Xe

2

- F_c²)²]/∑w(F_o)²}^1/for I > 2σ(I).

2

) 5431 Hz). No three-bond coupling to ¹⁵N (I ) /₂) was

1

3

¹⁵N-¹⁹

observed, presumably because the magnitude of J

is

F

Reaction of XeF₂with N₂O₄). The ¹⁹F NMR parameters were

found to be similar to those determined in SO₂ClF and CH₃CN

solutions (Table 1). Xenon diﬂuoride and NO₂F were also

identiﬁed in the ¹⁹F NMR spectrum, consistent with eq 16,

showing that the reaction of XeF₂with N₂O₄is slow at 30 °C.

2.7. X-ray Crystallography. Details of the data collection

parameters and other crystallographic information for FXeONO₂,

XeF₂·HNO₃, and XeF₂·N₂O₄are given in Table 2. The experi-

mental and calculated bond lengths and angles are summarized

in Table 3 and in Table S3 in the Supporting Information.

2.7.1. FXeONO₂. The FXeONO₂molecules (Figure 1a),

which possess C_ssite symmetry, are well-isolated, discrete

molecular units as inferred from the long intermolecular contacts

(Table 3 and Figure 1c) and the good agreement between the

experimental and calculated geometric parameters and vibra-

tional frequencies (see Computational Results and Raman

Spectroscopy). The majority of the Xe· · ·O/F contacts, which

range from 3.322(4) to 3.570(5) Å, are near the sum of the van

der Waals radii for xenon and ﬂuorine (3.63 Å)³¹and for xenon

and oxygen (3.68 Å)³¹and avoid the torus of electron lone pair

density about xenon.

The Xe-O bond length (2.126(4) Å) is similar to those found

for FXeOSO₂F (2.155(8) Å),⁴Xe(OTeF₅)₂(2.119(11) Å),³²and

Xe(OSeF₅)₂(2.09(3) Å).³²Correspondingly, the Xe-F bond

length (1.992(4) Å) and O-Xe-F angle (177.6(2)°) are similar

to those determined for FXeOSO₂F (1.940(8) Å, 177.4(3)°).⁴

The Xe-O-N angle (114.7(3)°) was found to be somewhat

larger than those determined for BrONO₂(113.8(4)°)³³and

ClONO₂(112.49(4)°),³⁴which is attributed to decreased oxygen

lone electron pair-oxygen bond pair repulsions and is in

accordance with the bond order trend Xe-O < Br-O < Cl-O

(Table S4, Supporting Information). The geometric parameters

of the ONO₂groups in FXeONO₂and XONO₂(X ) Cl, Br)

differ in that the N-O_Sbond is shorter than the N-O_Abond in

the halogen nitrates, but the opposite trend is observed for

FXeONO₂(A denotes anti and S denotes syn about the N-O

bridge bond and with respect to the O-Xe-F or O-X groups).

The same, but less distinct, bond length trend is found for the

relatively small and is obscured by the breadth of the ¹⁹F

1

resonance (∆ν _/) 13 Hz at 11.7440 T, ∆ν _/₂) 6 Hz at 7.0463

2

T).

The ¹²⁹Xe NMR spectrum of FXeO¹⁵NO₂(11.744 T) in

1

¹⁹F-¹²⁹Xe

SO₂ClF solvent consisted of a doublet arising from J

) 5429 Hz, which occurred at -1966 ppm in the xenon(II)

region of the spectrum. The ²J

_Xecoupling presumably was

¹⁵N-¹²⁹

not observed because it was obscured by the line width of the

129

1

Xe resonance (∆ν _/) 134 Hz). In an attempt to reduce

2

contributions to the ¹⁹F and ¹²⁹Xe line widths that result from

shielding anisotropy,²⁸the ¹²⁹Xe NMR spectrum was recorded

at a lower external ﬁeld strength (7.0463 T). The line width

1

(∆ν _/) 50 Hz), however, was not narrowed sufﬁciently to

2

¹⁵N-¹²⁹Xe

observe the J

coupling.

The ¹⁹F and ¹²⁹Xe NMR chemical shifts of FXeONO₂adhere

to a nearly linear empirical correlation between ¹⁹F and ¹²⁹Xe

NMR chemical shifts of oxygen-bonded FXe-R species (R )

OSO₂F, OS(O)(F)OMoOF₄, OS(O)(F)OWOF₄, OSeF₅, OTeF₅,

OWOF₅(WOF₄), OWOF₅(WOF₄)₂)²⁹and a correlation between

19

1

¹⁹F-¹²⁹Xe

the F chemical shifts and J

coupling constants that

encompasses the majority of xenon ﬂuoride species in the +2,

+4, and +6 oxidation states of xenon.³⁰No clear-cut correlation

exists among the chemical shifts and coupling constants of

FXeOTeF₅, FXeONO₂, XeF₂, and XeF⁺and the corresponding

group electronegativities other than their ¹²⁹Xe chemical shifts

(see Supporting Information).

The ¹⁴N NMR spectrum in SO₂ClF solvent consisted of a

1

singlet at -65.8 ppm (∆ν _/) 50 Hz). The resonance was

2

3

2

¹⁴N-¹⁹

F

¹⁴N-¹²⁹Xe

and J

quadrupole broadened to an extent that J

could not be observed. A ¹⁵N NMR study on a ¹⁵N-enriched

sample of FXeONO₂in the same spectral region was attempted

using relaxation delay times as long as 180 s, which did not

permit observation of a ¹⁵N resonance. The long relaxation time

likely results because the ¹⁵N atom is bonded to spinless oxygen

nuclei which do not provide a dipolar intramolecular spin-lattice

relaxation pathway.

The formation of FXeONO₂was also observed at low

concentrations by reaction of XeF₂with N₂O₄solvent (see

(28) Howarth, O. In Multinuclear NMR; Mason, J., Ed.; Plenum Press:

London, 1987; p 149.

(31) Bondi, A. J. Phys. Chem. 1964, 68, 441–451.

(32) Fir, B. A.; Mercier, H. P. A.; Sanders, J. C. P.; Dixon, D. A.;

Schrobilgen, G. J. J. Fluorine Chem. 2001, 110, 89–107.

(33) Minkwitz, R.; Hertel, T. Z. Naturforsch. 1997, 52b, 1307–1310.

(34) Obermeyer, A.; Borrmann, H.; Simon, A. J. Am. Chem. Soc. 1995,

117, 7887–7890.

(29) Schrobilgen, G. J.; Holloway, J. H.; Granger, P.; Brevard, C. Inorg.

Chem. 1978, 17, 980–987.

(30) Gerken, M.; Schrobilgen, G. J. Coord. Chem. ReV. 2000, 197, 335–

395.

9

J. AM. CHEM. SOC. VOL. 132, NO. 39, 2010 13827

A R T I C L E S

Moran et al.

Table 3. Experimental and Calculated^aGeometric Parameters for

FXeONO₂, XeF₂·HNO₃, and XeF₂·N₂O₄

FXeONO₂

exptl

calcd

exptl

calcd

Bond Lengths (Å)

Xe(1)-F(1)

Xe(1)-O(1)

O(1)-N(1)

1.992(4) 2.018 N(1)-O(2)

2.126(4) 2.148 N(1)-O(3)

1.365(7) 1.393

1.199(6) 1.200

1.224(6) 1.211

Bond Angles (°)

F(1)-Xe(1)-O(1) 177.6(2) 175.5 O(1)-N(1)-O(3) 118.4(5) 117.8

Xe(1)-O(1)-N(1) 114.7(3) 116.0 O(2)-N(1)-O(3) 127.1(5) 129.2

O(1)-N(1)-O(2) 114.5(4) 113.0

Contacts (Å)

Xe(1) · · · F(1A)

Xe(1) · · · O(1A)

Xe(1) · · · O(2A)

Xe(1) · · · O(3A)

N(1) · · · F(1A)

O(2) · · · N(1A)

3.420(4)

3.322(4)

3.478(4)

3.390(4)

2.780(6)

2.923(6)

Xe(1) · · · F(1B)

Xe(1) · · · O(2B)

Xe(1) · · · O(3B)

Xe(1) · · · O(3C)

O(1) · · · F(1A)

Xe(1) · · · O(3)

3.299(4)

3.545(4)

3.570(5)

3.518(4)

2.935(5)

2.965(7) 3.024

XeF₂·HNO₃

exptl

calcd

exptl

calcd

Bond Lengths (Å)

Xe(1)-F(1)

Xe(1)-F(2)

O(1)-N(1)

1.9737(8) 1.983 N(1)-O(2)

2.0506(8) 2.064 N(1)-O(3)

1.206(2) 1.193 O(2)-H(1)

1.368(2) 1.360

1.216(2) 1.220

0.83(2)

1.016

Bond Angles (°)

F(1)-Xe(1)-F(2) 178.98(3) 176.9 O(2)-N(1)-O(3) 117.2(1) 116.3

O(1)-N(1)-O(2) 114.6(1) 115.6 N(1)-O(2)-H(1) 106(2)

O(1)-N(1)-O(3) 128.2(1) 128.1

104.1

Contacts (Å)

H(1) · · · F(2)

1.86(2)

1.520 Xe(1) · · · F(1B)

3.3050(4)

3.4156(6)

3.456(1)

3.5284(6)

2.690(1) 2.536

Xe(1) · · · O(3)

Xe(1) · · · F(1A)

Xe(1) · · · O(1A)

Xe(1) · · · F(2A)

Xe(1) · · · O(2A)

3.317(1) 3.034 Xe(1) · · · F(1C)

3.4897(9)

3.4156(6)

3.4859(8)

3.5284(6)

Xe(1) · · · O(1B)

Xe(1) · · · O(1C)

Xe(1) · · · O(2B)

O(2) · · · F(2)

Figure 1. (a) X-ray crystal structure of FXeONO₂. Thermal ellipsoids are

shown at the 50% probability level. (b) Calculated geometry of FXeONO₂.

The atom numbering scheme corresponds to that used in Table 3. (c) Packing

diagram for FXeONO₂viewed along the a-axis.

XeF₂·N₂O₄

calcd^b

exptl

model A

model B

calculated N-O_Sand N-O_Abond lengths. These experimental

bond length trends agree with the relative, but opposite, bond

order trends calculated at the B3LYP/aug-cc-pVTZ(-PP) and

MP2/aug-cc-pVTZ(-PP) levels of theory. However, the bond

orders are nearly equal for FXeONO₂at the MP2 level of theory

(Table S4).

Bond Lengths (Å)

Xe(1)-F(1)

1.985(3)

Xe(1)-F(1)

Xe(1)-F(2)

N(1)-O(1)

N(1)-O(2)

N(2)-O(3)

N(2)-O(4)

N(1)-N(2)

2.025

2.003

1.187

1.184

1.187

1.184

1.792

2.015

2.010

1.186

1.185

1.799

N(1)-O(1)

N(1)-O(2)

1.194(6)

1.182(6)

1.738(8)

N(1)-N(1A)

2.7.2. XeF₂·HNO₃. The structure of XeF₂·HNO₃(Figure 2)

is only the second example of a hydrogen-bonded XeF₂adduct,

the other being [H₃O·2XeF₂][AsF₆].²⁰All atoms are coplanar,

allowing the molecules to form sheets in the a,c-plane comprised

of alternating XeF₂and HNO₃molecules that are congruently

stacked along the b-axis (Figure 3 and Figure S3 in the

Supporting Information). The packing arrangement in the a,c-

plane is largely determined by the short H· · ·F (1.86(2) Å) contacts

and weak Xe · · ·O(3) (3.317(1) Å) contacts, which are signiﬁcantly

less than the sums of the van der Waals radii (2.67 and 3.68 Å,

respectively).³¹The strong hydrogen bond distorts the local

symmetry of the XeF₂molecule from D_∞hto C_∞Vsymmetry. The

Xe-F bond length distortions (terminal (F_t), 1.9737(8) Å; bridge

(F_b), 2.0506(8) Å) parallel the distortions found in compounds

where XeF₂is homoleptically coordinated to a metal ion through

a single ﬂuorine bridge. The present Xe-F_band Xe-F_tbond

lengths are comparable to those in [Ca(XeF₂)₅][PF₆]₂,³⁵

[Cd(XeF₂)₅][PF₆]₂,³⁵[Cd(XeF₂)][BF₄]₂,³⁶[Mg(XeF₂)₄][AsF₆]₂,³⁷

[Mg(XeF₂)₂][AsF₆]₂,³⁷and [Ca₂(XeF₂)₉][AsF₆]₄,³⁸where the

Xe-F_bbonds range from 2.026(7) to 2.087(8) Å and the Xe-F_t

bonds range from 1.913(5) to 1.966(6) Å. The geometric

parameters obtained for HNO₃in the present structure (N-O_H,

Bond Angles (°)

F(1)-Xe(1)-F(1A)

O(1)-N(1)-N(1A)

180.0

F(1)-Xe(1)-F(2)

O(1)-N(1)-N(2)

O(2)-N(1)-N(2)

O(3)-N(2)-N(1)

O(4)-N(2)-N(1)

O(1)-N(1)-O(2)

O(3)-N(2)-O(4)

179.7

112.7

134.6

179.8

112.8

112.5

134.4

134.9

112.5(4)

113.0(4)

134.5(5)

O(2)-N(1)-N(1A)

O(1)-N(1)-O(2)

Contacts (Å)

F(1) · · · N(1B)

F(1) · · · N(1C)

Xe(1) · · · F(1B)

Xe(1) · · · F(1C)

Xe(1) · · · O(1B)

Xe(1) · · · O(1D)

Xe(1) · · · O(2)

Xe(1) · · · O(2C)

Xe(1) · · · O(2E)

Xe(1) · · · O(1C)

Xe(1) · · · O(1E)

Xe(1) · · · O(2B)

Xe(1) · · · O(2D)

Xe(1) · · · O(2F)

2.720(4)

2.834(5)

3.370(3)

3.516(4)

3.440(4)

3.490(4)

4.180(4)

3.440(4)

3.516(4)

3.435(4)

4.180(4)

F(1) · · · N(1)

2.910

3.245

F(1) · · · N(2)

Xe(1) · · · O(1)

Xe(1) · · · O(2)

Xe(1) · · · O(3)

4.003

3.854

^aB3LYP/aug-cc-pVTZ(-PP). ^bCalculated geometric parameters for

N₂O₄coordinated to XeF₂through two oxygen atoms bound to two

different nitrogen atoms (model A) and two oxygen atoms bound to the

same nitrogen atom (model B).

9

13828 J. AM. CHEM. SOC. VOL. 132, NO. 39, 2010

Xe₃OF₃⁺, Precursor to a Noble-Gas Nitrate

A R T I C L E S

Figure 4. X-ray crystal structure of the XeF₂·N₂O₄molecular addition

compound. Thermal ellipsoids are shown at the 50% probability level. The

atom numbering scheme corresponds to that used in Table 3.

2.7.3. XeF₂·N₂O₄. The XeF₂molecule of XeF₂·N₂O₄(Figure

4) is constrained by crystal symmetry to D_∞hsymmetry and

possesses an Xe-F bond length of 1.985(3) Å, which does not

differ signiﬁcantly from the bond length determined for crystal-

line XeF₂at -173 °C (1.999(4) Å).⁴⁰Although the N₂O₄

molecules were found to have local C_2Vsymmetry (N-O,

1.194(6) and 1.182(6) Å; N-N, 1.738(8) Å; O-N-O, 134.5(5)°),

the N-O bond lengths do not differ by more than (3σ, and

thus they may be regarded as having D_2hsymmetry. The absence

of additional bands in the Raman spectrum resulting from

symmetry lowering and destruction of the center of symmetry

(see Raman Spectroscopy) indicates retention of high (D_2h)

symmetry by N₂O₄. There is close agreement between the

geometric parameters of N₂O₄in XeF₂·N₂O₄and those deter-

mined from the crystal structure of N₂O₄at 100 K (N-O,

1.1855(9) Å; N-N, 1.7560(14) Å; O-N-O, 134.46(12)°).⁴¹

Figure 2. (a) X-ray crystal structure of XeF₂·HNO₃showing the H · · ·F

and Xe · · ·O contacts. Thermal ellipsoids are shown at the 70% probability

level. (b) Calculated geometry of XeF₂·HNO₃. The atom numbering scheme

corresponds to that used in Table 3.

Six N₂O₄molecules are coordinated to each XeF₂molecule

through their oxygen ligand atoms. Three coordination modali-

ties occur: a bidentate interaction with two oxygen atoms bound

to different nitrogen atoms (3.435(4) and 3.516(4) Å) of the

same N₂O₄molecule, a bidentate interaction with two oxygen

atoms bound to the same nitrogen atom (3.440(4) and 4.180(4)

Å), and an interaction with a single oxygen atom (3.490(4) Å)

(Figure 5a). Although the long contacts differ signiﬁcantly from

the two bidentate interactions calculated for the gas-phase adduct

(Figure 5b), the bidentate interaction resulting from each oxygen

atom bound to different nitrogen atoms was found to be 6.5 kJ

mol^-1more stable than the bidentate interaction with both

oxygen atoms bound to the same nitrogen atom. There are also

two long contacts between xenon and two ﬂuorine atoms of

different XeF₂molecules (3.370(3) Å) that are somewhat less

than the sum of the Xe and F van der Waals radii (3.63 Å).³¹

Two long F· · ·N contacts were also observed at 2.720(4) and

2.834(5) Å (cf. sum of the F and N van der Waals radii, 3.02

Å).³¹

2.8. Raman Spectroscopy. The low-temperature, solid-state

Raman spectra of FXeONO₂, XeF₂·HNO₃, and XeF₂·N₂O₄are

depicted in Figures 6 and 7. The experimental and calculated

frequencies and their assignments are listed in Tables 4-7 and

Tables S5-S7 in the Supporting Information.

Figure 3. Packing diagram of XeF₂·HNO₃viewed along the b-axis,

showing the interactions between the XeF₂and HNO₃molecules. Thermal

ellipsoids are shown at the 70% probability level.

1.368(2) Å; N-O_S, 1.216(2) Å; N-O_A, 1.206(2) Å; O_S-N-O_A,

128.2(1)°) are of higher precision when compared with the

previous structure of anhydrous HNO₃(N-O_H, 1.41(2) Å;

N-O_S) N-O_A, 1.22(2) Å; O_S-N-O_A, 130(5)°),³⁹although

they do not differ by more than (3σ.

ˇ

(38) Tramsˇek, M.; Benkicˇ, T.; Zemva, B. Angew. Chem., Int. Ed. 2004,

43, 3456–3458.

(39) Luzzati, P. V. Acta Crystallogr. 1951, 4, 120–131.

(40) Elliott, H. St. A.; Lehmann, J. F.; Mercier, H. P. A.; Jenkins, H. D. B.;

Schrobilgen, G. J. Inorg. Chem. 2010; DOI: 10.1021/ic101152x (in

press).

ˇ

(35) Bunicˇ, T.; Tavcˇar, G.; Tramsˇek, M.; Zemva, B. Inorg. Chem. 2006,

45, 1038–1042.

(41) Kvick, Å.; McMullan, R. K.; Newton, M. D. J. Chem. Phys. 1982,

76, 3754–3761.

ˇ

(36) Tavcˇar, G.; Zemva, B. Inorg. Chem. 2005, 44, 1525–1529.

ˇ

(37) Tramsˇek, M.; Benkicˇ, T.; Zemva, B. Inorg. Chem. 2004, 43, 699–

(42) Guillory, W. A.; Bernstein, M. L. J. Chem. Phys. 1975, 62, 1058–

1060.

703.

9

J. AM. CHEM. SOC. VOL. 132, NO. 39, 2010 13829

A R T I C L E S

Moran et al.

Figure 7. Raman spectra recorded at -160 °C using 1064-nm excitation:

(a) solid XeF₂·HNO₃and (b) solid XeF₂·N₂O₄recorded under frozen N₂O₄

solution. Symbols denote XeF₂(†), FEP (*), laser artifact (‡), and N₂O₄

(#).

Figure 5. (a) Local environment of the XeF₂molecule in the crystal

structure of XeF₂·N₂O₄, showing the three unique coordination pathways

to the xenon atom from N₂O₄. (b,c) Calculated geometries of N₂O₄

coordinated to the xenon atom of XeF₂through two oxygens bound to (b)

two different nitrogen atoms (model A) and (c) the same nitrogen atom

(model B). The numbering scheme corresponds to that used in Table 3.

at all levels (see Computational Results, Table 5, and Table S5

in the Supporting Information). Calculated values referred to

in the ensuing discussion are those obtained by the B3LYP

method. Assignments for XeF₂· HNO₃and XeF₂· N₂O₄were

made by comparison with the calculated frequencies and Raman

intensities (Tables 6 and 7 and Tables S6 and S7 in the

Supporting Information) of the energy-minimized geometries

(Figures 2b and 5b,c) and by comparison with those of XeF₂

(Table S8, Supporting Information), HNO₃(Tables S9 and S10,

Supporting Information), and N₂O₄(Tables S11 and S12,

Supporting Information).

2.8.1. FXeONO₂. The FXeONO₂molecule (C_ssymmetry)

possesses 12 fundamental vibrational modes that span the

irreducible representations 9A′ + 3A′′, which are Raman and

infrared active. The four FXeONO₂molecules in the crystal-

lographic unit cell have C₁site symmetry. A factor-group

analysis for FXe¹⁶ON¹⁶O₂(Table S13, Supporting Information)

predicts that each gas-phase Raman- and infrared-active mode

of FXeONO₂is split, as a result of vibrational mode coupling

within the centrosymmetric unit cell (C_2hunit cell symmetry),

into a maximum of two Raman-active (A_gand B_g) and two

infrared-active (A_uand B_u) components. Accordingly, two

components were resolved for every stretching mode of the

natural abundance molecule except for ν(O_Xe-N) + δ(NO₂).

Vibrational frequencies calculated at all levels of theory

(Table 5 and Table S5 in the Supporting Information) repro-

duced the experimental frequency and ^16/18O isotopic shift trends

for FXe¹⁶O¹⁴N¹⁶O₂, FXe¹⁶O¹⁴N(¹⁶O¹⁸O), and FXe¹⁶O¹⁵N¹⁶O₂

given in Table 4. The calculations clearly show that the

experimental ¹⁸O-enriched spectrum is not that of FXe¹⁸ON¹⁶O₂

Figure 6. Raman spectra of natural abundance FXe¹⁶ON¹⁶O₂(lower trace)

and 98.6% ¹⁸O-enriched FXe¹⁶ON(¹⁶O¹⁸O) (upper trace) recorded at -160

°C using 1064-nm excitation. Symbols denote XeF₂(†), FEP (*), laser

artifact (‡), [NO₂][AsF₆] (§), and [¹⁸ON¹⁶O][AsF₆] (+).

The spectral assignments for FXe¹⁶O¹⁴N¹⁶O₂, FXe¹⁶O¹⁴N-

(¹⁶O¹⁸O), and FXe¹⁶O¹⁵N¹⁶O₂were made by comparison with

the calculated frequencies and Raman intensities for all possible

^16/18O-isotopomers (Table 5). The spectra were calculated using

the aug-cc-pVTZ(-PP) basis set at the B3LYP, PBE1PBE, and

MP2 levels of theory and reproduced the experimental trends

9

13830 J. AM. CHEM. SOC. VOL. 132, NO. 39, 2010

Xe₃OF₃⁺, Precursor to a Noble-Gas Nitrate

A R T I C L E S

Table 4. Experimental Raman Frequencies and Intensities for FXe¹⁶O¹⁴N¹⁶O₂, FXe¹⁶O¹⁴N(¹⁶O¹⁸O), and FXe¹⁶O¹⁵N¹⁶O₂

^aThe Raman spectra were recorded in FEP sample tubes at -160 °C using 1064-nm excitation. Relative Raman intensities are given in parentheses.

The spectra of FXe¹⁶O¹⁴N¹⁶O₂and FXe¹⁶O¹⁴N(¹⁶O¹⁸O) were recorded for the solid products extracted from the product mixture using SO₂ClF solvent.

The spectrum of FXe¹⁶O¹⁵N¹⁶O₂was recorded for the nonextracted reaction mixture. Abbreviations denote shoulder (sh) and broad (br). ^bTwo bands

were also observed at 1448.3(<1) and 1410.3(2) cm^-1and were tentatively assigned to combination modes. ^cTwo bands were also observed at

1436.1(<1) and 1410.7(1) cm^-1and were tentatively assigned to combination modes. A third band at 1362.4(1) cm^-1was assigned to the in-phase

d

e

¹⁸ON¹⁶O stretch of [¹⁸ON¹⁶O][AsF₆]. ∆ν^16/18) ν(¹⁸O) - ν(¹⁶O). The abbreviation denotes average (av). ∆ν^14/15) ν(¹⁵N) - ν(¹⁴N). ^fAbbreviations

denote stretch (ν), bend (δ), twist (F_t), wag (F_w), and rock (F_r). Bond elongations and angle openings are denoted by plus (+) signs, and bond

contractions and angle closings are denoted by minus (-) signs. The band is coincident with the ν₁(A_1g) band of AsF₆

g

-

.

because, in this case, near-zero isotopic shifts are predicted for

ν₁(A′), ν(NO - NO) and ν₂(A′), ν(NO + NO) at 1576 and

1274 cm^-1, respectively. The FXe¹⁸ON(¹⁶O_A¹⁸O_S) and FXe¹⁸ON-

(¹⁶O_S¹⁸O_A) conformers (A denotes ¹⁸O anti and S denotes ¹⁸O

syn about the N-O bridge bond and with respect to the

O-Xe-F group) are also eliminated because, in contrast with

the observed average ^16/18O isotopic shifts of -9.9 and -10.9

cm^-1, respectively, a signiﬁcantly larger isotopic shift (-31

cm^-1) is predicted for both ν₄(A′) and ν₅(A′). The experimental

¹⁸O-enriched spectrum has therefore been assigned to an

equimolar mixture of FXe¹⁶ON(¹⁶O_A¹⁸O_S) and FXe¹⁶ON-

(¹⁶O_S¹⁸O_A) conformers and establishes that scrambling of ¹⁸O

among the bridge site and terminal sites does not occur.

enings on bands other than the symmetric and asymmetric NO₂

stretching bands.

In the case of FXe¹⁶ON(¹⁶O¹⁸O), additional line splittings are

observed for ν₁, ν₂, ν₄-ν₇, and ν₉, which are attributed to

lowering of the “vibrational unit cell symmetry” of FXe¹⁶ON¹⁶O₂

from C_2hsymmetry to lower noncentrosymmetric unit cell

symmetries. Symmetry lowering in the unit cell results from

the combinations and permutations of FXe¹⁶ON(¹⁶O_A¹⁸O_S) and

FXe¹⁶ON(¹⁶O_S¹⁸O_A) isomers that occur among the four vibra-

tionally coupled molecules of the crystallographic unit cell,

giving rise to 16 unit cells having the following symmetries

(relative weights are given in parentheses): C_2h(2), C_i(2), C_s

(2), C₂(2), and C₁(8). With the exception of the C_2hand C_i

symmetric cells, the remaining 12 unit cell symmetries are not

centrosymmetric with respect to the syn-/anti-conformations of

¹⁶O and ¹⁸O in their NO₂groups. As a consequence, the

counterparts of the formally Raman-inactive and infrared-active

A_uand B_ucomponents in the factor-group analysis of

FXe¹⁶ON¹⁶O₂are both infrared and Raman active under C_s(A′′),

C₂(B), and C₁(A) unit cell symmetries (Table S13, Supporting

Information), potentially doubling the number of factor-group

split lines that can be observed when compared with the

spectrum of FXe¹⁶ON¹⁶O₂. Although the additional factor-group

components that correspond to A_uand B_uunder C_2hsymmetry

are not resolved for ν₁, ν₆, ν₇, and ν₉, they are resolved for ν₂,

ν₄, and ν₅, which exhibit the predicted four-line splittings

The Raman spectra of FXe¹⁶ON(¹⁶O_A¹⁸O_S) and FXe¹⁶ON-

(¹⁶O_S¹⁸O_A) exhibit levels of complexity that are not reproduced

by the calculated spectra. The calculations show that the only

modes where the syn- and anti-conformers might be differenti-

ated are ν₁and ν₂, with calculated splittings of 7 and 6 cm^-1

,

respectively. However, these splittings are not resolved in the

experimental spectra but are manifested on the factor-group split

ν₁and ν₂bands as signiﬁcant line broadenings (Figure 6). The

attribution of these broadenings to equal populations of the

FXe¹⁶ON(¹⁶O_A¹⁸O_S) and FXe¹⁶ON(¹⁶O_S¹⁸O_A) conformers is

supported by the spectrum of FXe¹⁶ON¹⁶O₂(Figure 6), which

displays narrow factor-group split bands for ν₁and ν₂. Moreover,

¹⁸O isotopic enrichment does not result in similar line broad-

9

J. AM. CHEM. SOC. VOL. 132, NO. 39, 2010 13831

A R T I C L E S

Moran et al.

consistent with noncentrosymmetric unit cell symmetries. The

vibrational modes of the remaining two centrosymmetric unit

cells having C_isymmetry belong to A_gand A_urepresentations

(Table S13). Factor-group analyses of the C_isymmetric cells

predict that each vibrational band of free FXeONO₂will be split

into two Raman-active A_gand two infrared-active A_ucompo-

nents. As in the case of FXe¹⁶ON¹⁶O₂, the two remaining

centrosymmetric unit cells possessing C_2hsymmetry will give

rise to two Raman-active (A_gand B_g) and two infrared-active

(A_uand B_u) components (Table S13).

These observations establish that ¹⁸O is exclusively terminally

bonded to nitrogen and is not scrambled among the terminal

and bridge oxygen sites of FXeONO₂when the latter molecule

is synthesized by the low-temperature reaction of FXe¹⁸OXe---

FXeF⁺with NO₂F, in accordance with the reaction mechanism

proposed in Scheme 1. A weak band at 1362 cm^-1(Table 4

and Figure 6) also appears in the ¹⁸O-enriched spectrum of

FXe¹⁶ON(¹⁶O¹⁸O) and the products that result from Scheme 1.

The band is assigned to the in-phase ¹⁸ON¹⁶O stretch of the

mixed cation in [¹⁸ON¹⁶O][AsF₆] which arises according to

Scheme 2. The band is shifted 34 cm^-1to low frequency of

ν_s(NO₂) in [N¹⁶O₂][AsF₆] and is in good agreement with the

calculated ¹⁸O isotopic shift of -42 cm^-1for ¹⁸ON¹⁶O⁺(Table

S14, Supporting Information). Both the out-of-phase ¹⁸ON¹⁶O

stretch and δ(¹⁸ON¹⁶O) are predicted to be weak, with

δ(¹⁸ON¹⁶O) displaying a small isotopic shift (-6 cm^-1), but

neither was observed.

The highest frequency modes of natural abundance FXeONO₂

occur at 1571.7, 1578.8 cm^-1and 1262.1, 1285.5 cm^-1and are

assigned to the factor-group split ν(NO - NO) and ν(NO +

NO) stretches, respectively. These bands exhibit the greatest

sensitivity to ¹⁸O and ¹⁵N substitution, displaying substantial

isotopic shifts upon ¹⁸O or ¹⁵N enrichment (¹⁸O, -9.9, -11.4,

and -16.6 cm^-1, respectively; ¹⁵N, -35.1, -35.6, and -15.4

cm^-1, respectively), which are in good agreement with the

calculated isotopic shifts (¹⁸O, -13.6 and -24.2 cm^-^{1 15}N,

;

-38.0 and -14.8 cm^-1, respectively). The bands assigned to

ν(O_Xe-N) + δ(NO₂) (882.9 cm^-1), δ(O_Xe-N-O) + F_r(NO₂)

(725.5 cm^-1), and ν(O_Xe-N) - δ(NO₂) (685.4 cm^-1) also

exhibit signiﬁcant sensitivities upon ¹⁸O substitution (-9.5, -9.9

and, -10.9 cm^-1, respectively), in good agreement with the

calculated shifts (-11.2, -6.7 and, -7.1 cm^-1, respectively).

The band assigned to F_w(NO₃) (769.4 cm^-1) shifts only -3.3

cm^-1to lower frequency, in good agreement with the calculated

value of -4.4 cm^-1. The ν(Xe-F) stretching band is also factor-

group split (478.1, 503.8 cm^-1) and is readily assigned because

it is the most intense feature in the spectrum and because the

bands are insensitive to both ¹⁵N and ¹⁸O substitution.

The bands assigned to ν(Xe-O) + F_w(NO₃) (312.7, 318.9

cm^-1) exhibit a signiﬁcant ¹⁶O/¹⁸O isotopic shift (-5.5, -3.9

cm^-1), with a much smaller ¹⁴N/¹⁵N isotopic shift (-1.8, -1.0

cm^-1), in agreement with the calculated shifts (¹⁸O, -3.9 cm^-1;

¹⁵N, -0.6 cm^-1). The modes occurring below 221 cm^-1were

readily assigned with the aid of the calculated vibrational

frequencies and showed no or very small isotopic dependencies,

as expected for low-frequency bending and rocking modes.

2.8.2. XeF₂·HNO₃. The vibrational spectrum of XeF₂·HNO₃

was assigned under C_ssymmetry. A total of 18 fundamental

vibrational modes belonging to the irreducible representations

13A′ + 5A′′ are expected, all of which are Raman and infrared

active.

The four XeF₂·HNO₃molecular units occupy C_ssites in the

crystallographic unit cell. The factor-group analysis for the

9

13832 J. AM. CHEM. SOC. VOL. 132, NO. 39, 2010

Xe₃OF₃⁺, Precursor to a Noble-Gas Nitrate

A R T I C L E S

Table 6. Experimental and Calculated Vibrational Frequencies for XeF₂·HNO₃

^aInfrared spectrum of monomeric HNO₃in a N₂matrix (6-10 K); taken from ref 42. ^bThe Raman spectrum was recorded in a FEP sample tube at

-160 °C using 1064-nm excitation; relative Raman intensities are given in parentheses. Abbreviations denote shoulder (sh) and broad (br). A band

observed at 496(24) cm^-1was assigned to the ν₁(Σ_g⁺) band of free XeF₂. ^cB3LYP/aug-cc-pVTZ(-PP). Values in parentheses denote Raman intensities

(Å⁴u^-1), and those in square brackets denote infrared intensities (km mol^-1). ^dSubscripts indicate the ﬂuorine atom hydrogen-bonded (F_H) to HNO₃and

the terminal ﬂuorine atom on Xe (F_t). The abbreviations denote symmetric (s), asymmetric (as), stretch (ν), bend (δ), rock (F_r), wag (F_w), 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 HNO₃plane.

XeF₂·HNO₃is provided in Table S15 (Supporting Information)

and predicts that each gas-phase Raman- and infrared-active

band of XeF₂·HNO₃is split, as a result of vibrational-coupling

within the crystallographic unit cell (D_2hsymmetry), into a

maximum of four Raman-active (A_g, B_1g, B_2g, and B_3g) and four

infrared-active (A_u, B_u, B_2u, and B_3u) components. Three

components are observed for the δ(N-O-H) mode, six modes

including that of ν(Xe-F_H) (vide infra) are split into two

components, and no splittings are resolved for the remaining

eight modes in the Raman spectrum (Table 6).

coordinated to metal cations,^44,45where the higher frequency

bands range from 544 to 584 cm^-1and are assigned to Xe-F_t

stretching modes and the lower frequency bands range from

411 to 479 cm^-1and are assigned to bridging Xe-F_bstretching

modes. The observed difference is well reproduced by the

calculations (70 cm^-1).

The frequencies occurring between 624 and 3080 cm^-1are

associated with HNO₃modes and are not coupled to the XeF₂

modes of the adduct. With the exception of ν₃and ν₄observed

at 1291 and 1247 cm^-1, the mode descriptions are the same as

those of matrix-isolated HNO₃. The two modes that are the most

affected upon adduct formation are ν(OH) and F_w(NOH), which

shift to lower frequency by 410 cm^-1and to higher frequency

by 227 cm^-1, respectively, as a result of H-bonding. Signiﬁcant

shifts are also calculated for these modes (-306 and +264 cm^-1,

respectively). The shifts to lower frequency for the stretch and

to higher frequency for the wag are consistent with a H · · ·F

interaction that is stronger than the intramolecular H · · ·O

interaction in matrix-isolated HNO₃. The shifts to low (ν₂) and

high (ν_5-ν₇) frequencies are also reproduced by the calculations.

Although derived from ν₃and ν₄of HNO₃, the shifts for ν₃and

Hydrogen-bonding between the OH group of HNO₃and one

of the ﬂuorine ligands of XeF₂results in loss of the center of

symmetry in XeF₂and in the intense Raman bands at 529 cm^-1

and at 458, 468 cm^-1that are assigned to ν(XeF_t) (calcd 550

cm^-1) and ν(XeF_H) (calcd 480 cm^-1), respectively. Splitting of

the low-frequency mode arises from vibrational coupling within

the crystallographic unit cell. Unlike the Xe-F stretching modes

of free XeF₂(ν₁(Σ_g) ) 497 cm^-1, ν₂(Π) ) 213.2 cm^-1, ν₃(Σ_u)

) 555 cm^-1),⁴³the ν(XeF_t) and ν(XeF_H) modes in adducted

XeF₂are very weakly coupled. The magnitude of the ν(XeF_t)

and ν(XeF_H) frequency difference (66 cm^-1) is approximately

half of the average ranges observed for XeF₂homoleptically

+

ˇ

(44) Tavcˇar, G.; Tramsˇek, M.; Bunicˇ, T.; Benkicˇ, P.; Zemva, B. J. Fluorine

(43) Argon, P. A.; Begun, G. M.; Levy, H. A.; Mason, A. A.; Jones, C. G.;

Chem. 2004, 125, 1579–1584.

ˇ

Smith, D. F. Science 1963, 139, 842–843.

(45) Tramsˇek, M.; Zemva, B. J. Fluorine Chem. 2006, 127, 1275–1284.

9

J. AM. CHEM. SOC. VOL. 132, NO. 39, 2010 13833

A R T I C L E S

Moran et al.

Table 7. Experimental and Calculated Vibrational Frequencies for XeF₂·N₂O₄

^aRaman spectra of solid N₂O₄and XeF₂·N₂O₄were recorded in FEP sample tubes at -160 °C using 1064-nm excitation; relative Raman intensities

b

are given in parentheses. Abbreviations denote shoulder (sh) and not observed (n.o.). A combination band was also observed at 1337(10) cm^-1

.

^cModes associated with uncomplexed N₂O₄were observed at 1726(41), 1385(69), 1375(1), 1337(20), 812(55), 678(7), 498(43), 284(176), 281(181),

121(8), and 79(96) cm^-1. A combination band was also observed at 1354(8) cm^-1^dB3LYP/aug-cc-pVTZ(-PP). Values in parentheses denote Raman

.

intensities (Å⁴u^-1), and those in square brackets denote infrared intensities (km mol^-1). ^eCalculated frequencies correspond to the calculated geometry

in Figure 5b. ^fCalculated frequencies correspond to the calculated geometry in Figure 5c. ^gAbbreviations denote stretch (ν), bend (δ), wag (F_w), twist

(F_t), and rock (F_r). Bond elongations and angle openings are denoted by plus (+) signs, and bond contractions and angle closings are denoted by minus

(-) signs. The labels A (A′) and B (B′) denote oxygen atoms that are coordinated to the same N (N′) atom of O₂N-N′O₂, where A (B) and A′ (B′) are

cis to one other.

ν₄cannot be directly compared because the contributions to these

coupled modes differ signiﬁcantly from those of matrix-isolated

HNO₃.

Moreover, none of the N₂O₄modes of the adduct are signiﬁ-

cantly shifted with respect to those of free N₂O₄⁴⁶(Table 7 and

Tables S7 and S12 in the Supporting Information). Despite the

weak interactions between XeF₂and N₂O₄, the local D_∞h(XeF₂)

and D_2h(N₂O₄) symmetries are retained in the centrosymmetric

unit cell of XeF₂·N₂O₄. Consequently, the formally Raman

inactive ungerade modes corresponding to those of free N₂O₄

and XeF₂are also not observed in the experimental spectrum

of XeF₂·N₂O₄(Table 7). Overall, and unlike XeF₂·HNO₃, the

XeF₂portion of the XeF₂·N₂O₄Raman spectrum is consistent

with a weakly bound molecular addition compound (also see

Computational Results).

2.8.3. XeF₂·N₂O₄. The vibrational modes of XeF₂·N₂O₄were

assigned using the frequencies calculated for models A (Figure

5b) and B (Figure 5c) under C₁symmetry. For each model, a

total of 21 fundamental vibrations belonging to A irreducible

representations are expected, all of which are Raman and

infrared active. Because the unit cell (C_isymmetry) is occupied

by only one XeF₂·N₂O₄molecule having C₁site symmetry, the

Raman bands are not factor-group split.

Models A and B have very similar calculated frequencies

and relative intensities (Table 7 and Table S7 in the Supporting

Information), showing that the two coordination modes observed

experimentally would give rise to nearly coincident spectra. A

notable difference occurs for the mode assigned to ν_as(XeF₂),

which is predicted to have a moderate intensity in the Raman

spectrum for model A (Figure 5b). Only the seven bands that

are predicted to be intense in the Raman spectra of both models

were observed experimentally.

2.9. Computational Results. The electronic structures of

FXeONO₂, XeF₂· HNO₃, XeF₂· N₂O₄, BrONO₂, ClONO₂,

+

Xe(ONO₂)₂, NO₂, FXeON(O)F⁺, and FXeFNO₂were opti-

mized using the B3LYP, PBE1PBE, and MP2 methods and the

aug-cc-pVTZ(-PP) basis sets, resulting in stationary points with

all frequencies real (Tables 5-7 and Tables S3-S7, S14, and

S16 in the Supporting Information; also see Experimental

Section). The compounds XeF₂, HNO₃, and N₂O₄were also

calculated for comparison (Tables S8-S12, Supporting Infor-

mation). Values referred to in the following discussion were

calculated at the B3LYP/aug-cc-PVTZ(-PP) level of theory. The

Only one XeF₂stretching band is observed at 509 cm^-1

,

which is only 13 cm^-1higher in frequency than ν₁(Σ_g⁺) of free

XeF₂(497 cm^-1),⁴³indicating that the interactions with N₂O₄

are weak and that the local D_∞hsymmetry and the inversion

center of XeF₂are retained at the local symmetry level.

(46) Andrews, B.; Anderson, A. J. Chem. Phys. 1981, 74, 1534–1537.

9

13834 J. AM. CHEM. SOC. VOL. 132, NO. 39, 2010

Xe₃OF₃⁺, Precursor to a Noble-Gas Nitrate

A R T I C L E S

PBE1PBE and MP2 values are given in Tables S3-S12, S14,

and S16 of the Supporting Information.

symmetric arrangement of N₂O₄molecules found in the crystal

structure (Figure 5a), both optimized structures (Figure 5b,c)

are asymmetric. The vibrational frequencies of both calculated

structures proved instrumental in assigning the Raman spectrum

of solid XeF₂·N₂O₄(see Raman Spectroscopy).

2.9.2. Natural Bond Orbital (NBO) Analyses. The NBO

analyses were carried out for the MP2- and B3LYP-optimized

gas-phase geometries of FXeONO₂, XeF₂·HNO₃, XeF₂·N₂O₄,

Xe(ONO₂)₂, XeF₂, HNO₃, and N₂O₄. The NBO results are given

in Tables S8 and S17 (Supporting Information). The MP2 and

B3LYP results are similar; only the MP2 results are discussed

here.

(a) FXeONO₂and Xe(ONO₂)₂. As with the geometric

parameters, the charges, valencies, and bond orders were found

to be very similar for FXeONO₂and Xe(ONO₂)₂and are not

discussed in detail for the latter compound. The charge on Xe

is somewhat higher for FXeONO₂(1.15) than for Xe(ONO₂)₂

(1.08), indicative of the greater electron-withdrawing ability of

the ﬂuorine ligand when compared with that of the nitrate ligand.

This is corroborated by comparison with the xenon charge in

XeF₂(1.20). The negative charges on the ﬂuorine (-0.59) and

the oxygen directly bonded to xenon (-0.59) of FXeONO₂are

essentially equal to the ﬂuorine charges in XeF₂(-0.60). The

Xe-F (0.31) and Xe-O (0.36) bond orders of FXeONO₂and

XeF₂(0.34) are also very similar. In the case of the Xe-O-N

bridge, the ratio of the terminal N(1)-O(2,3) bond orders (1.21)

to that of N(1)-O(1) (0.79) is ca. 1.5, which is in accordance

with nearly equal weightings of dominant valence bond

structures I and II. The individual charges of the NO₃group

are 0.66 (N), -0.28 (O_A), -0.34 (O_S), and -0.56 (O bonded

to Xe). The greater negative charge on the oxygen syn to the

FXeO- group suggests a signiﬁcant interaction with the positive

xenon center which polarizes the NO₂group and leads to more

negative charge on the syn oxygen. The calculated Xe · · ·O_S

distance is 2.935 Å, which is signiﬁcantly less than the sum of

the van der Waals radii for xenon and oxygen (3.68 Å) and is

consistent with such an interaction. The NBO analysis also

shows that there is a small Xe · · ·O_Sbond order (0.02). The

overall charge of the NO₃group (-0.56) and the charge on Xe

(1.15) are also consistent with nearly equal contributions from

structures I and II.

2.9.1. Geometries. (a) FXeONO₂. The FXeONO₂geometry

optimized to C_ssymmetry (Figure 1b) starting from the

crystallographic geometry or from a C_2Vsymmetric FXeO₂NdO

molecule having a bidentate nitrate group bonded to xenon and

is in very good agreement with the experimental geometry,

underscoring that the FXeONO₂molecule is well isolated in

the crystal structure. This is particularly true for the bond angles,

which are often most affected by crystal packing. There is

excellent agreement between the calculated (2.148 Å) and

experimental (2.126(4) Å) Xe-O and Xe-F bond lengths

(2.018 and 1.992(4) Å, respectively). The calculations also

accurately reproduce the slight difference between the terminal

N-O bond lengths, which are 1.200 Å for the N-O bond anti

to xenon and 1.211 Å for the N-O bond syn to xenon

(experimental, 1.199(6) and 1.224(6) Å, respectively). Similarly,

all of the calculated angles are in good agreement with the

experimental values (Table 3 and Table S3 in the Supporting

Information).

(b) Xe(ONO₂)₂. Although Xe(ONO₂)₂was not synthesized

in the present work, the energy-minimized geometry was

calculated (Figure S4 and Table S16, Supporting Information).

Both the syn- (C_s) and anti- (C_2V) isomers of Xe(ONO₂)₂were

used as starting geometries, but both geometries optimized to a

lower energy C₁geometry, with a dihedral angle between the

two NO₃planes of 108.9°. The calculated geometric parameters

for FXeONO₂and Xe(ONO₂)₂are very similar, with the Xe-O

bond length being slightly longer for Xe(ONO₂)₂(2.168 Å) than

for FXeONO₂(2.148 Å). The Xe-O-N angle (115.7°) is very

similar to that in FXeONO₂(116.0°), with very similar angles

around nitrogen (Table S16).

(c) XeF₂· HNO₃. The energy-minimized geometry of

XeF₂·HNO₃is a planar structure, in agreement with the

experimental structure (see X-ray Crystallography). The ex-

perimental Xe-F_t, Xe-F_H, and N-O bond lengths, as well as

their associated bond angles, were accurately reproduced by the

calculations, whereas some discrepancies occurred for the

intramolecular contacts. The difference between the experimen-

tal and calculated Xe(1) · · ·O(3) contact distances (3.317(1) and

3.034 Å, respectively) is likely attributable to weak intermo-

lecular contacts from neighboring XeF₂and HNO₃molecules

within the crystal lattice (Table 3). Such contacts, which are

not taken into account by the present model, are expected to

weaken the Xe(1) · · ·O(3) contact. The calculated O-H (1.016

Å) and H · · ·F (1.520 Å) distances differ signiﬁcantly from the

experimental values of 0.83(2) and 1.86(2) Å, which is likely

attributable to the uncertainty in the location of the hydrogen

atom in the experimental electron density map (see X-ray

Crystallography). Although the experimental and calculated

O-H and H · · ·F distances differ signiﬁcantly from experiment,

the calculated O(2) · · ·F(2) distance (2.536 Å) is in reasonable

agreement with the experimental distance (2.690(1) Å).

(d) XeF₂·N₂O₄. An attempt to calculate the structure depicted

in Figure 5a from the crystal structure coordinates failed to

optimize. Therefore, three simpliﬁed models corresponding to

the three coordination modalities observed in the crystal structure

of XeF₂·N₂O₄(Figure 5a) were considered. The monodentate

model optimized to the bidentate geometry shown in Figure

5b. The bidentate models, A and B, were shown to correspond

to local minima (Figure 5b,c), with calculated bond lengths,

bond angles, and contact distances that are in reasonable

agreement with those of the experimental structure. Unlike the

(b) XeF₂·HNO₃and XeF₂·N₂O₄. The charges and valencies

calculated for HNO₃and N₂O₄in their XeF₂adducts are similar

to those calculated for free HNO₃and N₂O₄. In the case of

XeF₂·HNO₃, the H · · ·F bond order (0.04) is somewhat greater

than the Xe · · ·O bond order (0.01), with their small values being

consistent with the weak interactions observed in the crystal

structure. The negative charge on F(1) (-0.57) is somewhat

diminished by H-bond formation relative to that of F(2) (-0.63).

The small bond orders calculated for the O · · ·Xe interactions

in XeF₂·N₂O₄(0.02 for models A and B) are in accordance

with the X-ray crystal structure, where only very weak Xe · · ·O

interactions were observed (3.435(4)-4.180(4) Å, see X-ray

Crystallography). The xenon charges of XeF₂·HNO₃and

XeF₂·N₂O₄(A and B) (1.23 and 1.22, 1.21 respectively) are

essentially unchanged relative to that of XeF₂(1.20), which is

also consistent with a molecular addition compound.

9

J. AM. CHEM. SOC. VOL. 132, NO. 39, 2010 13835

A R T I C L E S

Moran et al.

3. Conclusions

Allied Chemical)²⁹and anhydrous HF (Harshaw Chemicals Co.)⁴⁹

were puriﬁed using the standard literature methods. Nitrogen dioxide

(Matheson) was puriﬁed by pressurizing a sample with high-purity

O₂(Canadian Liquid Air, zero grade) and allowing it to stand at

room temperature for 24 h, followed by removal of unreacted O₂

by pumping at -196 °C and three thawing, refreezing (-196 °C),

and pumping cycles. Red, fuming nitric acid (90%, Fischer

Scientiﬁc Co.) was dried in 30% oleum (1:4 v/v), followed by static

vacuum distillation at 0 °C⁵⁰into an intermediate glass vessel sealed

with a 4-mm J. Young glass/Teﬂon stopcock. Dinitrogen pentoxide

was prepared by dehydration of 100% HNO₃over P₄O₁₀as

previously described.⁵¹Sodium nitrite (BDH Chemical, g97%) and

¹⁵N-enriched NaNO₂(98+ atom % ¹⁵N, Isotec) were dried under

dynamic vacuum at 150 °C for 24 h and used without further

The synthesis and structural characterization of FXeONO₂

conﬁrms the ability of the nitrate ligand to stabilize the +2

oxidation state of xenon. The present synthesis of FXeONO₂

from [FXeOXeFXeF][AsF₆] and NO₂F is the only synthetic

route presently known that yields isolable amounts of FXeONO₂

+

and provides an interesting synthetic application of the Xe₃OF₃

cation, the only known oxide ﬂuoride cation of xenon(II). Raman

and NMR spectroscopic studies, as well as an X-ray crystal-

lographic study, show that FXeONO₂is strongly covalently

bound to xenon, which is corroborated by gas-phase quantum-

chemical calculations. The mechanism leading to the for-

mation of FXeONO₂has been explored using ¹⁸O-la-

beled [FXe¹⁸OXeFXeF][AsF₆]. The exclusive occurrence of

FXe¹⁶ON(¹⁶O¹⁸O) as the labeled product is consistent with the

formation of ¹⁸O(XeF)₂and FXeON(O)F⁺/FXeFNO₂⁺as reac-

tion intermediates. The absence of ^16/18O isotopic scrambling

among the oxygen sites of FXeONO₂was conﬁrmed by factor-

group analyses of the 16 isotopomeric crystallographic unit cells

that result from syn-/anti-isomerization, FXe¹⁶ON(¹⁶O_A¹⁸O_S)/

FXe¹⁶ON(¹⁶O_S¹⁸O_A), among the four FXeONO₂molecules of

the unit cell. A second pathway for oxygen exchange between

53

puriﬁcation. Arsenic pentaﬂuoride⁵²and XeF₂were prepared as

previously described and used without further puriﬁcation. The

syntheses of [Xe₃OF₃][AsF₆] and [Xe₃¹⁸OF₃][AsF₆] are described

in ref 20. Both NO₂F⁵⁴and ¹⁵NO₂F⁵⁵were prepared according to

the literature methods, with some modiﬁcations as described in the

Supporting Information.

4.2. Syntheses of FXeONO₂, FXe¹⁶ON(¹⁶O¹⁸O), and

FXeO¹⁵NO₂. In a typical synthesis, ca. 100 mg of [Xe₃OF₃][AsF₆]

([Xe₃¹⁸OF₃][AsF₆]) was synthesized as previously described²⁰in a

¹/₄-in. o.d. FEP reaction vessel outﬁtted with a Kel-F valve. The

1

reaction vessel also had a section of /₄-in. o.d. FEP tubing fused

to it, forming a side arm that was bent to form an h-shaped reactor.

FXe¹⁶ON(¹⁶O¹⁸O) and N¹⁶O₂led to the formation of a minor

+

1

The h-shaped reactor and a /₄-in. o.d. FEP transfer vessel ﬁtted

amount of ¹⁶ON¹⁸O⁺.

with a Kel-F valve were connected by means of 45° SAE

compression ﬁttings to a three-way FEP connector. A nickel vessel

outﬁtted with a 316 stainless steel valve and containing NO₂F

(¹⁵NO₂F) was connected to the remaining port of the three-way

The X-ray crystallographic and Raman spectroscopic study

of the XeF₂·HNO₃adduct is only the second known example

of a H-bonded adduct of XeF₂, while the molecular addition

compound, XeF₂·N₂O₄, shows that N₂O₄interacts in a bidentate

fashion with the xenon atom through two oxygen atoms bound

to two different nitrogen atoms or through two oxygen atoms

bound to the same nitrogen atom. Computational studies have

provided the geometric parameters of the presently unknown

Xe(ONO₂)₂molecule and have accurately reproduced the

experimental geometric parameters and trends in the vibrational

frequencies of FXeONO₂, XeF₂·HNO₃, and XeF₂·N₂O₄. The

calculated ∆H and ∆G values at 298.15 and 195.15 K show

that the gas-phase decompositions of FXeONO₂and Xe(ONO₂)₂

are spontaneous for all reaction channels considered and are

consistent with the experimental decomposition products.

Although the gas-phase thermochemical parameters for the

1

connector by means of a /₄-in. stainless steel Swagelok Ultratorr

union ﬁtted with Viton O-rings. All connections and the FEP

reaction vessels were passivated with F₂for several hours prior to

use. The nickel NO₂F sample vessel was cooled to -78 °C, and a

portion of NO₂F was condensed into a second ¹/₄-in. o.d. FEP vessel

at -196 °C, followed by warming to -78 °C to control the amount

of NO₂F added to [Xe₃OF₃][AsF₆]. The NO₂F in the auxiliary vessel

(-78 °C) was then condensed onto [Xe₃OF₃][AsF₆] at -196 °C,

followed by warming the reaction mixture to -50 °C, where, after

5 h, the solid slowly changed from a magenta or red-orange color

to a white suspension in excess liquid NO₂F. Excess NO₂F was

removed under vacuum at -110 °C to yield a white, microcrys-

talline, solid mixture of FXeONO₂, XeF₂, and [NO₂][AsF₆]. Sulfuryl

chloride ﬂuoride (ca. 0.75 mL) was then condensed onto the mixture

at -196 °C. The solution was warmed to -30 °C and agitated for

several seconds to effect dissolution of FXeONO₂and XeF₂. The

solid was allowed to settle (ca. 1 min), and the supernatant was

then decanted into the side arm of the reaction vessel at -78 °C.

After the solid had settled to the bottom of the side arm, the

supernatant was decanted into the main tube of the reactor, and

the process was repeated. Longer extraction times (ca. 5-10 min)

and/or more agitation in attempts to improve product yield resulted

in partial [NO₂][AsF₆] extraction, thereby contaminating the ﬁnal

product. The main portion of the reactor, containing [NO₂][AsF₆],

was removed by heat-sealing it off under dynamic vacuum, and

+

-

reaction of FXeONO₂and AsF₅to form XeONO₂and AsF₆

is very endothermic and nonspontaneous, the low-temperature

experimental reaction of FXeONO₂with liquid AsF₅is spon-

taneous, forming [NO₂][AsF₆], Xe, and O₂. The sought-after

[XeONO₂][AsF₆] salt was not observed as an intermediate by

low-temperature Raman spectroscopy. The ﬁndings are consis-

tent with volume-based thermodynamic calculations which

indicate that, while the reaction of FXeONO₂and AsF₅to form

[XeONO₂][AsF₆] is spontaneous under standard conditions, the

salt is unstable with respect to the observed decomposition

products.

(49) Emara, A. A. A.; Schrobilgen, G. J. Inorg. Chem. 1992, 31, 1323–

1332.

(50) Morton, J. R.; Wilcox, H. W. Inorg. Synth. 1953, 4, 52–53.

(51) Schenk, P. W. In Handbook of PreparatiVe Inorganic Chemistry;

Brauer, G., Ed.; Academic Press: New York, 1963; Vol. 1, pp 489-

491.

4. Experimental Section

4.1. Apparatus and Materials. All manipulations were carried

out under anhydrous conditions using glass and metal vacuum lines

as previously described.^44,47Sulfuryl chloride ﬂuoride (SO₂ClF,

(52) Emara, A. A. A.; Lehmann, J. F.; Schrobilgen, G. J. J. Fluorine Chem.

2005, 126, 1373–1376.

(53) Mercier, H. P. A.; Sanders, J. C. P.; Schrobilgen, G. J.; Tsai, S. S.

Inorg. Chem. 1993, 32, 386–393.

(47) Casteel, W. J., Jr.; Kolb, P.; LeBlond, N.; Mercier, H. P. A.;

Schrobilgen, G. J. Inorg. Chem. 1996, 35, 929–942.

(48) Casteel, W. J., Jr.; Dixon, D. A.; Mercier, H. P. A.; Schrobilgen, G. J.

Inorg. Chem. 1996, 35, 4310–4322.

(54) Faloon, A. V.; Kenna, W. B. J. Am. Chem. Soc. 1951, 73, 2937–

2938.

(55) Aynsley, E. E.; Hetherington, G.; Robinson, P. L. J. Chem. Soc. 1954,

1119–1124.

9

13836 J. AM. CHEM. SOC. VOL. 132, NO. 39, 2010

Xe₃OF₃⁺, Precursor to a Noble-Gas Nitrate

A R T I C L E S

1

the FXeONO₂/XeF₂mixture was dried under dynamic vacuum at

-78 °C and stored at that temperature until studied by Raman

spectroscopy.

that was fused to a /₄-in. o.d. FEP reaction vessel and functioned

as an NMR sample tube (see Syntheses of FXeONO₂,

FXe¹⁶ON(¹⁶O¹⁸O), and FXeO¹⁵NO₂). In all cases, the NMR sample

tubes were frozen at -196 °C and heat-sealed under dynamic

vacuum.

4.3. Reactions of XeF₂with HNO₃. On a glass vacuum line,

1

100% HNO₃(0.1047 g, 1.661 mmol) was condensed into a /₄-in.

o.d. FEP reactor ﬁtted with a Kel-F valve. Sulfuryl chloride ﬂuoride

(ca. 1.5 mL) was then condensed onto the sample at -196 °C. The

solution was frozen at ca. -145 °C inside a drybox, and XeF₂

(0.1406 g, 0.831 mmol) was then added to the reaction vessel. The

mixture was warmed to -30 °C and agitated for 1 h, after which

the solvent was removed under vacuum at -78 °C to give white,

microcrystalline XeF₂·HNO₃.

4.4. Attempted Syntheses of Xe(ONO₂)₂and FXeONO₂by

Reaction of XeF₂and [XeF][AsF₆] with N₂O₅. Dinitrogen

pentoxide (ca. 0.100 g, 0.93 mmol) was sublimed into a ¹/₄-in. o.d.

FEP reactor ﬁtted with a Kel-F valve on a glass vacuum line. Liquid

SO₂ClF was then condensed onto the sample at -196 °C (ca. 1.5

mL), followed by pressurizing the reaction vessel with 1 atm of

dry nitrogen. The reactor was transferred into a drybox, and XeF₂

(0.2351 g, 1.389 mmol) or [XeF][AsF₆] (0.3140 g, 0.926 mmol)

was added to the frozen mixture (-140 °C) and removed to a -78

°C dry ice/acetone bath. The mixture of XeF₂/N₂O₅was gradually

warmed to 0 °C to effect dissolution and monitored by low-

temperature Raman spectroscopy but showed no sign of reaction

after 24 h. Further warming to 10-15 °C for 2 days also showed

no sign of reaction.

4.7.2. NMR Instrumentation and Spectral Acquisition. Nuclear

magnetic resonance spectra were recorded unlocked (ﬁeld drift <1

Hz h^-1) on a Bruker DRX-500 spectrometer equipped with a

11.744-T cryomagnet and a 5-mm broad-band inverse probe. The

low-ﬁeld ¹⁹F and ¹²⁹Xe NMR spectra of FXeO¹⁵NO₂in SO₂ClF

were also recorded on a Bruker AV-300 spectrometer equipped

with a 7.0463-T cryomagnet using a 5-mm broad-band probe for

1

¹²⁹Xe and a 5-mm H/¹³C/¹⁹F/³¹P combination probe for ¹⁹F.

The ¹⁹F NMR spectra were recorded at 470.592 (282.363 at

7.0463 T) MHz in 64K memories, with a spectral width setting of

24 (28) kHz, yielding a data point resolution of 0.36 (0.43) Hz/

data point and acquisition times of 1.39 (1.16) s. A relaxation delay

of 0.1 s was applied, and 1500-2000 transients were typically

accumulated using a pulse width of 2.5 (6.0) µs, corresponding to

a bulk magnetization tip angle, θ, of approximately 90°. A line-

broadening of 0.1 Hz was used in the exponential multiplication

of the free induction decays prior to Fourier transformation.

The ¹²⁹Xe NMR spectra were recorded at 138.857 (83.307 at

7.0463 T) MHz in 32 (16)K memories, with a spectral width setting

of 100 (33.3) kHz, yielding a data point resolution of 3.05 (2.03)

Hz/data point and acquisition times of 0.16 (0.25) s. A relaxation

delay of 0.1 s was applied, and 50 000-100 000 transients were

typically accumulated using a pulse width of 12.2 (6.0) µs,

corresponding to a bulk magnetization tip angle, θ, of approximately

90°. A line broadening of 5-20 Hz was used in the exponential

multiplication of the free induction decays prior to Fourier

transformation.

The solid mixture of [XeF][AsF₆]/N₂O₅formed orange and white

clumps under SO₂ClF at -78 °C. When warmed to -30 °C, N₂O₅

appeared to dissolve, and further reaction took place. After ca. 30

min, the orange color dissipated, at which time the sample was

cooled to and stored at -78 °C until a low-temperature Raman

spectrum could be recorded.

The ¹⁴N (¹⁵N) NMR spectra were recorded at 36.141 (50.678)

MHz in 16 (32)K memories, with a spectral width setting of 29.0

(16.7) kHz, yielding a data point resolution of 1.76 (0.51) Hz/data

point and acquisition times of 0.28 (0.98) s. A relaxation delay of

0.05 (180) s was applied, and 100 000 (170) transients were

accumulated using a pulse width of 13.8 (28.0) µs, corresponding

to a bulk magnetization tip angle, θ, of approximately 90°. A line

broadening of 10 (5) Hz was used in the exponential multiplication

of the free induction decays prior to Fourier transformation.

The ^14/15N, ¹⁹F, and ¹²⁹Xe NMR spectra were referenced

externally at 30 °C to samples of neat CH₃NO₂, CFCl₃, and XeOF₄,

respectively. The chemical shift convention used is that a positive

(negative) sign denotes a chemical shift to high (low) frequency of

the reference compound.

4.5. Synthesis of XeF₂·N₂O₄. Xenon diﬂuoride (0.1510 g, 0.892

1

mmol) was added to a /₄-in. o.d. FEP reactor ﬁtted with a Kel-F

valve inside the drybox. Liquid N₂O₄was condensed onto the

sample at -78 °C (ca. 1.5 mL), followed by pressurization with 1

atm of dry nitrogen. The reactor was warmed to 10 °C to effect

complete dissolution of XeF₂and initially gave a pale brown

solution. Cooling to -10 °C resulted in precipitation of the

XeF₂·N₂O₄as a white solid. Adduct formation was veriﬁed by low-

temperature Raman spectroscopy of the compound in the presence

of frozen N₂O₄.

4.6. Raman Spectroscopy. The low-temperature (-160 °C)

Raman spectra of FXeONO₂, XeF₂·HNO₃, and XeF₂·N₂O₄were

recorded on a Bruker RFS 100 FT Raman spectrometer using 1064-

nm excitation and a resolution of 1 cm^-1as previously described.⁵⁶

Spectra were recorded using a laser power of 300 mW and a total

of 600 scans. The white solid mixture of FXeONO₂, XeF₂, and

[NO₂][AsF₆] resulting from the reaction of [Xe₃OF₃][AsF₆] with

NO₂F was recorded. Another spectrum of a mixture of FXeONO₂

and XeF₂extracted with SO₂ClF at -35 °C was recorded after

removal of SO₂ClF under vacuum at -78 °C. The spectrum of

XeF₂·HNO₃was recorded on a sample of XeF₂and HNO₃that

had been allowed to react in a 1:1 ratio in SO₂ClF at -30 °C for

1 h and was isolated by pumping off the solvent under vacuum at

-78 °C. The spectrum of XeF₂·N₂O₄was recorded on a sample

of XeF₂frozen in excess N₂O₄solution.

4.8. X-ray Crystallography. 4.8.1. Crystal Growth. All crys-

1

tals were grown at low temperatures in one arm of a /₄-in. o.d.

FEP T-shaped reactor consisting of a reaction tube heat fused to a

side arm and ﬁtted with a Kel-F valve as previously described.⁵⁷

The main arm of the reaction vessel containing the colorless solution

was placed inside the precooled glass dewar of a crystal growing

apparatus, and the temperature of the dewar and contents was slowly

lowered to induce slow crystal growth.⁵⁷The crystals were isolated

by decanting the solvent under dry nitrogen, at the temperature

achieved at the end of crystal growth, into the side arm of the FEP

vessel, which was initially immersed in a dry ice/acetone bath. The

side arm containing the decanted solution was then cooled to -196

°C, and the apparatus was evacuated. The crystalline products were

dried under dynamic vacuum at the appropriate temperature

(FXeONO₂, -78 °C; XeF₂·HNO₃, -80 °C; XeF₂·N₂O₄, -10 °C)

before the side arm containing the supernatant was heat-sealed off

under vacuum at -196 °C.

4.7. Nuclear Magnetic Resonance Spectroscopy. 4.7.1. NMR

Sample Preparation. Samples for NMR spectroscopy were prepared

in situ by carrying out the scaled-down versions of the previously

described reactions in FEP NMR tubes. The NMR tubes were

constructed from 4-mm o.d. FEP tubing fused to 5-cm lengths of

1

¹/₄-in. o.d. × /₃₂-in. wall FEP tubing which, in turn, were ﬂared

(a) FXeONO₂. Sulfuryl chloride ﬂuoride (ca. 1.5 mL) was

distilled at -78 °C onto a mixture of FXeONO₂, XeF₂, and

[NO₂][AsF₆] (vide supra) contained in a T-shaped FEP reaction

and ﬁtted to Kel-F valves with compression ﬁttings. In the case of

FXeONO₂, samples were prepared by extracting a mixture of

FXeONO₂and XeF₂with SO₂ClF into a 4-mm o.d. FEP side arm

(56) Gerken, M.; Dixon, D. A.; Schrobilgen, G. J. Inorg. Chem. 2000, 39,

(57) Lehmann, J. F.; Dixon, D. A.; Schrobilgen, G. J. Inorg. Chem. 2001,

40, 3002–3017.

4244–4255.

9

J. AM. CHEM. SOC. VOL. 132, NO. 39, 2010 13837

A R T I C L E S

Moran et al.

vessel. The reactor was pressurized with ca. 1 atm of dry nitrogen

and warmed to -30 °C, resulting in dissolution of XeF₂and

FXeONO₂. The colorless solution was cooled to -35 °C inside

the crystal growth apparatus, whereupon colorless plates formed

over a period of 2-3 h. The temperature was lowered to -50 °C

over a period of 3-4 h, allowing for more complete crystallization.

A crystal having the dimensions 0.22 × 0.16 × 0.04 mm³was

selected for low-temperature X-ray structure determination.

(b) XeF₂·HNO₃. Sulfuryl chloride ﬂuoride (ca. 1.5 mL) was

distilled at -196 °C onto a sample of solid XeF₂·HNO₃contained

in a T-shaped FEP reaction vessel, followed by warming to -30

°C to effect dissolution. After several minutes of intense agitation,

the adduct dissolved to give a clear, colorless solution. The solution

temperature was lowered to -40 °C, and after ca. 15 min clear,

colorless plates began to grow on the walls of the reactor. Over

the course of the ensuing 7 h, the temperature was lowered to -60

°C, providing a large quantity of crystalline material. A crystal

having the dimensions 0.20 × 0.13 × 0.04 mm³was selected for

low-temperature X-ray structure determination.

(c) XeF₂·N₂O₄. A solution of XeF₂in liquid N₂O₄was prepared

by condensing N₂O₄(ca. 1.5 mL) onto XeF₂(ca. 150 mg) at -3

°C, followed by dissolution of XeF₂at 10 °C. Crystal growth was

initiated at 3 °C, yielding long needles of XeF₂. Upon cooling to

-3 °C, plates began to form on the needles over the ensuing hour

and continued to grow over the next 6 h, after which time the N₂O₄

supernatant was decanted at -3 °C. A plate having the dimensions

0.16 × 0.08 × 0.04 mm³was selected for low-temperature X-ray

structure determination.

4.8.2. Collection and Reduction of X-ray Data. Crystals were

mounted at -100 ( 3 °C as previously described.⁵⁶Crystals of

FXeONO₂and XeF₂·N₂O₄were centered on a P4 Siemens

diffractometer equipped with a Siemens SMART 1K CCD charge-

coupled device (CCD) area detector that used the program

SMART⁵⁸and a rotating anode using graphite-monochromated Mo

KR radiation (λ ) 0.71073 Å). Diffraction data collection (-30

°C) consisted of a full φ-rotation at ꢁ ) 0° using 0.3° (1040 + 30)

frames, followed by a series of short (80 frames) ω scans at various

φ and ꢁ settings to ﬁll the gaps. The crystal-to-detector distance

was 4.999 cm, and the data collection was carried out in a 512 ×

512 pixel mode using 2 × 2 pixel binning. Processing of the raw

data was completed using SAINT+,⁵⁹which applied Lorentz and

polarization corrections to three-dimensionally integrated diffraction

spots. The program SADABS⁶⁰was used for the scaling of

diffraction data, the application of a decay correction, and an

empirical absorption correction based on the intensity ratios of

redundant reﬂections.

empirical absorption correction based on the intensity ratios of

redundant reﬂections.

4.8.3. Solution and Reﬁnement of the Structures. The

XPREP^63,64program was used to conﬁrm the unit cell dimensions

and the crystal lattices. The solution was obtained by direct methods,

which located the positions of the atoms deﬁning the FXeONO₂

molecule and the XeF₂·HNO₃and XeF₂·N₂O₄adducts. The

structure of FXeONO₂was solved as a racemic twin. The ﬁnal

reﬁnements for all structures were obtained by introducing aniso-

tropic thermal parameters and the recommended weightings for all

of the atoms except hydrogen. The maximum electron densities in

the ﬁnal difference Fourier maps were located near the heavy atoms.

All calculations were performed using the SHELXTL package^63,64

for the structure determinations and solution reﬁnements and for

the molecular graphics.

4.9. Computational Methods. Quantum-chemical calculations

were carried out using MP2, PBE1PBE, and B3LYP methods, as

implemented in the program Gaussian 09,⁶⁵for the geometry

optimizations and vibrational frequencies and intensities, and the

program Gaussian 03,⁶⁶for the NBO analyses. The standard all-

electron aug-cc-pVTZ basis set, as implemented in the Gaussian

program, was utilized for all elements except Xe, for which the

semi-relativistic large core (RLC) pseudopotential basis set, aug-

cc-pVTZ-PP, was used.⁶⁷The combined use of aug-cc-pVTZ and

aug-cc-pVTZ-PP basis sets is indicated by aug-cc-pVTZ(-PP). The

program GaussView⁶⁸was used to visualize the vibrational

displacements that form the basis of the vibrational mode descrip-

tions given in Tables 4, 6, and 7 and Tables S5-S8, S10, S12, and

S14 in the Supporting Information. Natural bond orbital analyses

were carried out using B3LYP and MP2 densities with the NBO

program (version 3.1).^69-71

Acknowledgment. This work is dedicated to Prof. Dr. Boris

Zemva on the occasion of his 70th birthday. We thank the

ˇ

Natural Sciences and Engineering Research Council of Canada

for support in the form of a Discovery Grant (G.J.S.) and the

award of postgraduate scholarships (M.D.M.), and 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 Re-

search Computing Network, www.sharcnet.ca).

Supporting Information Available: Calculated geometrical

parameters for FXeON(O)F⁺and FXeFNO₂(Table S1) and

+

their calculated geometries and relative zero-point energies

A crystal of XeF₂· HNO₃was centered on a Bruker SMART

APEX II diffractometer, equipped with an APEX II 4K CCD area

detector, a three-axis goniometer controlled by the APEX2 Graphi-

cal User Interface (GUI) software,⁶¹and a sealed-tube X-ray source

(Mo target) emitting KR radiation (λ ) 0.71073 Å) monochromated

by a graphite crystal. Diffraction data collection (-160 °C) consisted

of a full φ-rotation at a ﬁxed ꢁ ) 54.74° with 0.36° (1010) frames,

followed by a series of short (250 frames) ω scans at various φ

settings to ﬁll the gaps. The crystal-to-detector distance was 4.969

cm, and the data collection was carried out in a 512 × 512 pixel

mode using 2 × 2 pixel binning. Processing of the raw data was

completed using the APEX2 GUI software,⁶¹which applied Lorentz

and polarization corrections to three-dimensionally integrated

diffraction spots. The program SADABS⁶²was used for the scaling

of diffraction data, the application of a decay correction, and an

(62) Sheldrick, G. M. SADABS (Siemens Area Detector Absorption

Corrections), version 2.10; Siemens Analytical X-ray Instruments, Inc.:

Madison, WI, 2004.

(63) Sheldrick, G. M. SHELXTL, release 5.1; Siemens Analytical X-ray

Instruments, Inc.: Madison, WI, 1998.

(64) Sheldrick, G. M. SHELXTL, release 6.14; Siemens Analytical X-ray

Instruments, Inc.: Madison, WI, 2000-2003.

(65) Frisch, M. J.; et al. Gaussian 09, revision A.02; Gaussian, Inc.:

Wallingford, CT, 2009.

(66) Frisch, M. J.; et al. Gaussian 03, revision D.01; Gaussian, Inc.:

Wallingford, CT, 2004.

(67) 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.

(68) Dennington, R. I.; Keith, T.; Millam, J. M.; Eppinnett, K.; Hovell,

W. L.; Gilliland, R. GaussView, version 3.07 ed.; Semichem, Inc.:

Shawnee Mission, KS, 2003.

(58) SMART, release 5.054; Siemens Energy and Automation Inc.: Madison,

WI, 1999.

(59) SAINT, release 6.01; Siemens Energy and Automation Inc.: Madison,

WI, 1999.

(60) Sheldrick, G. M. SADABS (Siemens Area Detector Absorption

Corrections), version 2.03; Siemens Analytical X-ray Instruments Inc.:

Madison, WI, 1999.

(69) Reed, A. E.; Weinstock, R. B.; Weinhold, F. J. Chem. Phys. 1985,

83, 735–746.

(70) Reed, A. E.; Curtiss, L. A.; Weinhold, F. Chem. ReV. 1988, 88, 899–

926.

(71) Glendening, E. D.; Reed, A. E.; Carpenter, J. E.; Weinhold, F.; NBO,

version 3.1; Gaussian, Inc.: Pittsburgh, PA, 1990.

(61) APEX2, release 2.0-2; Bruker AXS Inc.: Madison, WI, 2005.

9

13838 J. AM. CHEM. SOC. VOL. 132, NO. 39, 2010

Xe₃OF₃⁺, Precursor to a Noble-Gas Nitrate

A R T I C L E S

(Figure S1); Raman spectra of FXeONO₂recorded at -160 °C

after 5 h at -50 °C and after 5 days -78 °C (Figure S2); gas-

phase ∆H°, ∆G°, ∆H_195.15, and ∆G_195.15for the synthesis and

frequencies for HNO₃(Table S10); calculated geometrical

parameters (Table S11) and vibrational frequencies for N₂O₄

(Table S12); factor-group analyses for FXe¹⁶ON¹⁶O₂and

FXe¹⁶ON¹⁶O¹⁸O (Table S13); calculated vibrational frequencies

+

decomposition reactions of FXeONO₂, Xe(ONO₂)₂, XeONO₂,

and N₂O₆(Table S2); experimental and calculated geometrical

parameters for FXeONO₂, XeF₂·HNO₃, and XeF₂·N₂O₄at

PBE1PBE and MP2 levels of theory using the aug-cc-

pVTZ(PP) basis set (Table S3); experimental and calculated

geometrical parameters and NBO analyses for FXeONO₂,

BrONO₂, and ClONO₂(Table S4); packing diagram for

XeF₂·HNO₃(Figure S3); calculated vibrational frequencies and

intensities for the ¹⁶O- and ¹⁵N-enriched isotopomers of

FXeONO₂(Table S5); experimental and calculated vibrational

frequencies and intensities for XeF₂·HNO₃at the PBE1PBE

and MP2 levels of theory (Table S6) and for XeF₂·N₂O₄(Table

S7); experimental and calculated vibrational frequencies, geo-

metrical parameters and bond orders, valencies, and NPA

charges from NBO analyses for XeF₂(Table S8); experimental

and calculated geometrical parameters (Table S9) and vibrational

for N¹⁶O₂

,

¹⁸ON¹⁶O⁺, and N¹⁸O₂(Table S14); factor-group

+

analysis for XeF₂·HNO₃(Table S15); calculated geometrical

parameters for Xe(ONO₂)₂(Table S16); calculated geometry

for Xe(ONO₂)₂(Figure S4); NBO analyses for FXeONO₂,

XeF₂·HNO₃, XeF₂·N₂O₄, Xe(ONO₂)₂, HNO₃, and N₂O₄(Table

S17); solid-state thermochemistry of [XeONO₂][AsF₆]; correla-

tion of ¹⁹F and ¹²⁹Xe NMR parameters and group electronega-

tivities; detailed syntheses of natural abundance and ¹⁵N-enriched

NO₂F; complete refs 65 and 66; and X-ray crystallographic ﬁles

in CIF format for the structure determinations of FXeONO₂,

XeF₂·HNO₃, and XeF₂·N₂O₄. This material is available free

of charge via the Internet at http://pubs.acs.org.

JA105618W

9

J. AM. CHEM. SOC. VOL. 132, NO. 39, 2010 13839

Article Doi

Xe3OF3+, a precursor to a noble-gas nitrate; Syntheses and structural characterizations of FXeONO2, XeF2·HNO3, and XeF2·N 2O4

DOI: 10.1021/ja105618w

Source and publish data:

Authors:

Article abstract of DOI:10.1021/ja105618w

Full text of DOI:10.1021/ja105618w

Products guided by the article

R&D Labs maybe for 1251536-19-2

Relevant to this article

Hot Product