THE CHEMISTRY OF KRYPTON, XENON AND RADON

6. T H E C H E M I S T R Y OF KRYPTON, X E N O N AND RADON NEIL BARTLETT University of California,

Berkeley

and F. O.

SLADKY

University of Innsbruck 1. G E N E R A L

FEATURES

OF

NOBLE-GAS

CHEMISTRY

1.1. H I S T O R I C A L I N T R O D U C T I O N

When xenon compounds were reportedi-3 in 1962, most chemists were greatly surprised; yet for more than 30 years there had been clear indications^ from the chemistry of iodine and the other halogens that the heavier noble gases might well form compounds with the more electro-negative ligands. However, the most promising previous attempts to prepare chemical compounds had failed. This and the key role that the noblegas valence electron configurations assumed in the electronic theories of valence, resulted (at least by the 1960's) in an unquestioning confidence in the chemical inertness of the gases. The discovery of argon (1894) by Rayleigh and Ramsay, was unexpected and was received with much skepticism, even by the great Mendeléevs. Of course, once the gases were established as a new group in the Periodic Classification, all chemists recognized that this new group beautifully completed Mendeleev's table. The inert elements fitted perfectly into the scheme, between the electro-negative elements (the halogens) and the electro-positive elements (the alkali metals). It is a mark of Ramsay's sure chemical awareness that he saw to it that his friend Moissan (see ref. 5, p. 146), the discoverer of ñuorine, was promptly supplied with 100 c.c. of the gas in order that he should attempt to prepare a ñuoride. Moissan was unsuccessful and concluded his account^ of his experiments (reported in 1895): " A la temperature ordinaire ou sous Taction d'une étincelle d'induction un melange de ñuor et d'argon n'entre pas en combinaison." Undoubtedly Ramsay, rightly, considered this the ultimate test for chemical activity, and this presumably led to his disagreement with the suggestion of Oddo^ that krypton and xenon halides should be preparable. 1 N . Bartlett, Proc, Chem, Soc, (1962) 218. 2 H. H. Claassen, H. Selig and J. G. Malm, / . Am, Chem, Soc, 8 4 (1962) 3593. 3 R. Hoppe, W. D á h n e , Η. Mattauch and K. M . Rödder, Angew, Chem, 7 4 (1962) 903. 4 L. Pauling, / . Am, Chem, Soc. 55 (1933) 1895. 5 A n excellent account of the discovery of the noble gases and the early controversy surrounding the argon discovery is given by M. W. Travers in his Life of Sir William Ramsay, Edward Arnold, London (1956). 6 H. Moissan, Bull. Soc. Chim. 13 (1895) 973. 7 G. Oddo, Gazz. chim. ital. 63 (1933) 380. 213

214

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY

The Bohr model of the atom (1913) quickly brought the noble gases to an established key position in the electronic theory of valence, particularly as first propounded by Lewis (1916) and Kossel (1916)8. The Lewis and Kossel theories had immediate impact, in rationalizing much of the chemical behaviour of the elements. The noble-gas valenceelectron configuration was clearly defined as the configuration to which other elements tended in their chemical bonding. As with all very successful theories, the exceptions and awkward cases tended, as time passed, to be ignored. A few chemists persisted, nevertheless, in attempts to bring the heavier gases into chemical combination. Ruif and Menzel (1937) again studied the argon/ñuorine system and also krypton/fluorine^; von Antropofl" (1932-3) examined krypton/chlorine or bromine mixturesio. N o lasting evidence for compounds was found. Of all the post-electronic theory-of-valence predictions of compounds of the gases, that of Pauling was the most accurate. He suggested^ in 1933, from considerations of ionic radii, that XeFe, KrFg, and perxenates should be preparable, and at his suggestion an attempt^ was made to synthesize a xenon fluoride. The attempt failed. Ironically a similar e x p e r i m e n t i 2 , carried out 30 years later but replacing the electric discharge of the early experiment with sunlight, provided a convenient preparative method for xenon difluoride. N o doubt if xenon had been as abundant as argon we should not have had to wait more than 60 years for noble-gas chemistry—it is conceivable that Moissan would have prepared the xenon fluorides in the last years of the nineteenth century, but surely Run" (the first to synthesize iodine heptafluoride) would have succeeded if Moissan had failed. The oxidation of xenon, 1962. The key to the eventual recognition that the noble-gas octet, in the heavier gases, is not chemically inviolate, lay in the long-known ionization potentials of the gases: /(E, g): He, 24.58; Ne, 21.56; Ar, 15.76; Kr, 14.00; Xe, 12.13; Rn, 10.75 eV. Bartlett and Lohmann had discovered^^ an oxyfluoride of platinum which proved to be the salt O2 [PtFö]". This formulation imphed that platinum hexafluoride was capable of oxidizing molecular oxygen, 02(g)+ PtF6(g) O J p t F ö ] "^, and indicated that the hitherto little investigated hexafluoride of platinum was an oxidizer of unprecedented power (with an electron affinity in excess of —160 kcal mole-i). Since the first ionization potentials of Rn and Xe were less than or comparable to the first ionization potential of molecular oxygen (12.2 eV), it was apparent that the heavier gases should be susceptible to oxidation (i.e. should depart from their octet-valence-electron configurations) in interaction with the strongly electron attracting PtFö molecule. As predicted, platinum hexafluoride vapour (deep red) spontaneously oxidized xenon gas (colourless) at ordinary temperatures in a visually dramatic, fast reaction, which deposited a yellow-orange, quinquevalent, platinum complex fluoridei» This and the subsequent discovery of the xenon fluorides^. 3 initiated the surge of activity, the essence of which is reported in this chapter. 8 G. N . Lewis, J. Am. Chem. Soc. 38 (1916) 762; W. Kossel, An. Phys. (Leipzig) 49 (1916) 229. 9 O. R u f f a n d W. Menzel, Ζ. anorg. Chem. 213 (1933) 206. 10 A. v o n Antropoff, K. Weil and H. Fraüenhof, Naturwissenschaften 20 (1932) 688; Α . v o n Antropoff, Η. Fraüengof and Κ. Η. Kruger, Naturwissenschaften 21 (1933) 315. 11 D . Μ. Yost and Α. L. Kaye, / . Am. Chem. Soc. 55 (1933) 3890. 12 L. V. Streng and A. G. Streng, Inorg. Chem. 4 (1965) 1370; J. H. HoUoway, Chem. Educ. 4 3 (1966) 202. 13 N . Bartlett and D . H. Lohmann, Proc. Chem. Soc. (1962) 115. 14 N . Bartlett and N . K. Jha, in Noble Gas Compounds (H. H. Hyman, ed.), University of Chicago Press, Chicago and London (1963), p. 23.

THE RELATIONSHIP OF NOBLE-GAS COMPOUNDS 1.2.

T H E

R E L A T I O N S H I P OF

O F

N O B L E - G A S

T H E

O T H E R

C O M P O U N D S

215 T O

C O M P O U N D S

E L E M E N T S

1.2.1. The Extent of Noble-gas Chemistry So far (1970), the experimental evidence suggests that only the heavier noble gases (Kr, Xe, Rn) can form chemical compounds. The favoured oxidation states are in harmony with the pattern of usual oxidation states of the other non-transition elements, obeying the general rule that the group number (8), or group number minus 2n (n = 0, 1, 2, 3), oxidation states are preferred. The compounds of xenon illustrate this pattern well, with the established positive oxidation states, + 8 (e.g. Xe04), + 6 (e.g. XeFe), + 4 (e.g. XeF4) and + 2 (e.g. XeFz) is. i6. The noble gases are brought into chemical combination only by the most powerfully oxidizing ligands and the known compounds all involve the association of a heavier (more readily oxidizable) noble-gas atom (Rn, Xe or Kr) with electro-negative ligands (e.g. F, O, OsOaF, etc.). Here we see the drive of the electro-negative ligand towards the light noblegas configurations (particularly that of neon). Clearly the drive of the ñuorine atom to attain the neon configuration exceeds the capability of the heavy noble gases to maintain their configurations. Of course, the ligands which provide stable noble-gas compounds are also ligands which promote unusually high oxidation states in other elements. There is a particularly close relationship of noble-gas compounds to non-transition element Group VII and Group VI compounds involving the same ligand. Thus the ñuorides of xenon and krypton show a familial resemblance to the fluorides of iodine, bromine and even tellurium, and the oxides and oxysalts of xenon conform to the pattern of antimony, tellurium and iodine behaviour. 1.2.2. Structural Features TABLE 1. COMPARISON OF X e F + A N D

XeF+

B o n d length (A) V (cm-i) Force constant (mdyn

A-i)

1.84" 621» 3.7»

IF

IF

1.906 610*» 3.6'

" V. M. M c R a e , R. D . Peacock and D . R. Russell, Chem. Communs. (1969) 62. » F. O. Sladky, P. A . BuUiner and N . Bartlett, / . Chem. Soc. (1969) 2179. L. G. Cole and G. W. Elverum, Jr., / . Chem. Phys. 2 0 (1952) 1543. R. A . Durie, Proc. Roy. Soc. A , 2 0 7 (1951) 388. ^ G. R. Somayajula, / . Chem. Phys. 3 3 (1960) 1541.

The simple noble-gas compounds are closely related structurally to analogous compounds of Groups VII and VI. Since each of the noble-gas cations is isoelectronic with its 15 J. G. Malm, H. Selig, J. Jortner and S. A . Rice, Chem. Revs. 6 5 (1965) 199. 16 R. Hoppe, Fortschr. chem. Forsch. 5 (1965) 213.

216

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY

neighbouring halogen atom, it was to be expected that the monofluoro noble-gas cations (e.g. XeF+, KrF+) would show a close relationship to the corresponding halogen monofluoride. Iodine monofluoride and XeF+ are compared in Table 1. Of course, the chemical binding must be similar in these related species, and it is appropriate to think of the bonding in terms of a classical electron-pair bond, each element achieving an octet conñguration. The bonding of two fluorine atoms to a neutral noble-gas atom, as in the generation of XeF2 or KrFa, must be very hke the bonding of two fluorine atoms to the halogen atom of halogen monofluoride: Cl(Br, I)F+2F->Cl(Br,I)F3. Certainly the approximately linear part of the T-shaped BrFa molecule resembles the KrFa molecule as Table 2 shows.

TABLE 2 . S T R U C T U R A L COMPARISON OF K r F 2 WITH B r F A N D BrFs

(Distances in A units)

Kr

Br^il2_F 1-721 86^1

Br-li^F C . Murchinson, S. Reichman, D . Anderson, J. Overend and F. Schreiner, / . Am. Chem. Soc. 9 0 (1968) 5690. D . W. Magnuson, Chem. Phys. 27 (1957) 2 2 3 . D . F . Smith, M . Tidwell and D . V. P. Williams, Phys. Rev. 77 (1950) 420.

The unique bond in BrFa is short, like the bond in the B r - F molecule. Undoubtedly, the K r - F bond in KrF+, like XeF+, will be shorter than in KrF2 and should resemble the B r - F molecule bond. The XeF4 molecule, which is square planar17, is similar to the approximately square IF4 group in the IF5 moleculei». Presumably IF4, like BrF4, is also of geometry. The well-characterized isoelectronic species XeF^, IF5 and TeF^" are very similar in shape, the bond lengths showing a decrease with increase in the atomic number (nuclear charge) of the central atom. The details are given in Table 3. The same kind of relationship occurs in XeOa and lOj, both of which are pyramidal, and Xe04 and IO4, both of which are tetrahedral. The perxenates are octahedral, like the paraperiodates and orthotellurates. 17 J. H . B u m s , P. A . Agron and H . A . Levy, Science 139 (1963) 1208; D . H . Templeton, A . Zalkin, J. D . Forrester and S. M . Williamson, / . Am. Chem. Soc. 85 (1963) 242; J. A . Ibers and W . C . Hamilton, Science 139 (1963) 106. 18 G. R. Jones, R. D . Burbank and N . Bartlett, Inorg. Chem. 9 (1970) 2264.

217

THE RELATIONSHIP OF NOBLE-GAS COMPOUNDS TABLE 3. COMPARISON OF X E N O N C O M P O U N D S W I T H RELATED MOLECULES

(Intemuclear distances in A units) EF5 (CAV symmetry)

Fe

>

F.

TeFj-

E-F. E-F. F.-E-Fe(°)

1.84 1.94 79

XeF+*

EO3 (Csv symmetry)

EO4 (Ta symmetry)

E-O

1.81 1.88 79

1.86 1.89 -81

lOi-

Xe04«

1.79

1.74

E-O 0-E-0(°)

01-3'

XeOa'

1.82 97

1.76 103

EOö (Οι, symmetry)

E-O

[SbOö]«»

[TeOö]»

[lOeP

PCeOel«^

1.97

1.83-1.95

1.93

1.86

• Α . J. Edwards and M . A . Mouty, / . Chem. Soc. A (1969) 703. ^ N . Bartlett, F . Einstein, D . F . Stewart and J. Trotter, / . Chem. Soc. A , (1967) 1190. ^ G. R. Jones, R. D . Burbank and N . Bartlett, Inorg. Chem. 9 (1970). 2264. ** R. W. G. Wyckoff, Crystal Structures, Vol. 3, Interscience, N e w York (1964). • G. Gunderson, K. Hedberg and J. L. Huston, Acta Cryst. A25, S3 (1969) 124. ' B . S. Garrett, Report 1745,97, Oak Ridge National Lab., 1954; Structure Reports 18 (1954) 393. » D . H . Templeton, A . Zalkin, J. D . Forrester and S. M . Williamson in Noble Gas Compounds ( H . H . H y m a n , ed.). University o f Chicago Press, Chicago and L o n d o n (1963), p p . 229-37. ^'Interatomic Distances, (L. E . Sutton ed.), Chem. S o c . Special Publ. N o . 11 (1958). * S. Raman, Inorg. Chem. 3 (1964) 634. ^ A . F. Wells, Structural Inorganic Chemistry, 3rd edn., Oxford University Press, Oxford (1962), p. 335. J. Ibers, W . C. Hamilton and D . R. MacKenzie, Inorg. Chem. 3 (1964) 1412; A . Zalkin, J. D . Forrester, D . H. Templeton, S. M . Williamson and C. W . K o c h , / . Am. Chem. Soc. 8 6 (1964) 3569; A . Zalkin, J. D . Forrester and D . H. Templeton, Inorg. Chem. 3 (1964) 1417.

218

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY

1.2.3. Thermochemical Relationships The structural relationships show that the chemical bonding in each series of related compounds across the Periodic Table cannot change sharply in character and, indeed, suggests that a common bonding description is appropriate. The thermochemical evidence further supports this view. As may be seen from the thermochemical cycle, for a gaseous xenon compound XeLy, formed from xenon atoms and gaseous diatomic molecules L2:

AH'f(XeLy, g) Xe(g) +

y/2L2(g)

-XeLy

yx Mean Thermochemical Bond Energy

the enthalpy of formation of the gaseous compound will be exothermic if the mean thermo­ chemical bond energy (TBE) exceeds the enthalpy of formation of the gaseous ligand atom L. In comparing fluorine oxygen and chlorine ligands it is evident from inspection of l\Hf{U g) 1 9 : F, 18.8; O, 59.2; CI, 29.0 kcal mole-i, that fluorine presents an unusually favourable case. Furthermore, the fluorine atom has a higher electro-negativity than oxygen and chlorine and is also the smallest^o. It is anticipated, therefore, that fluorine will form stronger bonds than chlorine or the other halogens, and that the heavier halides and oxides will be thermodynamically less stable. This is the usual situation, and the noble-gas com­ pounds prove to be no exception. Although all of the xenon fluorides are thermodynamically stable, the oxides and chlorides are not. Chlorides have been synthesized under high energy conditions and also by 1291 decay (see XeCb, and XeCU). Oxides have been obtained from the fluorides by metathetical reactions. This instability conforms with the character of oxides and chlorides of elements close to the heavier noble gases in the Periodic Table, as may be seen in Table 4. Although the mean thermochemical bond energies in the xenon oxides ( < 2 1 kcalmole-i) are lower than in the fluorides (^^30 kcal mole~i), structural and spectroscopic evidence suggests that the intrinsic X e - 0 bond energy is greater than the X e - F bond energy. This point is discussed further in section 1.3.3. This may account for the considerable kinetic stability of xenon(VI) and (VIII) oxides. As may be seen from the mean thermochemical bond energies listed in Table 5, there is, for fluorides, a general, smooth, trend of decreasing mean thermochemical bond energy, from left to right in each period of the Periodic Table. This is also the direction of increasing first ionization potential of the elements and, of course, increasing electro-negativity. If the mean thermochemical bond energy is traced across a series of geometrically related species (e.g. the octahedral or pseudo-octahedral set, T e F 6 , 82; IF5, 64; XeF4, 31 kcal mole-i) it is seen to fall sharply as the noble-gas compounds are approached. 19 JANAF Thermochemical Tables, D o w Chemical Co., Midland, Michigan (December 1960 to December 1969). 20 A. G. Sharpe, in Halogen Chemistry, Vol. 1 (V. Gutmann, ed.). Academic Press, L o n d o n and N e w York (1967), pp. 1-39.

THE RELATIONSHIP OF NOBLE-GAS COMPOUNDS

219

TABLE 4. M E A N THERMOCHEMICAL B O N D ENERGIES FOR C H L O R i D E s t A N D DISSOCIATION ENERGIES OF THE DIATOMIC O X I D E S OF THE X E N O N PERIOD

(Values in kcal m o l e - i )

Dissociation Energies

Geo 158

AsO 115

SeO 101

BrO 56

KrO <1

SnO 127

SbO 89

TeO 81

lO 47

XeO 9

t Lines connect molecules of related geometry. • National Bureau of Standards, Technical N o t e 270-3 (1968). ^ L. Brewer, G. R. Somoyajula and E . Brackett, Chem. Revs. 63 (1963) 111. ^ K. J . Frederick and J . H. Hildebrand, / . Am. Chem. Soc. 60 (1938) 2522. J. H. Simons, / . Am. Chem. Soc. 5 2 (1930) 3490. * R. H. Lamoreaux and W. F. Giauque, / . Phys. Chem. 7 3 (1969) 755. ' L. Brewer and G. M . Rosenblatt, Advances in High Temperature Chemistry, Vol. 2, Academic Press, N e w York (1969), pp. 1-83.

Most importantly, however, in the krypton and xenon periods the bond energy does not fall to zero and, indeed, the trend in bond energies, already available prior to the dis­ covery of xenon fluorides, was sufficiently clear to have provided the unbiased observer2i with an indication that xenon fluorides should not only be bound but also thermodynamically stable. The rapid decline in mean thermochemical bond energies in the fluoride sequence Ge -> Br presages the thermochemical instability of the krypton fluorides. The difluoride of krypton is the only known binary compound available in macroscopic quantities, and its low bond energy of 12kcalmole-i indicates that little hope can be held for higher fluorides since, if the usual pattern holds, the higher fluorides will have slightly lower bond energies. Of course, the entropy of formation becomes an ever more unfavourable feature towards thermodynamic stability as the number of atoms in the organized molecular unit 21 R. Hoppe, Angew. Chem. 76 (1964) 455.

220

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY

TABLE 5. M E A N THERMOCHEMICAL B O N D ENERGIES (kcal m o l e - i ) FOR

NoBLE-GAS

FLUORIDES A N D RELATED

FLUORiDEsf

THE RELATIONSHIP OF NOBLE-GAS

221

COMPOUNDS

TABLE 5 (cont.) Te

I

Xe [XeFg]

~21 est. SnF

t Geometrically related molecular species are cross-linked.

• JANAF Thermochemical Data, D o w Chemical Company, Midland, Michigan, U S A (1969). ^ National Bureau of Standards Report 10074 (July 1969). ^ P. A. G. O'Hare, J. Johnson, B. Klamecki, M. Mulvihill and W. N . Hubbard, / . Chem. Thermodyn. 1 (1969) 177. * P. A . G. O'Hare and W. N . Hubbard, / . Phys. Chem. 69 (1965) 4358. * National Bureau of Standards, Technical N o t e 270-3 (January 1968). ' M. Nichols, P h . D . thesis. University of Durham, 1958. « J. L. Margrave and co-workers. Rice University. *· L. Stein, in Halogen Chemistry, (V. Gutmman,ed.) Academic Press, London and N e w York (1967), Vol. 1, pp. 133-224. * L. Stein and P. L. Plurien, in Noble Gas Compounds (H. H . H y m a n , ed.). University o f Chicago Press, Chicago and London (1963), p. 144. J V. L Pepekin, Yu. A. Lebedev and A . Apin, Zh. Fiz. Khim. 4 3 (1969) 1564. S. R. Gunn, / . Am. Chem. Soc. 88 (1966) 5924; / . Phys. Chem. 71 (1967) 2934.

increases. Thus for the binary fluorides of xenon the entropy changes22 (cal deg-i mole~i) are:

LS}

XeF2

XeF4

XeF6

-26.5

-61

-96

22 B. Weinstock, Ε. Ε. Weaver and C. P. K n o p , Inorg. Chem. 5 (1966) 2189.

222

NOBLE-GAS CHEMISTRY.* NEIL BARTLETT AND F. O. SLADKY

KrF4

The combination of lower bond energy and less favourable entropy make and KrPe markedly less likely than KrPa. A glance at the bond energy data for the fluorides of the Si CI set of elements show that the prospect of obtaining neutral argon fluorides must be very low indeed. It is striking that for the addition of each pair of fluorine atoms in the sequence Xe XeFö the enthalpy changes are approximately constant: hHi

( X e ( , ) + 2 F ( , ) ->

XeF2(g)) =

- 6 5 kcal m o l e - i

(XeF2(,)+2Fc.) -> XeF4(,)) = - 6 1 kcal m o l e - i AÄ'a (XeF4(,)+2F(,) -5- XeFöt,)) = - 5 5 kcal m o l e - i

This is similar to the situation in the chlorine and bromine fluorides. The enthalpy change for addition of each pair of fluorine atoms in the sequence CIF CIF5 involves approxi­ mately the same energy: LHi

(C1F(,)+2F^,) -> C1F3(,)) = - 64 kcal m o l e - i

^H2 (ClF3(,) + 2Fc,) ^ ClFsi,)) = - 5 6 kcal m o l e - i this being approximately the same as for a single fluorine atom union with chlorine in CIF: hH

(Cl(,)+F(„

C1F(„) = - 6 1 kcal m o l e - i

The bromine fluoride enthalpy relationships are similar, but the BrF molecule bond energy (—60 kcal mole-i) is only 5 0 % greater than the enthalpy change {ca. 40 kcal g atom"i) for the addition of each fluorine hgand in the sequence BrF BrFs. The halogen monofluorides are classical "electron-pair bonded" compounds, whereas the higher fluorides must involve either higher valence orbitals of the central atom or less than electron-pair bonding, for some of the ligands at least. It is therefore reasonable that the halogen monofluoride TBEs should be higher. Cleariy, the isoelectronic noble-gas relatives XeF+, KrF+, ArF+, etc., should also show stronger bonding than the neutral fluorides. This is so for XeF+, where the bond energy is ca. 47 kcal mole-i (see section 3.2.1), in contrast to the mean bond energy in XeFa of 32 kcal mole-i. The KrF*^ bond energy must certainly be greater than 12 kcal mole-i and it is argued that ArF+ should also be a bound species22*. It should be appreciated, however, that the ArF+ bond is unlikely to be as strong as the bond in its isoelectronic relative C1F22»>. Unfortunately, the enormous electron affinity of the ArF+ ion, which must exceed —14.0 eV (see section 2.2.1), will demand an unusually oxidatively resistant counter anion if it is to be stabihzed. The change in bond energy in the iodine fluorides (TBE: I F , 67; IF3, 65 (est.); IF5, 63 kcal mole-i) is much less dramatic than in the chlorine and bromine series, but there is a sharp decline with IF7 (TBE 55 kcal mole-i). This represents an enthalpy change for the process: IF5(,) + 2F(,) -> IF7(„ = - 6 5 kcal m o l e - i which contrasts with Δ ^ , for each previous fluorine ligand pair addition, of ca, 125 kcal mole-i. It may be that this lower enthalpy derives in part from the ligand crowding 22- J. F. Liebman and L. C. Allen, Chem, Communs, 23 (1969) 1335. 22«» J. Berkowitz and W. A . Chupka, Chem. Phys, Letters, 7 (1970) 447, have determined the bond energy for ArF+ to be > 3 8 . 2 kcal m o l e ' i .

BONDING IN NOBLE-GAS COMPOUNDS

223

in molecular IF7. If so, we can be certain that the steric inhibition to XePg formation will be even more pronounced. The extrapolated mean bond energy given in Table 5 may therefore be over-optimistic of XePg formation. Since the fluorides are thermodynamically the most favourable compounds of the noble gases, it is evident from Table 5 that unless remarkable oxidatively resistant anion sources can be found to stabilize cations like ArF+ and NeF+, the chemistry of the noble gases will be limited to the heaviest elements—Rn, Xe and Kr.

1.3. B O N D I N G I N N O B L E - G A S

COMPOUNDS

1.3.1. Introduction Bonding in noble-gas compounds has attracted much attention from theorists as well as experimentalists. Unfortunately, there is still no definitive finding to resolve the central controversy, which is concerned with the degree of involvement of "outer" valence-shell Orbitals of the noble-gas atom in bonding. Obviously any acceptable theory should account for the following observations: (a) only the heavier, more readily ionizable, gases, form compounds, and (b) only the most electro-negative atoms or groups are satisfactory ligands for the noble gases. Clearly, no matter what the nature of the bonding, the noble-gas atom, in a compound, should bear a net positive charge and the ligands should be negatively charged. From the outset, the majority of bonding models have preferred to involve only s and ρ valence-shell orbitals of the noble gas in bondingi^. 23. Most of these models use only the ρ orbitals and preserve the octet as a bonding criterion. The proponents of these models argue that the promotional energy necessary for the involvement of "outer" noble gas orbi­ tals in bonding is greater than the possible bond-energy gain. Thus it has been pointed out that for the xenon atom the spectroscopic data24 indicate the 5p 5d promotion energy to be 10 eV. It appears then that the valence-state promotion energy for the pen­ tagonal bipyramidal valence-state xenon atom, appropriate for XeFe formation, would require 1100 kcal g a t o m " i valence-state promotional e n e r g y 2 5 . 25«. Most theorists have been unwilling to accept that this energy expenditure could be provided by the enhanced bond energy. This has not been the unanimous view, h o w e v e r 2 6 , and recent w o r k 2 5 suggests that outer orbitals may play a major role in the bonding, at least in the higher noble-gas oxidation states27.

The electron-pair-bond descriptions of the noble-gas halides, which models either implicitly or explicitly involve "outer" d (or / ) orbitals of the noble gas in bonding, are supported by the phenomenological evidence. Thus the noble-gas and other non-transition element compounds exhibit close physical relationships to their transition element relatives.

2 3 C. A . Coulson, / . Chem. Soc. (1964) 1442. 2 4 C. E. Moore, Atomic Energy Levels, Nat. Bur. Stand, Circular 467, Vol. 3, Washington (1958). 2 5 K. A. R. Mitchell, Chem. Soc. A (1969) 1637; K. A . R. Mitchell, / . Chem. Soc. (1969) 368. 2 5 · Valence-state promotion energies o f several hundred kcal m o l e - i are probably c o m m o n ; thus the tetrahedral carbon a t o m r e q u h - e s 2 8 /χ, 160 kcal g a t o m s - i in its valence-state promotion, but this is m o r e than adequately returned in enhanced b o n d energy. 2 6 L. C. Allen and W. D e W. Horrocks, / . Am. Chem. Soc. 8 4 (1962) 4344. 2 7 C. A . Coulson and F. A . Gianturco, / . Chem. Soc. A (1968) 1618. 2 8 L. H. Long, Quart. Revs. 7 (1953) 134.

224

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY

Furthermore, the very successful electron-pair repulsion rules^^ for molecule and ion shape are as effective for noble-gas compounds and their relatives as for classical "octet" compounds. 1.3.2. A Comparison of Non-transition Element with Related Transition Element Compounds It is generally accepted that all of the valence-shell orbitals of the transition elements are available for bond formation. Thus in tungsten hexafluoride a d'^sp^ tungsten atom orbital hybridization (or its equivalent) is assumed for the σ bonds—π bonding involving TABLE 6. A

COMPARISON OF SOME T R A N S F H O N A N D NON-TRANSITION ELEMENT C O M P O U N D S

Molecule

TeFö

OSO4

IF7

ReFy

Xe04

V Symmetry E - L (A units) /,(mdynÄ-i)

Vi (cm-i) TBE (kcal mole-i)

J

O.-

1.833* 5.1 * 769· 121 m

1.83' 5.01 J 701· 82»

— —

736« 100 est.

1.825^ 3-4 676« 55»

1.74» 7.14*» 971 127 p

1.74·» 5.75*» 906 est. 21a

• K. N a k a m o t o , Infrared Spectra of Inorganic and Coordination Compounds, Wiley (1963). ^ R. D . Burbank and N . Bartlett, Chem. Communs. (1968) 645. « E . W. Kaiser, J. S. Muenter, W. Klemperer, W. E . Falconer and W. A . Sunder, Bell Telephone Report, 1970. ^ W. A . Veranos, Bull. Soc. Chim. Beiges 7 4 (1965) 414. • M. Kimura, V. Schomaker, D . W. Smith and B. Weinstock, / . Chem. Phys. 48 (1968) 4001. ' Mean o f 1.84 and 1.82 A quoted respectively by L. Pauling and L. O. Brockway, Proc. Natn. Acad. Sei. 19 (1933) 68, and H. Braune and S. Knoke, Z. Phys. Chem. (Leipzig) B21 (1933) 297. « T . Ueki, A . Zalkin and D . H, Templeton, Acta Cryst. 19 (1965) 157. ^ C. Gunderson, K. Hedberg and J. L. Huston, Acta Cryst. A , 25, S3 (1969) 124. » H. H. Claassen, / . Chem. Phys. 30 (1959) 968. Í K. O. Christy and W . Sawodny, Inorg. Chem. 6 (1967) 1783. R. K. Khanna, / . Mol. Spectrosc. 8 (1962) 134. » H. H . Claassen, E . L. Casner and H . Selig, J. Chem. Phys. 49 (1968) 1803. " P. A . C. O'Hare and W. N . Hubbard, J. Phys. Chem. 70 (1966) 3353. " P. A . C. O'Hare, J. L. Settle and W. N . Hubbard, Trans. Faraday Soc. 62 (1969) 558. • J ANAF Thermochemical Tables, D o w Chemical C o . , Midland, Michigan, Supplement 32 (31 D e c . 1969). ρ Nat. Bur. Stand. Technical N o t e 270-4, U S Department o f Commerce, Washington, D C (May 1969), S. R. Gunn, / . Am. Chem. Soc. 87 (1965) 2290.

the O t h e r d orbitals may also be admitted. Similarly, rhenium heptafluoride is assumed to involve d^sp'^ rhenium σ orbital hybridization. Also, osmium tetroxide may be described in terms of four σ orbitals involving osmium sp^ hybrids, plus π orbitals, involving at least four osmium d orbitals. Now, as may be seen in Table 6, the analogous non-transition element compounds TeFe, IF7 and Xe04 are geometrically akin to their transition-element relatives. It should

29 R. J. Gillespie and R. S. N y h o l m , Quart. Revs. 11 (1957) 339.

225

BONDING IN NOBLE-GAS COMPOUNDS

also be noted that OsFs ^o, like XePg (see section 3.5.1), is unknown, all evidence pointing to both as very unstable species. Similarly, OSO2F4 and Xe02F4 (see section 3.5.3) are unknown and again appear to be very unstable species, whereas both OSO3F2 and Xe03F2 are known (the latter is kinetically stable to 'decomposition, although, thermodynamically unstable (see section 3.5.2)). These admittedly superficial relationships do give the impression that the bonding in the non-transition element compounds should be akin to that in their transition element relatives. Since the bond lengths and stretching force constants for WFe and TeFe are similar, we can assume that the intrinsic bond energy (which is represented in Fig. 1) must have Valence State

E'(,, (v.s.)

L'(,)(v.s;)

Valence State Promotion

X X Intrinsic .Bond Energy

AH(Ev.s.)

Energy Ground Slate

^(g)

X L , (g)

X Χ TBE

-ELx

F I G . 1. The interrelationship of intrinsic bond energy, valence-state promotion energy and mean thermochemical bond energy

approximately the same value in the two cases. It is a reasonable assumption that a state F atom is already in its valence state, but in any case it is unlikely that the valence state is very different for WFe and TeFö. Differences in the total valence-state-promotion energy for W F ö and TeFe can therefore be related to differences in the central-atom valence-state-promotion energy. Since the IBE for both W F ö and TeF^ appear to be similar, the differences in TBE presumably reflect differences in the W and Te valence-state-promo­ tion energies. It would appear then, that the valence-state-promotion energy for the tellurium atom, i.e. i\H (Te(g)(ground-state) Te(e)(valence-state), exceeds the corres­ ponding tungsten promotion by 240 kcal g a t o m " i . It is not inconceivable that this energy could represent the difference in the valence-state-promotion energies for d'^sp^ W and sp^d^ Te, each in an octahedral field of six fluorine hgands. The oxides Xe04 and OSO4 are also similar. The bond lengths are akin and the bond stretching force constants not too dissimilar. Again, the intrinsic bond energies should be ahke. This is in sharp contrast to the impression given by the mean thermochemical bond energies. It is therefore probable that the valence-state-promotion energy for Xe04 forma­ tion exceeds that for OSO4 formation by approximately 420 kcal mole-i. It is possible that this derives from d ( o r / ) orbital utihzation in the xenon bonding. It should also be noted that although the bond lengths decrease and the stretching force constant increases in the sequence XeF2, 2.00 A, 2.8 mdyn/A; XeF4, 1.95 A, 3.02 mdyn/A; XeFö, 1.89 A, the TBE is almost constant (see Table 5). Clearly, the TBE should increase in this sequence. Evidently, some valence-state promotioa must be involved in xenon fluoride formation (at least for the higher fluorides). 30 O. Glemser, Η. W. Roesky, K. H. Hellberg and H. U . Werther, Chem. Ber. 99 (1966) 2652; N . Nghi and N . Bartlett, Compíes Rendu, Acad. Sc. 269 (1969) 756.

226

NOBLE-GAS CHEMISTRY.* NEIL BARTLETT AND F. O. SLADKY

1.3.3. Bonding Without 'Outer" Noble-Gas Orbitals There is a considerable weight of expert opiniones, 23,31 that the bonding in noble-gas compounds does not involve outer, or higher valence-shell, orbitals of the noble-gas atom, at least not to an extent which could significantly affect the bond energy. Thus xenon is considered to use only its 5p and possibly 5^ orbitals in bonding, which is essentially of sigma type. The bonding has usually been discussed in molecular orbital terms, although Coulson has favoured the valence bond scheme for his discussion of the xenon fluorides23. In this model, XeF2 is represented as primarily involving resonance between F-Xe+F" and F-Xe+-F. This scheme preserves both classical concepts: the "octet" and the "electronpair bond".

00

00

00

anti-bonding non-bonding

bonding

F I G . 2. Simplified representation o f the ρσ m.o.s for XeFa.

Halides, The widely accepted simple m.o. model, proposed independently by Pimentel32 and R u n d l e 3 3 , applies well to noble-gas hahdes. It is of historical interest that in Pimentel's 1951 paper, which was devoted primarily to bonding in trihahde ions, he discussed the validity of his bonding scheme for noble-gas dihalides also. This model is best presented with a brief description of xenon difluoride. The krypton difluoride case can be taken as similar. Three three-centre molecular orbitals (represented in Fig. 2) generated from the Xe 5px, and a 2ρχ orbital of each ñuorine ligand (molecular axis being taken as x), are of most favourable energy, with a colinear disposition of the atomic orbitals from which they are derived. The best net bonding for the three atoms occurs when the arrangement is centro-symmetric. The observed D^H molecular symmetry of XeF2 (see section 3.2.1) is in harmony with these requirements. Since the noble-gas atom contributes two electrons and each of the ñuorine hgands only one electron to this ρσ m.o. system, the anti-bonding m.o. remains empty. In effect the filling of the non-bonding orbital "restores" the ñuorine ligand electron density since the orbital is largely concentrated on the ñuorine hgands. The bonding pair of electrons is, therefore, responsible for the binding together of all three atoms. Obviously the delocahzation of the electron pairs in this orbital must result in a net negative charge on each of the ñuorine ligands, thus leaving the xenon atom positively charged. Because of this, the binding has been termed "semi-ionic"i5. The best calculations place the net charge distribution to be close to the representation F^-Xe^+F^-. As will be shown (see section 3.2.1) this assignment fits most, if not all, of the physical and chemical 31 L. C. Allen, in Noble Gas Compounds, p. 317. (H, H . Hyman, ed.). University of Chicago Press, Chicago and London (1963), 32 G. C. Pimentel, / . Chem. Phys. 19 (1951) 446. 33 R. E, Rundle, / . Am. Chem. Soc. 8 5 (1963) 112.

BONDING IN NOBLE-GAS COMPOUNDS

227

properties of the compound. A similar view of the bonding in X e F i is given by Bilham and Linnett's representation^^:

\ / Xe

Each fluorine ligand possesses a closed spin quartet, the other quartet being shared with the central xenon atom, which, in effect then, provides the two electrons for bonding. Bonding in the xenon tetrafluoride molecule may be dealt with similarly, on the basis of two such three-centre systems33. The square planar geometry of the molecule is "pre­ dicted" by this approach as is the approximate charge distribution Xe2+ (F^"")4. A full molecular orbital treatment (see section 3.3.1) is needed to interpret the detailed spectroscopic (electronic) properties, however. The simple three-centre four-electron m.o. approach fails in the case of XePe, since it predicts an octahedral symmetry for the monomer. It is necessary to use a full m.o. description (see section 3.4.1) in order to account for the observed non-octahedral geometry, although again, in this description, higher orbitals are not necessarily involved and the bonding may simply amount to one-electron bonding. Note that the Linnett and Coulson models would allow for non-octahedral XeF^. We would expect the one-electron bonds in the noble-gas hahdes to be weaker than in related compounds, where electron-pair bonding must occur. In particular, we would expect the X e - F bond in XeFi to be weaker than in XeF+. This is supported by the experi­ mental findings (see section 3.2.6) given in Table 7. As has already been discussed (section 1.2.2), the relationship of XeF+ to XeF2 is very similar to that of the halogen monofluorides to their higher fluorides. Thus for bromine fluoride structures, shown in Table 7, the electron-pair bond length would be represented as 1.68-1.76 A and the one-electron bond as 1.78-1.88 A. It is apparent that the electron-pair bonds are not twice as strong as the one-electron bonds as the above bonding description implies. Clearly, either the repre­ sentation of the bonding in terms of assemblies of two-atom and three-atom centres is not adequate, or the total neglect of outer orbital contributions to the bonding is at fault. The bond shortening in the series of fluorides XeFz (2.00 A), XeF4 (1.95 A), XeFö (1.89 A) (see sections 3.2.1, 3.3.1 and 3.4.1) may be accounted for in terms of the increase in the charge on the xenon atom. TTie same feature is seen in the bromine fluorides and the other halogen fluorides^s. There is also the possibility of greater outer orbital participation of the central atom in bonding, the higher the oxidation state27, and this participation may contribute to the bond shortening. Oxides, The only noble-gas oxides established so far are the trioxides and tetroxides of xenon. Although the X e - O bonds are shorter (XeOa, 1.76; Xe04, 1.74 A; see sections 3.4.4 and 3.5.4) than X e - F bonds and the X e - O stretching force constants, e.g. fr (XeOa) = 5.66 mdyn A-i (section 3.4.4), are greater than for X e - F , e.g. /r(XeF4) = 3.02 mdyn A-i (section 3.3.1), the mean thermochemical bond energy for XeO is < 2 1 kcalmole-i, which contrasts with the X e - F value of ^^31 kcal mole-i. Evidently some oxygen or xenon valence-state excitation is involved in the oxygen-hgand 34 J. Bilham and J. W. Linnett, Nature 201 (1964) 1323. 35 A. B. Sharpe, in Halogen Chemistry (V. Gutmann, ed.). Academic Press, L o n d o n and N e w York (1967), Vol. 1, pp. 133-224.

228

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY TABLE

7.

COMPARISON OF BOND LENGTHS I N X e F + ,

XeF2

A N D BROMINE

FLUORIDES

(Intemuclear distances in A units) * Electron-Pair-Bond' Species: ^Br.

1-70

X e ^ F -

(^)

(C2v)

'Single-Elearon-Bond' Species:

-Br-

F

^Xe

2-00

(e) F

F (/>«>/.)

A Mixed Species:

F-^^—Br

F 1-72

• D . F . Smith, M . Tidwell and D . V . P. Williams, Phys, Rev, 7 7 (1950) 420. A . J. Edwards and G. R. Jones, Chem, Communs, (1967) 1304. « V. M . McRae, R. D . Peacock and D . R. Russell, Chem, Communs, (1969) 26. * W . G. Sly and R . E . Marsh, Acta Cryst, 10 (1957) 378. « S. Reichman and F . Schreiner, / . Chem. Phys, 5 1 (1969) 2355. ' D . W . Magnuson, / . Chem, Phys. 2 7 (1957) 223.

bonding. Again, it is commonly assumed that higher orbitals (e.g. Xe 4 / and 5d) play an insignificant role. Since the oxygen atom can receive two electrons, the simplest representa­ tion for the X e - O bond is as an electron-pair bond, both electrons, in effect, being provided by the xenon atom, i.e. a classical semi-ionic linkage X e : O. Of course, the appropriate valence state of the oxygen atom for such a bond would be which is 45.1 kcal mole-i above the ground state^ö. On this basis, then, the intrinsic X e - O bond energy would be the mean thermochemical bond energy {^21 kcalmole-i) plus the oxygen valence-state promotion energy ( ^ 4 5 kcal mole-i), i.e. 65 kcal mole-i, which is more in keeping with the vibrational spectroscopic data. Now the X e - O bond must have considerable polarity, since any oxygen ligand share of the bonding electron pair will contribute net negative charge to that ligand and corresponding positive charge to the xenon atom. For an ideally shared electron pair, the charge distribution should amount t o Xei+—Oi~. Valence-state promotional enhancement of the bonding in the oxides may well account for the kinetic stability of XeOa and Xe04. 36 G. Herzberg, Atomic Spectra and Atomic Structure,

D o v e r Publications, N e w York (1944).

229

BONDING IN NOBLE-GAS COMPOUNDS

1·3·4· Electron-pair Repulsion Theory Although, as has been remarked, there has been considerable reluctance on the part οΓ those providing bonding models for noble-gas compounds to admit the involvement of outer noble-gas orbitals in bonding, it is striking that the theory which, almost without exception, has correctly predicted the geometrical features of the noble-gas compounds, has been the electron-pair repulsion theory. This theory implies outer orbital involvement in the bonding. Electron-pair repulsion theory, which was first formulated independently by Tsuchida37 and Powell and Sidgwicks«, gained widespread appeal when it was rendered much more quantitative by Gillespie and Nyholm29. The theory assumes that each halogen-Hgand bond to the central atom involves an electron pair, and also usually assumes that the binding of a unidentate oxygen ligand involves four electrons (a double bond). Further­ more, all non-bonding valence electrons are assumed to have steric effect (i.e. be in direc­ tional orbitals). A basic tenet of the Gillespie-Nyholm rules is that the non-bonding electron pairs repel other electron pairs more than do bonding electron pairs (although the two pairs of an oxygen-ligand bond have approximately the same repulsive effect as a nonbonding pair). TABLE 8. SOME MOLECULE A N D I O N SHAPES PREDICTED BY ELECTRON-PAIR REPULSION THEORY

(Internuclear distances are in A units)

Species

XeF2

XeF4

Central atom coordination

5-trigonal bipyramid (3 n.-b.e.p.; 2 b.e.p.)

Predicted geometry

-Xe

-Xe-

6-pseudo octahedron (2 n.-b.e.p.; 4 b.e.p.) Xe F

XeFö

Observed structure

Either: 7-pseudo pentagonal bipyramid (1 n.-b.e.p.; 6 b.e.p.)

< ] F

F

X e - ! ^ F

F

XeFö m o n o m e r is not octahedral.* XeFöi^) undergoes rapid intramolecular rearrangement through Cap, Civ and Cs symmetry species.*

37 R. Tsuchida, Rev. Phys. Chem. Japan 13 (1939) 31 and 6 1 ; Bull. Chem. Soc. Japan 14 (1939) 101. 38 N . V. Sidgwick and H. M. Powell, Proc. Roy. Soc. 197 (1940) 153. 39 R. J. Gillespie, in Noble Gas Compounds (H. H. Hyman, ed.), University of Chicago Press, Chicago and London (1963), p. 333.

230

NOBLE-GAS CHEMISTRY: NEIL BARTLETT A N D F. O. S L A D K Y TABLE 8

Species

Central a t o m coordination

(cont.)

Predicted geometry

The cubic crystalline phase contains tetramers and hexamers which are essentially X e F j F " clusters. ·

or: 7-face occupied octahedron (1 n.-b.e.p.; 6 b.e.p.)

^

XeFj

Observed structure

( C „ )

5-pseudo trigonal bipyramid ( 2 n . - b . e . p . ; 3 b.e.p.)

^* F

X

e

unknown ^

^

< 9 0 · ν , ; Ta (Xe-F. >

XeFJ

XeFe)

F.

6-pseudo octahedron (1 n.-b.e.p.; 5 b.e.p.)

f

F

M-88

··

(C4v)

(C4,) (Xe-Fa
OXeF4

6-pseudo octahedron (1,4-e.b.; 1 n.-b.e.p.; 4 b.e.p.) F ^

··

^

F

F<^,,o^F

(C4,)

02XeF2

5-pseudo trigoml bipyramid (2,4-e.b.; 1 n^b.e.p.)

(C4.)

<180

^

X e )

~ 1 8 0 «

F

F (C2v)

OsXeFa

5-trigonal bipyramidal (3,4-e.b.;2b.e.p.)

Xe-

F

Molecular unknown

geometry

231

BONDING IN NOBLE-GAS COMPOUNDS TABLE 8 Species

Xe03

Central a t o m coordination

(cont,) Predicted geometry

Observed structure

J

4-pseudo tetrahedral (1 n.-b.e.p.; 3,4-e.b.)

~ 109· 28'

1-76

o

o 103· (C3,)

Xe04

4-tetrahedral

(4,4^.b.)

109*. 28'

,.4

o

I

109· 28'

Ν.

β ^ Ο o

O (Τ,)

Xe02F4

6-pseudo octahedral (2,4-e.b.;4e.p.b.)

Compound unknown ^ F

O

XeOFJ

Ν . >90·

6-pseudo octahedral (l,4-e.b.;5b.e.p.) F ^ l

^

(ion unknown)

F

F

e,, electron; p., pair; n.-b., non-bonding; 4-e., four electron. The ground state geometry is the geometry which provides the largest solid angle for the non-bonding electron pairs or multiple bonding pairs. N o n bonding electron pairs and multiple bonding pairs usually have maximum separation. • S. Reichman a n d F . Schreiner, / . Chem. Phys. 5 1 (1969) 2355. *» R. K. B o h n , K. Katada, J. V. Martinez and S. H . Bauer, in Noble-Gas Compounds ( H . H . H y m a n , ed.). University o f Chicago Press, Chicago a n d London (1963), p. 238. « K. Hedberg, S. H . Peterson, R. R. Ryan and B. Weinstock, / . Chem. Phys. 44 (1966) 1726. L. S. Bartell, R. M. Gavin, Jr., H . B. Thompson and C. L. Chemick, / . Chem. Phys. 43 (1965) 2547. " R. M. Gavin, Jr. and L. S. Bartell, / . Chem. Phys. 4 8 (1968) 2460. L. S. Bartell and R. M . Gavin, Jr., /. Chem. Phys. 4 8 (1968) 2466. R. D . Burbank and N . Bartlett, Chem. Communs. (1968) 645. « R . D . Burbank and G. R. Jones, Science 168 (1970) 248. ' N . Bartlett, F. Einstein, D . F. Stewart and J. Trotter, / . Chem. Soc, A (1967) 1190. « J. F. Martins and E. B. Wilson, Jr., / . Mol. Spectroscopy 2 6 (1968) 410. H. H. Claassen, E. L. Gasner and H . Kim, / . Chem. Phys. 4 9 (1968) 253. * J. L. Huston, Abstracts 160th A m . Chem. Soc. Meeting (1970) I N O R 149. ^ D . H. Templeton, A . Zalkin, J. D . Forrester and S. M . Williamson, / . Am. Chem. Soc. 8 5 (1963) 817. G. Gunderson, K. Hedberg and J. L. Huston, Acta Cryst. 2 5 (1969) 124.

232

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY

Shortly after the first reports of the noble-gas compounds, Gillespie^^ used the theory to predict the shape of a number of then unknown species. Those that are now known con­ form with his predictions. Some predicted and established geometries are compared in Table 8. It should be realized that the shape-predicting quantities of Rundle's three-centre fourelectron m.o. approach^o are as successful as the EPR approach for less than 6-coordination. Thus XeF^ is predicted to be a T-shaped species, the "linear" F - X e - F array being threecentre four-electron bonded (hence long, weaker X e - F bonds), the unique X e F bond being an electron pair bond (hence short). However, as we have seen, Rundle's approach fails for XeFö and IF7. Of course, it can be argued that the electron repulsion rules would be appropriate for bonds containing less than electron pairs. Thus IF7 could be represented as involving seven single-electron bonds (see Linnett, op. cit.) and XeFg as six single-electron bonds and a non-bonding electron pair. However, the similarity of the I F bonds in I F and IF5 (see section 1.2.3) should be borne in mind. 1.3.5. Covalent RadU of the Noble Gases Covalent radii for the noble gases have been given by several authors^i. 39, 42, their estimates usually being extrapolations from Pauling values^^ for the neighbouring elements or otherwise dependent on them. Sanderson^i has taken a somewhat different approach, 40 41 42 43

R. E . Rundle, Record Chem. Progress 23 (1962) 195. Bing-Man Fung, / . Phys. Chem. 69 (1965) 596. R. T . Sanderson, Inorg. Chem. 2 (1963) 660; R. T . Sanderson, / . Inorg. Nucl. Chem. 7 (1958) 228. L. Pauling, The Nature of the Chemical Bond, Cornell University Press, N e w York (1948).

TABLE 9. COVALENT R A D E OF THE NOBLE-GAS A T O M S

He

Ne

Ar

Kr

Xe

Rn

Ref.

0.4-0.6

0.70

0.94 0.95

1.09 1.11

1.30 1.30 1.31

1.4-1.5 2.12

a b c

Rn

Ref.

2.3-2.5 2.00

a d

E L E C T R O N E G A T I V r r Y C O E F H C I E N T S OF T H E

NOBLE-GASES

He

Ne

Ar

Kr

Xe

2.5-3.0 4.5

4.4 4.0

3.5 2.9

3.0 2.6

2.6 7 75

• Bing-Man Fung, / . Phys. Chem. 6 4 (1965) 596. R. J. Gillespie, Noble Gas Compounds (H. H. Hyman, ed.), University of Chicago Press, Chicago and L o n d o n (1963), p. 333. « R . T . Sanderson, Inorg. Chem. 2 (1963) 660; R. T . Sanderson, / . Inorg. Nucl. Chem. 7 (1958) 288. R. E . Rundle, / . Am. Chem. Soc. 85 (1963) 112.

ADSORPTION, ENCLATHRATON AND ENCAPSULATION OF THE NOBLE GASES

233

his values being based on his electro-negativity equalization principle and related concepts44 —his value for xenon is similar to that given by other authors. Values from the various authors are given in Table 9. 1.3.6. Electro-negativity Coefficients of the Noble Gases Most authors are agreed that the chemically reactive noble gases are of high electro­ negativity. F u n g 4 i has derived electro-negativity coefficients via a variant of the original Pauling method^^. Other estimates given by Rundle^^ were from the Mulliken formula. The various values are given in Table 9.

1.4. A D S O R P T I O N , E N C L A T H R A T I O N A N D OF THE NOBLE

ENCAPSULATION

GASES

The heavier noble-gas atoms are bigger than the lighter. The packing diameters derived for the gases in their crystal lattices and the diameters derived from viscosity measurements are in rough agreement, as may be seen from the data in Table 10, and show a smooth increase with atomic number. Of course, the size of the atom is a rough measure of the size of the outermost shell of electrons, which is the valence shell. Clearly the valence electrons of the larger atom noble gases, being further from the nucleus, are less firmly held. We have seen, earlier, that the heavier gases are active chemically. They are also the more polarizable. The static, atomic polarizability α for each of the noble-gas atoms is given in Table 10. The linear relationship of the heat of adsorption to the polarizability of the gas atoms was discovered by Chackett and Tuck (see ref. (d) of Table 10). There is also abundant evidence that the adsorption of the gases on zeolites and enclathration in water, hydroquinone, or other host cages, is primarily dependent upon the London dis­ persion energy. Thus, as discussed by Pauling^s for the case of the noble-gas hydrates, the energy of the electronic dispersion interaction between two molecules A and Β is Ir^iEA+EsY

In this equation a A and ae are the electric polarizabilities of the two molecules, Ε A and EB are their effective energies of electronic excitation, and r is the distance between their centres. The observed enthalpies of sublimation of crystals of the noble gases require that the effective excitation energy be taken as 1.57 times the first ionization energy. The heat of interaction of each of the noble gases with the water cage in the 8G-46H20 hydrates, is hsted in Table 11. The values of 6.5 and 9.3 kcal mole-i for krypton and xenon respectively are similar to the heats of adsorption for these elements on charcoal. 1.4.1. Noble-gas Clathrates Shortly after the discovery of argon, Villard prepared^ö a hydrate of the gas (1898). Hydrates of krypton, xenon and radon were prepared somewhat later47. All have the general ideal formula 8G-46H20. The gas atoms are held in cages in a pseudo-ice water lattice. 44 45 46 47

R. T. Sanderson, Chemical Periodicity, Reinhold, N e w York (1960). L. Pauling, Science 134 (1961) 15. p . Villard, Comptes Rendu 123 (1896) 377. M. V. Stackelberg and H. R. MuUer, Z. Electrochem, 58 (1954) 25.

234

1

< § ζ

o

I a*

1

I 9

I

?!

5Í?

I

m

Í i

I

.'s's's

«Γ rf «f

Ü 1^ Βχ

tu

osÜ

vo ON

I 12

ON rcn vo

I '

00 00
NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY

VO ON en en

ON

S S! 00 «o m

°

Ö

^

4>

I

ADSORPTION, ENCLATHRATION AND ENCAPSULATION OF THE NOBLE GASES

235

In 1949 Powell and Guter^s prepared a compound of argon and hydroquinone by crystallizing a solution of the latter in benzene under argon at a pressure of 20 atm. Hydro­ quinone compounds with krypton or xenon proved to be easier to prepare. These com­ pounds, like the hydrates, were shown from structural analyses by PowelH9 to be hydrogenbonded networks of the "host" species (jS-hydroquinone) containing cavities, which serve as cages for the "guest" species, i.e. noble-gas atoms, or other suitably sized species, e.g. methane. It is a usual characteristic of these cage compounds, which were given the name clathrates by Powell, that the lattice of the host shows a structural modification from that of the pure host^o. The modified host lattice contains fewer but larger cavities than the more stable, pure, host. This modification is energetically less favourable, but, of course, the energy of interaction with the guest atom (Van der Waals bonding) more than offsets that unfavourable feature. J. H. Van der Waals has used^i a statistical mechanical theory to account for the noblegas clathrates, particularly the hydroquinone compounds. From his theory. Van der Waals calculates the heat of formation, at constant volume, of the argon-hydroquinone clathrate to be 5.1 kcalg-atom-i (from solid )?-hydroquinone and gaseous argon), compared to a value of 5.4 kcal g-atom-i derivedsi from experimental findings of Evans and Richards. In this theory, the noble-gas atoms are assumed to be rotating as freely as in liquid argon. It is not essential for all suitable cavities in the clathrates to be filled to sustain the structure so long as there is sufficient interaction energy to maintain the host structural form. Accordingly, the clathrates are usually non-stoichiometricso. Apart from their theoretical interest, the clathrates are important as a means of con­ centrating and holding the heavier noble gases. The gases may be readily released either by dissolution or thermal decomposition. To illustrate52: argon amounts to ^^8.8 wt.% of the argon j8-hydroquinone compound (which has an "ideal" stoichiometry [C6H4(OH)2]3Ar). It may be preserved for several weeks in air at 1 atm pressure with an argon loss of less than 10%. When it is dissolved in ether, methanol or other solvents, or heated, the argon is Hberated. To achieve the same space concentration of the argon as in the clathrate, it would be necessary to compress pure argon to 95 atm at room temperature. A j!-hydroquinone clathrate has already proved useful53 as a carrier for radioactive »^Kr. Noble-gas hydrates. The crystal structures of the simple noble-gas hydrates were established54 by Claussen in 1951. Structural features of the clathrate hydrates generally have been reviewed by Jeffrey and McMullenss. Each noble gas, except helium and neon, forms a hydrate when mixed with water at ^0°C under a gas pressure exceeding the in­ variant decomposition pressure (see Table 11). Crystallization takes place at the phase boundaries within the water. The hydrates are usually non-stoichiometric. So-called "double hydrates" can be formed if a noble gas is mixed with another species hke acetic acid, chlorine, chloroform or carbon tetrachloride. The noble gas can also act 48 H. M. Powell and M. Guter, Nature 164 (1949) 240. 49 H. M. PoweU, / . Chem. Soc. (1948) 61. 50 Non-Stoichiometric Compounds (L. Mandelcom, ed.), Academic Press, N e w York (1964). 51 J. H . Van der Waals, Trans. Faraday Soc. 5 2 (1956) 184; J. H. V a n der Waals and J. C. Plattlenm, Clathrate solutions, in Advances in Chemical Physics (I. Prigogine, ed.). Vol. Π, Interscience, N e w York (1959). 52 H. M. Powell, / . Chem. Soc. (1950) 298, 300, 468. 53 D . J. Chleck and C. A. Ziegler, Intern. J. Appl. Radiation Isotopes 7 (1959) 141. 54 W. F. Claussen, / . Chem. Phys. 19 (1951) 259, 662, 1425. 55 G. A . Jeffrey and R. K. McMullen, in Progress in Inorganic Chemistry, Vol. 8 (F. A . Cotton, ed.), Interscience, N e w York, London and Sydney (1967), pp. 4 3 - 9 9 .

236

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY

as a so-called "help gas'* in enhancing the stability of a clathrate. Thus double hydrates of acetone with Ar, K r or Xe have been prepared from aqueous solution at — 30°C under gas pressures of 300, 30 and 1 atm respectivelysö. All noble-gas hydrates contain a common structural feature^s. This is a pentagonal dodecahedral arrangement of water molecules, shown in Fig. 3. In these H40O20 units.

o Sites for pentagonal dodecahedra • Sites for tetrakaidecahedra

Tetrakaidecahedron

Pentagonal dodecahedron XBL704-2ee6

F I G . 3. The 12A hydrate-clathrate structure." • L . Pauling, Science 134 (1961) 15.

each dodecahedron vertex is an oxygen atom site and each edge represents an O - H - O hydrogen bond. These units are linked differently in the two known noble-gas-containing clathrate hydrate structures. In the simpler structure, which has a cubic unit cell, λ ο « 12 A, the H40O20 units are linked by other water molecules to form a pseudo-body-centred array in which the two H40O20 polyhedra at the corner (0, 0, 0) and body centre ( i , i , ^) of the cubic unit cell (O^, Pnßn) are linked by six water molecules (two in each face of the cube). The polyhedra at Ϊ , i , i , are rotated through 90° with respect to those at 0, 0, 0 sites. Each water molecule is surrounded by four others, with which it forms hydrogen bonds. The O - H - O distance in the dodecahedron is 2.75 A, which is essentially the same as the value for ordinary ice, which is 2.76 A. The unit cell contains 46 water molecules in all and the water structure is approximately 12% less dense than that of ice I. The water structure contains fewer holes than ice I, but the holes are larger. The noble-gas atoms occupy the holes. The holes, or chambers, are of two types: two are defined by the dodecahedra and are smaller than the other six chambers which are defined by 24 water molecules at the corners of a tetrakaideca­ hedron (shown in Fig. 3). The tetrakaidecahedron has two hexagonal and 12 pentagonal 56 J. C. Waller, Nature 186 (1960) 429.

ADSORPTION, ENCLATHRATION AND ENCAPSULATION OF THE NOBLE GASES

237

faces and there are six tetrakaidecahedra in the unit c e l l . The hexagonal faces include those water molecules which link the dodecahedra. Pauling has discussed this hydrate structure in d e t a i l 4 6 . He points out that in the crystal 8Xe-46H20, two xenon atoms are in pentagonal dodecahedra chambers, and the 20 water molecules are at a distance of 3.85 Ä from the xenon atom. For each of the six xenon atoms in tetrakaidecahedral chambers, there are 12 water molecules at 4.03 Ä and 12 at 4.46 Ä from the xenon atom. Using the dispersion energy equation given above, and assuming ^Ή^ο « ^'xe » L 5 7 χ the first ionization potential of xenon (280 kcal mole-i), Pauhng derives the interaction energy for two molecules (Xe and H2O) to be = -aRxcRu^o/r^ in which Rx^ ( = 10.16 ml) and RH^O ( = 3.75 ml) are the mole refractions of Xe and H2O, α = 51 kcal mole-i, and r is the average X e - H 2 0 distance in Ä units (the mole refraction is ΛπΝβ times the polarizability; Ν is Avogadro's number). With inclusion of terms for more distant water molecules and other xenon atoms, his value for the interaction energy becomes - 1 0 . 3 kcal mole-i, which is in fair agreement with the experimental value of - 9.3 kcal mole-i given in Table 11. This interaction energy is large compared to the 0.16 kcal mole-i lower enthalpy of the clathrate "ice" structure compared with ice L In the other common hydrate structure, which is the one usually adopted by the "double hydrates", the cubic unit cell (a = ^17 A) contains 136 water molecules in a hydrogenbonded framework which defines 16 small chambers, with the pentagonal dodecahedron as the cage, and eight large chambers, each formed by 28 water molecules at the corners of a hexakaidecahedron. The hexakaidecahedron has four hexagonal and 12 pentagonal faces. The large chambers accommodate large molecules like chloroform and the smaller chambers, smaller molecules like xenon and krypton. Thus chloroform and xenon yield a hydrate of ideal composition CHCl3-2Xe- I7H2O, i.e., 8CHCl3- 16Xe- I36H2O. Pauling-^e and Miller^^, independently, have proposed that clathrate hydrate formation may be important in anesthesia. Certainly xenon is an effective anesthetic, and that role must involve London dispersion interactions and not chemical bonding. The recently reported^s binding of xenon to myoglobin is also, apparently, another example of enclathration. It may also be noted here that the large negative entropy of solution of argon in w a t e r 5 9 may be due to the formation of an orientated water sheath about the dissolved atoms. Addition of small quantities of dioxane leads to a rapid increase in the entropy of solution, presumably because of the destruction of the argon water sheath. ß'Hydroquinone and phenol clathrates. Argon, krypton and xenon )S-hydroquinone clathrates have been prepared and characterized by P o w e l l 5 3 . Clathrates involving other phenolic materials have also been described, principally those of phenol. It is of interest that Nikitin^o used the latter host-precursor in 1939 to attempt to prepare a phenol-radon clathrate. He was able to produce a mixed phenol clathrate of radon and hydrogen sulphide, but not a pure radon clathrate. In 1940 he prepared a xenon-phenol clathrate and, inci­ dentally, was the first to infer from this study that this and related compounds involved an inclusion type of association, von Stackelberg and his associates^i have elucidated the structure of this and a number of other phenol clathrates, several of which were first prepared by them. 57 S. L. Miller, Proc. Natn. Acad. Sei. 47 (1961) 1515. 58 Β. Schoenbom, / . Molecular Biology 45 (1969) 297. 59 A. Ben-Nairn and J. Moran, Trans. Faraday Soc. 61 (1965) 861. 60 B. A. Nikitin, Compt. Rend. Acad. Sei. USSR 29 (1940) 571. 61 M. V. Stackelberg, Ree. Trav. Chim. 75 (1956) 902; M. V. Stackelberg, A. Hoverath and Ch. Scheringer Ζ. Electroehem. 62 (1958) 123.

238

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY

The ideal composition of the j5-hydroquinone clathrates is G · 3j5-C6H4(OH)2. Typical compositions are represented in Table 12, where G is Ar, Kr and Xe. Structural analyses show that the noble-gas atoms are located in approximately spherical cages of hydrogen bonded hydroquinone molecules^i. The free diameter of the cage hole is ' ^ 4 . 2 Ä . The

TABLE 1 2 . NOBLE-GAS )9-HYDROQUINONE A N D PHENOL CLATHRATES

)9-Hydroquinone clathrates * ideal formula: G-3C6H4(OH)2

Phenol clathrates ** formula: / w G - n C e H s O H

C6H4(OH)2:G N o b l e gas

(ratio)

(kcal m o l e - i )

m

(kcal mole~i)

Ar Kr Xe

3:0.8 3:0.74 3:0.88

-6.0 -6.3

2.92 2.14 4.0

-9.85 -8.97 -8.8

¿\H% represents the enthalpy change for the process: host lattice 0? f o r m ) ( c ) + G(g)-» clathrate compound. « H . M. Powell, / . Chem. Soc. (1950) 298, 300 and 4 6 8 ; W . C. Child, Jr., Quart. Revs. (1964) 321. P. H. Lahr and H. L. Williams, / . Phys. Chem. 63 (1959) 1432.

j5-hydroquinone structure is 82 cal mole"i less favourable in free energy than the nonclathrate α form52. Although the hydroquinone clathrates are usually derived from solu­ tions, this is not essential. The argon clathrate has been prepared by subjecting solid a-hydroquinone to argon pressures of 300 atm and above, von Stackelberg63 has shown that the phenol clathrates contain 12 phenol molecules per unit cell. The argon, krypton and xenon phenol clathrates have also been made directly from the gases and molten phenol at > 4 0 X . Mössbauer studies of noble-gas clathrates. The value of Mössbauer studies in the further elucidation of the noble-gas clathrates has been briefly reviewed by Herber<52. Evidently the recoil-free fraction of the »^Kr-hydroquinone clathrate absorber is essentially constant between room temperature and '^ISO^K. The recoil-free fraction increases very rapidly below this temperature. The results for Τ > 150°K are what one would expect for a square well potential, but the low temperature results are open to several interpretations. The latter results are more typical of ordinary harmonic binding and suggest either the onset of phenomena normally associated with bulk materials (such as condensation and liquefac­ tion) or "sticky" collisions with the cavity wall. The work on both xenon and krypton hydroquinone clathrates shows that the noble-gas atoms are essentially in spherical sites— single resonances only are observed. The y-ray transitions in the xenon halides have excess energy relative to atomic xenon in the j5-hydroquinone clathrate, whereas the transitions in the tetroxide and perxenates have less energy. The isomer shifts are displayed as an energy level diagram in Fig. 4. The excess energy of the y-rays in the halides can be explained on the assumption that « R. H. Herber, as in ref. 55, pp. 1-42.

ADSORPTION, ENCLATHRATION AND ENCAPSULATION OF THE NOBLE GASES

239

these compounds involve xenon ρ orbitals in bonding. These findings have been accounted for by Perlow and Perlow (see Fig. 4) as follows: the transfer of 5p electrons from xenon increases the effective field acting on the xenon s electrons (55 and inner electrons)—hence, (the isomer shift) > 0. On the other hand, the oxygen-containing compounds, having XeF4

+0-4010-04

(XeCU)

+0-2510-08

{XeClj}

+01710-08

XcFj

+0-1010-12

Clathrate

Í

ε

^ 01910-02 _

0-221 002 0-221 002

F I G . 4. X e n o n isomer shifts ·. The energies shown are to be added to the transition energy for the neutral atomic species. <'>As given by G. J. Perlow and M. R. Perlow, / . Chem. Phys. 48 (1968) 958.

octahedral or tetrahedral symmetry, have appreciable 5^ admixture into their bonding orbitals, and the direct effect of the transfer of 5^ charge reduces the central charge density more than the shielding effect of 5/7 transfer increases it. Consequently, Δ^· < 0. Imphcit in this explanation is the assumption that the enclathrated xenon is essentially an un­ perturbed xenon atom. 1.4.2. Noble-gas Encapsulation in Zeolites The encapsulation of noble gases in zeolites^s is quite similar to the trapping of the gases in clathrates, but no simple stoichiometry appears to exist. Encapsulation does not appear to change the structure of the adsorbing zeolite. The process consists of forcing gas molecules into the pores of a suitable heat-stable material at elevated temperatures and pressures, and then trapping them by cooling the material in the presence of the gas, maintaining the high pressure until the cooling is complete. The most suitable host materials have so far proved to be the synthetic zeolites. These are dehydrated alumino-sihcates,

T . D . C.l.C. Vol. 1—I

240

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY

which are interlaced with regularly spaced channels which are of molecular dimensions. Pore and channel size in the synthetic zeolites can be controlled by varying the cation type of the zeolite. With this control, atoms or molecules above a given size can be excluded— hence the term "molecular sieves". Presumably in the "encapsulation" process, the increased vibrational motion of the atoms in the zeolite lattice lowers the potential energy barrier and the increased kinetic energy of the gaseous atom aids in overcoming this barrier. In any event, the channels fill up with gas atoms. With suitable choice of zeolite, the weight of gas trapped per unit weight of zeolite is greater than for any known clathrate, e.g. up to 20 wt. % argon has been encapsulated. Table 13 gives some representative findings.

TABLE 1 3 . E N C A P S U L A T O N OF A R G O N A N D K R Y P T O N I N SYNTHETIC ZEOLFFE T Y P E A *

Cationic composition of the zeolite ( K : N a ratio)

Initial encapsulation pressure (atm)

Ar

20:80 40:60 60:40

Kr

100% N a 60:40 92:8

Gas

Quantity encapsulated (g gas per 100 g zeolite) A t start

After 30 days

2500 2500 2500

16.7 19.4 19.0

< 3 0 % loss < 3 0 % loss < 3 0 % loss

4300 4300 4300

32.4 34.8 21.6

32.8t 30.2 20.1

t The apparent increase most probably arises from analytical inaccuracy. * Work o f L. H. Schaffer and W. J. Sesuy, quoted in Argon, Helium and the Rare Gases (G. A . Cook, ed.), Interscience, Vol. I, N e w York (1961), p. 230.

Encapsulation is not yet very effective for neon. The most effective zeolite pore diameter (at room temperature) for argon and krypton atoms are about 3.82 and 4.0 Ä respectively. There are indications that the «^Kr encapsulates may be as stable as the clathrates (see section 1.4.1).

1.5. G A S E O U S N O B L E - G A S

CATIONIC AND EXCITED ATOM

SPECIES

There are numerous reports of cationic species involving a noble-gas atom and some other atom or molecule. Two recent reviews have deak with them63» 64. These transient species are usually produced by high-energy radiation or bombardment with high-energy particles (e.g. electrons), and have invariably been detected by mass spectroscopy. Since the noble-gas cations are isoelectronic with the neighbouring halogen atoms, they are anticipated to be more electro-negative than them (because of their higher nuclear charge). It is to be expected, then, that the noble-gas cations should form bound « B. Brockelhurst, Quart. Revs. 22 (1968) 147. 6 4 G. V o n Bunan, Fortschr. chem. Forsch. 5 (2) (1965) 374.

241

GASEOUS NOBLE-GAS CATIONIC AND EXCITED ATOM SPECIES

species with other atoms and ligands. Now the electron affinity of these cationic species will, in many cases, approach the electron affinity of the noble-gas cation itself. This makes it unlikely that stable salts of cations containing helium and neon for use at ordinary temperatures and pressures will be isolated. There is more hope for Rn, Xe, K r and Ar species. As mentioned earlier (see section 1.2.3), certain argon cation species, e.g. ArF+, may be isolable as salts. However, KrF+ salts are more likely (see section 2.2.1) and XeF+ salts are known (see section 3.2.1). 1.5.1. Hydride Cations The simplest cation derivatives are the hydrides. They may be generated in ion-molecule reactions: G+ + H 2 - > G H + + H G + HJ->GH+ + H

The first reaction has been observed for Ar, Kr and Xe. The estimated proton affinities, i.e. A/^(G(g)-f H^^j -> (GH)+j) for the diatomic hydride cations are hsted in Table 14. TABLE 14. P R O T O N AFFINITIES*, G p ,

Gp(eV) / ( G ) (eV)

A N D IONIZATION POTENTIALS OF THE N O B L E G A S E S

He

Ne

Ar

Kr

Xe

1.8 24.586

2.2 21.563

3.0 15.759

>4 13.999

>6 12.129

" G. von Bunan, Fortschr. chem. Forsch. 5 (2) (1965) 374.

As may be seen from the thermochemical cycle in Fig. 5, the electron affinity of GH+^ is equal to the electron affinity of the proton ( - 1 3 . 6 e V ) minus the proton affinity of G. This is so only if A / / ( G H ( g ) G(g) + H(g)) is zero. Thus the electron affinity of ArH+ should be « - 1 0 . 6 eV.

AHdiss^OC?) FIG. 5. Relationship of ionization potential, proton affinity and hydride cation electron affinity.

Note that £ ( X e H + ) « - 7 . 6 eV. This is encouraging. Xenonium salts may be preparable. 1.5.2. Excited Atom Reactions Energy levels and radiative lifetimes of the noble gases are given in Table 15. The excitation energies of even the lowest excited states of the noble gases^are greater than the

242

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY TABLE 15. E N E R G Y LEVELS OF THE N O B L E GASES A N D THEIR R A D I A Ή V E LIFETIMES*

Metastables

He

Ne Ar Kr Xe

2 35 19.81 —

215 20.61 —

21P 21.21 0.56

16.61 — 11.55 — 9.91 — 8.31 —

16.67 21 11.62 8.1 10.03 3.4 8.43 3.8

16.71 — 11.72 — 10.56 — 9.44 —

3Po

Ions

— — 16.84 1.5 11.82 1.8 10.64 3.3 9.57 3.2

25 24.58 —

2P3/2

21.56 — 15.75 — 14.00 — 12,13 —

— —

eV nsec

2Pl/2

21.66 — 15.93

—

14.66 — 13.43

eV nsec eV nsec eV nsec eV nsec

" From B. Brocklehurst, Quart. Revs. 2 2 (1968) 147.

ionization energies of many molecules. Accordingly, excited noble-gas atoms will often bring about associative ionization63: G* + X - > G X + + e

The noble gases can therefore act as photosensitizers in much the same way as mercury. Thus krypton sensitized dissociation of N2 has been established. Excited xenon atoms generate the short-lived XeO molecule in the reaction63 Xe* + 0 2 - > X e O * + 0

2. K R Y P T O N

CHEMISTRY

So far, the chemistry of krypton has been limited to that of krypton difluoride and derivatives. All effOrts to conñrm the synthesis of a tetrafluoride^s or higher fluoride krypton have failed. These attempts have included the subjection of KrF2/F2 mixtures high energy irradiation at low temperatures^ö. Nor has any ñrm evidence appeared support the existence of oxides, oxysalts^^ or chlorides.

2.L

KRYPTON(I)

its of to to

C O M P O U N D S

N o Kr+ salts have been claimed, and the only established krypton(I) compound is the low-temperature species, krypton-monofluoride radical. 2.1.1. The Krypton Monofluoride Radical Although krypton monofluoride has only been generated in exceedingly small concentration's, in krypton difluoride crystals subjected to y-radiation (1.3 MeV), it is, 65 66 67 68

A. V. Grosse, A. D . Kirshenbaum, A. G. Streng and L . V. Streng, Science 1 3 9 (1963) 1047. D . R. MacKenzie and I. Fajer, Inorg. Chem. 5 (1966) 699. A. G. Streng and A . V. Grosse, Science 143 (1964) 242. W. E . Falconer. J. R. Morton and A . G. Streng, / . Chem. Phys. 41 (1964) 902.

KRYPTON(II) COMPOUNDS

243

nevertheless, of interest because of the information the unpaired-electron probe yields on the nature of K r - F bonding. The radicals, which colour the host crystals violet, persist indefinitely at - 1 9 6 ° but disappear on warming to - 1 5 3 ° . The K r F radical has one more electron than bromine monofluoride. Since this electron must be in an antibonding σ orbital, the K r - F bond can amount only to a one-electron bond. The bond strength may be comparable to the K r - F bond in KrF2 (mean thermochemical bond energy69 = 12 kcal mole-i) but certainly weaker and presumably, at most, only half the strength of the bond in the Kr-F+ ion (isoelectronic with BrF, for which the thermochemical bond energy = 60 kcal mole-i). In an e.s.r. study of the radical, hyperfine interaction of the electron spin with the i^F nucleus was observed, which was sufficient to show that the fluorine component of the antibonding σ orbital, containing the unpaired electron, is chiefly Έΐρσ (2ρσ population, Cl 2p = 0.61, whereas, 2s = 0.04). This is similar to the situation in XeF (see section 3.1.1). Furthermore, the findings indicated a lower fluorine character in the bonding orbitals of K r F , compared with XeF, which is in harmony with the greater electronegativity of krypton, than xenon.

2.2. K R Y P T O N ( I I ) C O M P O U N D S

2.2.1. Krypton Difluoride Synthesis. Krypton difluoride was first characterized by Turner and Pimentero who prepared it by the ultraviolet photolysis of fluorine, suspended in a sohd mixture of argon and krypton at 20°K. Although the first krypton compound to be prepared was described^s by its discoverers, von Grosse and his coworkers, as the tetrafluoride, the properties ascribed to this material have been shown to be those of the difluoride. Laboratory preparation. Because the difluoride is thermodynamically unstable69, all successful syntheses have used krypton and fluorine mixtures held at low temperatures (usually - 1 9 3 ° ) , and have involved either irradiation with y-rays, 1.5 MeV electrons^i, ultraviolet light^o, or 10 MeV protons66, or electric discharge of the gaseous mixturéis. 72. The last procedure is the simplest to reproduce and has been employed in several laboratories to make gram quantities of the fluoride. The essential apparatus is shown in Fig. 6. The cleanest synthesis, which is also moderately eflicient, involves the irradiation, at tempera­ tures of less than - 6 0 ° and approaching - 1 9 6 ° , of the gaseous krypton and fluorine mixtures with 10 MeV protons^ö. 69. Krypton difluoride is most conveniently identified by its vibrational spectra and par­ ticularly by its Raman active vi mode at 449 c m - i . The Raman spectrum is easiest t o obtain since sapphire, which is chemically resistant to K r F i , can be used as the container material. The very strong infrared band at 588 cm-i also serves for ready identification— silver chloride windows should be used. The instability of K r F i requires that the identifica­ tion be carried out quickly and at the lowest possible temperature. Thermodynamic properties. Krypton difluoride is colourless both in the solid and vapour phases. It decomposes spontaneously at temperatures well below room temperature, the 69 70 71 72

S. R. Gunn, J. Phys. Chem. 71 (1967) 2934. J. J. Turner and G. C. Pimentel, Science 140 (1963) 974. D . R, MacKenzie, Science 141 (1963) 1171. F. Schreiner, J. G. Malm and J. C. Hindman, / . Am. Chem. Soc. 87 (1965) 25.

244

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY To vacuum line

Direction of gas flow F I G . 6. Diagram o f apparatus used for the preparation o f krypton difluoride. A: Polychlorotrifluoroethylene container for the collection and storage o f the compound, attached t o the glass apparatus by compression fittings. Β and C: U-tubes o f Pyrex glass with break-seals. D: Electrical discharge reaction vessel made o f Pyrex glass (diameter 6 0 m m ; height o f wide portion 200 mm). T w o copper discs o f 2 0 m m diameter and 5 m m thickness, spaced 75 m m apart, serve as electrodes. The leads to the electrodes are silver-soldered into Kovar to glass seals. E: Valve manifold t o convert push-pull operation o f magnetic piston pump into unidirectional gas circulation as indicated. Each individual valve consisted o f a 10 m m glass tube ground flat at the end, protruding into a wider tube and closed with a thin square piece o f glass held in place by gravity. Application o f a small pressure head from below (0.1 m m ) permits gas t o flow upward. D o w n w a r d flow is inhibited b y the closure o f the ground end o f the glass tube by the square piece o f glass. Arrangement o f four valves in the way indicated in the figure permits use o f the pumping action o f each half stroke o f the piston. F: Magnetic piston pump after Brenschede*. G: Piston o f pump suspended from stainless steel spring. H: Solenoid. K l , K2, V3: Monel valves. With the reaction in progress valve 1 is kept closed while valves 2 and 3 are open. During the puriñcation and sublimation o f the product, ñrst t o tube C and then into tube 5 , valves 2 and 3 are closed t o separate the pump from the rest o f the system, a n d valve 1 is open t o establish a connection t o the vacuum line. Since the proton beam passes effectively through a 0.03 c m thick aluminium window, the synthesis o f this most oxidatively reactive o f all fluorides c a n be carried o u t in a n all aluminium vessel. (Aluminium is highly resistant t o oxidative fluorination.) A 1 hr irradiation at 5 μΑ yields ca. 1 g o f KTFZ. T h e G value for KrFa formation lies in the range 1 t o 1.5 m o l per 100 eV. 66 * W. Brenschede, Ζ . Physik.

Chem. (Leipzig)

A178 (1936-7) 74.

d e c o m p o s i t i o n rate for the v a p o u r at r o o m temperature being

1 0 % h r - i . The d e c o m p o s i ­

t i o n rate is substantially lower for the solid74 a n d evidently negligible at dry ice temperat u r e 7 2 . The s p o n t a n e o u s dissociation h a s prevented the accurate determination o f a n u m b e r o f t h e physical properties o f K r F i , b u t there is general agreement72.74 that its v a p o u r pressure is ca. 30 m m H g at 0"". The enthalpy o f vaporization AH^ub = 9.9 kcal m o l e " i has been derived from vapour-pressure measurements over the limited temperature range

- 1 5 . 5 t o 15° 74. 73 H . H . Claassen, G . L. G o o d m a n , J. G. Malm and F . Schreiner, / . Chem. Phys. 4 2 (1965) 1229. 74 S. R. Gunn, / . Am. Chem. Soc. 8 8 (1966) 5924.

245

KRYPTON(II) COMPOUNDS

Measurement of the heat of dissociation69, at 93°, of a gaseous sample of the difluoride (which is rapidly decomposed at this temperature) has given a standard heat of formation Δ / / ^ (KrFa^gj) = 14.4+0.8 kcal mole-i. The calorimetric measurements are supported by mass spectrometric, appearance potential data75. Although there is some ambiguity of interpretation, it is probable that the observed appearance potential A (Kr-*-, KrFi) = 13.21 ±0.25 eV, is appropriate for the process KrF2(e)+e -> Kr+ + F2(g)-l-2e. If so, since / ( K r ) = 14.00eV, it follows that Δ//(KrF2u, - Kr(,) + F2(g)) = 0.79±0.25 eV (i.e. 18±5kcalmole-i). The thermochemical average bond energy derived from the calorimetric data is 12 kcal mole-i. This is the lowest average bond energy of any known fluoride. Indeed, the atomization of KrF2 involves a lower enthalpy than the atomization of molecular fluorine. Krypton difluoride should, of all oxidative fluorides, be closest in activity to atomic fluorine. Although the KrF+ ion is not established, there is reason to beheve that it occurs in the complex KrF2, 2SbF5 (see section 2.2.2). As can be seen from Fig. 7, the electron afiinity of KrF+ equals the electron affinity of Kr+ ( - / (Kr, g) = 14.0 eV), less the difference in bond energy between KrF+ and K r F . It is reasonable to suppose that the BE (KrF+) is - 2 x B E (KrF). A recent value for the appearance potential76« for KrF+, A (KrF+, KrF2) = 13.39 eV. Since this corresponds to the process KrF2(g)-> KrF+ + F(g) + e, then from Fig. 7, A (KrF+, KrF2) = AH (KrF2(g, Kr(g)+2F(g))+/(Kr)-BE (KrF+). Therefore, the bond energy for the cation = - (13.39-14.00) eV-f Δ ^ . ι (KrF2) « 37 kcal mole-i. This is approximately three times the mean thermochemical t o n d energy of KrF2.

KrF2(g) KrF(g>+F(g)

A(iCrF-^,KrF2) = 13-39ev

l^í^lIísL

Ε (KrF^^)

KrF-^(g, + F(g)+e BE(KrF^;¡) KΓ^β) + 2F5^e

I (Kr) - 1400 eV F I G . 7. The electron affinity and bond energy o f KrF+. /(KrF<,)) =

/(Kr)-BE(KrFJ,)+BE(KrF,„)

B E (KrF + ) = / (Kr)-y4(KrF+, KrF2)4- AH.i

(KrFz)

Structural features. Like its relative, xenon difluoride, KrF2 is a symmetrical linear molecule (D^h) '^^' The K r - F interatomic distance has been determined for the gaseous species by electron diffraction^^ (1.889 ±0.010 A) and rotational infrared spectroscopy77 75 p . A. Sessa and H. A. McGee, / . Phys. Chem. 73 (1969) 2078. 76 "W, Harshbarger, R. K. Bohn and S. H. Bauer, / . Am. Chem. Soc. 89 (1967) 6466. 76" J. Berkowitz, Argonne National Laboratory, personal communication. 77 C. Murchison, S. Reichman, D . Anderson, J. Overend and F. Schreiner, / . Am. Chem. Soc. 90(1968) 5690.

246

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY

(1.875+0.002 Ä). The infrared and Raman data are in agreement with the D^H molecular symmetry and the force constants fr = 2.46,/rr = —0.20, and / a = 0.21 m d y n - i , o b t a i n e d 7 3 from the computed potential function 2 F = / r ( A r i + A^ + ^frrAnArz-^r^faAo^^, where Δ / Ί and Afi are changes in the bond distances and Δ α is the change in bond angle, have been interpreted^s (particularly the negative sign of fm the bond-bond interaction constant) in terms of a considerable weight of a no-bond F K r F structure in resonance admixture with (F-Kr)+F-, F-(Kr-F)+. The krypton difluoride molecule is related to the bromine trifluoride molecule. As may be seen from the molecular characteristics represented in Fig. 8, the K r - F internuclear distance in KrFa is slightly greater than the B r - F distance -Br J ü ^ F

Kr

\ - X e - ^ F

BrJi^F Force constants (mdyn A~i,

Frequencies ( c m - i )

Molecule

vi

V2

V3

fr

/rr

Λ

KrF2 (F)BrF2t XeF2

449 531 515

232.6

588 613 558

2.46 3.00 2.84

-0.20 0.15 0.13

0.21

213.2

0.20

F I G . 8. Comparison of KrF2 with BrF, BrFj and XeF2. t Only the approximately linear BrF2 part of the BrFs molecule is relevant for this comparison.

in the approximately linear part of the BrFa molecule. The unique B r - F bond distance in BrFa is comparable to the internuclear separation in the bromine monofluoride molecule. The bonding associated with two longer bonds of BrFa is presumably similar to that involved in K r F 2 . If four-electron three-centre ρσ m.o. bonding is assumed, then each fluorine ligand is, in effect, bonded to the central atom by one electron. The shorter, stronger bond in bromine monofluoride is consistent with electron-pair bonding. Presumably the bond length in KrF+ will be similar to that in BrF and it may, like XeF+ (see section 3.2.6), be slightly shorter than its halogen relative. Although crystalline krypton difluoride is not isostructural with xenon d i f l u o r i d e 7 9 it is, like the latter, essentially a molecular assembly. The structure is not known with high precision, but it is known that the molecules are oriented such that each fluorine ligand of one molecule is oriented towards a krypton atom of its nearest neighbour. The orientation and alignment of the molecules is consistent with an appreciable bond polarity. The effective volume of one molecule in crystalline KrFa which is 37.5 A3, is similar to the XeFz value of 39.2 A3. 78 C. A. Coulson, / . Chem. Phys. 4 4 (1966) 468. 79 S. Siegel and E. Gebert, / . Am. Chem. Soc. 86 (1964) 3896.

KRYPTON(II) COMPOUNDS

247

Bond polarity and bond type. Both i^p n.m.r. studies of KrF2 solutions in anhydrous hydrogen fluoride72 and Mössbauer studies^o on the pure sohd have yielded data which have been interpreted in terms of the charge distribution in the linear molecule. In both studies close relationship of KrF2 to the much more extensively studied XeF2 molecule was assumed. The i^F chemical shielding value σρ relative to F2, for KrF2 (σρ 370 χ 10"6) is much less than for XeF2 (σρ = 629 χ 10-6). The interpretation of these findings was based on the assumption that the bonding in KrF2, as in XeF2, is of the three-centre, fourelectron ρσ m.o. type, and that the shielding arises from the paramagnetic term. On this basis, it has been argued that ^ F ( K r F 2 ) = - 0 . 4 5 e , whereas for XeF2, qr = - 0 . 7 3 ^ . The lower polarity of the K r - F bond, relative to the X e - F bond, is readily rationalized in terms of the higher electro-negativity of krypton than that of xenon (see section 1.3.5). It is of interest that the i^F n.m.r. study of KrF2 solutions in H F showed that there is no ñuorine exchange with the solvent, even at 25°, in contrast with XeF2 (see section 3.2.1). It is to be expected that ionization to form KrF+ is less favourable than for XeF+ formation. The KrF2 charge distribution derived from the Mössbauet findings^o is similar to that from the n.m.r. data. Since the experimentally derived quadrupole interaction energy for KrF2, e^Q (where e is the charge on the electron, Q the quadrupole moment and eq the field gradient) = 9 6 0 + 3 0 mHz, is a consequence of an electric field gradient at the position of the krypton nucleus, there must be a non-spherical electron distribution about that nucleus. This is assumed to arise from the sharing of some krypton 4p electrons with the fluorine atoms. On the assumption of that only ρσ orbitals are involved in the bonding, the interaction energy (for one electron transferred from krypton) e^Q (KrF2) = +f^^ß<''~^>i i-e. the field gradient due to one Αρσ electron is eq = + f K ' ' " ^ > . However, atomic beam work had shown that for the state of K r (i.e. for a Ap^Ss configuration, where one electron has been removed from a ρ orbital) i e 2 ß < r - 3 > = 452.2 mHz. If is approximately the same for KrF2 and Kr, the interaction energy for the one ρ electron deficiency for the two cases should be in the ratio 2 : 1 ; they are approximately so (960:452.2). This indicates, supposing the assumption to tíe valid, that the electron transfer from the krypton atom to the fluorine ligands would be ^\e. The authors of the Mössbauer study have pointed out that this finding can be simply rationalized (in valence bond terms) on the assumption that KrF2 is a resonance hybrid of F - K r + F - and F K r + - F . They point out that if the electro-negativity of Kr+ and F are similar this would result in equal sharing of the bond electron pair in KrF+. Consequently, the krypton atom in KrF2 would have a net deficiency of one Αρσ electron. An electron transfer of \e from the K r atom to the ligands, however, seems rather high, particulariy since a charge distribution of that size has been well substantiated for XeF2 (see section 3.2.1), and xenon is less electro-negative than krypton. Furthermore, Coulson's interpretation's of the peculiar nature of the b o n d bond interaction force constant requires considerable weight of the no-bond structure F K r F in the resonance admixture with F(Kr-F)+ and ( F - K r ) + F - , which requirement, of course, reduces the Kr charge to less than + 1 . The form F-Kr^+F* presumably does not make significant contributions to the bonding. It may be that inclusion of outer orbital character (e.g. Ad) of the krypton atom in the bonding model used to interpret the Mössbauer findings would have yielded a lower charge distribution. The physical properties of krypton difluoride are summarized in Table 16. The chemistry of krypton difluoride. As the thermodynamic instability of the compound 80 S. L. Ruby and H. Selig, Phys, Rev, 147 (1966) 348.

248

N O B L E - G A S C H E M I S T R Y . * N E I L B A R T L E T T A N D F. O . S L A D K Y TABLE 16. PHYSICAL PROPERTIES OF K R Y P T O N DIFLUORIDE

Thermodynamic

*

Ai/sub (kcal m o l e - i ) Vapour pressure ( m m H g ; r ° C ) Δ ^ } , 298.15° (g) (kcal m o l e - i ) (KrF2(g) -> Kr„) + 2F(,)) (kcal m o l e - i ) Mean thermochemical bond energy (kcal m o l e - i )

-9.9 1 0 ± 1 ( - 1 5 . 5 ° ) ; 2 9 ± 2 (0°), 7 3 ± 3 (15.0°) 14.4±0.8 23.4 11.7

Solubility Anhydrous H F dissolves KrFa to

16 moles per 1000 g H F at 20°

i^F n.m.r. data * ( N o F exchange for H F solutions). Chemical shielding values σ%\ relative to F2, σψ^ = 0 : 374 χ 1 0 - 6 (4.6 moles per 1000 g H F ) , 362 X 1 0 - 6 (16,4 moles per 1000 g H F ) at 0° Infrared and Raman data ^ Infrared bands ( c m - i ) R a m a n bands ( c m - i )

232.65

580,596 vs

1032 m

449 (vapour) 462.3 (solid)

Assignments Fundamentals (cm-1) Force constants (mdyn A - i ) Molecular

V2 232.6 / „ 2.46

Vl 449 / „ 0.20

V3 588 / a , 0.21

V1 + V3

structure

jDooh symmetry, K r - F internuclear distance (A) Crystallographic

1.889 ± 0 . 0 1

data *

Tetragonal unit cell a = 6.533; c = 5.831 A; Κ = 248.9 A3; Ζ = 4 ; D d c , = 3.24 g c m - 3 Molecular

energetics ^

Appearance potentials

Possible process A (Kr+, KrF2) = 13.21 ± 0 . 2 5 eV (KrF2 + e - > K r + + F 2 + 2 ^ ) A (KrF+, KrF2) = 13.71 ± 0 . 2 0 e V ( K r F 2 + e - > K r F + + F + 2 e )

• S. R. Gunn, / . Am. Chem. Soc. 88 (1966) 5924; / . Phys. Chem. 71 (1967) 2934. F. Schreiner, J. G. Malm and J. C. Hindman, / . Am. Chem. Soc. 87 (1965) 25. H. H. Claassen, G. L. G o o d m a n , J. G. Malm and F. Schreiner, / . Chem. Phys. 4 2 (1965) 1229. W. Harshbarger, R. K. B o h n and S. H. Bauer, J. Am. Chem. Soc. 89 (1967) 6466. « S. Siegel and E. Gebert, / . Am. Chem. Soc. 86 (1964) 3896 ' P. A . Sessa and H. A . McGee, / . Phys. Chem. 73 (1969) 2078.

towards dissociation suggests, it is a powerful oxidative fluorinator. In keeping with an anticipated fluorinating ability approaching that of atomic fluorine, it oxidizes chloride (of silver chloride infrared cell windows) to chlorine trifluoride and chlorine p e n t a f l u o r i d e 7 3 . Its interaction with water generates krypton and o x y g e n 7 2 : KrF2+H20

Kr+0.5O2+2HF

Although clearly an unusual fluorinator, it has failed to oxidize xenon trioxide to the jtrioxide difluoride (see section 3.5.2) or indeed to any other xenon(VIII) oxyfluoride or fluoride. Furthermore, it has not proved possible to prepare XeFg (see section 3.5.1) by bringing XeFg into interaction with it.

XENON CHEMISTRY

249

There have been claims^' that the hydrolysis, by ice, of krypton difluoride (incorrectly as K r F 4 ) at - 3 0 ° to - 6 0 ° yields 2 - 3 % of an acid, and that hydrolysis by 0.35 Ν barium hydroxide at 0-50° results in a yield of approximately 7 % of the barium salt. These claims have not been substantiated. The difficulty of preparing quantities of KrFa and problems of handling have resulted in meagre study. Undoubtedly this scant attention will be remedied when the remarkable potential of the compound as an oxidative reagent is more generally appreciated. It is important to note the greater stability of the complex of KrF2 with SbFs (see section 2.2.2). Because of this, the complex may prove to be a more useful reagent than the parent fluoride. identified72

2.2.2. Krypton Difluoride Complexes The only established derivative of krypton difluoride is the briefly described complexSi with antimony pentafluoride, KrF2, 2SbF5. There are also indications that an arsenic pentafluoride complex forms at —78°, but this compound dissociates readily at low tempera­ tures. The antimony compound was formed by treating KrF2 with SbFs in glass or Kel-F containers. The components interact completely at —20°. The compound dissolves in excess antimony pentafluoride, which can be removed at 25°, to leave a colourless solid, KrF2, 2SbF5, m.p. - 5 0 ° (decomp. Kr + F2+SbF5). The solution in SbFs decomposes slowly at 25° although the decomposition of the sohd at this temperature is very slow. Aqueous hydrolysis (either basic or slightly acid) liberates krypton, oxygen and fluorine monoxide. Although the only structural information available on this compound is an infrared spectrum with strong bands at 813 and 600-700 c m - i , it is quite probable that it is the salt KrF+[Sb2Fii]-, analogous to XeF+[Sb2Fii]- (see section 3.2.6). Unfortunately, the stretching frequency anticipated for the cation (approximately that of the BrF molecule^s, i.e. 672 cm-i) hes in the region of S b - F stretch. It will probably be necessary to solve the crystal structure to confirm the salt formation. The salt formulation accounts for the observa­ tion that the thermal stability of the complex is greater than that of isolated krypton difluoride. Note that the bond in the Kr-F+ ion (section 2.2.2) appears to be considerably stronger (TBE - 30 kcal mole-i, see above) than in the molecule (TBE 12 kcal mole-i).

3. X E N O N

CHEMISTRY

The chemical behaviour of xenon is appropriate for an element of Group VIII of the Periodic Table. The normal oxidation states are even numbered and range from + 2 to + 8 . The + 8 oxidation state is known only in the oxide Xe04, the perxenates XeOg", and the trioxide difluoride Xe03F2. The octafluoride is unknown. This pattern resembles that of osmium, where the octafluoride is unknown, but the tetroxide OSO4 and trioxide difluoride OSO3F2 are well-estabhshed compounds. Octafluorides would surely be ligand "crowded" molecules, and it may well be that the lack of success in attempts to make these compounds is associated with a large kinetic barrier (see sections 1.2.3 and 3.5.1). So far the compounds of xenon all involve highly electro-negative ligands (e.g. F, O, -OSO2F, -OTeFs). The fluorides are readily preparable from the elements and are thermodynamically stable, and the other known compounds (e.g. oxides, oxyfluorides and 81 H. Selig and R. D . Peacock, / . Am. Chem. Soc. 86 (1964) 3895.

250

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY

oxysalts) have usually been derived from them. Xenon dichloride, which can be preserved only at low temperatures, has been made from the elements by trapping the products from a glow discharge at 20°K, but the only other established synthetic route to the xenon halides (see section 3.2.2) has involved the β decay of corresponding 1291 anions (thus: 129ΙΒΓ2 λ i29XeBr2). 3.1.

XENON(I)

3.1.1. Xenon(I) Fluoride, the XeF Radical The radical XeF- has been detected by Morton and Falconer82» 83 in an e.s.r. study of a single crystal of XeF4 y-irradiated at 77 °K. It is probable that the blue colour of the ir­ radiated XeF4 crystal was due to the XeF- radical since the colour and the e.s.r. spectrum were observed83 to fade simultaneously and rapidly at 140°K. It has been inferred, from kinetic data, that the XeF radical is an intermediate in water oxidation84 and N O and NO2 oxidations^ by XeFa. It is also probable that the decomposi­ tion of FXeOSOzF involves an XeF- intermediate (FXeOSOzF-> X e F - + SOaF-; 2XeFXe + XeFz; 2SO3F SzOeFi) since the overall chemical changese is 2FXeS03F Xe + XeF2 + S206F2. Similarly, the decomposition of XeF+OsF¿ is thought to involve charge transfer (XeF+iOsF^]--> X e F ' + OsF^) with subsequent XeF- disproportion. This accounts for the observeds^ products: 3XeFnOsF6]-->

Xe2Fí[OsF6]-+

Xe+20sF6.

Johnston and Woolfolk have shown^^ that the first bond dissociation energy, Di = (XeF2(g) XeF(g) + F(g)), of XeF2 is greater than the second. They suggest Ώχ = 54 kcal mole-i. Since the total bond energy of XeF2 (see section 3.2.1) is 64 kcal mole-i, it follows that Dz (XeF2), i.e. BE (XeF-), would be 10 kcal mole-i. However, should the bond energy in the XeF radical be less than Δ ^ / (F) = 18.7 kcal mole-i, then dissociation 2XeF -> 2Xe-|-F2 would be favoured, but there is no evidence to indicate that this occurs. It is therefore more likely that the value for Di is closer to 44 kcal mole-i, since then BE (XeF) Ä 20 kcal mole-i. This value is still consistent with dispro­ portion 2XeF XeF2 + Xe, for which, since AS ^ 0, AG ^ -24 kcal mole-i. It is obvious that the XeF radical should be a very effective F atom source. The abundance of xenon isotopes, of differing nuclear spin, has provided for a much better definition of the XeF radical than was possible for K r F (see section 2.1.1). The e.s.r. data given in Table 17 indicate that the unpaired electron occupies an orbital possessing axial symmetry about the internuclear axis. Neglecting inner-shell polarization, the data indicate that this σ molpcular orbital is largely derived from Xe 5p and F 2p, since the spin population corresponds to 4.9% Xe 5^, 3 6 % Xe 5p, 2.6% F 2s and 4 7 % F 2p, To a first approximation then XeF- can be described in classical molecular orbital terms with the unpaired electron in the highest antibonding sigma orbital (σ*). Departure of the 82 W. E. Falconer and J. R. Morton, Proc. Chem. Soc. (1963) 95. 83 J. R. Morton and "W. E. Falconer, / . Chem. Phys. 39 (1963) 427. 84 V. A. Legasov, V. N . Prusakov and B. B, Chaivanov, Russ. J. Phys. Ch. 42 (1968) 610. 85 H. S. Johnston and R. W^oolfolk, J. Chem. Phys. 41 (1964) 269. 86 N . Bartlett, M. Wechsberg, F. O. Sladky, P. A . Bulliner, G. R. Jones and R. D . Burbank, Chem. Communs. (1969) 703; N . Bartlett and F. O. Sladky, The Second European Fluorine Chemistry Symposium, Göttingen, August 28-31, 1968. 87 F. O. Sladky, P. A. Bulliner and N . Bartlett, / . Chem. Soc. A (1969) 2179.

251

XENON(I) TABLE 1 7 . X e F RADICAL D A T A

Bond energy (kcal m o l e - i )

^^20

AH (2XeFc,) - > XeFacg) + Xe(g>) (kcal m o l e - i )

- -24/

\

Source for e.s.r. characterization:

y-irradiation o f XeF4 single crystal" ^

^^^^

Principal values of the hyperfine interaction tensorsf and g tensorst* Species

Xe.

Xe^

F,

Fx

i32XeF i29XeF i3iXeF

2368 701

1224

2649 2637 2653

540 526

—

1.9740 1.9740 1.9740

—

2.1251 2.1264

t Units are M c , errors ± 10 M c . t Errors ± 0 . 0 0 0 8 .

Comparison o f experimental isotropic A a n d anisotropic Β tensor parameters with one-electron parameters theoretically derived ^

[Sngßym Nucleus

^obs

i9F(« = 2)

1243 1605

i29Xe (η = 5)

47,900 M c 33,030 M c

% ns spin popn.

3 5

% np spin popn.

{igßym xpn

Nucleus

19F in = 2) 129F (« = 5)

703 382

1515 M c 1052 M c

47 36

• W. E . Falconer and J. R. Morton, Proc. Chem. Soc. (1963) 95. ^ J. R. M o r t o n and W . E . Falconer, / . Chem. Phys. 3 9 (1963) 427.

principal g-values of XeF from 2.(X)23 (free spin) must be associated with spin-orbit inter­ action of the ground state with Π excited states of the molecule. The ground state con­ figuration is written (σ2/?ρ+σ5/?χ., σι)2; (π2/?ρ+π5/7χ., πι, 2 ) ^ ; (.nlpp-nSpx^,

7 : 3 , 4 ) ^ ; σ2ρρ-σ5ρχ^,

oiS^\ or briefly

Of course σι and πι, 2 are bonding orbitals and πι, 4 and ^2 antibonding orbitals. The g shifts are attributed to the transitions (^\^\ T^\,i^\ ^3,4^; ö-2i-> σι2; πι,2^; π3,43; oi?-

ci^; 7^1,2^*; π3,4^; σ2ΐ-> σ·Ρ·\ πι,2^; π^Α^Ι

3.1.2. Xenon(I) Complexes Several complex salts have been r e p o r t e d ^ ' T h e stoichiometry of these materials, e.g. XePtFö and XeRhFö, implies either xenon(I) or (11)2 species. The first compound is of particular interest since it was the first xenon compound to be reported in which the xenon valence electron configuration was unequivocally different from the supposed "ideal" octet. This and the related XeRhFe are formed spontaneously at ordinary temperatures by interaction of xenon gas with the hexafluoride vapour: Xe(g) + MF6(g) XeMFöcc). F o r the 1:1 stoichiometry it is essential to maintain a large excess of xenon over hexafluoride.

252

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY

The infrared data support the formulation of these materials as platinum(V) and rhodium(V) compounds and their interaction with alkali fluoride in iodine pentafluoride (which is not an oxidizing solvent towards platinum(IV) or rhodium(IV)) generates A'M^Fe salts. Of course, the Xe+ species should be paramagnetic. So far there is no reliable magnetic or structural data to support the designation of the xenon in the solid materials as Xe+ rather than Xe2"^ or XeaF^ (which would be appropriate for the formulation XezF^ [FsPt-F-PtFs]-). It is noteworthy that the material of composition Xe(PtF6)2 which is formed under conditions of twofold excess of PtFe over Xe, is87 an equimolar mixture of XeF+ [PtFe]- and PtFs (see section 3.2.6). Considering, the estabhshedi^ capabihty of PtFe to oxidize O2 to OJ, the rapid forma­ tion of the XePtFö material from the gas phase, and the quinquevalence of the platinum in the soHd product, it is probable that this and XeRhFö represent true xenon(I) compounds.

3.2.

XENON(II)

3.2.1. Xenon Difluoride Xenon difluoride was first detected independently in two laboratories^* »8^ and several eflective syntheses were quickly reported89-9i. Laboratory preparation. The static thermal method of preparation employing large excess of xenon over fluorine is the best way to make large quantities of the fluorideis; indeed, as Falconer and Sunder have shown^^, high yields of pure material can be obtained with Xe:F2 molar ratios of ^^2:1. The gaseous mixture is heated in a nickel or Monel vessel at 400°, quenched to room temperature, and the difluoride isolated by vacuum sublimation. A convenient preparationi2.93^ which avoids the special metal equipment used in the last procedure, simply involves exposure to sunlight of Xe/F2 mixtures (-^1 atm total pressure) contained in dry Pyrex glass vessels, at room temperature. Other syntheses. As might be expected, fluoride decomposition or excitation to provide atomic fluorine, when carried out in the presence of xenon, has been shown to yield xenon fluorides. When the technique allows for the separation of crystalline fluoride, either by continuous circulation of the energized gaseous mixture through cold traps (at 0° or below), or by providing energy to the F - X e system at room temperature or below, the xenon fluoride product has invariably proved to be XeF2. Weeks, Chemick and Matheson^^ were the first to exploit a cold trap in a circulating gas system to effect the preparation of high purity material. Others have exploited this feature in successful application to hot tube syntheses9i.94, Xhe Weeks et al. photosynthesis^^ employed irradiation of Xe/F2 gaseous mixtures with light from a high-pressure mercury arc (2500-3000 A). Photo­ synthesis, with sunhght, involving X e / 0 F 2 mixturesi2 at ^^25° and XQ/O2F2 mixtures^s at — 118°, are also effective. 88 89 90 91 92 93 94 95

C. L. Chemick, H . H . Claassen et al.. Science 138 (1962) 3537. J. L. Weeks, C. L. Chemick and M . S. Matheson, / . Am. Chem. Soc. 8 4 (1962) 4612. R. Hoppe, Η . Mattauch, Κ. Μ . Rödder and W. Dähne, Ζ . anorg. Chem. 324 (1963) 214. D . F . Smith, / . Chem. Phys. 3 8 (1963) 270. W. E. Falconer and W. A . Sunder, / . Inorg. Nucl. Chem. 2 9 (1967) 1380. J. H . Holloway, Chem. Communs. (1966) 22. p . Gróz, L Kiss, Α . Révész and Τ. Sipos, / . Inorg. Nucl. Chem. 2 8 (1966) 909. S. I. Morrow and A . R. Young, II, Inorg. Chem. 4 (1965) 759.

XENON(II)

253

Gamma-ray irradiation at 4 χ lO^ rad h r - i of ^ 1:2 Xe/Fa mixtures at 64° has been shown^ö to yield ^^^l:! XeF2/XeF4 mixtures with a G value (for Xe consumption) of 3.4 atoms per 100 eV absorbed in the gas mixture. The observations of Gard, Dudley and C a d y 9 7 that XeF2 is formed in interaction of Xe with O F 2 (>187°), CF3OF (220-250°), and FSO3F (170-180°), are consistent with syntheses involving, simply, interaction of Xe with F2 derived from dissociation of the other reagent. The fluorination of xenon, at 200°, by iodine heptafluoride^s, Xe(.)+IF7(.)->XeF2,IF5(c,

is similarly dependent upon F2, generated by pyrolysis IF7 IF5 + F2. It is possible that xenon difluoride has also been o b s e r v e d 9 9 as a product of the fission of a U O 2 - L Í F mixture. Of particular interest is the formation of XeF2 as a product of the interaction of xenon with carbon tetrafluoride in a high voltage discharge^oo and by the interaction of excited xenon (^Ρχ) with perfluoro-cyclobutaneioi: Xe+ + c-C4F8 -> XeF2 + c-C4F6. Thermodynamic properties. Xenon difluoride is colourless as solid, liquid or gas. The vapour pressure of the solidio2 at 25° is 4.55 m m ; accordingly the solid develops large crystals easily at room temperatures. Although the melting point of supposedly pure XeF2 has been variously given at 140, 130+0.6, 1 3 4 + 2 and 139.6 + 0.2° (ref. 84), the most rehable value is 129.03+0.05°, as given by Schreiner et al.^^'^. The vapour pressure data given by the last workers for the temperature range 0-115° is also the most rehable, from which A ^ s u b = 13.2 kcal mole-i. The enthalpy of sublimation is in close agreement with the prior value of 13.3 kcal mole"i, calculated by Jortner et al.^^^, assuming an electrostatic stabihzation of the solid involving a point charge of —0.5e on each F ligand and + le on the xenon atom. Three values have been given for the enthalpy of formation of the difluoride: (XeF2, g) (kcal mole-i), - 2 5 . 9 from equihbrium constant data22, - 2 8 . 2 + 0 . 6 from a calorimetric study 104, and — 3 7 + 1 0 from appearance potential studiesios. The second value is preferred, from which the total thermochemical bond energy is 64 kcal mole-i. But Johnston and Woolfolk have evidence from kinetic studies^s, involving XeF2 and X e F 4 interactions with N O and N O 2 , that the first bond dissociation energy for XeF2(g) -> XeF-(g) + F(e) is much greater than the second. As has been discussed under the XeF- radical (section 3.1.1), reasonable values for the bond dissociations are Di 44 and D2 20 kcal mole-i. It will be appreciated from the thermochemical cycle given for XeF2 in Fig. 9, that the total bond energy of XeF2 is related to the bond energy of XeF+ by the equation T B E (XeF2,g) = B E ( X e F + ) + ^ (XeF+, X e F 2 ) - / ( X e ) .

Since the last two experimentally observed quantities are 12.8 and 12.1 eV respectively, TBE (XeF2, g) = 64 = BE (XeF+) + 0 . 7 e V (16 kcal mole-i). Therefore the bond energy 96 D . R. MacKenzie and R. H. Wiswall, Jr., Inorg. Chem. 2 (1963) 1064. 9 7 See ref. 14, p. 109. 98 N . Bartlett and D . E. McKee, unpublished observation. 99 Y. Kamemoto, / . Inorg. Nucl. Chem. 27 (1965) 2678. 100 D . E. Milligan and D . R. Sears, / . Am. Chem. Soc. 85 (1963) 823. 101 G. H. Miller and J. R. Dacey, J. Phys. Chem. 69 (1965) 1434. 102 F. Schreiner, G. N . McDonald and C. L. Chemick, / . Phys. Chem. 7 2 (1968) 1162. 103 J. Jortner, E. G u y W^ilson and S. A . Rice, / . Am. Chem. Soc. 85 (1963) 814. 104 V. I. Pepekin, Y. A. Lebedev and A . Y . Apin, Zh. Fiz. Khim. 43 (1969) 1564. 105 H. J. Svec and G. D . Flesch, Science 142 (1963) 954.

254

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY

for XeF+ is 48 kcal mole-i, which is compatible with the greater strength and shortness of the X e - F bond in the cation, compared with XeFz (see section 3.2.6). A(XeF, XeFj)

AHat (XcFj) BE(XeF;)

(Total Β Ε (XeFj))

(g)

I(Xe) F I G . 9. Thermochemical cycles for XeF2 and derived species B E (XeF+) = / ( X e ) - / i (XeF+, X e F 2 ) + t o t a l B E (XeF2, g) / ( X e F ) = A (XeF+. X e F 2 ) - i ) i /(XeF) =

(XeF2,g)

/(Xe)+BECXeF)-BE(XeF+)

TABLE 18. PHYSICAL PROPERTIES OF X E N O N DIFLUORIDE

Colourless crystals, liquid and vapour Triple p o i n t ' A//.„b*

129.03°C 13.2±0.2kcalmole-i

Vapour pressure (solid) ·

l o g P „ „ = ^ ^ ^ ^ - 1 . 2 3 5 2 1 log Γ-f 13.969736

S^(gas)* SMsolid)*

25^C, 62.057; 250°C, 69.715; 50VC, 57°C, 29.4 cal m o l e - i d e g - i

Thermodynamics

of XQ¥2

75.345 cal mole^i d e g - i

formation

Δ/r.tom (XeF2(„ — X e ( „ + 2F(„) Mean thermochemical bond energy First bond dissociation energy **

25°C, - 2 5 . 9 0 3 ; 5 0 r C , - 2 5 . 4 9 1 kcal m o l e - i 25°C, - 1 7 . 8 5 8 ; 5 0 r C , - 5 . 2 2 2 kcal m o l e - i 64 kcal m o l e - i 32 kcal m o l e - i - 5 4 kcal m o l e - i (but see text)

Solubility Relative solubilities: BrFs very g o o d ; BrFa very g o o d (but complex formation); IF5 g o o d (adduct formation); C H 3 C N g o o d ; H F fair; SO2 fair; W F Ö poor; N H 3 very slight · Solubility in: 25 ge 1-1 at 0° H20' 12.25 -2.0 29.95 Temp (°C). HF« 7.82 6.38 9.88 C (moles per 1000 g) 21 0 Temp. (°C) CH3CN'» 320 168 C(gl-i) 80.0 61.0 49.2 33.2 16.8 Temp. ( ° Q CNF, 3HF * 30.55 25.38 22.17 16.3 19.56 C (moles per 1000 g) Electrical conductivity measurements support the essentially molecular nature o f XeF2 in all o f the solutions; look under appropriate headings for i.v., u.v. and n.m.r. spectra of these solutions.

255

XENON(II) TABLE 1 8 (cow/.) Thermodynamics

and kinetics ofX^Fi

hydrolysis

AG° (XeF2(.„ + H2O •Xecg, + i02(g) + 2HF(.,)) est.J « - 5 3 . 4 kcal m o l e - i ; hence ΑΓ., « 1040 Hydrolysis conditions

First order rate constant

Ref.

4.2xl0-4sec-i(25^)

0.01 Μ HC104

19.6

r 2.52 2.83 ± ± 0.01 0.02 X X 11 00 -- 54 sec-i s e c - i (25°) (0")

Only XeF2 in H2O

—

18.4±2.1

L 1.2 X1012 e x p ( - 1 8 , 4 0 0 / Ä r ) m i n - i

Vibrational

-8.1

—

-

\ /

k 1 J

spectra

XeF2, solid: Bands ( c m - i ) 547 « ( v s ) " 496 R (1.0)"·« 108 R (0.33) "· Assignment lattice m o d e VI V3 XeF2, vapour: Bands ( c m - i ) 1070 » ( w ) " ' « » 213.2 IR(vs)*» 560 IR (vs) ^ Assignment V1 + V3 V2 V3 C H 3 C N solution: H. Q. R Bands ( c m - i ) 1235 IR (w) 509 R (s) 533 IR (s) '· · Assignment VI V3 V3 ( c m - i ) in other solvents: CH3NO2, 532; dioxane 530; C C I 4 , 538 Ultraviolet

spectrum ofXéFz

vapour *· "·

(see also photoelectron spectra)

Wavelength A„»x(A)

Half width Δ ν (cm-i)

2300 1580 1425 1335 1215 1145

8249 8060 (1000) (1290) (2070) (2730)

Est. extinction coefficient e m o l e - i cm-2

Est. oscillator strength if)

0.86x102 1.12x104 0.4x104 0.4x104 0.4x104 0.6x104

0.0033 0.42 0.02 0.02 0.03 0.06

For assignments, see references t, u, x, v, y. Rydberg bands ^'

"· *

Allowing for a systematic error o f - 1 8 A in the Rydberg data o f refs. t and v, this and the helium I a n d U photoelectron data o f ref. y can be fitted by the t w o series riOO,160\ 109,737 = \ l 0 3 , 9 5 0 / - ( ; i - 4 . 0 4 ) 2 ^ " ^ " ' > « = 6, 7, . . . and The V, Rydbergs are

,737 _ /100,160\ 109,7 \ 1 0 3 , 9 5 0 / - ( ^ ,42)2 c m - i ;

Λ =

the spin orbit split components o f a 5π„

The va Rydbergs are probably the spin orbit components o f a 5π„

5, 6 , . . .

• 6s transition. • 5d transition.

256

N O B L E - G A S C H E M I S T R Y : N E I L B A R T L E T T A N D F. O . S L A D K Y TABLE 18

Photoelectron

(cont.)

spectra (see also U V spectrum) ^* Term energies in X e and XeF2 ^ Xe«'

XeF2

Upper S t a t e

Term ( c m - i )

Upper S t a t e

Term ( c m - i )

5p(2Pi,2)6s

30,400 31,433 19,322 19,317 16,628 16,567 12,551 12,590

5π„(2Ιΐ3/2)6ί 5π„(2ΙΙι/2)65

30,865 30,080

5π„(2Π3/2)5ί/ 5π„(2ΙΙ,/2)5ί/

17,860 16,600

5pmn)6s 5ρ(2Ρ3/2)βρ 5pm/2)6p

5pmi2)5d

5p(2Pi,2)5d

5pm¡2)7s Spm,2)ls Appearance potentials

of ions

^ XeF2 XeF+

Observed and calculated

ionization potentials Adiabatic obs.

12.35 ± 0 . 0 1 1 12.89 ± 0 . 0 1 ca. 13.5 14.00 ± 0 . 0 5 15.25 ± 0 . 0 5 16.80±0.05

AP(eV) 12.28 12.78

(eV) of XéF2 ^ Vertical obs.

12.42 ± 0 . 0 1 12.89 ± 0 . 0 1 13.65±0.05 14.35 ± 0 . 0 5 15.60±0.05 16.00±0.05 17.35 ± 0 . 0 5 ca. 22.5

0.92 K T cale.

12.51 (5π„) 11.79 (ΙΟσ,) 14.71 Qng) 15.92 (4π„) 16.93 (6σ„) 25.24 (9σ,) 37.10 (5σ„) 37.20 (8σ,)

t There is a n indication o f a weak shoulder centred at 12.30 e V which may correspond to the true adiabatic i.p. for this band. However, it seems more likely that this lowest band is a hot one, arising from the appreciable excitation o f the l o w frequency bending m o d e in the ground state o f the neutral molecule. XeF2 n.m.r. data Microcrystalline solid, 25°; i^Fa, 612 p p m ; line width 6.0 G ; H F solution (1 mole per 1000 g), - 1 9 . 5 °

ΐ9Ρσ Ref.

(ppm)

129Χβσ (ppm)

b' c'

629 630.3

-3930

/l9P^129Xe

(c/s)

5600 5690

257

XENON(II) TABLE 18

(cont.)

O N F , 3 H F solutions ( - 1 m o l e per 100 g), 40° Ref.

C H 3 C N solution

i h

631 610±2

5640 ± 2 0 5450 ± 2 0

—

Nuclear magnetic resonance single crystal data ^' (Rigid-lattice second moment M2 = Μ 2 + 4 / 4 5 Δ σ 2 / ί ο » where and σ , are the shielding _L and || to symmetry axis). Λ/2 obs. = 3.25 G2, M2 calc. = 2.85 G^—the difference (see crystal structure data) Δ σ | = (105 ± 10) χ lO^.

Mössbauer

spectrum

= applied field strength Δ σ = σ ^ . - σ „ ,

probably represents thermal

displacements

(see sections 1.4.1, 2.2.1 and 3.2.2)

Splitting (mm s e c - i )

eiqQ (Mc sec-i)

Isomer shift (mm s e c - i )

(4.2°K) 3 9 . 0 0 ± 0 . 1 0

2490

0.10±0.12

Κ

Electron transfer per b o n d

1.43 t

1.43 Í

0.711

t Up is the quadrupole coupling in units of the quadrupole coupling of a single pt hole, t hp is the hole in the 5p shell assuming that the bonding involves ρσ orbitals only.

Molecular

structure

Vapour: / ) o o * geometry established by I R and R data"* Rotational I R (V3 band) ^ gives, OL^ = 3 . 3 1 x 1 0 - 4 c m - i , BQ = 0.11350 c m - i , and -1.44cm-i F r o m Bo the X e - F bond length is 1.9773 ± 0 . 0 0 1 5 A

Single crystal structure (see Fig. 10) Tetragonal unit cell«': 0 = 4 . 3 1 5 ± 0 . 0 0 3 ; c = 6 . 9 9 0 ± 0.004 A K= 130.15A3;z = 2 /)c.ic = 4.32 g c m - 3 . S G /4/mmm (2)i¡)

Structural parameters '

Xe F

/

ζ

β

ßll

(fermi units)

(0.0) 0.2838 ± 0 . 0 0 0 4

0.0341 ± 0 . 0 0 2 0 0.0635 ± 0.0022

0.0083 + 0.0006 0.0087 ± 0.0004

(0.476) (0.550)

(Values in parentheses were not varied; / a r e neutron scattering amplitudes.)

X2i

258

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY TABLE 18

(cont,)

Intramolecular distance X e - F , 2 . 0 0 ± 0 . 0 1 A Each X e atom has eight "non-bonded" F neighbours at 3.41 A Each F atom has one F neighbour at 3.02 A (along c) and 4 F at 3.09 A Magnetic

susceptibility

"

= 4 0 - 5 0 X 1 0 - 6 e.m.u. Bonding: Refs. u, g', j - v ' . Proposed energy levels: Refs. u - w , y. " F. Schreiner, G. N . M c D o n a l d and C. L. Chernick, / . Phvs. Chem. 7 2 (1968) 1162. " B. Weinstock, Ε. Ε. Weaver and C. P. K n o p , Inorg. Chem. 5 (1966) 2189. ^ M. Karplus, C. W. Kern and D . Lazdins, / . Chem. Phys. 4 0 (1964) 3738. ^ H, S. Johnston and R. Woolfolk, / . Chem. Phys. 41 (1964) 269. • N . Bartlett and F. O. Sladky, Chem. Communs. (1969) 1046. ' E. H. Appelman and J. G. Malm, / . Am. Chem. Soc. 86 (1964) 2297. ' H. H. Hyman and L. A . Quarterman, in Noble Gas Compounds (H. H . H y m a n , ed.). University o f Chicago Press, Chicago and L o n d o n (1963), p. 275. H. Meinert and S. Rüdiger, Ζ. Chem. 7 (1967) 239. * A. V. Nikolaev, A . S. Nazarov, A . A . Opalovskii and A . F. Trippel, Dokl. Akad. Nauk SSSR 186 (1969) 1331. 1 V. A. Legasov, V. N . Prusakov and B. B. Chaivanov, Russ. J. Phys. Ch. 4 3 (1968) 610. ^ E. H. Appelman, Inorg. Chem. 6 (1967) 1305. ' I. Fehér and M. Lörine, Inorg. Rem. Folz. 7 4 (1968) 232. ·" J. J. Turner and G. C. Pimentel, see ref. (g), p. 101. " D . F. Smith, see ref. (g), p. 295. ° P. A. Agron, G. M. Begun, H. A . Levy, A . A . Mason, G. Jones and D . F. Smith, Science 139 (1963) 842. ^ S. Reichman and F. Schreiner, / . Chem. Phys. 51 (1969) 2355. H. Meinert and G. Kauschka, Z . Chem. 9 (1969) 70. ' H. Meinert and G. Kauschka, Z . Chem. 9 (1969) 114. " N . Bartlett and D . E. McKee, unpublished observation. ' E. G. Wilson, J. Jortner and S. A . Rice, / . Am. Chem. Soc. 8 5 (1963) 813. " E. S. Pysh, J. J. Jortner and S. A . Rice, / . Chem. Phys. 4 0 (1964) 2018. " J. J. Jortner, E. G. Wilson and S. A . Rice, see ref. (g), p. 358. ^ J. G. Malm, H. Selig, J. J. Jortner and S. A . Rice, Chem. Revs. 65 (1965) 199. " Y . J. Israeli, Bull soc. chim. France 3 (1963) 649. ^ C. R. Brundle, M. B. Robin and G. R. Jones (in press). ' J. O. Morrison, A. J. C. Nicholson and T. A . O'Donnell, / . Chem. Phys. 4 9 (1968) 959. C. E. Moore, Atomic Energy Levels, Nat. Bur. Stand. Circular 467, Vol. 3, Washington (1958). J. C. Hindman and A . Svirmickas, see ref. (g), p. 251. ^' T. H. Brown, E. B. Whipple and P. H. Verdier, see ref. (g), p. 263. D . K. Hindermann and W. E. Falconer, / . Chem. Phys. 50 (1969) 1203. C. L . Chernick, C. E. Johnson, J. G. Malm, G. J. Perlow and M. R. Perlow, Phys. Letters 5 (1963) 103. G. J. Perlow, C. E. Johnson and M. R. Perlow, see ref. (g), p. 279. »' S. Siegel and E. Gebert, / . Am. Chem. Soc. 85 (1963) 240. H. A. Levy and P. A . Agron, see ref. (g), p. 221. R. Hoppe, Η. Mattauch, Κ. Μ. Rödder and W. D ä h n e , Ζ . anorg. Chem. 3 2 4 (1963) 214. R. E. Rundle, / . Am. Chem. Soc. 85 (1963) 112. '^' Κ. S. Pitzer, Science 139 (1963) 414. " C. A . Coulson, / . Chem. Soc. (1964) 1442. «>' C. A . Coulson, J. Chem. Phys. 4 4 (1966) 468. C. J. Jameson and H. S. Gutowsky, / . Chem. Phys. 40 (1964) 2285. L. C. Allen, see ref. (g), p. 358. R. Bersohn, / . Chem. Phys. 38 (1963) 2913. D . Lazdins, C. W. Kern and M. Karplus, / . Chem. Phys. 39 (1963) 1611. " L . L . Lohr, Jr. and W. N . Lipscomb, / . Am. Chem. Soc. 85 (1963) 240. ·' K. A . R. Mitchell, / . Chem. Soc. A (1969) 1637. J. Jortner, E. G. Wilson and S. A . Rice, / . Am. Chem. Soc. 8 5 (1963) 814. D . F. Smith, / . Chem. Phys. 38 (1963) 270; and see ref. (g), p. 295. R. C. Catton and K. A . R. Mitchell, Chem. Communs. (1970) 457.

XENON(II)

259

The kinetics of XéFi formation. Rate studies of the interaction of xenon with fluorine have shown that the reaction is zero order in Xe io6, lo?^ los. These same investigations indicate that the reaction is primarily heterogeneous. Weaver^o? has noted a first-order dependence of the reaction on F2, which may be due to a slow step involving the dissociation of adsorbed fluorine molecules into adsorbed fluorine atoms. Evidently the walls of nickel or Monel vessels have marked catalytic activity, and C0F3, NÍF2 and CaF2 have been shown to be effective catalysts. However, metal fluorides cannot be catalytic agents in the photochemical synthesis carried out in Pyrex vessels. Again, wall reactions may be involved, but there is no clear evidence for this at this time. Sinel'nikov et al.^^^ have shown that atomic fluorine (generated in a glow discharge) is capable of converting condensed xenon (at 77°K) to xenon difluoride (45% yield in 75 min). From this, and the efficiency of the gas discharge synthesis, they have concluded that xenon activation is not necessary for xenon difluoride formation. Structural features. Infrared^o, 1 1 1 and R a m a n i i 2 , 1 1 3 , 1 1 4 , 1 1 5 , 1 1 0 spectroscopy have established the symmetrical linear {D^t) geometry of the XeF2 molecule. The vibrational spectroscopic data is given in Table 18. A high-resolution infrared study^^ of the V 2 stretching mode has provided a bond length of 1.9773+0.0015 A for the vapour phase molecule. This bond length is similar to that in the molecule in the crystalhne phase. The value given by Levy and Agronii^, u s from their single-crystal neutron diffraction study is 2.00+0.01 A. In crystalline XeF2, the hnear molecules are ahgned parallel in a body-centred tetragonal array, the unit cell of which is shown in Fig. 10. It is clear that there are strong interactions between molecules since each xenon atom has not only its two bound fluorine atoms at 2.0 A but also eight fluorine atoms at 3.41 A, the latter being fluorine hgands of the eight nearest XeF2 neighbours. This structural arrangement is compatible with the high enthalpy of subhmation of XeF2 (AHnub = 13.2 kcal mole-i), and Jortner et al.^^^ have convincingly accounted for these features. They point to the considerable bond polarity in the XeF2 molecule and account for A i ^ s u b in terms of an electrostatic stabilization energy of 11.31 kcal mole-i (assuming the charge distribution in each molecule to be -o.5F-Xe+i-F-o-5) and a dispersion energy of - 2 kcal mole-i—a total of 13.3 kcal mole-i. It is evident from the packing arrangement that the region close to the equatorial plane of each XeF2 molecule is avoided by neighbouring molecule fluorine ligands. This may indicate that the "non-bonding" valence electrons of the xenon atom provide very effective shielding of the xenon positive charge in this plane. It seems also that the effective volume of the xenon atom is considerable in crystalline XeF2 (and even in its derivatives, see 106 B. G. Baker and P. G. F o x , Nature 204 (1964) 466. 107 E. E. W^eaver, B. Weinstock and C. P. K n o p , / . Am. Chem. Soc. 85 (1963) 111. 108 B. H. Davis, J. L. Wishlade and P. H. Emmett, / . Catalysis 10 (1968) 266. 109 S. M. Sinel'nikov, I. V. Nikitin and V. Y . Rosolovskii, Izv. Akad. Nauk Ser. Khim. (1968) 2806; (English trans.), p. 2655. 110 P. A . Agron, G. M. Begun, H, A . Levy, A . A . Mason, G. Jones and D . F . Smith, Science 139 (1963) 842. 111 W. A . Yeranos, Molec. Phys. 12 (1967) 529. 112 See ref. 14, p. 39. 1 1 3 See ref. 14, p. 101. 1 1 4 See ref. 14, p. 295. 115 See ref. 14, p. 304. 116 S. Reichman and F. Schreiner, / . Chem. Phys. 51 (1969) 2355. 117 H. A . Levy and P. A . Agron, / . Am. Chem. Soc. 85 (1963) 241. 118 See ref. 14. p. 211.

260

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY 4-315Ä

-

Q

F I G . 10. The unit cell o f xenon difluoride

sections 3.2.6 a n d 3.2.7). The molecular v o l u m e for XeF2 is 65.1 A3. But in m a n y fluorides (e.g. XeFö, IF7, IF¿-AsF¿-) the effective v o l u m e o f each fluorine hgand is 17.8 A3. Therefore ^29 k} o f the effective molecular v o l u m e o f XeFa can be attributed t o a n effective xenon(II) volume. This effective volume m a y largely derive from the shielding properties (or steric activity) o f the "non-bonding" x e n o n valence electrons. Findings from nuclear magnetic resonanceii^, 1 2 0 , 1 2 1 , 1 2 2 and Mössbauer studiesi23, 1 2 4 (see sections 2.2.1 and 3.2.2) have generally been interpreted in terms o f considerable b o n d ionicityi25.119, 123. Recent broad-hne n.m.r. studiesi20 s h o w that the previous quantitative evaluations m a y b e considerably in error, since m u c h o f the earlier n.m.r. experimental work w a s apparently carried o u t using rather impure XeFa samples. Nevertheless, b o n d polarity in XeFz must be high and is probably n o t very different from the value derived by Rice and his coworkersi04. From the core-electron chemical-shift derived from X-ray electron spectroscopy (ESCA) o f gaseous XeF2, Karlsson et alA^^ have concluded that the positive charge o n 119 C. J. Jameson and H. S. Gutowsky, / . Chem. Phys. 4 0 (1964) 2285. 120 D . K. Hindermann and W . E. Falconer, / . Chem. Phys. 50 (1969) 1203. 121 See ref. 14, p. 251. 122 See ref. 14, p. 263. 103.123 C. L. Chemick, C. E. Johnson, J. G. Malm, G. J. Periow and M . R. Periow, Phys. Letters 124 See ref. 14, p. 279. 125 D . Lazdins, C. W. Kern and M. Karplus, / . Chem. Phys. 39 (1963) 1611. 126 S.-E. Karisson, K. Siegbahn and N . Bartlett, / . Am. Chem. Soc. (1970) (in press).

5 (1963)

XENON(II)

261

the xenon atom is + 1 — a value in fair agreement with n.m.r. and Mössbauer findings. A high resolution helium I and II photoelectron spectroscopic study of the diñuoride (involving valence electron promotion or ionization) has been reported by Robin and his coworkersi27. They have shown that the vertical ionization potentials of the first eight ionizations in the X e F i molecule compare well with the results of Gaussian type orbital calculations. The first two ionic states of X e F 2 are the (12.42 eV) and (12.89 eV) spin-orbit components, formed by ionization from the highest filled π orbital (5π„). They conclude that the X e F 2 molecule in these two states is essentially of the same size and shape as the ground state neutral species. This indicates that the 5π„ orbital is essentially nonbonding. The photoelectron spectrum of X e F 2 yields characteristic Rydberg energies which correlate with Rydberg excitation energies derived by Rice and his coworkersi^s, 1 2 9 , 1 3 0 from the X e F 2 ultraviolet spectral data. The interpretation given by the latter workers for the ultraviolet spectrum of X e F 2 is in terms of a semi-empirical LCAO molecular orbital description. The bonding in X e F 2 . The bonding models for X e F 2 have been presented earlier (see sections 1.3.3-4). Perhaps, because of the wealth of data on the molecule and its simplicity, it has been the subject of many theoretical papers. The majority of these have depended upon molecular orbital models which have excluded Xe outer orbitals {5d and 4/) from significant involvement. Coulson's review23 gives the essence of these models. On the other hand, more recently, M i t c h e l l 2 5 has concluded on the basis of model calculations, following the extended valence-bond method of Hurley, Lennard-Jones and Pople, that the X e F 2 structure with lowest energy is the locahzed electron pair structure F - X e - F , which involves a contracted Sda Xe orbital in the bonding. Mitchell, furthermore, argues that Coulson's favoured valence-bond structure F - X e + F " is energetically unfavourable. However, the F - X e 2 + F - structure is claimed to be lower in energy than F-Xe+F", and hence to be more hkely to contribute to the bonding description. Clearly, it will be some time before the question of outer Xe orbital participation in the bonding in the Xe compounds can be satisfactorily resolved. Non-aqueous XéFi chemistry. Since X e F 2 is easily made and is thermodynamically stable at ordinary temperatures and pressures, it is a convenient source of other xenon(II) compounds. Furthermore, since the X e - F mean thermochemical bond energy is one of the lower bond energies of known fluorides, being comparable to the mean bond energy of CIF5, X e F 2 is, potentially, a strong oxidizer and fluorinator. Both of these aspects have been explored. The fluorine ligands of the X e F 2 may be substituted, by highly electro-negative hgands, using the ligand protonic acids: XeF2+HL->EXeL+HF FXeL+HL->XeL2+HF

The high exothermicity of AHf (HF) is a major factor in bringing about the forward reaction. Sohd products have been obtained for L = SO3F and CIO4 8«, OTeFs i3i» i32^ 127 c. R. Brundle, M. B. Robin and G. R. Jones (in press).

128 See ref. 14, p. 358. 129 130 131 132

E. E. F. F.

G. Wilson, J. Jortner and S. A . Rice, / . Am. Chem. Soc. 85 (1963) 813. S. Pysh, J. Jortner and S. A . Rice, / . Chem. Phys. 4 0 (1964) 2018. O. Sladky, Angew. Chem., Intn. Edn. 8 (1969) 373. O. Sladky, Angew. Chem., Intn. Edn. 8 (1969) 523.

262

NOBLE-GAS CHEMISTRY! NEIL BARTLETT AND F. O. SLADKY

and possibly O2CCF3133^ 134. The Perchlorates and trifluoroacetates are detonatable. All of these compounds are described later (see sections 3.2.4 et seq). Efforts to generate FXeCl or X e C b by interaction of X e F 2 with HCl or BCI3 have failed, xenon being ehminated quantitatively 135. Although xenon difluoride is potentially a strong oxidizer it is frequently unreactive for kinetic reasons. Its stability in aqueous solution (see later) is typical of this kinetic inertness, although solutions in other solvents, e.g. CH3CN i36,137 and other organic solventstes, are also stable if certain catalysts, particularly fluoroacids, are absent. The difluoride is also soluble in the halogen fluorides IF5, BrFs and BrFs i39 and in anhydrous hydrogen fluoridei40. Physical evidence, where available, has always shown the dissolved XeF2 to be mono-molecular in solutioni4i.i42.i36,138,143 and to be geometrically similar to the gas phase species. Occasionally the solvents form complexes with the difluoride (e.g. IF5 and BrFs), and this limits their usefulness in preparations. Either BrFs i39 or acetonitrilei37,142, i44, i36 (depending on the particular application) is a convenient solvent for XeF2, but it is possible that O N F , 3HF (b.p. 94°) will also prove to be a very useful solvent for iti43.

The oxidizing capability of X e F 2 can be exploited by using a fluoride ion acceptor as catalysti39. Thus a solution of iodine and X e F 2 in acetonitrile may be kept indefinitely (particularly if CsF is present, to absorb fluoroacids), but on introduction of H F or BF3 the X e F 2 oxidizes the iodine to form iodine fluorides: l2+XeF2

F- acceptor F" acceptor > Xe+2IF; IF+XeF2 > X e + I F 3 , etc.

Oxidizable strong fluoride ion acceptors interact rapidly with the difluoride: X e F 2 + 2 S 0 3 -> X e t +S2O6F2

Solutions of xenon difluoride and benzene or other aromatics are stable until hydrogen fluoride is introduced, at which point the solutions become coloured, xenon evolves, and H F and fluoroaromatics are formed. Thus 4.1 χ 10-2 moles of benzene in interaction with 1.26x10-2 moles of X e F 2 f o r 2 i h r yields a distillate of composition: 88.72% benzene, 10.28% fluorobenzene, 0.7% /7-difluorobenzene and 0.3% o-difluorobenzene. A tarry residue (0.640 g) contained monofluorobiphenyl, biphenyl, difluorobiphenyl and trifluorobiphenyl. Xenon difluoride fails to interact with perfluoro-olefins, even after several days, whereas it interacts with olefins to yield difluoro-olefinsi^s. This may be due to the presence 133 134 135 136 137 138 139 140

141 142 143 (1969)

J. L Musher, / . Am. Chem. Soc. 90 (1968) 7371. M. Eisenberg and D . D . DesMarteau, Inorg. Nucl. Chem. Letters 6 (1970) 29. M. Wechsberg and N . Bartlett, to be published. H. Meinert and G. Kauschka, Z. Chem. 9 (1969) 70. H. Meinert and S. Rüdiger, Ζ . Chem. 7 (1967) 239. Η. Meinert and G. Kauschka, Z. Chem. 9 (1969) 114. N . Bartlett and F. O. Sladky, Chem. Communs. (1968) 1046. See ref. 14, p. 275. E. H. Appelman and J. G. Malm, / . Am. Chem. Soc. 86 (1964) 2297. H. Meinert and G. Kauschka, Z . Chem. 9 (1969) 35. A. V. Nikolaev, A. S. Nazarov, A . A . Opalovskii and A . F . Trippel, Dokl. Akad. Nauk SSSR 1331.

144 H. Meinert and S. Rüdiger, Ζ. Chem. 9 (1969) 35. 145 T.-C. Shieh. N . C. Yang and C. L. Chemick, / . Am. Chem. Soc. 86 (1964) 5021.

186

XENON(II)

263

of some H F in the latter reaction. It is of interest that the vie difluorides formed in this reaction isomerize to the gem difluorides: CH2=CH2 XeR

^CHjFCHiF

^

CH3CHF2

+

CH3—CH=CH2—^ C H3CH — C H F 2 It seems likely that in all of these fluorination and oxidation reactions, the reaction takes place first by ionization of XeFi to XeF+ (or related species) followed by electron transfer to give XeF-. As discussed above, the X e F species is very weakly bound and must be a very effective F atom source (see section 3.1.1). The X e - F bond in XeF+ (see above and section 3.2.6) is considerably stronger than in XeF2, as is to be expected, since here we have essentially electron-pair bonding involving Xe 5p (and possible 5^) orbitals in the bonding (see sections 1.3.1-4). This must contribute to the ready ionization of XeFa. A number of salts of the XeF+ ion have now been estabhshed and they are discussed in section 3.2.6. The ionization enthalpy (XeF2(g)XeF+j + F - j ) is « + 2 1 5 kcal mole-i, which compares with that for O N F , (ONF(,) V n O ¿ + F - ^ + 2 0 8 kcal mole-i. There is no reason why a large number of XeL+ salts should not be derivable from the XeF+ salts or XeF2 itself. It will probably be essential to use rather electro-negative ligands L, and it will be necessary, for stable salts, t h a t ] ^ (latticeenergy—ionization potential of the anion), should be more exothermic than the electron aflSnity of the cation. The compounds which XeF2 forms with and XcF^ (see section 3.2.7) preserve the essential form of XeF2. The intermolecular bonding in these compounds is primarily coulombic and like that im solid XeF2 itself. It is a consequence of the considerable bond polarity of XeF2. The aqueous solution chemistry of XeF2. Xenon difluoride dissolves in water (^^25 g l"i at 0°) with only very shght decomposition. The presence of molecular XeF2 in the solution has been established by ultraviolet spectroscopy!'*! and electrical conductancei46. The difluoride may be recovered by extraction or fractional distillationi^i. Clearly, X e F 2 has remarkable kinetic stability since

XeF4, IF5, XeOF4

CCI4

Δ G° (XeF2(.o + H2O -> X e ( „ + i 0 2 ( . , + 2HF(..,))

has been estimated^^ to be —53.4 kcal mole-i, from which ATea = [Xe][O2]nHF]2/pCeF2][H2O]«[HF]2/pCeF2]«1040

Neutral or acid solutions decompose rather slowly, the half-hfe being ' ~ 7 h r at 0° i^i. So far, XeF2 is the only established xenon(II) aqueous solution species. Decomposition in basic solution is very fast, the base catalytic effect being roughly in the order of base strength. The hydrolysis products are xenon, oxygen, fluoride ions and hydrogen peroxide. E v i d e n t l y i 4 7 ^ in 0.01 μ perchloric acid, X e F 2 oxidation of water proceeds with a firstorder rate constant of 4.2 χ 10-4 sec-i at 25° with an activation energy of 19.6 kcal mole"i and = - 8 . 1 eV. Two independent studies of the kinetics of X e F 2 hydrolysis, in water alone, have been reported. Fehér and Lörine studiedi-*» the reaction at 0° and 25°

A'S'act

146 E. H. Appelman, Inorg. Chem. 6 (1967) 2168. 147 E. H. Appelman, Inorg. Chem. 6 (1967) 1305. 1 4 8 1 . Fehér and M . Lörine, Inorg. Kem. Folz. 7 4 (1968) 232.

264

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY

a n d found the reaction to be first order, their rate constants being 2.83+0.02 χ 10*5 sec-i at 0° a n d 2.52+0.01 χ 10-4 at 25°. Legasov et alM similarly found first-order kinetics with = 1.2 x 1012 exp ( - - 1 8 , 4 0 0 / Ä r ) m i n - i , for the temperature range 10-40°. From this study the activation energy is 18.4+2.1 kcal mole-i. Since the first bond dissociation of XeFa is known to exceed the second, the latter authors concluded t h a t their observed activation energy w a s in keeping with XeF- radical formation in the first stage. Oxygen atom formation,, in aqueous XeFa decomposition is consistent with some of the oxidizing properties of aqueous XeFa. The specific conductance of 4 x . l 0 - 4 o h m - i c m - i , f o u n d i 4 6 for a saturated XeFa solution at 0°, is probably t h a t of hydrogen fluoride formed in the oxidation of water by the XeFa. This finding, in a n y case, is consistent with the spectroscopic findings a n d suggests t h a t at least 9 7 % of the dissolved XeFa is initially present as molecular XeFa. Furthermore, since i ^ F exchangei^^ between XeF2 a n d aqueous H F proceeds only to the extent of ^^0.8% after 2 h r at 0°, it is clear t h a t reversible F " ion addition (e.g. XeF2(.<„ + F ^ τ ± XeFa^"^) or F - ion abstraction (e.g. X e F 2 ( . c i ) + H F ( a , ) τ± XeF^+J + H F 2 ) a r e not significant processes in acid aqueous media. The interaction of XeF2 with water is not only catalysed by base but also by species which have an affinity for fluoride i o n s i 4 9 . Of a variety of metal ions investigated, the order of the accelerating effect is the same as the order of the stability constants of their monofluoro complexes: Th4+ > A13+ > Be2+ > L a 3 + . These findings parallel the oxidizing behaviour of XeF2 in non-aqueous solvents in which the stronger fluoride ion acceptors have greater effect in promoting oxidation by XeF2. These findings suggest t h a t the hydrolysis of XeF2 m a y involve XeF+ formation. It is significant t h a t , in spite of the considerable investigation of the chemistry of X e F 2 , no evidence for a kinetically stable monoxide h a s appeared. It m a y be t h a t the fleeting yellow colour reported by several i n v e s t i g a t o r s 9 4 . i4i to accompany alkaline hydrolysis of XeF2 is associated with the X e F radical, but it is more likely to be an unstable hydroxy species or oxide: OHHzO+XeFa

2HF+Xe(OH)2

or

Χ6θ·>Ή2θ

It is possible t h a t XeO, like its iodine analogue I F , could exist as an unstable polymer. Evidently, if XeO is formed as an intermediate in aqueous hydrolysis, its hfetime must be very short since there is no evidence for higher oxide formation such as might be anticipated from mutual oxidation reduction reactions of the type 2 X e O X e + X e 0 2 . It is of interest t h a t the disproportionation of hypoiodite, I O - + I O I - + I O J , is very slow compared with the iodide catalysed reactioni^o involving an activated complex { I - + 2 I O - + H + } . The xenon counterpart of the latter would be unlikely to exist, since X e would not be retained. Evidently XeOa i n t r o d u c e d i 4 7 into an aqueous XeF2 solution is extensively con­ sumed in the course of the interaction of X e F 2 with water, although an aqueous solution of X e 0 3 itself c a n be kept almost indefinitelyi^i. It h a s been presumed t h a t the XeOa consumption results from reduction by an XeO intermediate: X e O ( a q ) + X e 0 3 ( a q ) 2Xe02; Xe02 X e + 0 2 , but even an oxygen atom could serve as an equally effective reducer [ 0 ] + X e 0 3 -> X e 0 2 + 0 2 . The interaction of XeF2 with water is in effect an oxygen atom 149 M . T. Beck and L. D o z s a , / . Am. Chem. Soc. 89 (1967) 5413. 150 J. Sigalla, / . chim. phys. 5 8 (1961) 602. 151 E. H . Appelman and J. G . Malm, / . Am. Chem. Soc. 8 6 (1964) 2141.

XENON(II)

265

source for brómate oxidation to perbromatei52. This is perhaps the most dramatic example of the oxidizing capability of XeFa since perbromates were previously unknown and had been considered by some to be impossible to synthesize. Aqueous XeF2 solutions also oxidize^^i. i53 chloride to chlorine, iodide and iodate to periodate, Ce(III) to Ce(IV), Cr(III) to Cr(VI), Co(II) to Co(III) and Ag(I) to Ag(II) and the fluoride has even been suggested as an analytical reagent for I " and Cr(III) i53. Alkaline solutions of Xe(VI) are oxidized to Xe(VIII). The oxidation potentials have been estimated for XeF2/Xe in acid solution to be 2.2 V and for XeO?/Xe in alkahne solution to be 1.3 V. A Polarographie s t u d y i 5 4 shows that acidified X e F i solutions are reduced in a single step to xenon at a potential of - Ό V with respect to the HgzSO^Hg reference electrode. Analysis and identification. The analysis of XeFa samples may be carried out conv e n i e n t l y 9 2 by transferring known weights of the material to a small nickel weighing bottle (provided with a valve) containing an excess of degassed mercury. By keeping the bottle warm and the contents agitated, complete reduction of the difluoride occurs within a few hours: Hg+XeF2

HgF2(Hg2F2)+Xe

Xenon production may be determined by gas measurement or by weight loss. The fluorine content is simply given by the increase in weight due to mercury fluoride formation. This method is easier than the earlier one^^ which used hydrogen to reduce the difluoride. Hydrolysis in baseiss^ with gas collection ( X e + i 0 2 ) and mass spectrometric characteriza­ tion, together with acidimetric analysis for the H F formed in the hydrolysis H20+XeF2 -> 2 H F + X e + i 0 2

or iodimetric titration, on the basis of the reaction XeF2(.q) + 2 I - - > X e + 2 F - + l 2

have also been used e f f e c t i v e l y 9 4 . Perhaps the most sensitive test for the presence of X e F 4 and XeF^ in the difluoride sample is the melting point (129.02°) 102. The distinctive infrared absorptions of X e F i (555 cm-i), X e F 4 (590 cm-i) and XeFa (broad band, 530-610 cm-i) serve to detect the presence of a few per cent of each of the fluorides in a sample of any one—down to 1% of X e F 4 in XeFa can be detected. The advent of laser Raman spectroscopy has now made identification or detection of very small quantities or concentrations of XeF2 rather easy. A dry glass container is satisfactory for sample holding, and the band due to symmetric XeF2 stretch (vi) at 497 cm-i is extremely strong and well removed from bands attributable to the other xenon compounds. An X-ray powder photograph will readily confirm whether the bulk of a solid sample is or is not XeF2—a 10 % abundance could well be missed however. 3.2.2. Xenon Dichloride and Dibromide Synthesis. The synthesis of xenon dichloride was first claimed by M e i n e r t i 5 6 who subjected a 1:1:1 mixture of Xe, F2 and SÍCI4 or CCI4 to a high-frequency discharge (25 MHz, 150-350 mA), at - 8 0 ° . This yielded colourless crystals which decomposed at 152 E. H. Appelman, / . Am. Chem. Soc. 90 (1968) 1900. 153 A . Schineer Erdeyne and K. Kozumlza, Inorg. Kem. Folz. 7 5 (1969) 378. 154 B. Jaselskis, Science 146 (1964) 263. 155 See ref. 14, p. 167. 156 H. Meinert, Z. Chem. 6 (1966) 71.

266

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY

^^+80°. Mass spectroscopy of the reaction product gave a strong XeCl+ spectrum. The compound was not characterized further. Presumably interaction of F2 with the tetra­ chloride (of Si of C) generates chlorine atoms, essential for X e C b formation. The compound has also been prepared, using the matrix isolation technique, by Nelson and P i m e n t e l i 5 7 . in this preparation, a Xe:Cl2 mixture of 200-100:1 was passed through a microwave discharge (2450 Mc, R K 5609, Raytheon Corpn.) and then condensed upon a cesium iodide optical window, maintained at or close to 20 °K. Infrared spectra were recorded in the range 400-200 c m - i . An absorption centred at 313 cm-i was shown to be due to XeCb. Xenon dichloride (and a number of other xenon compounds) have been detectedi^s by Mössbauer spectroscopy, as products of the β decay of their 1 2 9 1 relatives: 129IC12-

-β

> 129XeCl2

Although of value in the study of structure and bonding, the last technique does not, of course, lend itself to the preparation of macroscopic quantities of the xenon compounds. Thermodynamic features. As discussed in section 1.2.3, the greater bond energy of CI2 relative to F2 and the lower thermochemical bond energy of chlorides relative to fluorides, together indicate that xenon chlorides should be thermodynamically unstable (see Table 4, p. 219). The failure to prepare the xenon chlorides from the fluorides in metathetical reactions, e.g. -70« X e F 2 + 2 H C l (or BCI3)

^ X e + a 2 + 2 H F (or BF„Cl3_„)i35

and the evident instability of the dichloride, show that this is so. Structure and bonding. The infrared absorption at 313 c m - i , observed in the spectrum of the matrix isolated material at 20°K, has been convincingly attributed to the V 3 (asym­ metric Xe-Cl stretch) mode of XeCl2. Since no other absorption (attributable to the symmetric stretch, vi) was observed, the molecular symmetry is evidently D^H (the bending mode V2 is expected to be < 2 0 0 c m - i , i.e. below the hmit of detection in this study). The asymmetric stretching force constant K-Kr, given in Table 19, shows that the Xe-Cl bond is weak compared with the difluoride X e - F and K r - F bonds. The Mössbauer effect is uniquely suited to study the process of xenon compound forma­ tion by β decay of 1 2 9 1 compounds. The new molecules are formed one at a time in the decay and each molecule signals its formation, and the details of its structure, through its contribution to the hyperfine spectrum of the y radiation. This radiation is emitted from the i 2 9 X e nucleus, the 39.6 keV first excited state of which is populated in the β decay of 1291. This state decays with a mean life τ = 1.46+0.06 χ 1 0 - 9 sec, usually by internal conversion, but by y emission in 8 % of the cases. The excited i 2 9 X e nuclear state (spin and parity | + ) has a quadrupole moment eQ, which results in a doublet absorption spectrum if the electric field gradient eq ( = Θ^Κ/θ^ζ) does not vanish. Therefore less symmetrical hgand arrangements (i.e. non-spherical, non-O^ or ηοη-Γ^) give rise to a resonance sphtting. The splitting is proportional to qQ. On the basis of a cahbration against spectroscopic and atomic-beam studies of the quadrupole coupling caused by a single hole in the 5p electron shell of i 3 i X e , and comparison of i 2 9 X e and i 3 i X e Mössbauer sphttings in X e F 4 i s » , a quadrupole splitting of 27.3 mm sec-i is assigned to the loss of one 5pz electron (symmetry 157 L. Y. Nelson and G. C. Pimentel, Inorg. Chem. 6 (1967) 1758. 158 G. J. Perlow, and M. R. Perlow, / . Chem. Phys. 48 (1968) 955.

267

XENON(II)

axis ζ) or a pair of 5p^ or 5py electrons. This interpretation assumes that the only xenon orbitals involved in the bonding are the 5^ and 5p, the latter having prime importance in forming σ molecular orbitals.

TABLE 1 9 . COMPARISON OF X e C h W I T H O T H E R X e D I H A L I D E S A N D KrF2

I (mdynesA-1) t Mössbauer Γ Splitting (mm s e c - i ) data < e2qQ ( M H z ) (i29Xe) (^Electron transfer per bondi

XeF2

XeCl2

XeBr2

KrF2

5472.60»»

313 1.317»»

—

580« 2.59»»

39.0 ± 0 . 1 2490 0.72

28.2 ± 0 . 1 1800 0.52

22.2 ± 0 . 4 1415 0.41

960±30· 0.5·

t See also Fig. 4 . t The electron transfer per b o n d , from the noble-gas a t o m t o each ligand, is derived o n the supposition that outer orbitals of the noble-gas atom, e.g. X e 5d, are not involved in the bonding, and that the bonding is primarily ρσ. ' J. J. Turner and G. C. Pimentel, Preparation o f inert-gas c o m p o u n d s by matrix isolation: krypton difluoride, in Noble Gas Compounds (H. H . Hyman, ed.). University o f Chicago Press, Chicago and L o n d o n (1963), p. 101. »»L. Y . Nelson and G. C. Pimentel, Inorg. Chem. 6 (1967) 1758. ^ G. J. Perlow and H . Yoshida, / . Chem. Phys. 4 9 (1968) 1474. G. J. Perlow and M . R. Perlow, / . Chem. Phys. 4 8 (1968) 955. " S. L. Ruby and H . Selig, Phys. Revs. 147 (1966) 348.

The observed data for X e C b are given in Table 19. It is clear, from the comparison with the XeF2 and XeBri data, that the bond polarity decreases in the sequence XeFa > XeCla > XeBri. This is in harmony with the decrease in electro-negativity of the ligands F Br. It is doubtful, however, if any of the bonds are as polar as the figures given in Table 19 indicate (see section 2.2.1). The figures probably represent upper hmits for bond polarity. If there is significant Xe 5d orbital participation in the bonding as Mitchell asserts25% this must have appreciable influence on the electric field gradient, in which case the sphtting calibration just referred to will not be valid. Other properties. Nothing is known of the long-term stability of the dichloride and heavier hahdes, n o r the highest temperature at which they may be safely stored. N o r is anything known of the reaction chemistry of X e C b , although it is clear from thermo­ dynamic considerations that it is potentially a powerful oxidizer and chlorinator. Attempts to make XeCl+ salts have f a i l e d i 3 5 . Xenon dibromide. The dibromide of xenon has been detected by Mössbauer spectroscopy as a product of the β decay of its i 2 9 i relative: IBr¡—lxeBr2

It is to be expected that XeBri will be much less stable than the chloride. Nothing further is known about the compound.

268

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY

3.2.3. Xenon(n) Oxide and Hydroxide It is of interest that X e O ( g ) was detected s p e c t r o s c o p i c a l l y i 5 9 as a bound gas phase species prior to 1962. The bond energy for this species is given as 9 kcal mole-i (see Table 4 and section 1.2.3). Xenon(II) oxide has not been isolated in the condensed phase, however— not even by matrix isolation techniques. 1 5 9 a There are indications that the hydrolysis of X e F 2 gives rise to xenon(II) oxide or hydroxide, this being a plausible explanation for the fleeting yellow colour observed, occasionally, to accompany the X e F 2 decomposition (see section 3.2.1, aqueous X e F 2 chemistry). Since 1 0 " is a kinetically rather stable speciesiso, similar stability would have been expected for XeO. It is probable that monomeric XeO would be much less favourable energetically than the oxygen bridged polymer. (It is conceivable that the brown solid reported to be (IF)„ by Schmeisser and Scharfi^o is a fluorine bridge polymer.) Polymeric XeO would pre­ sumably be a helical polymer with linear or near linear - ( 0 - X e - O ) - groups and a non­ linear X e - O - X e bond. The fleeting yellow product of X e F 2 hydrolysis could be such a polymer. Of course, (XeO),, is anticipated to be thermodynamically unstable. 3.2.4. Xenon(n) Fluoride Fluorosulphate and Related Compounds Although, so far, it has not proved possible to derive xenon(II) oxide or xenon dichloride by metathesis from the difluoride, several ligands have been successfully substi­ tuted for fluoride. One or both of the fluorine ligands of X e F 2 may be substituted. Com­ pounds of this type were first reported independently by Bartlett and Sladkys^ and by M u s h e r i 3 3 . i n the monosubstitution derivativess^. 1 3 1 , 1 3 3 ^ the known ligands are - O S O 2 F , - O C I O 3 , -OTeFs and - 0 C ( 0 ) C F 3 i34. i t is probable that other highly electro-negative hgands (e.g. - O C F 3 , - O S F 5 and even - S F 5 ) will also prove to be effective. General synthetic procedure. The preparation of the monosubstituted X e F 2 derivatives has generally involved interaction of the fluoride with the appropriate anhydrous acid: XeF2+HL^FX-L+HF

The considerable exothermicity of Δ ^ ; (HF) ( = - 6 4 . 9 2 kcal mole-i) i9 provides the main driving force for these reactions. In the cases of the fluorosulphate and perchlor a t e 8 6 . 1 3 5 , the difluoride is treated with an equimolar amount of acid at temperatures of — 75° or lower and the hydrogen fluoride generated is removed as thoroughly as possible at the lowest possible temperature. The Perchlorate and trifluoroacetate are kinetically unstable and detonate easily. The pentafluoro tellurate is the most stable^^i of the established compounds. General structural features. The available structural data show that the linear threecentre atomic feature of X e F 2 is maintained when one of the fluorine hgands is substituted. So far, all effective ligands are highly electro-negative - O R groups. The X e - O bond is longer (and presumably weaker) than the X e - F bond. These features are exemplified by the structure of the fluorosulphate^ö which is shown in Fig. 11. Evidently the X e - F bond is shorter than in X e F 2 itself, and it is as though the bonding were tending to XeF+SOsF(see section 3.2.6), i.e. the resonance form F-Xe+SOsF" has greater weight than 159 c . D . Cooper, G. C. Cobb and E. L. Tolnas, / . Mol, Spectrosc. 7 (1961) 2 2 3 . 159« T h e observed bond energy may be for X e O X e + i D (O) rather than X e + 3P (O). 160 M . Schmeisser and E. Scharf, Angew.

Chem. 7 2 (1962) 324.

269

XENON(II)

- F X e + - 0 S 0 2 F . The similarity of the X e - F features of the vibrational spectra shown in Table 20 indicate that the X e - F bond in all of the known FXeL compounds is similar to that in FXeOSOzF. F(l)

FIG. 11. The molecular structure of FXeOSOzF. Precision of bond lengths is ca. 0.01 A (un­ corrected for thermal motion). The angles F ( l ) - X e - 0 ( 1 ) and X e - 0 ( 1 ) - S are 177.5 ± 0 . 4 ° and 123.4 ± 0 6 ° .

Stability and reactions. The fluorosulphate and Perchlorate are thermodynamically unstable and the latter is dangerously explosive. The fluorosulphate has been kept for many weeks at 0 ° but has a half-life of only a few days at ^ 2 0 ° and decomposes smoothly when molten according to the equations^ 2 F X e S 0 3 F -> X e F 2 + X e + S 2 0 6 F 2

Presumably this involves radical formation, FXeSOaF-> FXe + SOaF, and the yellowgreen colour gives some support to this. The equilibrium 2 S O 3 F τ± SaOeFz is well knowni^i and, as has been discussed under XeF- (section 3 . 1 . 1 ) , all evidence indicates that the radical disproportionates: 2XeF -> Xe + XeFi. So far it has not proved possible to obtain ( € 1 0 4 ) 2 from the Perchlorate, the decomposition usually being complex (yielding mainly C I O 3 , some C I 2 O 7 , O2, Xe and CIO2) and occasionally proceeding with explosive v i o l e n c e i 3 5 . In contrast, the pentafluoro-orthotellurate is thermally quite stable (to 1 2 0 ° in perfluorinated vessels) and can b e i 3 i distilled (hquid) unchanged at 53° (0.01 mmHg). These compounds have considerable potential as oxidizers in reactions in which the substituted ligand is transferred to a more electro-positive centre, but little has, so far, been reported on these aspects. Interaction of the FXeOR compounds with one mole of the anhydrous acid H O R generates the bis compounds^. 132,134 (see section 3.2.5) F X e O R + H O R -> Xe(OR)2 161 p . M. Nutkowitz and G. Vincow, / . Am. Chem. Soc. 91 (1969) 5956.

270

TABLE 20. COMPARISON OF FXeOS02F AND RELATED COMPOUNDS

FXeOSOiF · Colourless

Unit cell

Solvents CH3CN ii>F n.m.r. data <^F(ppm)

36.6 Sublimable 1 week

16.5 -24 (Detonates) 53 (0.01 torr) Rapid decomposition oo of melt a = 9.88, b = 10.00, Powder patterns show — c = 10.13 A (all ±0.01 A), that FXeOClOj is ζ = 8, space group similar to FXeOS02F = ptca · but not isomorphous * with it • CH3CN CH3CN, CCI4 — —

(6cF3CooH==0)

— —

(Detonates) (Detonates) 10 hr _

37(F(Te)) 66.3(F(Xe))

— (inCHjCNsoln.)

Infrared and Raman bands (cm-i) and assignments FXeOSOiF ·· -


FXeOClOs«'

FXeOTeFs»

FXeOCOCFj (in CH3CN) ^

f {

253 s R

ż(F-Xe-O)

202 m R

I

243 m R

ż(Xe-O-Cl)

258 s

395 mw R

r

505 vsIR

L

525msR

FXeOCOCFs ^

NOBLE-GAS CHEMISTRYNEIL BARTLETT AND F. O. SLADKY

Melting point CC) Boiling point (°C) Half-life at '^20°C

FXeOClOs" FXeOTeFj Very pale yellow Pale yellow ?

Colourless

Infrared and Raman bands (cm-i) and assignments (Cont.) FXeOSOaF · "

'

F

X

e

O

C

l

O

a "·

FXeOTeFs«

FXeOCOCFa (in CH3CN) ^

531 m R 536 m R 540 s R

f 586 SIR 593 w R ^g}vsIR

614 m I R 616 m w R

ż(C103)+a(C103) J +v(a-0-) Ί

638 mw R 722VSIR 726 w R

^ f 623 s ^ ^.JTIJYÍ I 704vsUR +vCre-0) ^ ^ J

758 sh IR

/ 970 vs IR ^ 970 w R

u^—n\^^^ / KS=0)sym I

1197wR 1210 vsIR

{

f 1014 mw R 1018 vsIR v(Cl=0)sym { 1032 vw R I 1048 sh IR

χενον(π)

ri ^ Hb-U-)

KC-F)

970-1220 IR

f 1202 mw R 1215 vs IR V(C1=0) asym < 1243 vw R L 1295 w IR

1390 w R 1393 SIR

V(C=0)

1730 IR

271

• N. Bartlett, M. Wechsberg, F. O. Sladky, P. A. BuUiner, G. R. Jones and R. D. Burbank, Chem. Commurts. (1969) 703; N. Bartlett and F. O. Sladky, The Second European Fluorine Chemistry Symposium, Göttingen, 28-31 August 1968. * F. O. Sladky, Angew. Chem,, Int. Edn., 8 (1969) 373. * M. Eisenberg and D. D. DesMarteau, /. Inorg, NucL Chem. Letters 6 (1970) 29. " P. A. BuUiner, Ph.D. thesis, Princeton University (1970). • F. O. Sladky, unpublished observation.

272

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY

Attempts to prepare mixed compounds (e.g. by interaction of have so far failed^^a: FXeOTeFs+ HOSO2F

HSO3F

with FXeOTeFs)

H F , S2O6F2, FjTeOTeFs, X e , O2

It is possible that the F ligand in the F X e - O R compounds can be donated to strong fluoride ion acceptors (like AsFs and SbFs): F X e - O R + A s F s -> [XeORJ+EAsFé]?

since FXeOTeFs forms a 1:1 complex with AsFs i3i of which prehminary studies indicate the salt f o r m u l a t i o n i 6 2 . The interaction of FXeOTeFs with CsF i 3 i , FXeOTeFc + r^F ^ TeF6+(FXeO-Cs+) ^ ^ ^ " ^ ^ ^ ^ ^ ^ ' ^ - > X e F 2 + CsOTeF5

CsF+Xe+i02

raises the possibility of isolating F X e O " salts by careful control of reactions of this kind. Much can probably be achieved, in exploring the chemistry of these compounds, by exploiting the solvent properties of C H 3 C N (which is oxidatively rather inert). 3.2.5. Xenon(n) bis Fluorosulphate and Related Compounds Stability and reactions. The disubstituted derivatives of X e F i are less well characterized than their monosubstituted relatives. The compounds are obtained^^. 132,134 by treating X e F 2 with two moles of the anhydrous acid or by allowing the FXeOR relative to interact with one mole of acid. As discussed in section 3.2.4, attempts to prepare unsymmetrical R O - X e - O R compounds have failed. The bis fluorosulphate decomposes more readily than F X e O S 0 2 F : Xe(OS02F)2

spontaneous > Xe+(S03F)2

and the meagre available evidence indicates that this lower stability of the bis compounds will prove to be general. The decomposition of the fluorosulphate provides very pure (S03F)2 but, unfortunately, this mode of decomposition is not general and the perchlorate^ö, o r t h o t e l l u r a t e i 3 2 and t r i f l u o r o a c e t a t e i 3 4 decompose as indicated (the first and last being dangerously explosive materials): Xe(C104)2

> Xe+(C103),+02 (+some

ChOy)

120" Xe(OTeF5)2 ^ X e 4 - F 5 T e - 0 - T e F 5 + 0 2 ( + s o m e TeFö) -20« Xe(OCOCF3)2

> Xe+C2F6+2C02 i i 17hr.

Physical data on the bis compounds is given in Table 21. Fluorosulphonic acid will displace the perfluoro-orthotelluric a c i d i 6 2 : X e ( O T e F 5 ) 2 + H O S 0 2 F -> X e ( O S 0 2 F ) 2 + 2 H O T e F 5

but attempts to displace the telluric acid with

H O O C C F 3 102

have led to decomposition:

-20« Xe(OTeF5)2+2HOOCCF3 162 F . O. Sladky, unpublished information.

^ Xe+2HOTeF5 + 2C02+C2F6

TABLE 21. PHYSICAL PROPERTIES OF Xe(OS02F)2 AND RELATED COMPOUNDS Xe(OS02F)2 · Pale yellow

Melting point (°C) Solvents

CH3CN

Unit cell

43-45 (decomp.) CH3C:N

a = 7.82 ^ b = 13.5, c = 6.78 A, β = 96°, Γ = 4 See Table 22

—

35-37 CH3CN,

—

Xe(OTeF5)2' Colourless Pale yellow

Xe(OCOCF3)2'

—

CCI4

CH3CN (explosive with C2H5OH, acetone, benzene)

α = 15.6 \ b = 8.7, c = 12.9 A, ζ = 4 Space group Cmca IR R IR 780 w 705 vs 701 w 628 vs 692 m 647 m v(Te-F)+ y(Te-0-)

-

510 KXe-O

XENON(II)

Vibrational spectra (cm-i)

Xe(OC103)2' Pale yellow

970-1240 475 m

KC-F) 1730 v(C=0)

N. Bartlett, M. Wechsberg, F. O. Sladky, P. A. Bulliner, G. R. Jones and R. D. Burbank, Chem. Communs. (1969) 703. M. Wechsberg and N. Bartlett, to be published. F. O. Sladky, Angew. Chem., Int. Edn., 8 (1969) 523. F. O. Sladky, unpublished information. M. Eisenberg and D. D. DesMarteau, Inorg. Nucl. Chem. Letters 6 (1970) 29.

273

• • « ^ •

434 m KXe-O-) 300 w 320 w 237 m 178 vw 131 vs ^O-Xe-O)?

274

592 _^_b

1287 ^ n97w

FXeOSO^F 1390W

y.in^n Xe(OS02F)2

1082 119^^ 970w

I497mw

/ 959mw 946mw I

/

\ 823w 815w \

I

1251vs

^

'

970^ 800w 616mw 536m 540s ^ ^^^^^ ^^^^ ^^Ws

/ I425w 1238mw j^j^^ J2i9mw /

FO,SOOSO,F JoF^ ^^i^^Fz)

786

433s 395mw 253s I 243m \ +v(Xe-0) 436s 386mw 257^5 / ' I

I \

880m 824s ^(S-O)

I 601s 54Iw v(Xe-O)?

I

^(Xe-O-S)

/ 527mw 485mw

I

.

/

1 798vs 598mw v(O-O) bridge

566

409

392mw

299s ?

w = weak, m = medium, s = strong, ν = very, ν = stretching, δ = deformation, py, = wagging. • P. A. BuHiner, Ph.D. thesis, Princeton University (1970). K. Nakamoto, Infrared Spectra of Inorganic and Coordination Compounds, Wiley, New York (1963).

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY

TABLE 22. RAMAN FREQUENCIES AND ASSIGNMENTS FOR FXeOSOzF, Xe(OS02F)2 AND S2O6F2'

XENON(II)

275

Structural features. The available structural evidence, which is primarily infrared and Raman data, suggest that the xenon atom is symmetrically placed between the two O R groups. The Raman data for Xe(OS02F)2 and related SO3F compounds given in Table 22 supports the R - O - X e - O - R formulation but suggests that the molecule is not centrosymmetric. 3·2·6. XeF2 a s a Fluoride Ion Donor Compounds of composition XeF2, 2MF5 (where Μ is a quinquevalent metal atom) have been known since 1963 i4, i63, but several years passed before they were shown to be salts containing the XeF+ i64, i65,87 ion. Indeed, strong fluoride ion acceptors, such as AsFs and the metal pentafluorides, readily withdraw a fluoride ion from XeF2 and three classes of salt have been e s t a b h s h e d 8 7 : XeF+IMFö]" (XeF2, M F 5 ) , X e F + [ M 2 F n ] - (XeF2, 2MF5) and X e 2 F i [MFö]"" ( 2 X e F 2 , M F 5 ) . It is quite possible that the range of salts is more extensive than this since the phase study by Maslov et αΙΛ^^ of the X e F 2 - S b F 5 system established the compounds XeF2, S b F s ; XeFz, 1 . 5 S b F 5 ; XeF2, 2 S b F 5 (provedi^s to be XeF+[Sb2Fii]-) and XeF2, 6SbF5.

Laboratory preparation. It is possible to prepare the XeF2 salts by simply fusing X e F 2 / M F 5 mixturesi65.166, i67, but the fluoroarsenates cannot be made in this way. It is better to prepare the compounds from a solvent, and bromine pentafluoride (b.p. 41.3°) has proved to be excellent for this p u r p o s e 8 7 . Typically, the difluoride and appropriate pentafluoride are weighed (by diffierence) in the desired molar ratio, into a Kel-F tube which is then attached to a vacuum line. Bromine pentafluoride is distilled on to the X e F 2 / M F 5 mixture, which is allowed to dissolve at room temperature. By removing the BrFs slowly, well-crystalhne homogeneous material may be obtained. Structural features. The m o l e c u l a r s t r u c t u r e i 6 5 of X e F + [ S b 2 F i i ] - is represented in Fig. 1 2 . The structure estabhshes the existence of a short-bonded X e F species, the properties

F I G . 12. T h e molecular structure o f X e F + [ S b 2 F i i ] - . 163 A . J. Edwards, J. H . Holloway and R. D . Peacock, Proc. Chem. Soc. (1963) 275. 1 w F . O. Sladky, P. A . Bulliner, N . Bartlett, B. G. D e B o e r and A . Zalkin, Chem. Communs. (1968) 1048. 165 V. M . McRae, R. D . Peacock and D . R. Russell, Chem. Communs. (1969) 62. 166 O. D . Maslov, V. A . Legasov, V. N . Prusakov and B. B. Chaivanov, Zh. Fiz. Khim. 4 1 (1967) 1832. 167 J. H . Holloway and J. G. Knowles, / . Chem. Soc. A (1969) 756.

276

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY

of which, as may be seen from the comparison with I F molecule in Table 1 (p. 215), are consistent with its designation as XeF+. It is clear from the Raman spectra87 of the compounds X e F 2 , 2MF5 that they are structurally similar to X e F + [ S b 2 F i i ] - . All are characterized by strong bands in the 600-621 cm-i region assigned to the XeF+ stretch. The rather short Xe . . . F distance of 2.35 Ä between the cation and anion in the antimony salt indicates a small Van der Waals radius for the positively charged xenon atom. The Raman data for the X e F 2 , M F 5 compounds show the formulation XeF+iMFg]" to be appropriate, the vibrational characteristic of the XeF+ species being very Hke that of the X e F + [ M 2 F i i ] - . N o t only is there evidence of a weak interaction of the XeF+ ion with the M F 5 but some interaction between the XeF+ ions is also indicated. It should be noted that I-Cl forms a chain polymerias but evidently i«»* the XeF+ species does not polymerize (at least not in these salts).

2 14(003)Ä 1-90 (0 03)Ä

F i o . 13. A cx)mparison o f the Xe2F^, H 2 F ¡ and I5 ions. The c r y s t a l s t r u c t u r e o f 2XeF2, AsFs appropriate.

provesi64 t h e f o r m u l a t i o n X e 2 F í [ A s F 6 ] ~

to be

The c a t i o n i s p l a n a r a n d V - s h a p e d a n d i s r e p r e s e n t e d i n Fig. 13. The Raman

s p e c t r a o f t h i s s a l t a n d t h e o t h e r 2XeF2, M F 5 c o m p o u n d s e s t a b h s h t h a t t h e l a t t e r a l s o contain the Xe2FJion87.

This i o n h a s a s i m i l a r s h a p e t o t h a t o b s e r v e d i n t h e H2FJ a n d I j

168 R. W . G . Wyckoff, Crystal Structures, Vol. I, Interscience, N e w York (1963), p . 174. 168« N . Bartlett, D . Gibler, M . Gennis and A . Zalkin, t o b e published.

XENON(II)

277

i o n s . The n e a r h n e a r i t y of t h e FXe . . . F a s s e m b h e s i n d i c a t e s t h a t t h e c a t i o n r e s e m b l e s t w o XeF2 m o l e c u l e s s h a r i n g a c o m m o n however, that the nificance that the

five-centre

fluorine

a t o m . The

system is tending t o

b o n d length disparity

shows,

F-Xe+F-Xe+-F. It i s p e r h a p s o f s i g ­

XeiFJ s a l t s s h o w h t t l e e v i d e n c e o f i n t e r a c t i o n o f t h e c a t i o n w i t h t h e a n i o n XeF+ i o n s a r e a d e q u a t e l y " n e u t r a l i z e d " b y t h e b r i d g i n g fluoride

—highly polarizing hgand.

Reactions, The fluoride i o n d o n o r a b i l i t y o f X e F a i s l e s s t h a n t h a t o f n i t r o s y l and the available thermochemical data indicate that

(XeF2(,)

XeF^^^ + F =

fluoride + 2158

k c a l m o l e ~ i ( s e e s e c t i o n 3.2.2 a n d Fig. 9). The i n c r e a s e i n t h e X e - F b o n d e n e r g y (48 k c a l m o l e - i i n t h e c a t i o n , 32 i n X e F 2 ) i n c a t i o n f o r m a t i o n c o n t r i b u t e s t o

fluoride

ion donor

a b i l i t y o f t h e d i f l u o r i d e . It i s a b e t t e r F~ d o n o r t h a n X e F 4 b u t i n f e r i o r t o X e F ö i^^.

An i m p o r t a n t a s p e c t o f t h e XeF+ s p e c i e s i s i t s h i g h e l e c t r o n aflSnity. Since (XeF+) = BE (XeF+)~-BE ( X e - F > - / (Xe) ( s e e Fig. 9), t h e e l e c t r o n affinity m u s t b e i n t h e r a n g e 10-11 eV. This v a l u e i s c o n s i s t e n t w i t h t h e d e c o m p o s i t i o n o f t h e s a l t XeF+iOsFö]" t o OsFg a n d x e n o n : Ε

3XeF+[OsF6]--^ [ X e z F a l + i O s F o l ' + l O s F o + X e In t h e [ X e 2 F 3 ] + i o n , e a c h X e F + s p e c i e s r e c e i v e s e l e c t r o n d e n s i t y f r o m t h e a s s o c i a t e d

F - i o n ( f o r w h i c h / ( F - ) = 83 k c a l m o l e - i ) . Thus t h e a s s e m b l y b e c o m e s a m u l t i c e n t r e b o n d e d s y s t e m w i t h a l o w e r e l e c t r o n affinity t h a n X e F + . and

So f a r t h e s a l t s h a v e n o t b e e n u s e d a s r e a g e n t s , b u t t h e a c i d c a t a l y s i s o f XeFa o x i d a t i o n fluorination r e a c t i o n s , r e f e r r e d t o i n s e c t i o n 3.2.2, i m p l i e s t h a t t h e XeF+ o r Xe2F^

i o n s are t h e effective oxidizers.

The k n o w n s a l t s a r e b r i e f l y d e s c r i b e d i n Table 23. 3.2.7. Molecular Adducts of XeF2 Shortly a f t e r t h e p r e p a r a t i o n o f t h e first x e n o n c o m p o u n d s a

fluoride

of xenon was

isolated w h i c h w a s initially t h o u g h t t o b e a s e c o n d crystalhne modification o f X e F 4 .

A

t h r e e - d i m e n s i o n a l X - r a y s t r u c t u r a l analysis^^o q u i c k l y s h o w e d , h o w e v e r , t h a t t h e c o m p o u n d w a s a 1:1 m o l e c u l a r a d d u c t o f X e F 2 a n d X e F 4 . The m o l e c u l a r d i m e n s i o n s o f t h e m o l e c u l e s are n o t significantly different f r o m t h o s e o f t h e p u r e c o m p o n e n t s ( s e e

Table 28). Several

o t h e r a d d u c t s o f XeF2 h a v e s i n c e b e e n p r e p a r e d i n w h i c h t h e XeF2 m o l e c u l e i s e s s e n t i a l l y indistinguishable from t h e molecule in crystalline X e F 2 .

Estabhshed 1:1 c o m p o u n d s a r e

X e F 2 , X e F 4 i 7 0 ; XeF2, I F 5 I 8 ; X e F 2 , X e O F 4 i 3 5 a n d X e F 2 P C e F 5 ] + [ A s F 6 ] " - 1 3 5 ; a n d t h e

1:2 c o m p o u n d s X e F 2 , 2IF5142, XePj, 2 ( [ X e F 5 ] + ^ [ A s F 6 ] - ) Table 24.

have been reported.

Some

physical properties o f the c o m p o u n d s are given i n

Preparation,

The a d d u c t s m a y b e v e r y c o n v e n i e n t l y p r e p a r e d b y m i x i n g t h e c o m -

p o n e n t s i 3 5 , 9 8 , n i i n t h e a p p r o p r i a t e m o l a r p r o p o r t i o n s . By f u s i n g t h e m i x t u r e o r d i s s o l v i n g it i n a suitable s o l v e n t ( e . g . acetronitrile)i42 a h o m o g e n e o u s s a m p l e i s p r o d u c e d .

The

1:1 XeF2, IF5 a d d u c t m a y a l s o b e m a d e ^ s b y h e a t i n g a x e n o n / I F ? m i x t u r e t o 200°: 200« X e + I F 7 — » ^ X e F 2 , IF5

Structure and bonding. Attention h a s a l r e a d y b e e n d r a w n t o t h e s e m i - i o n i c n a t u r e o f 1.3.3 a n d 3.2.2), a n d t h e e x c e l l e n t a g r e e m e n t , w i t h t h e

t h e b o n d i n g i n XeF2 ( s e e s e c t i o n s

N . Bartlett and F. O. Sladky, / . Am, Chem, Soc, 9 0 (1968) 5316. 170 J. H . B u m s , R. D . Ellison and H . A . Levy, Acta Cryst, 18 (1965) 11. 171 P. AUamagny, M . Langignard and P. Dognin, CR, Acad, Set, Paris C (1968) 226.

278

TABLE 23. SOME PHYSICAL PROPERTIES OF XeF+[MF6]-, XeF+[M2Fii]-, AND XeiV^lMFé]'

SALTS

XeF+[MF6l- Salts Ru*

Ir'

Pale yellow-green

Pale yellow-green

Melting point CQ — 110-111 Xe-F stretch -614 604,599 (Raman, cm-i) · Unit cell* <

Os

—

Brownf

152-153 608,602

—

Pt*

Ta*

Nb* Pale Orange-red

Yellow-green

82-83 52-53 608,602

30-35 —

Pale yellow

—

isomorphous — • — a « 8.0, b « 11.1, c « 7.3, 5=90.7°, z=4 Space group Plijn

—

t Unstable

XeF+[M2Fii]- Salts

Sb

Ru

Yellow' Melting point (°C) Xe-F stretch 621 (Raman, cm-i)' Unit cell

Ir

Bright green*

63«·* 49-50 * 69-70* 604,598 612,601 -<

Pt Orange-yellow*

— — isomorphous

82-83 * —

Ta

Nb

Dark red*

Pale yellow*

Pale yellow*

42-47* — • <

isomorphous

•

yellow

NOBLE-GAS CHEMISTRY: NEIL BARTLETt AND F. O. SLADKY

Sb'

279

XENON(II) TABLE 2 3 .

(cont.)

X e z F j [MF6-] Salts

As*'*

Ru«

Os*

Ir*

Pale yellow-green

Pale yellow-green

Pale yellow-brown

Pale yellow-green

98-99 · 593,579

593,582

92-93 · 592,578

Melting point ("Q 99-100«·' X e F stretch 600,588 (Raman, c m - i ) · α = 1 5 . 4 4 3 , ¿, = 8.678, Unit cell < c=20.888A,i?=90.13, 2 = 12 Space group 72/a

isoinorphous

• F . O. Sladky, P. A . Bulliner and N . Bartlett, / . Chem, Soc, A (1969) 2179. J. H . HoUoway and J. G . Knowles, / . Chem, Soc, A (1969) 756. « V. M . McRae, R. D . Peacock and D . R. Russell, Chem, Communs, (1969) 62. • O. D . Maslov, V. A . Legasov, V . N . Prusakov and B. B . Chaivanov, Zh, Fiz, Khim, 141 (1967) 1832. • F. O. Sladky, P. A . Bulliner, N . Bartlett, B. G . D e Boer and A . Zalkin, Chem, Communs, (1968) 1048. ' N . Bartlett and R. Mews, unpublished observation.

TABLE 2 4 . SOME PHYSICAL PROPERTIES OF X e F a A D D U C T S

XeF2, XeF4 ·

Melting point (^'Q vi(XeF2)cm-i X e ( I D - F bond length ( A ) t Unit cell

— 2.010 ( σ = 0.006) α = 6.64 A

i5 = 9 2 ° 4 0 ± 5 ' ζ = 2 Space group P 2 i / c

XeF2,IF5* Colourless solids

98 495 2.007 ( σ = 0.009)

XeF2,XeÖF4

29 494

—

::íí^}±o.oiA ζ = 4 Space group /4//n

Z= 4 Space group / 4 / m

t Bond length in XeF2(c) = 2.00 ± 0 . 0 1 A . * J. H . B u m s , R. D . Ellison and H. A . Levy, Acta Cryst. 18 (1965) 11. * G. R. Jones, R. D . Burbank and N . Bartlett, Inorg. Chem, 9 (1970) (in press). * N . Bartlett and M . Wechsberg, Ζ, anorg. v. allgem. Chemie, in press (1971).

experimental enthalpy of vaporization, given by the lattice energy calculation for XeFaccryto, based on coulombic interactions between - * F - X e + i - F - * molecules. The crystal structurei^o of X e F 2 , X e F 4 shows that similar coulombic interactions between the XeFz and X e F 4 molecules are responsible for the ordered close-packed structure. Undoubtedly all of the XeFa molecular complexes owe their existence t o appreciable coulombic interactions between the interacting species. The crystal structure of the 1:1 XeF2, I F 5 compound's, illustrated in Fig. 14, nicely illustrates the nature of the intermolecular interactions. Each molecule is surrounded by

280

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY

A portion o f the structure selected to indicate the basal to basal and apical to apical environment o f IF5 molecules. The "cells" which have been drawn through x e n o n atoms t o assist in spatial perception should not be confused with the unit cell.

Projection o n c axis o f two successive layers o f XeF2 molecules. Iodine atoms in intermediate layer indicated by small light circles. F I G . 14. The crystal structure o f XeF2, IF5.

XENON(II)

281

Interatomic distances (A) and angles (") with estimated standard deviations in parentheses (a) In I F 5 molecule I-F(l) 1.817 (10) 1.862 (10)t I-F(2) 1.873 (5) 1.892 (5)t F ( l ) . . . F(2) 2.395 (9) F ( 2 ) . . . F(2') 2.615 (8) F(l)-I-F(2) 80.9 (0.2) (b) In XeF2 molecule Xe-F(3)

2.007 (9)

2.018 (9)t

(c) Attractive interactions between XeF2 and I F 5 F(3)...I F ( 2 ) . . . Xe"

3.142(7) 3.361 (6)t

F ( 2 ) . . . Xe

3.516(5)

F(2)-F(2"') F(2)-F(3) F(2)-F(3")

2.929 (12) 3.418 (8) 2.901 (7)

(d) Intermolecular F . . . F contacts F(l)... F(2).., F(2)... F(3)...

F(l) F(2") F(3'") F(3'^)

2.621 2.953 3.214 2.961

(19) (8) (8) (14)

t Corrected for thermal motion, assuming that fluorine a t o m rides o n heavy atom. All other distances were uncorrected. t R o m a n numbers refer to atoms at symmetry equivalent positions, where I = (y, - x , z), II = ( 1 / 2 - Λ 1/2-fx, 1 / 2 + 2 ) , III = (Iß-x, \l2-y, 1 / 2 - 2 ) , IV = (\-y, x, 2).

an approximately cubic arrangement of molecules of the other kind. The detailed arrange­ ment is consistent with close packing of the molecules and with electrostatic attraction of the negatively charged fluorine ligands of one molecular species for the positively charged central atom of the other. The attraction of the fluorine hgands of XtFz for the iodine atoms of the I F 5 molecules is particularly important. The disposition of the fluorine ligands in a layer of XeFa molecules is determined by the orientation of the nearest I F 5 molecules as illustrated. Where superimposed I F 5 molecules, in adjacent layers, are base to base, the sandwiched X e F 2 molecules orient to make short I - F contacts. On the other hand, the XeFa molecules are oriented away from the I F 5 molecules where they abut apex to apex. This arrangement suggests that the iodine atom bears an appreciable positive charge which is effectively shielded by fluorine hgands but not by the non-bonding valence electron pair. Presumably, the non-bonding pair is concentrated significantly on the ζ axis rendering the I F 5 molecule pseudo-octahedral. Consequently the centres of the faces of the pseudooctahedron on the "pair" end of the molecule would effectively possess positive charge (the "pair" screening being poor in this direction). The fluorine hgands of the X e F 2 molecules align approximately as this model dictates. The X e - F interatomic distance of 2.007 (σ 0.009) A in the X e F 2 molecules is not sig­ nificantly different from that observed in crystalline X e F 2 (2.01 ±0.01 Ä). Although I F 5 and BrFs are known to be geometrically s i m i l a r 2 o , the latter does not form a stable adduct with X e F 2 . Presumably, this is because the charge on the (more electro-negative) bromine atom is less than on the iodine atom and perhaps also because the central-atom charge screening by the fluorine ligands in BrFs is more effective than in I F 5 . It is of interest that the X E O F 4 adduct with X e F 2 is much less stable than is X e F 2 , I F 5 adduct. This is compatible with the greater screening of the positive charge on the xenon(VI) atom in X E O F 4 compared with the iodine charge screening in I F 5 . In X E O F 4 , the basal fluorine ligands are in the same plane as the xenon atom, i.e. Z . O E - X e - F | , = 91°, whereas in I F 5 , Z F A - I - F * = 80°.

282

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY

The interaction of X e F 2 with the XeFJ^ ion is not surprising in view of the close simi­ larity of XeF5' and I F 5 (see section 3.4.5).

3.3. X E N O N ( I V ) C O M P O U N D S

Xenon tetrafluoride is the only xenon(IV) compound, available in macroscopic quanti­ ties, so far characterized unambiguously. The tetrachloride of xenon has been detected by Mössbauer spectroscopy as the product of the β decay : i ^ P i c i ^ i29XeCl4. 3.3.1. Xenon Tetrafluoride Historical and preparative. Xenon tetrafluoride was first reported^ in 1962 by Claassen, Selig and Malm. They prepared it by heating mixtures of xenon and fluorine, at about 6 atm, in a 1:5 ratio at 400° in a sealed nickel can. These conditions are close to optimum for X e F 4 preparation, and the static synthesis remains the most convenient one for the preparation of gram lots. A hot-tube flow method J 72 is claimed to yield good quaUty tetrafluoride if the Fa/Xe molar ratio is 4:1 and the residence time in the hot zone of a nickel tube, at 300°, is 1 min ΐ7ί>. A flow system, designed for continuous production of the tetrafluoride, using a F2/Xe molar ratio 3:1 and a furnace temperature of 560°, has also been describedi^a. There are indications, however, that considerable care must be exercised if high purity X e F 4 is to be obtained by the flow method. Thus the earlier erroneous report that "xenon tetrafluoride" (prepared by the hot-tube flow method) interacts with SbFs: X e F 4 + 2 S b F s XeFa, 2SbFs!, was undoubtedly due to the sample of supposed tetrafluoride having been largely difluoride. However, even the static method is far from perfect. As Weinstock and his c o w o r k e r s 2 2 have demonstrated, it is not possible to prepare X e F 4 pure (see Fig. 15 and thermodynamic features) by the thermal method; therefore, if high purity material is desired, it is necessary either to resort to chemical purificationi69 or else submit the sample to tedious f r a c t i o n a t i o n i o 2 . An essentially quantitative yield of X e F 4 has been claimed for the electric discharge m e t h o d i 7 4 . A Fa/Xe molar ratio of 2:1 (operating pressure 2-15 mm) in a reaction vessel at - 7 8 ° yielded X e F 4 quantitatively. Like the hot-tube flow method, this lends itself to continuous operation. The tetrafluoride also f o r m s i 7 5 when Fi/Xe mixtures (2.6-2.08:1) are irradiated with Co y-rays or 1.5 MeV electrons without coohng of the sample. When the reaction vessel is cooled well below 0° so as to freeze out any XeFa (the first reaction product), the yield of X e F 4 is very small. Utilization of absorbed energy must be eflScient since initial G values, based on xenon consumption, are in the range 5-15 atoms per 100 eV absorbed. X e F 4 is relatively stable to y radiation with an initial G value of 0.6-1.8 at 45°. An independent study involving 1.6 MeV electrons agrees with the above findings. Laboratory preparation. Xenon tetrafluoride is difficult to separate from the difluoride by physical means (the two have similar vapour-pressure relationships and also form a 1:1 adduct—see section 3.2.7), and to circumvent this it has often been the practice to employ Fi/Xe ratios (2:1) and temperatures of 450-500°, which conditions generate 172 J. H. Holloway and R. D . Peacock, Proc. Chem. Soc. (1962) 389. 173 E. Schumacher and M. Schaefer, Helv. Chim. Acta 47 (1964) 150. 174 A . D . Kirshenbaum, L. V. Streng, A . G. Streng and A . V. Grosse, / . Am. Chem. Soc. 85 (1963) 360. 175 See ref. 14, p. 81. 176 G. K. Lavrovsksys. V. E. Skurat and V. L. Tal'roze, Dokl. Akad. Nauk SSSR 154 (1964) 1160.

283

XENON(IV) COMPOUNDS

minimal XeFa but XePe as a major impurity. The XePg (approximately one-third of the product) is removed by treating the mixture with sodium fluoride, which forms a complex with it at room temperaturesi. A more general chemical puriñcationi^p, which effectively ehminates XeF2 and XeF^ simultaneously, exploits the inferior fluoride ion donor abihty of X e F 4 compared to XeF2 or XeFö. With this puriñcation, conditions which maximize the X e F 4 production may be employed. Such a product, which will contain small but significant quantities of XeF2 and XeFö (in approximately equimolar amounts), is dissolved in bromine pentafluoride and an excess of arsenic pentafluoride is condensed upon the mixture. Since XeF2 and XeFg form involatile salts, and since the solvent and AsFs may be removed quantitatively at 0° or below (at which temperatures the X e F 4 has a very low vapour pressure), the X e F 4 can be isolated by subsequent vacuum subhmation at 20°.

'USu^"

XeF2 XeF4 + AsF5(excess) XeFö.

Xe^FanAsFd-

• XeF4 -AsF,.-BrF,

in BrFs soln.

vacuum

• XeF4 t

20"

XeFj+EAsFo]-

Samples purified in this way melt at the same temperature (117°) as those obtained by repeated fractional subhmation and give identical broad hne i^p n.m.r. spectral^?. It is best to make the crude X e F 4 in a nickel or Monel vessel after the method of Claassen et aL^, but the purification is best carried out in Kel-F vessels. Thermodynamic features, A thorough study of equilibrium conditions in the Xe/F2 system, as a function of temperature^z, has defined the conditions for maximizing the yield of each of the binary fluorides. The tetrafluoride is the most difficult to obtain in

TABLE 25. E q i n u B R i u M CONSTANTS FOR THE X e / F 2 SYSTEM22

Temp. Γ Κ )

Ki0ítF2) J^2(XeF4) ^3(XeF6)

298.15

[1.23xlOi3]t [1.37x1011] [8.2x105]

523.15

573.15

[8.80 xlO^] 1.43x103 0.944

[1.02x104] 1.55x102 0.211

Ki = (XeF2)/(Xe)(F2); K2 = (XeF4)/(XeF2)(F2); K2 t Values in square brackets are calculated.

623.15

673.15

773.15

[1670] 27.2 0,0558

[360] 4.86 0.0182

29.8 0.50 0.0033

(XeF6)/(XeF4)(F2).

high purity, as perusal of the equilibrium constant data in Table 25 and Fig. 15 reveal. It will be appreciated that low temperatures and high fluorine pressures will favour XeFe formation and low F2 pressures and high temperatures favour XeF2 formation. The tetrafluoride cannot, of course, be made pure by the thermal method (see under prepara­ tion). The /\,H^ (XeF4(e)) = — 57.6 kcal mole-i obtained from calorimetry agrees with that from photoionization studies (Table 26) and is therefore the most rehable value. Kinetic studies, involving interaction with N O or N O 2 , have shown that the first bond dissociation energy of X e F 4 is considerably greater than the second. This is similar t o the XeF2 case. Values for Di = 48 and D2 = 15 kcal mole-i have been indicated^^. 177 D . E. Hindermann and W. E. Falconer, / . Chem, Phys, (in press).

284 NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY

«O

uivß '3jnss3Jd uinuqiimbg

vo

uijB 'ajnssajd u m u q i i m b g

-4S H

Temperature,

Temperature,

FIG. 15. Pressure and temperature influence on XeF2, XeF4, and XeFö formation (b) (a) Equilibrium pressures of xenonfluoridesas a function of temperature. Equilibrium pressures of xenonfluoridesas a function of temperature. Initial conditions: 125 mmoles Xe, 275 mmoles F2 per 1000 ml. Initial conditions: 125 mmoles Xe, 1225 mmoles F2 per 1000 ml.

XENON(IV)

285

COMPOUNDS

The tetrafluoride, unhke the xenon(IV) oxide and hydroxide systems, is stable with respect to disproportionation. This is consistent with the enthalpies and entropies of formation:

AH} (kcal m o l e - i ) ASf (cal d e g - i m o l e - i )

XeF2(„

XeF4(.)

XeF6(.,

-28.2 -26.5

-57.6 -61

-82 -96

TABLE 26. SOME PHYSICAL ΡΚΟΡΕΚΉΕ8 OF X e F 4

Colourless crystals, liquid and vapour Triple-point 390.25°K (117.10°C) · Thermodynamic

features

Vapour pressure (solid) (275-390.25°K) (390.15°K) ΔΑ.„5

log

= - 3226.21 - 0.43434 log Γ + 1 2 . 3 0 1 7 3 8 ·

817.97 m m (only point for liquid than XeF4)

XeF4) (XeF2 is

1 4 . 8 ± 0 . 2 kcal m o l e - i -f 1 5 . 3 ± 0 . 2 kcalmole-i»»

Heat capacity: v a p o u r ' (cal d e g - i m o l e - i ) at 298.15°K solid«

21.5 28.334 73.0 ± 2 /75.7±0.4 \75.6±0.4

Entropy: vapour • (cal d e g - i m o l e - i ) at 298.15°K calc. from molecular data

35.0t

solid Heat o f formation Γ(°Κ) 393 298 423 298.15

- 5 7 . 6 kcal m o l e - i « -48' -53±5« -51.5^*

-58'

t Preferred value. t KeF4 sample may not have been pure. Mean thermochemical

bond energy (¿"xe-r)

31.7 kcal m o l e - i (based o n Δ / Γ ° = - 5 1 . 5 , Δ / ί ° (F, g) = 18.8 kcal m o l e - i ) Density (X-ray) (g c m - 3 ) 4.10 ·»; 4 . 0 4 ^ Solubility In anhydrous H F " r (°C)

20 27 40 60

Mole

XeF4

1000 g H F

0.18 0.26 0.44 0.73

Mole H F Mole

XeF4

278 192 114 68

more volatile

286

N O B L E - G A S CHEMISTRY: NEIL B A R T L E T T A N D F. O. S L A D K Y TABLE 26

Diamagnetic

{cont.)

susceptibility

χ mole (c.g.s. units)

- 5 0 . 6 x 10-6 ( - 2 0 ° ) » - 52 ± 3 χ 1 0 - 6 (77-293°Κ) ·" - 5 3 Χ 1 0 - 6 (calculated)"

Ultraviolet absorption spectrum ö/XeF4(g) *

IR, Raman

Wavelength (A)

Half-width (cm-i)

Estd. extinction coefficient (mole-i c m - 2 )

Est. oscillator strength

2280 2580 1840 1325

7000 10,000 11,200

398 160 4.75 X 1 0 3 1.5 x l 0 4

0.009 0.003 0.22 0.80

o/XoF^it) Symmetry

a\g aiu bu bly bu Cu Cu Force k krr Kr

Crystal

cm-i

Fundamentals

VIR V2IR V 3 R V 4 R

V 5 inactive V 6 l R V7

IR

543 291 235 502 221 586 182

vs s w vs vs (calculated)

constants 3.00 mdyn A-i ^ 0.12 0.06

structure X-ray diffraction ^

b

: 5.922 ± 0.003 A

c • : 5.771 ± 0 . 0 0 3 A β

: 99.6±0.r

·· *

Xe-F 1.953 ± 0.002 A ^FXeF 90.0±0.r See Fig. 16 (p. 289)

Monoclinic Pliln

a · : 5.050 ± 0 . 0 0 3 A

Neutron-diffraction

r = 2 Z) = 4.04 g m l - i X e - F = 1.93 A ^F-Xe-F

89.7° (ίτ = 0.9^) 90.3

XENON(IV)

287

COMPOUNDS

TABLE 26

(cont,)

19F /i.m.r. data '

Coupling constant /(Hz)

Chemical shift (ppm; σρ, = 0)

σψ Solid Liquid H F solution

482 445 456

448 452

3836 3860

450

3864

3860

Broad line n.m.r. findings *

XeF4, microcrystalline solid

ÍTABST

(ppm)

σΛ ay σ.

Experiment

Theory

218±5 0±8 261 ± 2 5 or 3 9 4 ± 2 5 § 394 ± 2 5 or 261 ± 2 5

168 -33 -33 570

Rigid lattice second m o m e n t

Exp. r.l.s.m. = 6 . 1 i ± 0 . 1 6 G 2 Total calc. r.l.s.m. = 5.62 G2 (Intermolecular interactions accoimt for 3.87 G2 o f the total calc. r.l.s.m.) Field independent s.m. (300*'K) = 1.83 G2

t The absolute scale values for chemical shifts are referred to the bare i»F nucleus } σχ is the out o f plane shielding (Z)4jkXeF4), is the shielding || X e - F bond, and σ, is the in plane shielding perpendicular to the X e - F bond. § If <7, = 394 ppm, then σ, = 261 ppm or vice versa. Mass

spectra«

Abundance Appearance potential (eV)

XeFj

XeFj

XeFJ

XeF+

Xe+

7 12.9 (0.1)

100 12.1 (0.1)

60 14.9 (0.1)

67 13.3 (0.1)

-800 12.4(0.1)

Photoionization * spectra give AH° (XeF4(,)

XeF^^,) + F ^ ) ) = 9.66 eV

Electron diffraction ^ Radial distribution maxima: 1.94, 2.77 and 3.88 A. X e - F (DAH) = 1 . 9 4 ± 0 . 0 1 A Mössbauer spectrum: See Table 29. • F. Schreiner, G. N . M c D o n a l d and C. L. Chemick, / . Phys. Chem. 7 2 (1968) 1162. * J. Jortner, E. G. Wilson and S. A . Rice, / . Am. Chem. Soc. 85 (1963) 814. ' W. V. Johnston, D . Philipovich and D . E. Sheehan, in Noble Gas Compounds (H. H . H y m a n , ed. University of Chicago Press (1963), p. 139. B. Weinstock, Ε. Ε. Weaver and C. P. K n o p , Inorg. Chem. 5 (1966) 2189. • See ref. c, p. 144; also supported b y work o f J. Berkowitz, ref. x. ' S. R. G u n n and S. M. Wüliamson, Science 140 (1963) 177; see ref. c, p. 133. « H. J. Svec and G. D . Flesch, Science 142 (1963) 954. J. A . Ibers and W . C. Hamilton, Science 139 (1963) 106. * S. Siegel and E. Gebert. / . Am. Chem. Soc. 85 (1963) 240.

288

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY TABLE 26 (cont.)

J D . H . Templeton, A . Zalkin, J. D . Forrester and S. M . Williamson, / . Am. Chem. Soc. 8 5 (1963) 242. See ref. c, p. 275. » J. Slivnki et al., Croat. Chem. Acta 3 4 (1962) 187. " S. Mariöiö, Z. Veksli, J. Slivnik and B . VolavSek, Croat. Chem. Acta 35 (1963) 77. ° E . A . Bourdeaux, / . Chem. Phys. 4 0 (1964) 229. • J. G. Malm, H . Selig, J. Jortner and S. A . Rice, Chem. Revs. 65 (1965) 199; Ε. S. Pysh, J. Jortner and S. A . Rice, / . Chem. Phys. 4 0 (1964) 2018. ^ W. A . Yeranos, Molec. Phys. 9 (1965) 449. ^ H . H . Claassen, C. L. Chemick and J. G. Malm, / . Am. Chem. Soc. 85 (1963) 1927. ' See ref. c, p. 287. • J. H . Burns, P. Α . Agron and Η . Α . Levy, Science 139 (1963) 1208. * R. Hoppe, Fortschr. chem. Forsch. 5 (1965) 335. " D . Κ. Hinderman and W . E. Falconer, / . Chem. Phys. (in press). * D . Lazdins, C. W . Kern and M . Karplus, / . Chem. Phys. 3 9 (1963) 1611; M . Karplus, C. W . Kern and D . Lazdins, / . Chem. Phys. 4 0 (1964) 3738. ^ D . K. Hindermann and C. D . Comwell, / . Chem. Phys. 4 8 (1968) 4148. ' J. Berkowitz, Argonne National Laboratory, personal communication to N . Bartlett. ' See ref. c, p. 238.

Thus for the disproportionation 2 X e F 4 ( g ) -> X e F 2 ( g ) + X e F 6 ( g ) , AH° = + 5 k c a l m o l e - i and A'S' « 0 cal deg-i mole-i, and for 3XeF4(g) Xe(g)+2XeF6, = + 9 kcal mole-i and AS « —9 cal deg-i mole-i. The enthalpy of vaporization of X e F 4 , as in the case of XeF2, is indicative of strong electrostatic interactions between X e F 4 moleculesio^. Kinetic features. Since X e F 2 is an intermediate in the formation of X e F 4 , the kinetics of formation of the former (see section 3.2.1) are relevant to the latter. From a studyi^s of the thermally excited F 2 + X e reactions, with F2:Xe ratios of 16 or more and total pressures of 10-20 mmHg in the temperature range 190-250°, the XeF2 formation was found to be zero order in F2 and first order in Xe (the reverse of the findings for the condi­ tions which favour exclusive X e F 2 formation—see section 3.2.1) and the X e F 4 formation, zero order in F2 and first order in XeF2. The thermal reactions were shown to be hetero­ geneous and a mechanism involving adsorption and dissociation of F2 on the nickel fluoride surface of the container has been proposed. The activation energy for F2+XeF2 X e F 4 was deduced to be 13 kcal mole-i. Structural features. Vibrational spectroscopici79.180,181 and electron difiraction d a t a i 8 2 have established the XeF4(g) molecule to be square planar (D4h)i the latter study giving X e - F to be 1.94+0.01 A . The crystallographic findingsi^ show that the size and shape of the isolated molecule is not significantly different in the crystal (see Fig. 16). A neutrón difi'raction studyi7« has shown the X e - F bond length in the sohd to be 1.953 (0.004) A , the ^ F X e F = 90.0 (0.02)°, and the amphtudes of vibration normal to the bond directions to be greater than in the bond direction. Thermal motion has also been indicated by the broad-hne i ^ p n.m.r. studiesi^?. Bond polarity and bond type. Nuclear magnetic resonance, Mössbauer and electron spectroscopy all indicate considerable electron migration Xe F in the X e - F bond. 178 c . F . Weaver, P h . D . thesis. University of California, Berkeley, September 1966. 179 H . H . Claassen, C. L. Chemick and J. G. Malm, / . Am. Chem. Soc. 85 (1963) 1927. 180 See ref. 14, p. 287. 181 W . A . Yeranos, Molec. Phys. 9 (1965) 449. 182 See ref. 14, p. 238.

XENON(IV) COMPOUNDS

(b) The molecular structure of XeF4 as determined by X-ray diffraction *. •Seeref. 17b.

XeF4

289

(a) The view of the XeF4 crystal structure • along the b axis. The numbers give the elevation of the atoms in units of ^/lOO above the plane of projection. • See ref. 17b. Fio. 16. The crystal and molecular structure of XeF4

290

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY

The 19F Chemical shift and i 2 9 X e - i 9 F coupling c o n s t a n t i 8 3 , i 2 i , i 8 4 ( s e e Table 26) do n o t differ substantially from those obtained for the F - X bonds in the other xenon fluorides a n d t h e related fluorides BrFa, IF5 a n d TeFe. The chemical shifts have b e e n interpreted by Karplus a n d h i s c o w o r k e r s i 2 5 i n terms of a fluorine ligand charge of —0.50. This evaluation assumed t h e bonding to involve primarily t h e F 2p a n d Xe 5p orbitals. On t h e other hand, Gutowsky a n d h i s coworkersii^^ on t h e basis of a locahzed bond descrip­ tion, using spd hybrid xenon orbitals, have concluded t h a t t h e fluorine ligand charge is 0.49. A b r o a d - h n e i^F n.m.r. s t u d y i 7 7 h a s shown t h a t prior, similar, studies a r e erroneous, probably as a consequence of XeF2 contamination. Experimental shielding values (see Table 26) a r e in quantitative a n d e v e n , in some cases, quahtative disagreement with theo­ retical values obtained using a semi-empirical localized orbital bonding scheme. It s e e m s t h a t more complete theoretical treatments, including delocalized orbitals, w i l l be necessary to account f o r t h e observations. The Mössbauer s p e c t r u m i 2 3 . i 2 4 of X e F 4 , which is discussed further in section 3.3.2, interpreted^ss on t h e assumption of bonding involving Xe 5p orbitals. A fluorine hgand charge of —0.75 is assigned. This s e e m s rather large, a n d it m a y w e l l be t h a t t h e fault derives from t h e over-simplified bonding model. The chemical shift observed in t h e X - r a y photo-ionization s p e c t r u m i 2 6 h a s b e e n accounted f o r on t h e basis of a coulombic model, which assumes t h e central atom a n d ligands to be charged spheres. A fluorine hgand charge in t h e range —0.3 to —0.5 is compatible with t h e findings. (See section 3.4.2 a n d Table 32.) The most popular bonding descriptions have followed t h e lead of P i m e n t e l 3 2 a n d R u n d l e 3 3 a n d generally emphasize t h e importance of t h e Xe 5p a n d F 2p orbitals. AUen^i", C o u l s o n 2 3 a n d Jortner a n d Ricéis have reviewed t h e bonding models. Several authors have given i3o, i s s , 1 8 6 , i s ? energy l e v e l diagrams to account f o r t h e observed spectroscopic features. There a r e persistent claims also f o r t h e involvement of outer orbitals, particularly Xe 5J, in t h e b o n d i n g 2 5 . (See sections 1.3.1-4 a n d 3.2.2.) Chemical properties. A number of t h e earlier studies involving X e F 4 a r e suspect because of t h e likehhood t h a t t h e samples w e r e contaminated with X e F i or XeFe or both (see under "Preparation" a n d "Thermodynamic features"). The tetrafluoride c a n be kept in thoroughly dried glass or quartz a n d c a n be stored indefinitely in Kel-F nickel or Monel containers. The tetrafluoride undergoes instantaneous hydrolysis, with t h e formation of a transitory yellow speciesiss.i89. The latter h a s b e e n trapped at -SO"* a n d , on t h e basis of infrared a n d e.s.r. examinationi^o, formulated as XeOFz (see section 3.3.3). The ultimate products of hydrolysis areiss Xe, O2, H F a n d XeOa (see section 3.4.4). If t h e hydrolysis is carried o u t in strong base, perxenates are formedi^i. 1 5 5 . i s s ( s e e section 3.5.5), a n d if t h e hydrolysis occurs in aqueous KI very little free O2 is h b e r a t e d i 7 2 , X e F 4 + 4 I Xe+4F-+2l2. As with X e F 2 , the l o w bond energy of X e F 4 causes it to be a strong oxidative fluorinator a n d it should compare with BrFa in oxidative capabihty. It s e e m s , however, h k e X e F z , to has been

183 184 185 186 187 188

T. H . Brown, E. B. Whipple and P. H . Verdier, Science 140 (1963) 178. A . C. Rutenberg, Science 140 (1963) 993. J. Jortner, E. G. Wilson and S. A . Rice, / . Am, Chem, Soc. 85 (1963) 815. E. A . Bourdeaux, / . Chem, Phys. 4 0 (1964) 246. Y . J. Israeli, Bull. soc. chim, France 3 (1963) 649. S. M . Williamson and C. W . K o c h , Science 139 (1963) 1046. 189 See ref. 14, p. 158. 190 J. S. Ogden and J. J. Turner, Chem. Communs. (1966) 693.

XENON(IV) COMPOUNDS

291

be kinetically rather inert. Solutions of X e F 4 in anhydrous H F (with which the fluoride does not undergo fluorine hgand exchangeis^) are strongly oxidizing and fluorinate platinum metal t o P t F 4 ^73. The neat compound oxidizes SF4 t o SEe i^^, oxidizes N O to O N F , but does not interact, at a measurable rate, with N O 2 The interaction of X e F 4 with H2 does n o t proceed at room temperature in the absence of a catalyst, but a slow reaction occurs at 70° and this proceeds rapidly at 130° 1^2. The tetrafluoride, in contrast with XeF2, fluorinates perfluoropropane at room temperaturei^s but gives similar products to XeF2 in interaction with the hydro-olefins (see section 3.2.1). Attempts t o prepare other xenon(IV) compounds by metathetical reactions have met with little success. The interaction of X e F 4 with BCI3 at - 7 8 ° 1^3 yields xenon and chlorine quantitatively: 3 X e F 4 + 4 B C l 3

4BF3 + 3 X e + 6 C l 2 . There are indications, however, that

(as in the XeF2 case) the fluorine ligands may be substituted by other highly electro­ negative ligands, since X e F 4 dissolved in trifluoroacetic anhydride is reported t o yield a crystalhne compound c o n s i d e r e d i 9 4 to be X e ( O O C C F 3 ) 4 . Clearly, hgands such as - O - C I O 3 , O S O 2 F and OTeFs are possible, but there is no certainty that the compounds will be stable to disproportionation or spontaneous reduction of the noble-gas atom. The fluoride ion donor ability of X e F 4 has been shown^«^ to be less than that of X e F i or XeFö, and this forms the basis of a chemical purification for X e F 4 . These findings are in harmony with the enthalpies of ionization derived from photoionization studies by Berkowitzi95:

Process

XeF2(,) XeF4(,i XeF6(,)

• X e P i +F(;) •XeFj+tt, + Ρ^,, •XeFii.)+Fr.)

(eV)

9.45 9.66 9.24

The best fluoride ion acceptor (SbFs) forms a crystalhne sohdi^^, 197 (or solids) with X e F 4 , probably containing XeFJ, and there is also evidencei^7 for weaker fluoride ion acceptors (PF5, AsFs) forming compounds in BrF3 solution. Reports that X e F 4 interacts with SbFs or TaFs to form XeF2 derivativesi63 or t o yield XeF2, 2IFs on dissolutioni^s in IPs, are certainly erroneous. It is probable that the X e F 4 used in those studies was grossly con­ taminated with XeF2. Analysis of X e F 4 . The sohd is most conveniently tested for purity by a melting point determination (117.1°), but its Raman spectrum is also highly characteristic a n d readily reveals the presence of XeF2 or XeFö (see Table 26). The infrared spectrum readily characterizes the vapour. Samples of X e F 4 have been analysed by reduction with hydrogen at 130° .· XeF4+2H2->Xe+4HF 191 192

193 194 195 196 197 198

See ref. 14, p. 73. See ref. 14, p. 144. N . Bartlett, Endeavour 22 (1964) 3. A . Iskraut, R. Taubenest and £ . Schumacher, Chimia 18 (1964) 188. J. Berkowitz, Argonne National Laboratory, personal communication. N . Bartlett and N . K. Jha, unpublished observation. D . Martin, C.R. Acad. Sei. Paris C (1969) 1145. H . Meinert, G. Kauschka and S. Rüdiger, Ζ . Chem. 7 (1967) 111.

292

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY

A more convenient analytical technique involves reduction with mercury: XeF4+4Hg -> Xe+2Hg2F2 (or 2HgF2)

The xenon may be measured tensimetrically or gravimetrically, and the fluorine content is obtained from the weight of mercury fluoride formed. As has been remarked above, it has also been claimed that X e F 4 may be analysed iodimetrically by dissolution in aqueous KI. 3.3.2. Xenon Tetrachloride EffOrts to prepare macroscopic quantities of XeCU and X e B r 4 from X e F 4 by meta­ thetical reactions have failed, but the former has been detectedi^s, 199 by Mössbauer spectroscopy as a product of the β decay of its i 2 9 i analogues: 1291CI4 i 2 9 X e C l 4 . There is no evidence for the existence of X e B r 4 . The conditions for the observation of XeCU are essentially the same as for the dihahdes described in section 3.2.2. The chemical shift observed for XeCU is shown in Fig. 4 (p. 239) relative to shifts for the other xenon compounds. The Mössbauer data for the xenon halides are compared in TABLE 27. MÖSSBAUER D A T A FOR THE X E N O N HALIDES200. 199

Halide

XeF4 XeF2 XeCU XeCl2 XeBr2

Splitting (mm s e c - i )

(MHz)

5p electron transfer

electron transfer per b o n d

41.04 ± 0 . 0 7 39.0 ± 0 . 1 25.6 ± 0 . 1 28.2 ± 0 . 1 22.2 ± 0 . 4

2620 2490 1640 1800 1415

3.00 1.43 1.88 1.03 0.81

0.75 0.72 0.47 0.52 0.41

27. T h e quantities in the table have the same meaning and were derived in the same way as those discussed in section 3.2.3. A g a i n , the bonding model assumes that the only xenon orbitals participating in the bonding are the X e 5p orbitals, hence the indicated bond polarities are probably not quantitatively reliable; the trends are probably correct. Table

3.3.3. X e n o n O x i d e D i f l u o r i d e O f the several claims for X e O F 2 in the literaturei72. i9i, 190 only one is supported with experimental evidence. T h e compound identified as X e O F 2 is the bright yellow solid formed by hydrolyzing X e F 4 at —80°. T h i s product gives neither an e.s.r. spectrum nor an infrared D shift (D2O in place of H2O), and contains only one atom of O . T h e observed infrared absorption bands have been assigned on the basis of Czv symmetry: 747 (vi, au X e - O str.), 520 (v2, ai, X e - F sym. str.) and 490 cm-i (V4, ¿2, X e - F asym. str.). It is argued that these data are compatible with a structure O

I

F-Xe-F

with < F - X e - 0 of about 90°, thus resembhng CIF3 and B r F s (see section 1.4.4). T h e X e - O stretching force constant is 4.7 mdyn A - i on the basis of this model and assignments. 199 J. G. Perlow and H . Yoshida, / . Chem. Phys. 49 (1968) 1474. 200 B. Jaselskis and J. P. Warriner, Anal. Chem. 38 (1966) 563.

293

XENON(IV) COMPOUNDS

This appears a httle low, even allowing for a lower value with the lower oxidation state, since the force constant for XeO in X e O F 4 is 7.08 mdyn Ä " i (see section 3.4.2). 33.4.

Xenon(IV) Oxide

Although XeOa has been postulated as an intermediate in the hydrolysis of XeF4 (see section 3.4.4), there is no firm evidence for its existence. 3.3.5. F4-X Xe(OR)x Compounds Apart from the r e p o r t i 9 4 of the synthesis of Xe(OOCCF3)4 (see section 3.3.1), no other compounds in this class have yet been reported. 3.3.6. X e F 4 as a Flaoride Ion Donor and Acceptor There is n o evidence that X e F 4 can accept fluoride ion (or any other ion), and this provides for the ready removal of XeFg contaminant by absorption of the latter with alkah fluoridesi. Although X e F 4 is a poorer F~ donor than either X e F i or XeFö (see section 3.3.1) it does form compounds with the best F - acceptor SbFs i^?. These may contain the X e F J ion, but there is presently n o evidence to support this. (See also sections 3.4.1 and 3.4.5.) 3.3.7. Molecular Adducts of X e F 4 Only one molecular adduct of X e F 4 has been established, XeFi, X e F 4 i70. The crystal structure of this adduct shows it t o be an ordered arrangement in which, as may be seen from Table 28, each molecule possesses essentially the same size and shape as in the com­ ponent sohds. The bonding is presumably a consequence of the coulombic interactions

TABLE 2 8 . INTRAMOLECULAR DISTANCES I N X e F i , X e F 4 A N D THEIR COMPONENTS

X e - F (A) in XeF2 X e - F (A) in XeF4 F - X e - F in XeF4

XeF2,XeF4*

XeFi»»

(2) 2.010(0.012) (2) 1.972(0.014) (2) 1.945(0.014) ( 2 ) 8 9 . r (0.8)

(2) 2.00 (0.02)

XeF4''

1.953 (0.004) 90.0 (0.02)

• J. H . B u m s , R. D . Ellison and H . A . Levy, Acta Cryst, 18 (1965) 11. H . A . Levy and P. A . Agron, / . Am, Chem, Soc, 85 (1963) 2 4 1 . ^ J. H . B u m s , P. A . Agron and H. A . Levy, Science 139 (1963) 1208. XeF2, XeF4 intermolecular contacts XeflD—F-XeaV) distances (2) 3.28, (2) 3.42, (8) in range 3.61-3.69 A. X e a V ) — F - X e a i ) (6) in range 3.35-3.37 A. Compare with data in Figs. 19 and 25.

between positive charges on the xenon atoms and negative charges on the fluoride hgands of the surrounding molecules. The compound resembles other X e F 2 adducts (see section 3.2.8).

294

NOBLE-GAS CHEMISTRY: NEIL BARTLETT A N D F. O.

3.4.

XENON(VI)

SLADKY

COMPOUNDS

T h e c h e m i s t r y o f x e n o n ( V I ) i s h m i t e d t o t h e fluoride X e F g , t h e o x y f l u o r i d e s X e O F 4 a n d X e O z F a , t h e o x i d e X e O a , a n d c o m p l e x e s o f these c o m p o u n d s . T h e trioxide is t h e r m o ­ d y n a m i c a l l y unstable with respect t o the elements i n their standard states, a n d there are indications that this is also s o for X e 0 2 F 2 . T h e oxide is a powerful explosive. T h e o x y ­ fluoride

X e O F 4 and the

fluoride

are thermodynamically stable at ambient temperatures.

All o f t h e c o m p o u n d s , a t l e a s t p o t e n t i a l l y , a r e s t r o n g o x i d i z e r s .

3.4.1. X e n o n H e x a f l u o r i d e

Preparation. T h e p r e p a r a t i o n o f X e F e w a s first d e s c r i b e d i n f o u r i n d e p e n d e n t a n d a l m o s t s i m u l t a n e o u s reports^o^, 201-3. All p r e p a r a t i o n s a r e c a r r i e d o u t i n n i c k e l o r M o n e l v e s s e l s a n d , i n g e n e r a l , h i g h F2 p r e s s u r e s a n d l o w e r t e m p e r a t u r e s f a v o u r X e F ö f o r m a t i o n . A 95 % c o n v e r s i o n t o X e F e i s o b t a i n e d w i t h F 2 / X e r a t i o s o f 20:1 a t 50 a t m p r e s s u r e ( s e e s e c t i o n 3.31 a n d T a b l e 25). T h e h e x a f l u o r i d e h a s a l s o b e e n r e p o r t e d ^ ^ i a s a p r o d u c t o f t h e e l e c t r i c discharge o f a

3:1 F 2 : X e m i x t u r e , t h e p r o d u c t b e i n g t r a p p e d a t —78°. F u r t h e r m o r e , O2F2

is reportedi9i t o oxidize X e F 4 t o X e F e b e t w e e n -133°

and

-78°.

Laboratory preparation. In o r d e r t o o b t a i n p u r e X e F ö i t i s best204 t o f o r m t h e N a F - X e F g adduct by mixing with N a F , the other xenon

fluorides

(XeF2, X e F 4 and XeOF4) hkely to

be present being unable t o form stable complexes with the N a F . T h e a p p a r a t u s a n d e x p e r i m e n t a l p r o c e d u r e s f o r X e F e synthesis205 are essentially a s given under X e F 2 a n d X e F 4 , but since higher

fluorine

pressures are desirable, the nickel or

M o n e l vessels used t o contain the h o t mixtures should be strong e n o u g h t o safely withstand

400 a t m p r e s s u r e o f F2. A X e / F a r a t i o o f 1:20 i s s a t i s f a c t o r y . T h e m i x t u r e i s h e a t e d t o 300° f o r 16 h r . E x c e s s fluorine i s r e m o v e d u n d e r v a c u u m a t - 1 9 2 ° o n t o N a F (previously

fluorinated

and the crude X e F e condensed

a n d in appreciable excess). This mixture is w a r m e d t o

50° f o r 2 h r t h e n left a t r o o m t e m p e r a t u r e o v e r n i g h t . T h i s s e r v e s t o c o m p l e x a l l o f t h e X e F ö as a N a F - X e F ö adduct. U n c o m b i n e d impurities (XeOF4, XeF2, XeF4) are removed b y p u m p i n g o n t h e N a F m i x t u r e a t t e m p e r a t u r e s u p t o 50° t o c o n s t a n t w e i g h t ( a n h o u r o r s o s h o u l d s u f i i c e ) . T h e X e F ^ i s t h e n r e t r i e v e d b y h e a t i n g t o 125° u n d e r v a c u u m , w h e n t h e gas is rapidly e v o l v e d a n d m a y b e collected in c o l d traps (nickel, M o n e l o r K e l - F at

—196°).

T h e u s u a l care s h o u l d b e t a k e n t o g u a r d a g a i n s t X e O a f o r m a t i o n (EXPLOSIVE), a n d it should always be assumed that the oxide m a y have formed in apparatus used for XeFg synthesis a n d handling.

Thermodynamic features.

E q u i l i b r i u m studies22 o f t h e X e / F z s y s t e m h a v e d e f i n e d t h e

o p t i m u m conditions for X e F g formation. T h e equilibrium constant data from these studies a r e g i v e n i n T a b l e 25 a n d t h e t h e r m o d y n a m i c d a t a a n d o t h e r p h y s i c a l d a t a f o r X e F g a r e g i v e n i n T a b l e 29. I t i s o f i n t e r e s t t h a t t h e e x p e r i m e n t a l e q u i l i b r i u m c o n s t a n t d a t a s u g g e s t a n e n t r o p y 5° f o r X e F e c e ) w h i c h r e q u i r e s a m o l e c u l a r s y m m e t r y l o w e r t h a n OH ( s e e b e l o w ) . H e a t capacity a n d vapour pressure measurements206 have provided accurate physical c o n ­ stants a n d also indicated structural c h a n g e s i n solid X e F e at

253.8° a n d 291.8°K. T h e

201 See ref. 14, p. 64. 202 J. G. Malm, I. Sheft and C. L. Chemick, / . Am. Chem. Soc. 85 (1963) 110. 203 F. B. Dudley, G. Card and G. H . Cady, Inorg. Chem. 2 (1963) 228. 204 I. Sheft, T. M . Spittler and F. H . Martin, Science US (1964) 701. 205 c . L. Chemick, J. G. Malm and S. M. Williamson, in Inorganic Syntheses (H. F . Holtzclaw, Jr., ed.), Vol. VIII, McGraw-Hill, N e w York (1966), p. 258. 206 F. Schreiner, D . W. O s b o m e , J. G. Malm and G. N . McDonald, / . Chem. Phys. 51 (1969) 4838.

XENON(VI)

295

COMPOUNDS

TABLE 2 9 . SOME PHYSICAL PROPERTIES OF

XeF6

The solid is colourless below m.p., the liquid and vapour are yellow-green Melting point ( ° C ) « 4 9 . 4 8 Phase changes ("C)*'

18.65,

Boiling point ( ° C ) «

75.57

Thermodynamic

-19.35

features

Vapour d e n s i t y " M o l e c u l a r weight^' (25.6°), 249.6; (24.8°), 245.5, 248 » Theor. XePe, 245.3 Vapour pressure ( m m ) 2 . 7 (0.04°); 23.43 (22.67°) Vapour pressure equations log Ρ„„ = -XIT-h Y Solid Temp, range (°K)

X

Y

273.19->295.82 254 ->291.8*' 291.8 ->322.38*=

3400.12 3313.5 3093

12.86125 12.5923 11.8397

Liquid: 3 2 2 . 3 8 - > 3 5 0 «

Second virial coefficient Β (PV = RT^BP) AH^ub (kcal m o l e - i ) AHtus (kcal m o l e - i ) Δ 5 ν . ρ (cal d e g - i m o l e - i ) Cp, solid (at 2 9 8 . 1 5 ° K ) (cal d e g - i m o l e - i ) " Δ / Γ Α , , (kcal m o l e - i ) ( 2 9 8 . 1 5 ° K ) AS}^^^ (cal d e g - i m o l e - i ) S ° (cal d e g - i m o l e - i ) S o H d ( 2 9 8 . 1 5 ° K ) Liquid ( 3 3 5 ° K ) Gas (335°K)

A l s o f r o m Xe/Fa equilibria, gas ( 2 9 8 . 1 5 ° K ) Δ^°(XeF6(.,^XeFÍ,„+F,;,)'

logP™„ = - - 6 1 7 0 . 8 8 Γ - 2 3 . 6 7 8 1 5 l o g r + 8 0 . 7 7 7 7 8 -955cm3mole-i 15.6 ^ 1 3 . 2 1.37« 32.74 41.03 70.4 ^ 8 2 . 9 * ; 8 1 ' -97* 50.33 61.10 96 27 88.84 or 91.87 9.24 e V (213 kcal m o l e - i )

Solubility Anhydrous H F »

Γ(°0

M o l e s XeFö per 1000 g H F

M o l e s H F per m o l e XeFö

15.8 21.7 28.5 30.25

3.16 6.06 11.2 19.45

15.8 8.25 4.46 2.57

Molar conductivity' o f H F solutions ( o h m - i c m - 2 ) ¡n range 60-147 o h m - i cm~2 (0.75-0.02 mole l - i ) XeFe gives yellow-green solutions in WF^, IF5 and BrFs " Dielectric

constant»

Dipole moment ^: Magnetic

susceptibility

Specific conductance Ultraviolet

(55°C):

4.10±0.05

< 0.03 D χ„ =^ - (44.5 ± 0.5) x 1 0 - 6 cm3 m o l e - i ' (50°C):

arui visible absorption

1.45 ± 0.05 x IO-I6 o h m - i c m - 2 spectrum

3300 A, strong, half-width 580 A; very intense absorption below 2750 A

296

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY TABLE 29

First ionization potential:

(cont)

I ( X e F 6 ( , ) ) ' (1016 A) = 281.5 kcal m o l e - i

Infrared and Raman spectrum: non-octahedral "· " monomer Solid (unknown crystalline f o r m ) " : R (cm-i): Liquid": R (cm-i):

204?, 300 w, 404 vw, 583 s, 636 sh, 656 vs (54°Q: (92°C):

205 ?, 295 w, 370 vw, 403 vw, 506 vw, 574 sh, 585 s (P), 637 sh, 654 vs (P) 205 ?, 295 w, 370 vw, 403 vw, 506 ms, 577 s, 616 sh, 650 vs.

Vapour: IR ( c m - i ) »: 400 ms, 520 m, 563 mw, 616 s, 1036 vw, 1075 w, 1118 vw, 1238 w R ( c m - i ) ( 9 4 ° C ) " : 520 s, 609 m s (P), also 2 0 6 ? Electron diffraction: Non-octahedral molecular symmetry °· *'· X e - F bond length: 1.890 A » Crystallographic data: Unit cells:

Density*

C u b i c ' (stable between 103° and 3 0 Γ Κ ) ·

Monoclinic

e o ( - 8 0 ° Q = 25.06 (0.05) A Space group Fm3c; r = 144 XeF« dX-ny (g cm--3) ( - 8 0 ° C ) = 3.73 (0.02) (See Fig. 17)

ao 9.33 (0.03) A 00 10.96(0.03) CO 8.95 (0.03) β 91.9 (0.2)°

<

solid

Γ(°Κ) 77.22 d ( g c m - 3 ) 3.848((Í.006)

194.42 3.751(0.007)

242.97 3.668(0.014)

> 293.11 3.465(0.013)

297.55 3.411(0.015)

liquid 328.34 3.173(0.03)

• J. G. Malm, B. D . Holt and R. W. Bane, in Noble Gas Compounds (H. H . H y m a n , ed.). University of Chicago Press, Chicago and L o n d o n (1963), p. 167. * B. Weinstock, Ε. Ε. Weaver and C. P. K n o p , Inorg, Chem, 5 (1966) 2189. * F . Schreiner, D . W. Osborne, J. G. M a l m and G. N . M c D o n a l d , / . Chem, Phys, 51 (1969) 4838. * See ref. a, p. 39. * See ref. a, p. 144. ' J. Berkowitz, Argonne National Laboratory, personal communication. * See ref. a, p. 275. N . Bartlett and F . O. Sladky, / . Am, Chem, Soc, 9 0 (1968) 5316. * H. Selig and A . Mootz, Inorg, Nucl, Chem, Letters 3 (1967) 147. J R. F . Code, W. E. Falconer, W. Klemperer and I. Ozier, / . Chem, Phys, 47 (1967) 4955. Β. Volavsek, Mh,f Chemie 97 (1966) 1531. » J. G. Malm, I. Sheft and C. L. Chemick, / . Am, Chem, Soc, 85 (1963) 110. " E. L. Gasner and H . H. Claassen, Inorg, Chem, 6 (1967) 1937. " H. Kim, H . H. Claassen and E. Pearson, Inorg, Chem, 7 (1968) 616. * R. M. Gavin, Jr. and L. S. Bartell, / . Chem, Phys, 48 (1968) 2460. ^ K. Hedberg, S. H. Peterson, R. R. R y a n and B. Weinstock, / . Chem, Phys, 4 4 (1966) 1726. R. D . Burbank and N . Bartlett, Chem, Communs. (1968) 645. ' P. A . Agron, C. K. Johnson and H . A . Levy, Inorg. Nucl. Chem. Letters 1 (1965) 145. • R. D . Burbank and G . R. Jones, Science 168 (1970) 248. * J. Serpinet and O. Rochefort, Bull. soc. chim. France 10 (1968) 4297. interpretation o f these regions o f a n o m a l o u s heat capacity, in terms o f changes previously

identified207

crystalhne modifications

h i g h e n t r o p y o f v a p o r i z a t i o n o f the hquid206

of

(32.74

XeFö,

between

has been questioned^os.

The

e.u.) is indicative o f polymerization in

207 p . A . Agron, C. K. Johnson and H . A . Levy, Inorg. Nucl. Chem. Letters

1 (1965) 145.

208 G. R. Jones, R. D . Burbank and W. E. Falconer, / . Chem. Phys. (in press).

XENON(VI)

COMPOUNDS

(b) F I G . 1 7 {see over for

caption).

297

298

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY

the hquid state. Two independent vapour density measurements22. 209 show the hexa­ fluoride to be predominantly monomeric in the gas phase. The hexafluoride is much more volatile than either XeF2 or X e F 4 , although much less volatile than other hexafluorides. There is considerable disparity between the enthalpy of formation from the equilibrium studies ( - 7 0 . 4 kcal mole-i) and the value ( — 82.9 kcal mole-i) from the heat of combustion of X e F ö ( X e F 6 + 3 H 2 -> Xe + 6HF). Since the latter value is also in agreement with the value from photo-ionization studies^^s^ it is preferred. This infers a mean thermochemical bond energy for X e F ö of 32-33 kcal mole-i, which is, within the experimental uncertainty, the same as the values derived for X e F 2 and X e F 4 . Although equilibrium studies22 showed no evidence of XeFs formation it is of interest that the rate of exchange of I8F2 with XeFö at 150° was found to be a hnear function of 1 8 F 3 concentration, indicating an associative mechanism2io. Crystal and molecular structure. Although both a cubic and monochnic form of crystal­ hne XeFö have been i d e n t i f i e d 2 0 7 , only the cubic structure has been described in detail2ii (Fig. 17). This structure indicates that XeFö is eSectively X e F ^ F " in the cubic phase. The existence of both tetramers and hexamers in the same unit cell indicates that the way in which the XeF^ ion is "bridged" by the F " ions is not of prime importance. It should be noted, however, that the "bridging" F " ions are not close to the fourfold axis of the X e F J groups. This is consistent with the location of the non-bonding xenon(VI) valence electron 209 J. Serpinet and O. Rochefort, Bull, soc, chim, France 10 (1968) 4297. 210 See ref. 14, p. 68. 211 R. D . Burbank and G. R. Jones, Science 168 (1970) 248.

FIG. 17. The structural units of XeFe (cubic) * (previous

page)

The cubic unit cell (space group Fm3c), Ö0 ( - 8 0 ° Q = 2 5 . 0 6 ± 0 . 0 5 A, contains 144 "XeFe units". Ions of X e F j and F~ are associated in tetrameric and hexameric units (a and b in the figure). There are 2 4 tetramers and 8 hexamers in the unit cell. Both right- and left-handed conformations o f both tetramers and hexamers occur. The X e F j ions have similar size and shape in the tetramers and hexamers. T h e F - ions "bridge" the XeFJ ions in each cluster type. T h e F5Xe+ . . . F - . . . XeFJ bridges are u n symmetrical in the tetramer but symmetrical in the hexamer, the chemically important bond lengths and angles are. Tetramer

Hexamer

fXe-F.p(7,l)(A) XeF5+

Xe-Fb.. (8 3) (A) F.p-Fb.. (A)

F5.-F5« (A) -^F.p-Xe-Fb..

Fbms—XC—Fbai F5Xe+ . . . F - "bridge" distance (A) FsXe^ . . . F - . . . X e F ^ "bridge" angle

1.84(4) 1.86 (3) 2.29(6) 2.54 (13) 77.2 (1.8)° av. 87.2 (4.5)° av.

1.76 (3) 1.92(2) 2.33 (3) 2.63 (3) 80.0 (0.6)° 88.3 (0.2)° 2.56 (2)

Γ2.23 (3) \ 2 . 6 0 (3)

118.8(0.3)°

The F - bridges 2 X e F ^ groups in the tetramer and 3 in the hexamer. • R. D . Burbank and G. R. Jones, Science 168 (1970) 248.

120.7 (1.2)°

XENON(VI) COMPOUNDS

299

pair on the fourfold axis, trans to the unique X e - F bond of the XeF^ ion. Thus the X e F j would appear to be pseudo-octahedral. Note the similarity of the structure of ( X e F 6 ) 4 to the structure of [XeFsl+iAsFe]" shown in Fig. 21 (p. 316) (discussed in section 3.4.6). The X e F j unit in (XeFö), is almost indistinguishable from that observed in the X e F J salts. From the limited structural information available on the monochnic p h a s e 2 0 7 it is probable that tetrameric units, rather like those occurring in the cubic phase, form the structural unit. The entropy of vaporization indicates that the liquid is polymeric but the colour is similar to that of the vapour. It is possible that an appreciable proportion of the liquid is monomeric. The polymeric nature of the liquid is again attributable to F"" bridging between XeFf units. The solutions of XeFö in H F 2 1 2 and W F 0 2 1 3 show a dependence of i^F n.m.r. chemical shift upon the temperature. These observations are simply explained by polymerization-depolymerization equilibria. The high conductivity of solutions in H F 2 1 4 suggests the ionization: XeFgieoin)+HF - > X e F j ( , ο ΐ η ) + H F ^ ( , o i n )

Molecular XeFö. The molecular structure of x e n o n hexafluoride in the gas phase has been the subject of much work and many papers. We still do not have a clear view of the structural features of isolated XeFö molecules. It is certain that the vapour at room tempera­ ture is not a collection of octahedral molecules. It is also evident that a large proportion of the molecules must be of non-octahedral symmetry. Electron diffraction data^is have estabhshed that XeFö(g) has a very different symmetry from that of TeFö, and the datadlo. 2 1 7 have been interpreted2i8.2i9 on the basis of non-centric molecular symmetry. Bartell et al 217f2is concluded that the instantaneous molecular configurations encountered by incident electrons are predominantly in the broad vicinity of C^v structures conveniently described as distorted octahedra, in which the xenon(VI) non-bonding valence electron pair avoids the bonding pairs. Burbank et al found that a mixture of geometries of Civ, Czv and Cs symmetry, each different symmetry having equal weight in the mixture, accounted for the observed electron diffraction data, within the limits of the experimental error. Bartell d e r i v e s 2 i 7 X e - F to be 1.890+0.005 A . Facile molecule inversion or intra­ molecular rearrangement is certainly compatible with the findings. Evidently the major gas species are (or is) appreciably distorted, but Goodman22o has persistently maintained, from theoretical considerations, that an OH ground state for XeFö would be separated by only very small energy from a triplet state of D^d symmetry. Goodman predicts this triplet because his OH structure has two electrons populating an antibonding aig orbital (see Fig. 18). The excitation of one of these electrons to generate a triplet requires httle energy, but the state would be Jahn-Teller deformed to remove the orbital degeneracy. Goodman considers that a centro-symmetric distortion is required, and since a D4H distortion is not compatible with the observed electron diffraction data, concludes that D^a symmetry is appropriate. Bartell a l l o w s 2 i 8 that a 2)34 species could be consistent with the electron 212 213 214 215 216 217 218 219 220

T. E. H, K. L. R. L. R. G.

H. Brown, P. H. Kasai and P. H . Verdier, / . Chem. Phys. 4 0 (1964) 3448. J. Wells, L. Reeves, S. P. Beaton and N . Bartieft, unpublished observation. Selig and A . M o o t z , Inorg. Nucl. Chem. Letters 3 (1967) 147. Hedberg, S. H. Peterson, R. R. Ryan and B. Weinstock, / . Chem. Phys. 4 4 (1966) 1726. S. Bartell, R. M. Gavin, Jr., H. B. Thompson and C. L. Chemick, / . Chem. Phys. 4 3 (1965) 2547. M. Gavin, Jr. and L. S. Bartell, / . Chem. Phys. 4 8 (1968) 2460. S. Bartell and R. M . Gavin, / . Chem. Phys. 4 8 (1968) 2466. D . Burbank and N . Bartlett, Chem. Communs. (1968) 645. L. G o o d m a n , Bull. Am. Phys. Soc. (1967) 296.

300

NOBLE-GAS CHEMISTRY.* NEIL BARTLETT AND F. O. SLADKY

diffraction findings. Unfortunately, a magnetic deñection molecular-beam e x p e r i m e n t 2 2 i has provided no evidence for a paramagnetic XeFö species, although this does not necessarily deny the existence of a triplet species (if the spin-orbit coupling is sufficiently strong). It should also be noted, however, that an electrostatic deflection molecular-beam experim e n t 2 2 2 has provided no evidence for a dipolar XePe species. Indeed, pQíéFe) must be <0.03D, and this is difficult to reconcile with non-centric structures. Claassen and his c o w o r k e r s 2 2 3 , from millimetre waveband studies, were unable to obtain evidence to support Bartell's inverting or pseudo rotator m o d e l 2 2 4 . Raman and infrared d a t a 2 2 5 indicate that either the ground-state vapour molecules possess a symmetry lower than OH or they have some very unusual electronic properties that markedly influence the region of the spectrum usually considered the vibrational-rotational region. Clearly more work must be done to settle this important structural problem. Bonding and bond polarity. Alone, of the first theoretical papers on the bonding in noble-gas compounds, the valence-shell electron-pair repulsion model predicted^^ that the XeFö molecule would be non-octahedral (see section 1.3.4). The three-centre four-electron "molecular orbital" model predicted an octahedral molecule (see section 1.3.3). It now appears, however, that the "non-bonding valence electron pair" in XeFöig) does not have the "bulk" (or steric activity) usually associated with such a "pair"2i8. Evidently, even a Hückel molecular orbital m o d e l 2 i 8 . 226 containing only Xe 5^, 5p atomic orbitals and ρσ orbital in each fluorine, generates essentially the same structural predictions as the electron-pair repulsion theory, and like it, predicts a larger distortion from OH symmetry than that observed. The non-octahedral geometry and exceptional flexibihty of the XeFec,) molecule imphed by the electron diffraction findings have been interpreted in terms of a pseudo­ Jahn-Teller e f f e c t 2 i 8 . This explanation is probably best understood with the aid of Fig. 18. This model supposes that the Xe 5d orbitals partake in the bonding, the d orbital degeneracy

6Fs F I G . 18. Schematic correlation diagram illustrating m.o. energy levels for an OH molecule 221 222 223 224 225 226

R. w. E. L. H. L.

F. Code. W. E. Falconer, W. Klemperer and I. Ozier, / . Chem, Phys. 47 (1967) 4955. E. Falconer, A. Buchler, J. L. Stauffer and W. Klemperer, / . Chem, Phys, 48 (1968) 312. L. Gasner and H. H . Claassen, Inorg, Chem, 6 (1967) 1937. S. Bartell, / . Chem. Phys. 46 (1967) 4530. Kim, H. H. Claassen and E. Pearson, Inorg. Chem, 7 (1968) 616. S. Bartell, / . Chem, Ed. 4 5 (1968) 754.

XENON(VI) COMPOUNDS

301

having been removed by the ligand field. This model places a pair of electrons of the OH symmetry XeFö molecule in the antibonding orbital a*^. This orbital is supposedly very close in energy to the triply degenerate tf^ set of orbitals and indeed so close that certain deformations of the molecule can bring about considerable mixing of the ground and excited states. The ri„ deformation mode ought to have maximum effect in this regard and therefore be particularly favoured energetically. This is described as a pseudo-Jahn-Teller effect. The amphtudes of vibration of the tu deformations indeed do appear to be large. A tiu deformation is equivalent, geometrically, to a "lone-pair" pushing aside the hgands and protruding into the coordination sphere. The same correlation diagram (Fig. 18) is appropriate for Goodman's description. Goodman considers the represented XcFeiOn) species to be the ground state species, but allows that the a*¿-tf^ energy gap is so small that there is an extensive population of the triplet state (suitably distorted to lift the orbital degeneracy of the i,* set—a first-order Jahn-Teller effect). Thus, according to this view, the XeFö population should contain octahedral and D^a symmetry species, the concentration of the latter increasing with increasing temperature. As in the case of the other fluorides, the i ^ p n.m.r. data has been interpreted on the basis of bond polarities approximating to Xe3+-(Fo-5-)^. Both extreme theoretical a p p r o a c h e s i i 9 , i 2 5 (one involving Xe 5d orbitals, the other not) yielded this high bond polarity. Soft X-ray photoionization s t u d i e s i 2 6 (involving the ejection of core electrons from the xenon atom) have shown that the charge withdrawn from the xenon atom in XeF^ is a httle less than three times the charge withdrawn inXeFi. A charge somewhat less than 3-H on the xenon atom is consistent with the findings (see Table 32 (p. 306) and section 3.4.2). The ready ionizationi^s XeF6(., -> X e F 5 + ( . ) + F - ( , )

AH = 9.24 e V

compared with X e F 4 and X e F a , the corresponding enthalpies of which are 9.66 and 9.45 eV respectively, is of considerable chemical importance and surely of significance to bonding theory. The enthalpy of ionization is approximately 0.6 eV less than anticipated on the basis of the X e F 2 and X e F 4 data. This is compatible with the greater stability of X e F ^ salts compared with X e F J and even X e F + salts (see section 3.4.6). Evidently, the pseudooctahedral geometry of the X e F J ion must be especially favourable. The IF¿" ion is also a very favourable species (thus IF¿^ salts have greater stabihty than IF^" salts227 and I O F 5 , which is nearly octahedral, is not known to form l O F ^ salts)!«^. As we have seen, the preference for the "octahedral" geometry also shows up in the crystal structure of XeFg (cubic). It is possible that even in the molecular state XeFe and I F 7 are close to X e F J F - and IF¿"F- ion pairs. Chemical properties. As befits its higher oxidation state, XeFe is a much more powerful oxidizer and fluorinator than XeFi or X e F 4 . It seems that XeFö has little of the kinetic stability noted in XeFa and X e F 4 chemical behaviour. Thus, unlike XeFa and X e F 4 , it is not possible to store XeFe in glass or quartz. There is presumably a stepwise interaction, but these reactions are only effective for the formation of X e O F 4 and XeOa 228.· 2XeF6+Si02 -> 2 X e O F 4 + S i F 4 2XeOF4+Si02 ^ 2 X e 0 2 F 2 + S i F 4 2Xe02F4+Si02 ^ 2 X e 0 3 + SiF4 227 N . Bartlett and D . D . Gibler, unpublished observation. 228 H. D . Frame, J. L. Huston and I. Sheft, Inorg. Chem. 8 (1969) 1549.

302

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY

The interaction with water229, 2 3 0 , 2 3 1 . 2 3 2 . 2 3 3 , 2 0 3 similarly yields X e O F 4 a n d X e O s a n d the latter creates a hazard in XeFö handling. A large excess o f water yields X e 0 3 ( a q ) as described in sections 3.4.4 a n d 3.4.5. Hydrolysis in strong base leads t o the formation o f perxenates, the "ideal" disproportionation being 2XeF6+ 16NaOH -v N a 4 X e 0 6 + X e + 0 2 + 1 2 N a F + 8 H 2 0

The "disproportionation" c a n b e m u c h m o r e c o m p l e x than this (see section 3.4.5). Hydrolysis in the presence o f o z o n e generates perxenate m u c h more efficiently. The hexa­ fluoride interacts violently with H2 t o yield x e n o n a n d HF202 a n d with mercury t o give x e n o n a n d mercury fluoridesio?. The greater reactivity o f XeFö is illustrated by its interaction with perfluoropropene^^s which cleaves the m o l e c u l e : 2XeF6+3CF3-CF = CF2

3C2F6+3CF4+2Xe

The tetrafluoride yields perfluoropropane a n d XeFz d o e s n o t interact with the olefin. Cleavage o f the carbon skeleton is n o t characteristic o f XeFö, however, since it interacts with perfluorocyclopentene t o generate the cyclopentane a n d lower x e n o n fluorides234. Efforts t o prepare derivatives o f XeFö other than oxyfluorides a n d oxides have s o far failed. Thus HCl a n d N H 3 interact with XeFö according t o the following equations234: XeF6+6HC1 -> X e + 3 C I 2 + 6 H F XeF6+8NH3

Xe+6NH4F+N2

It is reasonable t o suppose that substitution b y highly electro-negative groups, as in the case o f XeF2 (see section 3.2.4), c a n occur. It is n o t surprising, in the light o f the physical evidence o n the considerable fluoride i o n d o n o r ability o f XeFö, that the fluoride should form X e F ^ salts with fluoride i o n acceptors. The salts are described in section 3.4.6. The greater fluoride i o n d o n o r ability o f XeFö a n d XeFa relative t o X e F 4 provides for a chemical purification o f the X e F 4 (see section 3.3.1). Although there is n o firm physical evidence t o support the c o m p l e x a n i o n X e F 7 - the alkali salts A X C F T are k n o w n . Of the A 2 X e F 8 salts, ( N O ) 2 X e F 8 h a s been s h o w n t o b e t h e salt (NO+)2(XeF8)2-. These "salts" are described in section 3.4.7. The s o d i u m salt, 2 N A F . X e F ö , forms readily w h e n the c o m p o n e n t s are mixed at 50!, but d e c o m p o s e s under v a c u u m at 125!. This p r o v i d e s 2 0 4 for the purification o f XeFö (see laboratory preparation). Analysis and characterization. The hexafluoride is best characterized b y its v a p o u r infrared spectrum. It is necessary t o use AgCl w i n d o w s t o withstand t h e chemical attack. Care should be taken t o scan the 928 c m ~ i region for signs o f XeOF4 (the m o s t c o m m o n impurity). Analysis c a n b e carried o u t in m u c h the same w a y as for X e F 4 b y using H2 or Hg as the reducing agent202,107. The latter is preferred.

229 230 231 232 233 234

D . F . Smith, Science 140 (1963) 899. J. Shamir, H . Selig, D . Samuel and J. Reuben, / . Am, Chem, Soc, 87 (1965) 2359. B. Cohen and R. D . Peacock, / . Inorg, Nucl, Chem, 28 (1966) 3056. J. G. Malm, F . Schreiner and D . W . Osborne, Inorg, Nucl. Chem, Letters 1 (1965) 97. D . F . Smith, / . Am, Chem, Soc, 85 (1963) 816. G. L. Gard and G, H . Cady, Inorg, Chem, 3 (1964) 1745.

XENON(VI) COMPOUNDS

303

3.4.2. Xenon Oxide Tetrafluoride Preparation. The oxyfluoride X e O F 4 was ñrst detected by mass spectroscopyss among the xenon fluorides prepared by thermal excitation, and was soon isolated in macroscopic quantities by the partial hydrolysis of XeFö 229, 2 3 0 : XeFö + HzO ^ XeOF4 + 2 H F

Unreacted XeFe and H F are removed by treatment with N a F (which forms compounds with both)23i.232. The best reported procedure is that of Smith229^ who used a circulating loop incorporating an infrared cell for monitoring the interaction. In this arrangement, air saturated with water vapour is bled into the circulating loop, ñlled with XeFe near its saturation vapour pressure, and the XeFö consumption is monitored by following the intensity of the XeF^ band at 520 c m - i . Yields of 80%, based on XeFg consumption, have been obtained. The static methods, involving interaction of XeFe with H 2 O or S i 0 2 235»ii2^ are hazardous if not carried out with great care, and it is probably better, if efl&cient and largescale synthesis are not important factors, to prepare the compound by heating Xe/Fi/Oa mixtures to 235°, the Xe:F2 ratio being ' ^ 1 : 4 and the oxygen in considerable excess ( 2 i times the F2 content)i35. The last procedure yields X e F 4 as a major impurity, but the much greater volatility of the X e O F 4 permits ready separation by vacuum distillation at ^0°. This method is no more hazardous than XeFe synthesis. Whenever XeF^ is handled in an apparatus, which has.not been previously fluorinated or "pickled" with XeFg, the oxyfluoride X e O F 4 is produced. This occurs so readily that it was a source of some confusion in the early studies involving XeFö 236. Some physical properties. The hazards associated with the preparation of X e O F 4 have restricted its study. The hmited physical data are summarized in Table 30.

TABLE 3 0 . SOME PHYSICAL PROPERTIES OF X e O F 4

Colourless solid, liquid and vapour" Melting point f C ) :

- 2 8\ - 4 1 ^ - 4 0 S - 4 6 . 2 ^

Vapour pressure ^ Τ CO

P^^

0 0

7.0 8

23 Density

(g c m - 3 ) : d =

Optical

29

3.168-0.0032Γ«

properties

Refractive index *:

(Cauchy relation) η = 1 . 4 0 7 5 3 + - —

(Measurements at 4 3 5 8 and 4 4 7 1 A ) temp, coeff. o f n: Molar refraction RD ( 2 5 ° C ) = 1 8 . 2 4 c c . m o l e - i Dipole moment ^: 0 . 6 5 ± 0 . 0 9 D Dielectric

constant ( 2 4 ° C ) : 2 4 . 6

Volume susceptibility

235

See ref. 1 4 , p. 1 0 6 .

236

See ref. 1 4 , p. 5 0 .

T . D . C.I.C. Vol. I — L

= 0.82

± 0.0003

-0.00049

304

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O, SLADKY TABLE 30 (com.)

Infrared and Raman spectra '· CAV Symmetry IR R 920(2) Ρ Γ VI 926 s 567 (10) Ρ 576 m ai] 294 s 285 ( 0 + ) Ρ i 527 (4) 230 (calc.) bz 608 vs e^ 365 (2) 361 s 161 ( 0 + ) V9

ι

Molecular

dimensions

Microwave spectroscopy '· * Xe-O 1.703 ± 0 . 0 1 5 A Xe-F 1.900 ± 0 . 0 0 5 ^ O - X e - F 91.8 ± 0 . 5 ° * Nuclear magnetic resonance

spectra

170 shift: (7o = - 3 1 3 ± 2 ppm (σ, Hji^O = 0), / ( i 2 9 X e - i 7 0 coupling) = + 6 9 2 ± 10 cps ^ 19F shift: ffF = - 1 0 0 . 2 7 ppm (σρ, CCI3F = 0 ) = 330 ppm (σρ, F2(g) = 0) ·'· '· " y (i29Xe-i9F coupling) = 1127 \ 1124*^ i29Xe shift: σχβ = - 5 5 1 1 ppm (σχ., Xe(g) = 0)·" Photoelectron

spectroscopy

"

AE (XeOF4(g)) = 7 . 0 2 ± 0 . 1 3 eV (relative to Xe(,>) Electrical conductivity ^ Specific conductivity (at 24°C): 1.03 χ 10-5 o h m - i cm-2, increases markedly with C s F or R b F added: specific conductance (0.29 Μ C s F soln. in XeOF4): 8.5 χ lO-i o h m - 3 cm-2 Mass spectrum ** Positive ions: all unipositive ions observed but X e O F j / X e O F J ratio ^ 100:1 Negative ions: XeF", XeF^, X e F ] , XeF^, X e O F j • C. L. Chemick, H. H. Claassen, J. G. Malm and P. L. Plurien, in Noble Gas Compo unds (H. H . Hyman ed.), University of Chicago Press, Chicago and L o n d o n (1963), p. 106. " D . F. Smith, Science 140 (1963) 899. H. D . Frame, J. L. Huston and Irving Sheft, Inorg. Chem. 8 (1969) 1549. ^ H. Selig, Inorg. Chem. 5 (1966) 183. ' H . Selig, C. L. Chemick and C. W. Williams, Inorg. Nucl. Chem. Letters 1 (1965) 17. ' J, F. Martins and E. B. Wilson, Jr., / . Molec. Spectrosc. 2 6 (1968) 410. « G. M. Begun, W, H. Fletcher and D . F. Smith, / . Chem. Phys. 42 (1963) 2236. " See ref. a, p. 287. • J. Martins and E. B. Wilson, Jr., / . Chem. Phys. 41 (1964) 570. ^ J. Shamir, H. Selig, D . Samuel and J. Reuben, / . Am. Chem. Soc. 87 (1965) 2359. •'See ref. a, p. 251. • A. C. Rutenberg, Science 140 (1963) 993. " See ref. a, p. 263. " S. E. Karlsson, K. Siegbahn and N . Bartlett, / . Am. Chem. Soc. (1970) (in press). ° See ref. a, p. 47.

The c o m p o u n d i s c o l o u r l e s s i n a l l p h a s e s , i s low m e l t i n g (—46.2°) a n d e a s i l y v o l a t i l e 2 3 7 . It i s a n t i c i p a t e d t o b e t h e r m o d y n a m i c a l l y stable238 a n d a l l o b s e r v a t i o n s i n d i c a t e t h a t it i s s o .

The n . m . r . d a t a ( i ^ F , n o , i 2 9 X e ; s e e Table 30) a r e c o n s i s t e n t w i t h t h e l i q u i d b e i n g 237 H. Selig, Inorg. Chem. 5 (1966) 183. 238 s . R. Gunn, / . Am. Chem. Soc. 87 (1965) 2290.

XENON(VI) COMPOUNDS

non-associated (unlike

XeFö),

a n d t h e l o w electrical conductivity o f t h e pure liquid s h o w s

that autoionization is very limited.

Nevertheless,

c o n s t a n t (24.6 a t 24*") a n d d i s s o l v e s t h e a l k a l i the electrical conductivity237. of

8.5 X 10-3

o h m - i cm-2 at

305

t h e liquid h a s a moderately high dielectric

fluorides,

with considerable enhancement o f

Thus a 0.29 Μ C S F s o l u t i o n 24° 2 3 7 . Although XeOF4

p o s s e s s e s a specific dissolves

XeFe

conductivity

a n d is

miscible

w i t h HF, t h e electrical c o n d u c t a n c e is n o t m a r k e d l y affected b y their addition.

Structural features. Like t h e " i s o e l e c t r o n i c " h a l o g e n p e n t a f l u o r i d e s 3 5 a n d XeFs" ( s ^ ^ 3.4.1 a n d p a r t i c u l a r l y s e c t i o n 3.4.6 f o r a s t r u c t u r a l c o m p a r i s o n ) , XeOF4 i s k n o w n f r o m n . m . r . i 2 i , vibrationaU^o» a n d m i c r o w a v e s p e c t r o s c o p y 2 3 9 to b e s q u a r e - b a s e d section

p y r a m i d a l i n s h a p e , w i t h t h e o x y g e n l i g a n d a p i c a l (C41,). The m i c r o w a v e d a t a i n d i c a t e t h a t the xenon a n d four

The X e - F b o n d (1.890 A) 217 a n d (1.703 + 0.015 A)

fluorine

length

239

significantly shorter t h a n f o r is

shorter

( ^ O - X e - F = 91.8+0.5°). is v e r y s i m i l a r t o t h a t f o u n d f o r XeFö XeF4 (1.95 A) π». The X e - O b o n d l e n g t h i n e i t h e r XeOa (1.76 A) 240 o r X e 0 4

hgands are very nearly coplanar

(1.900 ±0.005 A) than

observed

TABLE 3 1 . COMPARISON OF B O N D STRETCHING F O R C E CONSTANTS ( m d y n A - I ) OF

XeOF4 WITH

THOSE OF RELATED MOLECULES

XeOF4 k r (Xe-F) k r (Xe-O)

3.21» 7.11-

XeF4 3.26« 7.08*^

XeOa

Xe04

5.66*»

5.75 « (6.7)t

3.00«»

—

• D . F . Smith, Science 140 (1963) 899. H. H . Claassen, C. L. Chernick and J. G. Malm, / . Am. Chem. Soc. 8 5 (1963). G. M . Begun, Ψ. H . Fletcher and D . F . Smith, / . Chem. Phys. 4 2 (1963) 2236. D . F . Smith, in Noble Gas Compounds ( H . H . H y m a n , e d . ) . University o f Chicago Press, Chicago and L o n d o n (1963), p . 295. * W. A . Veranos, Bull. soc. chim. Beiges 7 4 (1965) 414. This value was derived using a Urey-Bradley force field and an assumed value for vi. t Converting this to a valence- force field value yields a k r ( X e - O ) o f 6.7 m d y n A " i .

(1.74 A) 241. The force constants given in Table 31, derived from the vibration data, are consistent with these observations, the X e - O bond evidently being stronger than in XeOa or Xe04. Bonding and bond polarity. The observed geometry is as predicted by valence-electronrepulsion t h e o r y 3 9 , and by the three-centre four-electron bond description32.33 (see section 1.3). In the former representation the molecule is pseudo-octahedral, the non-bonding valence electron pair being on the four-fold molecular axis trans to the oxygen a t o m 2 i 4 » . The X-ray photo-electron spectrumi26 of XeOF4(g) yields a xenon core-electron chemical shift intermediate between XeF4 and XeFg as shown in Table 32. These data 239 J. F . Martins and E, B. Wilson, Jr., / . Mol. Spectrosc. 2 6 (1968) 410. 2 4 0 D . H . Templeton, A . Zalkin, J. D . Forrester and S. M . Williamson, / . Am. Chem. Soc. 8 5 (1963) 817. 241 G. Gundersen, K. Hedberg and J. Huston, Acta Cryst. 25 (1969) 124; / . Chem. Phys. 5 2 (1970) 812. 241« The steric activity o f a n oxygen ligand appears t o be comparable t o that o f a non-bonding valence electron pair, thus XeOs (see section 3.4.4) is pseudo-tetrahedral with -^O-Xe-Oav = 130°.

306 TABLE

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY 32. X-RAY

PHOTO-ELECTRON

CHEMICAL SHIFTS FOR XeMy

CORE-ELECTRONS I N GASEOUS

COMPOUNDS *

XeF4(,>

XeOF4(„

XeF6(g)

- 2 . 9 5 (13)

- 5 . 4 7 (18)

-7.02(13)

- 7 . 8 8 (18)

-1.48

-1.37

XeF2(g)

X e M v chemical shift (eV) Shift per fluorine atom

0

—

» S.-E. Karlsson, K. Siegbahn and N . Bartlett,

-1.31

—

Am. Chem. Soc. (1970) (in press).

show that the ligands remove electron density from the xenon atom (the expelled coreelectrons being more bound than in atomic xenon). The XeOF4 shift is seen to be closer to that of XeFö than that of XeF4. The shifts have been interpreted quantitatively in terms of a simple coulombic modeU26. This assumes a spherical, positively charged xenon atom of charge -i-q and radius r«, and spherical, negatively charged at a hgands distance R L from the xenon atom. Thus for XeOF4 A W . =

-(í-9o)(¿-¿)-.o(¿-¿)

and by equating the first term with

AEXCF^

or f

A^XCF^,

Thus the dependence of qo upon has been evaluated and compared with a similar interdependence for the xenon fluorides. The findings are as follows:

X e n o n valence shell radius (A) Oxygen ligand charge go Fluorine ligand charge ^pf t Essentially constant for all xenon

1.5 1.5 0.63

1.4 0.91 0.48

1.3 0.64 0.37

1.2 0.47 0.30

1.1 0.36 0.20

1.0 0.28 0.16

0.9 0.22 0.13

fluorides.

They show, no matter what the choice of the xenon valence-shell radius, that the oxygen ligand withdraws more electron density than a fluorine ligand. This is compatible with the valence-electron-pair theory, if multiple bonding is allowed for the X e - O bond, but is also in harmony with the three-centre orbital model, which, in its simplest representation, yields a charge distribution Xe+3(F^)40-i for the oxyfluoride. In the latter view the X e - O bond is a semi-ionic bond Xe+ :0~. The considerable polarity of the X e - O bond predicts a large dipole moment for XeOF4 ( > 4 D) if the non-bonding valence electron pair is sterically inactive (i.e. in the Xe 5^ orbital). Since the observed dipole moment239 is only 0.65+0.09 D , at least considerable xenon valence-shell polarization occurs, or else the non-bonding "pair" resides in a sterically active orbital (e.g. an sp hybrid).

XENON(VI) COMPOUNDS

307

The

n.m.r. spectrum o f i 7 0 X e F 4 s h o w s a resonance (σ, - 3 1 3 p p m relative t o d o w n field from aqueous XeOa 2 4 2 , 2 4 3 which has been i n t e r p r e t e d 2 3 0 i n terms o f greater double b o n d character for X e - O in X e O F 4 than in X e 0 3 ( a q ) . T h e resonance is, however, t o higher field than m a n y " d o u b l e - b o n d " oxygen c o m p o u n d s (σ, —500 t o - 6 0 0 p p m relative t o Hi^^O). A locahzed orbital model (employing X e 5d crystals) has been givenii^ t o account for the observed n.m.r. chemical shifts and c o u p h n g constants. Ligand exchange in the system X e 0 2 F 2 / X e O F 4 . A i ^ p n.m.r. study a n d i s F radiotracer i n v e s t i g a t i o n 2 2 8 i n the system X e 0 2 F 2 / X e O F 4 has s h o w n the half hfe for ñuorine ligand exchange t o be < 7 m i n at 0° and > 4 s e c at 70°, but a detailed kinetic study w a s n o t carried out. Chemical properties of XeOF4. T h e oxyñuoride hydrolyses further, evidently in stepwise H2I7O)

fashion235: X e O F 4 + H 2 0 -> X e 0 2 F 2 + 2 H F Xe02F2+H20

Xe03 + 2HF

but it has n o t proved possible t o control the hydrolysis t o generate macroscopic yields o f X e 0 2 F 2 . T h e usual product is X e O a 2 2 9 . T h e oxyfluoride interacts similarly with SÍO2, especially at elevated temperatures. T h e ready formation o f X e O a renders investigations o f X e O F 4 (also X e F 4 and X e F ö ) hazardous, particularly if carried o u t in oxide-containing apparatus. Adducts o f X e O F 4 with alkah fluorides (CsF, R b F , K F , b u t n o t N a F ) m a y contain the XcOFj i o n , or polymers o f it (see section 3.4.9). T h e adducts formed with strong fluoride ion acceptors (see section 3.4.8) probably contain the X e O F J ion. T h e 1:1 molecular adduct formed between X e F 2 and X e O F 4 has been discussed in section 3.2.7. T h e formation and structure o f this adduct are consistent with the high b o n d polarities discussed above. Analysis and characterization. T h e c o m p o u n d is m o s t readily characterized a n d detected by infrared or Raman spectroscopy (the X e - O stretch at 926 (IR), 920 ( R ) , is very characteristic, particularly in conjunction with the X e - F stretch at 608 vs (IR), 567 v s ( R ) ) . Analysis has been accomplished235 by interaction o f X e O F 4 with H2 in a nickel can at 300°: X e O F 4 + 3 H 2 -> X e + H 2 O + 4 H F

but it is probable that similar reduction with mercury w o u l d prove t o be m o r e convenient.

3.4.3.

X e n o n D i o x i d e Difluoride

Preparation, Mass spectroscopy gave the first indications o f the existence o f X e 0 2 F 2 . It w a s subsequently prepared b y H u s t o n 2 4 4 b y mixing X e O a with X e O F 4 . T h e latter is distilled o n t o the former which is cooled t o dry-ice temperature. (Dry ice is used t o minimize detonating the X e O a by thermal shock.) T h e X e O a is allowed t o dissolve i n the X e O F 4 (overnight). T h e resulting mixture o f X e 0 2 F 2 , X e O F 4 a n d X e F 2 is fractionally distilled i n a K e l - F apparatus. T h e X e O F 4 , being more volatile, is readily removed. T h e difluoride is slightly more volatile than X e 0 2 F 2 . It is possible that chemical purification could be achieved by complexing the X e F 2 with AsFs (see section 3.4.1), but this supposes that X e 0 2 F 2 does n o t form a salt or adduct with AsFs. 242 J. Reuben, D . Samuel, H. Selig and J. Shamir, Proc. Chem. Soc. (1963) 270. 243 H. H. Claassen and G. Knapp, / . Am. Chem. Soc. 86 (1964) 2341. 244 J. L. Huston, / . Phys. Chem. 71 (1967) 3339.

308

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY

Physical properties. The sohd, hquid and vapour are colourless, and the solid low melting (29.5-30.5°). Although thermodynamically unstable with respect to XeFa and O2, it can be kept at room temperature for several days in preconditioned Kel-F containers. TABLE 33. SOME PHYSICAL ΡΚΟΡΕΚΉΕ8 OF X e O i F a

Colourless solid, liquid and vapour" Melting point: 30.8°C« Δ ^ ; (Xe02F2(e)) estimated »»: + 5 6 kcal m o l e - i Volume susceptibility Xv = - 0 . 8 6 « Infrared and R a m a n (liquid and solid) **

Raman

Solid

Liquid

205 m s 224 w 315 v s

198 w 223 vw 313 m s

350 ms 537 vs

333 m s 490 s

769 814 850 882

Mass spectra • Positive ions:

w w vs s

578 w 788 vw

Infrared (Ar matrix)

Assignment C2v symmetry

V4(ai) 317 m s 324 s

537 550 574 585

vgibi) viibi) V3(öl) V2(öl) V5 + V9 54ί(Βι) V5 + V7 547(^2)

vs w w vs

?

vsibz)

? 845 vs 905 w

848 m s 905 s 1023 w 1444 w 1496 vw

Vl(öl)

V6(bú

?

V1 + V8 1433(^2)

?

X e 0 2 F i (0.89), X e 0 2 F + (0.66), X e O F J (4.8), X e O j (1.6), X e F j (3.3), X e O F + (2.3), XeO+ (2.3), XeO+ (1.0), XeF+ (4.1)

Negative ions: X e F " , X e F 2 , X e O F " ; n o evidence for heavier ions • J. L. Huston, / . Phys, Chem. 71 (1967) 3339. S. R. G u n n , / . Am, Chem, Soc. 87 (1965) 2290. " H. D . Frame, J. L. H u s t o n and I. Sheft, Inorg. Chem. 8 (1969) 1549. H. H. Claassen, E. L. Gasner and H. K i m , / . Chem. Phys. 49 (1968) 253.

The vibrational spectra, represented in Table 33, indicate that the molecule is of Czv symmetry. Such an assignment is compatible with the observation that the isoelectronic species IO2FJ is of Czv symmetry245, being pseudo-trigonal bipyramidal ( ^ O - I - O = 180°; ^ F - I - F = 110°). Chemical properties. The compound interacts with XeFö, with hquefaction, to yield XeOF4 XeF6+Xe02F2 245

See ref. 168, Vol. III.

2XeOF4

XENON(VI)

309

COMPOUNDS

It rapidly hydrolyses in moist air to yield XeOs, but a faint ozone-like odour reminiscent of X e 0 4 can be DISCEMED244.

Analysis, The compound has been analysed by decomposing it in a quartz container at 300°. The mixtures of Xe, O2 and SÍF4 were SÍO2

X e 0 2 F 2 — • 1.0Xe+1.5O2+0.45SiF4 1.0Xe+1.5O2+0.5SiF4

Found Calculated

analysed mass spectrographically, with relative sensitivities to the three gases calibrated by means of a known mixture. 3.4.4. Xenon Trioxide Historical note and preparation. In their first report^ on the synthesis of X E F 4 , Claassen, Selig and Malm noted that the hydrolysis of the solid yielded initially a yellow sohd (now considered to be X e O F 2 ; see section 3.3.3) which dissolved to yield a clear, pale yellow solution. Cady and his coworkers, in their first REPORTAOS of XeFö, noted that the hydrolysis of the ñuoride yielded a solution containing an oxidizing xenon species which they assumed to be xenic acid Xe(OH)ö. Simultaneous with the latter investigation, SMITH233 observed that XeFö exposed to moist air yielded a sohd product, which proved to be xenon trioxide. Independently, Wilhamson and Koch discoveredi»« that the sohd recovered by evaporation of a hydrolysed X E F 4 solution was also XeOa 24o. Since XeOa is a powerful explosive, great care must be exercised in its preparation (and, indeed, in the handling ö / X e F 4 and XeFö since the oxide is formed when these interact with moisture). The oxide is most efficiently prepared from XeFö, and two detailed procedures have been GIVEN246,247, The method given by Huston and his coworkers is probably the safer. Figure 19 illustrates the experimental arrangement for the controlled hydrolysis. -Dry

Kel-F tube _ Ö 1

N2

Hoke A431 valve / 1

^

ΐ^,ΤβΠοη bottle , Monel U tube containing XeF^'

-Water

F i o . 19. Apparatus for hydrolysis o f x e n o n hexafluoride. 246 E. H. Appelman and J. G. Malm, in Preparative Inorganic Reactions (W. L. Jolly, ed.). Vol. Π , Interscience, N e w York (1965), p. 349. 247 B. Jaselskis, T. M. Spittler and J. L. Huston, / . Am, Chem, Soc, 88 (1966) 2149.

310

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY

The XeFö, contained in a U-shaped Monel vessel, has a vapour pressure of ' ^ 3 0 m m at room temperature, and this vapour is swept in a stream of dry nitrogen into water con­ tained in a Teflon bottle: XeF6(,) + 3H20(,) -> 6HF(aq) + X e 0 3 ( a q )

The hydrolysis of gram amounts takes several hours. It is essential that a stream of dry nitrogen be maintained at aU times, since water should not gain entrance to the XeFö container. If it is necessary to remove the H F from the aqueous solution, this can be achieved by treating the solution with magnesium oxide (MgO + 2 H F - > M g F 2 + H 2 0 ) to make a TABLE 34.

SOME PHYSICAL ΡΚΟΡΕΚΉΕ8 OF XeOa

Colourless, hygroscopic, detonatable solid with low vapour pressure at 20° Thermodynamic

features

A i / s u b (est.) ¿\H} (298.15°K) (s) Mean thermochemical

3 0 ± 10 kcal m o l e - i « + 9 6 ± 2 kcal m o l e - i

bond

energy

17.5±4kcal5°, Xe03(g) 68.69 cal d e g - i m o l e - i ^ Index of refraction: α = 1.79 Vibrational spectra * Infrared s p e c t r u m ' (solid) v
v<>| v^V v'J>

770cm-i 311 820 298

X e - O stretch force constant k,\ 5.66 mdyn A~i Raman spectrum (of 2.0 Μ aqu. soln.) C3. Vl 780 c m - i V2 344 V3 833 V4 317 Crystal structure (X-ray) Orthorhombic 7 = 4 fl = 6 . 1 6 3 ± 0 . 0 0 8 A V = 2 6 2 A3 b = 8.115±0.010 c = 5.234 ± 0 . 0 0 8 ^ m o i . = 3 9 . 4 cm3 i/x-r.y: 4 . 5 5 g cm-3 Space group: Ρίχΐχΐχ Xe-O

1.74 ± 0 . 0 3 1.76 1.77

^O-Xe-O

108 ± 2 ° 100 101

See Fig. 20

A

I av. J iy av. J

1.76

{p)

A

103°

« S. R. Gunn, / . Am. Chem. Soc. 87 (1965) 2290. ^ S. R. Gunn, in Noble Gas Compounds (H. H. Hyman, ed.). University of Chicago Press, Chicago and London (1963), p. 149. G. Nagarajan, Bull. soc. chim. Beiges 7 3 (1964) 665. ^ D . F. Smith, / . Am. Chem. Soc. 86 (1963) 816. « H. H. Claassen and G. Knapp, / . Am. Chem. Soc. 85 (1964) 2341. ' See ref. b, p. 295. » D . H. Templeton, A . Zalkin, J. D , Forrester and S. M. Williamson, / . Am. Chem. Soc. 85 (1963) 817.

XENON(VI)

COMPOUNDS

σ

F I G . 20. The molecular shape and packing in XeOa (cryst). (See Table 33 for unit cell data.)

311

312

NOBLE-GAS CHEMISTRY.* NEIL BARTLETT AND F. O. SLADKY

slightly alkaline slurry. Following filtration through a sintered glass filter of medium porosity, the magnesium in solution is removed by passage through a column of hydrous zirconium oxide, converted to the nitrate form by exhaustive washing with 0.1 Μ H N O 3 . The aqueous XeOa solutions may be kept indefinitely if oxidizable impurities are excluded. It is important to exercise the greatest care if the solution is evaporated to dryness since the solid oxide may detonate. Physical properties (Table 34). The oxide is colourless. The sohd deliquesces at humidities greater than ^25%. The enthalpy of subhmation has been estimated from mass spectro­ metric observations238 t o be 3 0 + 1 0 kcal mole-i. Calorimetry has given 248 AH^ = + 9 6 k c a l m o l e - i from which an X e - O bond energy of ~ 17.5 kcal mole-i has been derived. The XeOa molecule is pyramidal and almost identical in shape and size to the lOj i o n 2 4 9 , 240. The bond angles 0 - X e - O are closer to angles than right angles. The structure is shown in Fig. 20. Bonding and bond polarity. The pseudo-tetrahedral geometry of the XeOa molecule conforms to the representation of the bonding as xenon electron-pair donation, Xe+: O", the xenon orbitals being sp^ hybrids. This model also imphes that the oxygen ligand valence state might be a singlet (see section 1.3). The shortness of the X e - O bonds relative to X e - F bonds and the higher force constants of the former (see Table 31) indicate that the intrinsic Xe-O bond energies are greater than the intrinsic X e - F bond energies. As has already been remarked (see section 1.3,2), there are indications that the intrinsic energies of X e - O bonds are not significantly different from transition metal oxygen hgand bonds, usually considered multiple, e.g. O s - 0 in O S O 4 . The greater strength of the X e - O bonds, relative to X e - F , may therefore be a consequence of the greater electron affinity of singlet oxygen (}D) or, alternatively, a result of multiple bonding of the oxygen to xenon. The latter explanation implies the involvement of "outer" xenon orbitals (e.g. Xe 5d). The considerable kinetic stability of the xenon oxides must be of significance to bonding theory, but this point has not been given serious theoretical consideration. This kinetic stability contrasts with the evident kinetic instability of the chlorides and "heavier" halides. Presumably the difference lies in a readier formation, on the part of the halogen ligands, of an energetically favourable ligand diatomic, by electron transfer back to the xenon atom:

Xe

^

Xe

There must be considerable polarity in the X e - O bond and the three β near-neighbour Xe . . . O distances of 2.8, 2.89 and 2.90 Ä must represent coulombic interactions -Xe+* . . . O-^-. The low volatihty of the compound is compatible with such interactions. Chemical properties. The thermodynamic instability of the solid oxide and its high solubility and kinetic stability in aqueous solution has resulted in most work on the oxide being involved with aqueous solutions. The chemistry of the aqueous solutions is discussed in the following section. 248

See ref. 14, p. 149.

249

See ref. 14, p. 229.

XENON(VI) COMPOUNDS

313

The oxide also forms salts with the alkali hydroxides (see section 3.4.11) and with alkali fluorides and chlorides (see section 3.4.12). Cationic derivatives are not known. There is also some evidence^oo, 2 5 0 for ester-like XeOa-alcohol intermediates, in the oxidation of alcohols. A solution of XeOa in t-butyl a l c o h o l 2 5 0 has been titrated with potassium or rubidium t-butoxide, using a p H meter with glass-calomel electrodes. The titration curves are similar to those given by glacial acetic acid i n the same solvent and have end points at 1:1 molar ratio. Unstable precipitates formed during the titration, analysis of which indi­ cated empirical formulae t - B u O - X e 0 2 - O M t - B u O H or MHXe04-2t-BuOH (M = Κ or Rb). Attempts to isolate a XeOa-t-butyl alcohol ester failed and XeOa was lost on con­ centration beyond 0.4 M—vapour phase decomposition of ester-like species is possible. Analysis, The trioxide is most conveniently analysed by iodometric titrationisi. Excess sodium iodide is added to an aqueous acid: X e 0 3 + 6 1 - + 6 H + -> X e + 3H2 O + 3 I 2

The liberated I2 is titrated with standard thiosulphate to an amylose end point. It is of interest that if acid is added before the I", any perxenate decomposes to xenon(VI) and oxygen, whereas if I " is added first, all of the oxidizing power is captured as triiodide. 3.4.5. Aqueous Xenon Trioxide C'Xenic Acid'') Preparation. Aqueous XeOs solutions are preparedi^i. 2 4 7 as described under XeOa. Solutions more concentrated than 2 Μ in XeOs may be obtained. Physical properties. The physical properties of aqueous XeOa are given in Table 35. Solution calorimetric measurements have y i e l d e d 2 5 i ¿^^Hf (Xe03-ooH20) at 298.15°K to TABLE 35.

Solubility of X e 0 3 in wateri

SOME PHYSICAL PROPERTIES OF A Q U E O U S X e 0 3

> 2 Μ•

Thermodynamic features AH^m 3.9 ± 2.0 kcal m o l e - i AH} (Xe03-oo H2O, 2 9 8 . 1 5 ° K ) 9 9 . 9 4 ± 0 . 2 4 kcal m o l e - i AS} - 7 0 ± 4 cal d e g - i m o l e - i AG}(XeOyoo H2O, 298.15°K) + 1 2 0 . 8 ± 1.2kcal m o l e - i Electrode potentials of the X e / X e 0 3 couple Acid: EX ( X e ( , ) + 3 H 2 0 ( i ) -> X e 0 3 ( . q ) + 6H(íq, + 6 e - ) = 2 , 1 0 ± 0 . 0 1 V Base: E^ (Xe(g) + 70H(rq) -> H X e 0 4 ( . , ) + 3 H 2 0 ( i , t 6 e - ) = 1 . 2 4 ± 0 . 0 1 V Equilibrium constant': ( H X e O ^ o , ) -> Xe03(aq) + 0H(¡"q)) = ( 6 . 7 ± 0 . 5 ) x 1 0 - 4 Raman spectrum: see Table 33 Electrical conductivity Molal freezing-point

·:

Molar conductance (0.02 Μ Xe(VI) soln.):

depression":

< 0 . 0 4 at 25°

1.95±0.15°

• E. H. Appelman and J. G. Malm, / . Am. Chem. Soc, 86 (1964) 2141. ^ P . A . G. O'Hare, G. K. Johnson and E. H. Appelman, Inorg. Chem, 9 (1970) 332.

be +99.94+0.24 kcal mole-i. This implies A ^ s o m = 3 . 9 + 2 kcal mole-i. The Xe/Xe(VI) redox potentials, derived from the aqueous solution calorimetric data, show XeOa to be one of the strongest oxidants in aqueous media: Acid: Xe(,) + 3H20(i) -> Xe03(aq) + 6H+(aq) + 6 e - , E^ = 2 . 1 0 ± 0 . 0 1 V Base: Xe(,) + 7 0 H - ( . „ -> H X e 0 4 - ( a q ) + 3 H 2 0 + 6 e - , E^ = 1 . 2 4 ± 0 . 0 1 V 250 B. Jaselskis and J. P . Warrier, / . Am. Chem. Soc. 9 1 (1969) 201. 251 p . A . G. O'Hare, G. K. Johnson and E. H. Appelman, Inorg. Chem. 9 (1970) 332.

314

NOBLE-GAS CHEMISTRY: NEIL BARTLETT A N D F. O. S L A D K Y

These potentials may be compared with those o f the well-known oxidizers cerium(IV) (E° = —1.61 V) and O 3 (E" = —2.07 V). Xenic acid is reduced at the dropping mercury electrode in a single step to x e n o n 2 5 2 . The Raman spectrum of 2 Μ solution of XeOa e s t a b l i s h e s 2 4 3 that the primary solution species is molecular XeOa. This is further supported by the low electrical conductivity of the solutionisi. A i^o n . m . r . s t u d y 2 4 2 shows that the oxygen hgands undergo fast exchange with the solvent, equilibrium being established within 3 min at 23 + 10°. The i^o chemical shift (—278+2 ppm with respect to water) is in the same range as those of perchloric acid ( - 2 8 8 ) and BrO^, ClOj and C I O 4 ( - 2 9 7 , - 2 8 7 , - 2 8 8 ) . This study also confirmed that XeOs may be extracted from aqueous solution with C H C I 3 iss. An equihbrium constant, 6.7 + 0.5 χ 10-4, has been reported^si for the process HXe04-(aq)

Xe03(aq) + OH-(aq)

This accounts for the variation in the ultraviolet spectrumisi as a function of p H . Shoulders at 210 and 250 mμ which are evident in acid media are presumably associated with X e 0 3 , and a shoulder which appears at 265 ιημ in strong base may be associated with H X e O j . The isolation of salts, e.g. C s H X e 0 4 (see section 3.4.11), supports this rational. Disproportionation. In strong base X e 0 3 ( H X e 0 4 ? ) d i s p r o p o r t i o n a t e s 2 5 3 , 1 5 1 . J h e yield o f perxenate at - 2 5 ° is - 3 3 % for N a O H solutions of 0.25 -> 4.2 M and K O H solutions of 2 3.6 Μ 253. Higher O H " concentration and higher temperatures increase the yield. With LiOH the yield of perxenates is claimed to be 5 0 % or higher. The dis­ proportionation is complex and there are probably several routes to perxenate. In at least one experiment, howeverisi, the reaction appeared to b e : 4HXe04- + 5 0 H -

BHXeOe-^ + X e + 3H2O

Complexes of xenon(VI) and (VIII) do play a role in K O H (where a yellow sohd appears) and N a O H (yellow solution), but must have small influence in LiOH solution where no appreciable amount of Xe(VI)-Xe(VIII) complex appears to be formed. A complex of composition K4Xe06-2Xe03 has been isolated from the K 0 H / X e 0 3 ( a q ) s y s t e m 2 5 4 but other complexes probably occur. Evidence has been presented to show that these Xe(VI)-Xe(VIII) complexes decompose predominantly according to the e q u a t i o n 2 5 3 [ X e u , O 4 + 3 . ] -> [ X e 0 4 ] + : ^ X e + 1.5j^02

It appears that the higher the perxenate concentration the less likely is the X e 0 3 to dis­ proportionate as follows: 2 X e 0 3 -> [ X e 0 2 ] + [ X e 0 4 ] ; [XcOi] ^ X e + O i

The better than 5 0 % yield of perxenate obtained in certain N a O H solutions and the LiOH system indicate that X e 0 3 ( H X e 0 4 ) also interacts with a lower oxidation state than xenon(VI). The sequence X e 0 3 + [XeOzl -> [ X e 0 4 ] + [XeO]; [XeO]

appears more likely than X e 0 3 + [ X e O ]

[Xe04]

Xe+i02,

+ Xe, since

XeF2(aq)

Xe03(aq,

252 Β. Jaselskis, Science 143 (1964) 1324. 253 c . W. K o c h and S. M. W^illiamson, J. Am. Chem, Soc. 86 (1964) 5439. 254 T. M . Spittler and B. Jaselskis, / . Am. Chem. Soc. 8 8 (1966) 2942.

is known to reduce

XENON(VI) COMPOUNDS

315

Oxidation of inorganic ions. The oxidizing capabihty o f XeOacaq) is i n accord with the oxidation potentials cited above. Iodide is oxidized253, i s i rapidly in acid solution (but this is slow above p H 7), and calorimetry gives^si AÄ^(Xe03(aq) + 9I-(aq) + 6H+(aq) ->Xe(g) + 3l3-(aq) + 3H20(i)) = - 2 1 9 . 6 3 ± 0 . 0 6 kcal mole-1 Bromide and chloride are also oxidized t o the free h a l o g e n i s i . Acidic manganese(II) solutions are oxidized t o Μ η θ 2 over several hours, and after a day or t w o M n O j is detect­ able. In 2 Μ HCIO4, h is oxidized t o iodate a n d the rate is greater in 6 Μ acid. Under the latter conditions, Br2 is oxidized t o BrOj, The kinetics o f the interaction o f Xe03(aq) with plutonium(III) solutions have been e x p l o r e d 2 5 5 . The reaction is 6Pu(III) + X e 0 3 + 6H+ -> 6Pu(IV) + X e + 3 H 2 0 (in HCIO4 acid)

and the rate law is -Í/[PU(III)]/¿// = /:[Pu(III)][Xe03]. The activation enthalpy, free energy, and entropy are AHÍ = 15.3 + 2.1; Δ ^ ί = 20.2 + 0.1 k c a l m o l e - i a n d ASt = 16 + 6.9 cal d e g - i m o l e - i . Although the studies did n o t provide for a decisive mechanism for the reaction, a t w o electron change producing plutonium(V), which then reacts with plutonium(III) t o form plutonium(IV), appears m o s t plausible. A photochemically induced oxidation o f neptunium(V) by XeOsiaq) has been r e p o r t e d 2 5 6 : 6Np(V) + Xe(VI) -> 6Np(VI) + X e

The reaction is first order in XeOs, the rate expression being —d[Np(Y)]/dt = A:i[Xe03] for which Ä:ixl06 ( s e c - i ) = 6.28 + 0.58. The formation o f excited Xe03:Xe03+Av UV XeO*, appears t o be the rate-controlling step. It is noteworthy that the thermal reaction is very slow. Since w e m a y expect Xe03 t o be a "two-electron oxidizer", t w o neptunium(V) ions w o u l d need t o be oxidized simultaneously (Np(VII) is n o t a reahstic species). (The effective oxidizer here m a y n o t contain x e n o n but be a derivative o f a Xe03 + H 2 0 i!t reaction.) Oxidation of organic compounds. Since periodate is highly specific for the oxidation o f viC'diols, their oxidation by Xe03(aq) has also been investigated257. The X e 0 3 solution interacts readily with y/c-diols and primary alchohols in neutral or basic solution but there is n o interaction in acid. The i;/c-diols yield carboxylic acids or C O 2 from the terminal - O H group. This contrasts with IO4 oxidations, which yield aldehydes (I0¡" being the reduction product). It m a y be that this difference in behaviour has t o d o with the absence o f stable aqueous oxidation states o f the x e n o n below xenon(VI). Xenon trioxide has been recommended2oo as an analytical reagent for the determination of primary and secondary alcohols in aqueous solution, the products being C O 2 a n d H 2 O . The oxidation o f tertiary alcohols is slow. Similarly, carboxylic acids m a y also be quantitatively oxidized t o C O 2 and H2O 258. 3.4.6. Complexes of XeFö with F " Acceptors A number o f adducts involving XeFg in c o m b i n a t i o n with recognized F" acceptors have been reported. They include the following 1:1 adducts (m.p. (°C) given in parentheses): 255 256 257 258

J. M. Cleveland, Inorg. Chem. 6 (1967) 1302. J. M. Cleveland and G. J. Werkema, Nature 215 (1967) 732. B. Jaselskis and S. Vas, / . Am. Chem. Soc. 86 (1964) 2078. B. Jaselskis and R. H . Krueger, Talanta 13 (1966) 945.

316

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY

XeFö· AsFs (130.5°) 259,26o, 2 6 i ; XeFö · B F 3 (90°) 2 5 9 ; XeFö· SbFs 234; XeFg· G e F 4 ( s u b . ) 2 0 2 ; XeFö-PtFs ( - 1 0 0 ° ) 263.264; XeFö-IrFs (116°) i69; XeFö-RuFs (118°) 227. J h e X - r a y s t r u c t u r e s o f t h e 1 Ί a d d u c t s w i t h PtFs a n d AsFs h a v e s h o w n t h e m t o b e XeFJsalts 264,265 (Fig.

21).

The

iridium and ruthenium c o m p o u n d s

are isostructural w i t h the

platinum

XeFö is e s s e n t i a l l y X e F J F " ( s e e Fig. 17) it s e e m s p r o b a b l e t h a t a l l o f t h e 1:1 a d d u c t s a r e XeFj" s a l t s . However, t h e

compound227. i n view of the observation that the cubic form o f

FIG. 21. The molecular structure of X e F s + [AsFöJ- ·. • F. Hollander, D . Templeton, M. Wechsberg and N . Bartlett, unpublished observation.

XeFö a d d u c t s ( w i t h F " a c c e p t o r s ) m a y a l s o c o n t a i n X^F^ i o n s i n c l u s t e r s w i t h w i t h F " i o n s : 2XeFö-AsF5 (110°) 2 6 I ; 2XeFö-PF5 2 6 I ; 2XeFö-SbFs 234; 2XeFö-PtFs 263; 2XeFö-IrFs (135°) i69; XeFö-2SbFs (108°) 234; 4XeFö-GeF4; 2 X e F ö - G e F 4 262; 4 X e F ö - S n F 4 ; a n d 2XeFö-SnF4 266. Preparation. The c o m p l e x e s m a y b e p r e p a r e d s i m p l y b y f u s i n g t h e n e a t c o m p o n e n t s o r b y d i s s o l v i n g t h e m i n n o n - r e d u c t i v e s o l v e n t s (e.g. BrFs o r HF). Alternatively, XeFö m a y b e f o r m e d in situ, t h u s XeFJ[PtFö]- a n d 2XeFö-PtF5 h a v e b e e n i s o l a t e d 2 6 3 from Xe/Fi/PtFs m i x t u r e s w h e r e t h e fluorine p r e s s u r e s a n d t h e r e a c t i o n t e m p e r a t u r e f a v o u r e d XeFö f o r m a ­ t i o n . The s a m e m i x t u r e s y i e l d e d XeF+ a n d X e i F ^ p t F ö ] " s a l t s ( s e e s e c t i o n 3.2.6) a t l o w

following bridging

fluorine

concentrations.

Some physical and chemical properties. The c o m p o u n d s a r e c o l o u r l e s s i f t h e a c c e p t o r fluoride

is a n o n - t r a n s i t i o n e l e m e n t

m e t a l d e r i v a t i v e . All

fluoride

a n d a p p r o p r i a t e l y c o l o u r e d if it is a transition

o f the c o m p o u n d s are rather l o w melting a n d evidently

dissociate

readily266,262. 259 H. Selig, Science 144 (1964) 537. 260 N . Bartlett, S. P. Beaton and N . J. Kha, Abstracts, 148th National Meeting of the American Chemical Society, Chicago, Illinois; August-September 1964, N o . K 3 . 261 K. E. Pullen and G. H. Cady, Inorg. Chem. 6 (1967) 2267. 262 K. E. PuUen and G. H. Cady, Inorg. Chem. 6 (1967) 1300. 263 N . Bartlett, F. Einstein, D . F. Stewart and J. Trotter, Chem. Communs. (1966) 550. 264 N . Bartlett, F. Einstein, D . F. Stewart and J. Trotter, / . Chem. Soc. A (1967) 1190. 265 F. J. Hollander, D . H. Templeton, M. Wechsberg and N . Bartlett, unpublished observation. 266 K. E. Pullen and G. H. Cady, Inorg. Chem. 5 (1966) 2057.

XENON(VI)

COMPOUNDS

317

Although the crystal structures o f X e F j [ P t F 6 ] - a n d X e F ^ A s F o ] - clearly indicate the ionic formulation, they s h o w the cation t o possess considerable polarizing capability. The X e F J [ A s F 6 ] - structure, represented in Fig. 21, s h o w s the AsF¿" i o n appreciably distorted as a consequence o f those fluorine ligands o f AsF^" near the x e n o n a t o m s , being attracted t o the x e n o n . This is consistent with the charge o f — + 3 which the xenon(VI) a t o m is considered t o bear. Hydrolysis o f these XeFe c o m p o u n d s occurs very readily. An almost quantitative yield o f xenon(VI) in solution has been r e p o r t e d 2 6 6 . The salts have considerable potential as oxidizers a n d fluorinators, b u t there are n o reports o n these aspects o f the c o m p o u n d s .

3.4.7. XeFö Adducts with F " Donors A number o f c o m p o u n d s have been reported involving XeFö in c o m b i n a t i o n with recognized i o n donors. The following alkah fluoride c o m p o u n d s have been r e p o r t e d 2 6 7 . 268: CsF-XeFö (yellow, ρ 4.72 g c m - 3 ) ^ 2CsF-XeF6 (cream coloured) -0" -400" R b F - X e F e (colourless) > 2 R b F X e F 6 (colourless) > RbF <250" 2KF-XeF6 (colourless) > KF <250" 2 N a F - X e F 6 (colourless)

> CsF

^ NaF

Lithium c o m p o u n d s d o n o t form. Other r e p o r t s 2 0 4 ,

232

mention N a F - X e F e a n d give a

d e c o m p o s i t i o n temperature (under v a c u u m ) o f 125°. This d e c o m p o s i t i o n temperature presumably apphes t o the 2 N a F - X e F 6 c o m p o u n d since there is n o evidence for a 1:1 compound268.

As mentioned in section 3.4.1, the reversible c o m p o u n d formation o f XeFö

with N a F provides a convenient m e t h o d for the puriñcation o f XeFö since XeFi, X e F 4 and X e O F 4 (the c o m m o n impurities in an XeFö preparation) d o n o t c o m p l e x with N a F . The enthalpies o f dissociation o f the 1:1 adducts, 2Cs(Rb)F-XeF6 -> Cs2(Rb2)F2-XeF6+XeF6

have been derived from vapour pressure-temperature measurements t o b e 14.0 a n d 8.7 kcal m o l e - i respectively.

A c o m p o u n d with nitrosyl fluoride269^ 2 N 0 F - X e F ö , is presumably closely related t o the alkali fluoride c o m p o u n d s . The occurrence o f infrared a n d R a m a n bands at 2310 a n d 2305 c m - i indicate the presence o f N 0 + cations. A crystal structure o f this compound269a h a s estabhshed the salt formulation ( N O + ) 2 ( X e F 8 ) 2 - . The a n i o n is a shghtly distorted archimedian antiprism, suggesting very little evidence o f steric activity o f the n o n - b o n d i n g valence electron pair. Presumably, XeFy-, like XeFö itself, will exhibit only subtle steric activity o f the non-bonding-valence electron pair. The adducts are extremely reactive chemically a n d react violently with water. The hydrolysis presumably gives a quantitative yield o f xenon(VI) in solution, since x e n o n i s n o t evolved. Hydrolysis o f CsXeFv in moist air yields CsXeOaF (see section 3.4.11). Dissolution o f the latter yields CsHXe04 268.

3.4.8. X e O F 4 Complexes with Fluoride Ion Acceptors As in the case o f XeFö, a n t i m o n y pentafluoride forms a c o m p l e x 2 3 7 with X e O F 4 which is stable at r o o m temperature. Excess SbFs yields a material o f c o m p o s i t i o n X e O F 4 ' 2 S b F 5 267 R. D . Peacock, H . Selig and I. Sheft, Proc. Chem. Soc. (1964) 285. 268 R. D . Peacock, H, Selig and I. Sheft, / . Inorg. Nucl. Chem. 2 8 (1966) 2561. 269 G. J. M o o d y and H . Selig, Inorg. Nucl. Chem. Letters 2 (1966) 319. 269« s . Peterson, J. H. Holloway, J. Williams, B. Coyle, Science, 1971.

318

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY

(m.p. '-ΤΟ^). A complex, 2 X e O F 4 - V F 5 , is also r e p o r t e d 2 7 0 . Arsenic pentafluoride forms an a d d u c t 2 3 7 with X e O F 4 at - 7 8 ° , but this does not exist at room temperature, even under pressures of one or two atmospheres of AsFs. This contrasts with the XeFö behaviour and it is clear that complex formation using AsFs may be used as a convenient chemical method for separating X e O F 4 and XeFö. There is no ñrm structural information on the X e O F 4 complexes with the fluoride acceptors, but it is probable that they are salts of the X e O F ^ ion, which is presumably pseudo-trigonal bipyramidal (the oxygen ligand being in an equatorial position). 3.4.9. X e O F 4 Complexes with Fluoride Ion Donors The oxide tetrafluoride complexes readily with CsF, R b F , K F 237, and N O F 269. The trend in thermal stabilities of the alkah fluoride complexes is similar to that observed in the XeFö complexes, namely Cs > Rb > K. The N O F complex, O N F - X e O F 4 , dissociated readily (m.p. 40°, v.p. 30 mmHg at 23°) 269. Thermogravimetric a n a l y s i s 2 3 7 indicates the following stoichiometrics: ~125» CsFXeOF4

^ 3CsF-2XeOF4

-275' -400> 3CsFXeOF4 ^ CsF

50*' 3RbF-2XeOF4

> 3RbFXeOF4(?)

90» 3KFXeOF4

-400° ^ RbF

-250* > 6KFXeOF4

> KF

The MF->^XeOF4 adducts are chemically reactive, and bulk samples dissolve exothermically in w a t e r , but xenon is not e v o l v e d . When allowed to stand in moist air, H F is evolved and MXeOsF salts (see section 3.4.12) are formed. The X e O F j ion may occur in these X e O F 4 complexes, but it is more probable that polymeric species involving X e O F 4 molecules "bridged" by F " ions, like the XeFö tetramers and hexamers (see Fig. 17), will be found. 3.4.10. Xe02F2 Adducts Adducts of Xe02F2 have not been reported, but it is possible that both XeOiF^ and XeOiF^ salts will be preparable. The cation should be structurally akin to XeOs and the anion pseudo-octahedral. 3.4.11. Xenates(VI) Preparation. Monoalkali xenates of empirical formula M H X e 0 4 - I . 5 H 2 O (M = Na, K, Rb, Cs) have been p r e p a r e d 2 7 i by lyophilization of 0.1 Μ XeOa and alkah hydroxide in 1:1 ratio. The cesium salt has also been prepared by interaction of XeOacaq) in the presence of F " 247. A barium salt has also been c l a i m e d i 4 i and questionedisi. Physical properties. The infrared spectra of the salts are represented in Table 36. The salts are colourless and are considerably more stable (kinetically) than XeOa, particularly when anhydrous. They are susceptible to detonation, particularly if they contain excess XeOs. The salts are insoluble in methyl, ethyl and t-butyl alcohols, CHCI3 and CCI4. The infrared spectra show that the salts do not contain p e r x e n a t e 2 7 i . Evidently the sodium salt is structurally different from the heavier alkali salts. Infrared absorptions at 270 G. J. M o o d y and H. Selig, / . Inorg. Nucl. Chem. 28 (1966) 2429. 271 T. M. Spittler and B. Jaselskis, / . Am. Chem. Soc. 87 (1965) 3357.

XENON(VI) COMPOUNDS

319

3500 and 1600 cm-i indicate that the former contains hydroxy! groups. All are characterized by a strong band or bands in the region 770-810 cm-i—presumably associated with Xe-O stretch. TABLE 36. INFRARED SPECTRA ( c m - i ) OF M O N O A L K A L I X E N A T E S ( V I )

Empirical Formula NaHXe04« K(Rb)HXe04" CsHXe04

3500 w 3500 w 3120 vw

1600W 1360 m 770-800 s 1600w 1360 m 770-800 s 1430 vw. 783 s 741 s

730 sh 730 sh 721 s

625 s -480 s 699 s

469 s 386 s

340-370m 340-370m 347 s, 316 s

« T. M. Spittler and B. Jaselskis, / . Am, Chem. Soc. 87 (1965) 3357. B. Jaselskis, T. M. Spittler and J. L. Huston, / . Am. Chem. Soc. 88 (1966) 2149.

Chemical properties. The xenates disproportionate on dissolution in water to xenon(0) and ( V n i ) . Like the oxide itself, the xenates(VI) oxidize I " to h and this has served as the basis for their a n a l y s i s 2 7 i . 3.4.12. Complexes of XeOa with F - and CI" and Br- Donors Alkali fluoro-xenates(VI) MXeOaF 272 and chloro-xenates(VI) 273 have been prepared. Brief mention has also been made of CsXeOaBr 272. Preparation. The MXeOaF salts (M = K, Rb, Cs) are prepared from aqueous solution. The Cs+ and Rb+ salts have been made by neutralizing the XeFe hydrolysis product (0.5 M) with 2 Μ alkali hydroxide, to p H 4. Evaporation of the solution yields crystals. These are washed with ice-cold water and are best stored in a desiccator. All of the salts may also be prepared by evaporating a solution prepared by mixing equal volumes of 0.5 Μ XeOacaq) and 1 Μ KF(aq) and containing a few drops of H F . The chloro-xenate CsXeOaCl is p r e p a r e d 2 7 3 similarly, although acetronitrile may be used as the solvent. A white crystalline precipitate is obtained after 3 hr following mixing of an ice-cold solution of —2.0 ml of 1.5 Μ CsCl with 0.4 ml of 1.5 Μ X e O s í a q ) . Physical properties. The fluoro-xenates(VI) are claimed to be272 the most thermally stable of the solid oxygenated xenon(VI) compounds. Even the chloro-xenate273 is considerably more stable than XeOa. The infrared spectra of the MXeOsF salts are characterized by bands at 812 (s), 761 (m), 380 (w) and 333 (w) cm-i, but no bands were observed in the usual X e - F stretching region (500-600 cm~i). This finding is compatible with the crystal structure of the potassium salt. This structure274 shows XeOa units linked in infinite chains by bridging ñuorine atoms. The geometry of the XeOa moiety is very similar to that of XeOs itself, as may be seen from the representation of the xenon coordination group in Fig. 22. The xenon first coordination sphere is a grossly distorted square-based pyramid with one of the oxygen ligands apical. To a first approximation the ñuorine ligands are F™ species—the X e - F bonds are certainly much longer and (from the infrared evidence) weaker than in the ñuorides. Presumably the non-bonding valence electron pair occupies 272 B. Jaselskis, J. L. Huston and T. M. Spittler, J. Am. Chem. Soc. 91 (1969) 1874. 273 B. Jaselskis, T. M. Spittler and J. L. Huston, J. Am. Chem. Soc. 89 (1967) 2770.

320

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY

Angles Oi-Xe-02 03-Xe-02 Oi~Xe-03 03-Xe-F2 02-Xe-Fi F2-Xe-Fi Oi-Xe-Fi Oi-Xe-F2

101.1 (8)° 97.8 (7)° 100.5 (1.2)° 85.3 (5)° 77.2 (6)° 98.7 (2)° 87.6 (1.1)° 85.8 (7)°

FIG. 22. The xenon coordination geometry in KXeOßF 274

the sixth apical position of the pseudo-octahedral complex. The F~ hgands presumably lower the positive xenon charge and lower its polarizing power—so enhancing the stability of the XeOa group. Since the infrared spectrum of the chloro-xenate273 shows some similarities to those of the fluoro-xenates (818 (s), 793 (s), 766 (m), 749 (m), 663 (w) and 4 0 0 ( m ) c m - i ) the structure may be similar to that of KXeOaF. Chemical properties. The fluoro-xenates do not decompose thermally below 200° but loose substantial quantities of xenon and oxygen above 260°. Some samples exploded above 300°. The cesium chloro-xenate is stable to ^^150°, and explodes at 205° in vacuo leaving CsCl. Addition of concentrated H2SO4 to CsXeOaCl yields CI2, chlorine oxides, O2 and Xe.

3.5. X E N O N ( V I I I )

COMPOUNDS

The o c t a f l u o r i d e o f x e n o n i s u n k n o w n , a n d a v a i l a b l e e v i d e n c e i n d i c a t e s t h a t s h o u l d i t b e p r e p a r a b l e it w i l l n o t b e t h e r m o d y n a m i c a l l y s t a b l e u n d e r o r d i n a r y c o n d i t i o n s o f t e m p e r a ­ ture a n d pressure (see section

1.2). The t r i o x i d e d i f l u o r i d e a n d Xe04 a r e t h e o n l y k n o w n

m o l e c u l a r c o m p o u n d s o f x e n o n ( V I I I ) . The p e r x e n a t e s a r e t h e b e s t c h a r a c t e r i z e d x e n o n ( V I I I ) c o m p o u n d s a n d the alkali metal salts h a v e considerable thermal stabihty k i n e t i c ) . The p a t t e r n o f k n o w n c o m p o u n d s o f o s m i u m ( V I I I ) w i t h i s s i m i l a r , OSO4 a n d OSO3F2 b e i n g k n o w n a n d

fluorine

(presumably

oxygen hgands

OsFs, OsOFe a n d O.SO2F4 u n k n o w n 2 7 5 .

3.5.1. Xenon Octafluoride Although i n a n e a r l y report276 a c l a i m w a s m a d e f o r t h e p r e p a r a t i o n o f XeFf, t h i s h a s n o t been substantiated a n d m u c h further w o r k h a s failed t o support t h e existence o f this 274 D . J. Hodgson and I. A. Ibers, Inorg. Chem. 8 (1969) 326. 275 N . Bartlett and N . K. Jha, / . Chem. Soc. (1968) 536. 276 J. Slivnik, B. VolavSek, J. Marsel, V. Vr§öaj, A . Smalc, B. Frlec and Z. Zemljiö, Croatia Chemica 3 5 ( 1 9 6 3 ) 81.

Acta

XENON(VIII) COMPOUNDS

321

compound. In a thorough study of equihbria in the Xe/Fz system, Weinstock et al.^^ found no evidence for XeFs, even at high Fi/Xe ratios and moderate temperatures (see section 1.2). 3.5.2. Xenon Trioxide Difluoride Xenon trioxide difluoride has been prepared by interaction of XeFe with solid sodium perxenate contained in a Kel-F tube. This interaction produces a larger quantity of other xenon compounds, principally X e O F 4 . The compound was detected mass spectrometrically, XeOaFJ being observed. Xenon tetroxide is formed in small quantities along with XeOsFa. These two compounds have comparable volatihties, both being sufiiciently volatile at — 7 8 ° to yield characteristic mass spectra. Since Xe03F2 is more volatile than XeOaFa (see section 3 . 4 . 3 ) the former is probably symmetrical and non-polar. This indicates that the D^h geometry shown is probable. This would be consistent with all theoretical predictions (see section 1.3).

F

Xe

F

3.5.3. Other Xenon(Vffl) Oxyfluorides A l t h o u g h , in view of the six energetically favourable X e - F bonds XeOFe at first sight appears to be a thermodynamically favourable formulation, the low enthalpy of formation of I F 7 from I F 5 for the process IF5(g) + F2(e) -> I F t c s ) (see section 1.2) suggests that the compound may be thermodynamically unstable and hence demanding of a special synthetic approach. N o such objections apply to the OXQF^ ion, and it may be that the parent oxyfluoride can be prepared from salts of this ion. T h e oxyfluoride X e 0 2 F 4 may well be preparable. T h e AH^ may be no more unfavour­ able than + 1 0 to 2 0 kcal mole-i.

3.5.4. Xenon Tetroxide Preparation. X e n o n tetroxide was first prepared by the interaction of concentrated sulphuric acid with xenates(VIII) at room temperature277, 278. S o d i u m or barium perxenate ( N a 4 X e 0 6 ; B a 2 X e 0 6 ) , dried in a vacuum desiccator, is contained in the side-arm of an all P y r e x apparatus. I t is tapped slowly into cold ( - 5 ° ) reagent grade sulphuric acid, contained in a bulb below the side-arm. W i t h care an approximately 3 4 % yield of the tetroxide is obtained. T h e barium salt is more satisfactory than the sodium. T h e gaseous tetroxide is condensed in liquid nitrogen cooled traps as a yellow solid. I f the mixing of the reactants is too fast or massive, the tetroxide decomposes with "flashes of fire"277 and negligible Xe04 yields are obtained. T h e tetroxide is readily purified by vacuum sublimation into a trap at - 7 8 ° . Physical properties. T h e limited data available on X e 0 4 are given in T a b l e 3 7 . T h e vapour pressure of X e 0 4 is ^^25 mm at 0 ° and the volatility at — 7 8 ° is suflicient to provide for mass spectrometric detection. U s u a l l y , decomposition to X e and O2 occurs before the sample reaches 0 ° , and the decomposition can be violent—samples have exploded at 277 H. Selig, Η. Η. Claassen, C. L. Chemick, J. G. Malm and J. L. Huston, Science 143 (1964) 1322. 278 J. L. Huston, M. H. Studier and E. N . Sloth, Science 143 (1964) 1161.

322

NOBLE-GAS CHEMISTRY.* NEIL BARTLETT AND F. O. SLADKY

- 4 0 ° 277. This instability is an accord with the heat o f formation obtained238 by detonating several gaseous samples at ^^25°:

PCe04(g) -> Xe(g) + 202(g))*= - 1 5 3 . 5

kcal m o l e - i

This enthalpy indicates a mean thermochemical b o n d energy of 21.1 kcal, which is a little greater than the value for XeOs. TABLE 37.

SOME PHYSICAL PROPERTIES OF X e 0 4

The solid is yellow Thermochemical features AH} (g) (298.15°K): + 1 5 3 . 5 kcal · Mean thermochemical bond energy: 21.1 kcal * Vapour pressure, (mm) CQ 3 ( - 3 5 ° ) , 10 ( - 1 6 ° ) , 25 (0°) Vibrational

spectra Obs.»>

Infrared

VI

Supports T¿ symmetry

V2 V3 V4

Urey-Bradley force c o n s t a n t s « ( m d y n A-i): Molecular

—

877 (PQR) 870 885 306 (PQR) 298 314

Calc.«

906 301 876 305

K, 5.75; H, 0.10; F, 0.5; A:, 0.05

geometry

X e - O bond length: PR separation** yields X e - O = 1.6 ± 0 . 2 A Electron diffraction confirms Ta symmetry and gives X e - O = 1.736 ± 0 . 0 0 2 Ä " S. R. Gunn, / . Am. Chem. Soc. 87 (1965) 2290. ^ H. Selig, Η. Η. Claassen, C. L. Chernick, J. G. Malm and J. L. Huston, Science 143 (1964) 1322. W. A . Yeranos, Bull. soc. chim. Beiges 7 4 (1956) 414. ^ G. Gunderson, K. Hedberg and J. Huston, Acta Cryst. 25 (1969) 124; see also R. J. Gillespie, in Noble Gas Compounds (H. H, Hyman, ed.). University o f Chicago Press, Chicago and L o n d o n (1963), p. 333.

The infrared spectrum of gaseous samples were obtained using nickel cells having either silver chloride or polyethylene windows. The findings and their interpretation are given in Table 37. The vibrational data indicate the molecule to be tetrahedral and this is borne out by the electron diffraction findings24i. The X e - O stretching force constant is lower than in the case of which suggests that the X e - O bond in is weaker. This is sup­ ported by the bond length of 1.736 A obtained from electron diffraction data, this being -.0.04 A longer than the Xe-O bond in 239. Since the Xe-O bond strength in appears to be greater than in it may be that the charge removal from the xenon atom by the four oxygen ligands in is less than the charge removal by the four fluorine ligands and one oxygen hgand in the oxide tetrafluoride. If each X e - O bond involves a net charge distribution Xe+-0-"(as in X e : -> 0·), this would yield a central-atom charge of + 4 for Xe04 and only + 3 for (assuming three-centre four-electron bonds -^F-Xe+i-F"^). However, the ligand repulsions for the oxide would be greater than for the oxyfluoride. The chemical shift derived from the Mössbauer spectrum of Xe04 has been interpreted!58 as indicative of a charge of + 1 . 7 or

XeOF4,

Xe04

XeOF4 XeOF4

Xe04, Xe04

XeOF4

XENON(VIII) COMPOUNDS

323

the xenon atom, whereas the same authors argue that the x e n o n a t o m positive charge in + 3 . These values seem t o o disparate and the bonding assumptions u p o n which these numbers are based are probably at fault (the only x e n o n orbitals assumed t o be involved are the 5s and 5p). Chemical properties. Although potentially an oxidizing agent o f great power, nothing has been reported o n the chemical properties o f the tetroxide. Analysis and identification. The tetroxide is probably m o s t readily identified by its strong infrared bands at 877 and 305.7 c m - i . Mass spectra278 s h o w the typical isotope pattern o f x e n o n repeated every 16 mass units up t o X e 0 4 . Conventional analysis has been a c h i e v e d 2 3 8 by decomposing a sample with a spark (thermal decomposition w o u l d sufiice) followed by cooling o f the sample t o —196° (to retain the xenon), the oxygen then being measured with a Toepler p u m p . Subsequently, the x e n o n w a s measured in a hke manner.

X e F 4 is

3.5.5. The Xenates(Vffl)—"Perxenates" So far, attempts t o prepare perxenates and other oxysalts o f x e n o n by oxidizing x e n o n with powerful oxidizers in aqueous media have failed. All o f the oxysalts are derived from the fluorides. Preparation. Perxenates were first described b y Malm, Holt and Banei^s. They found that the hydrolysis o f XePe in strong sodium hydroxide generated x e n o n gas and a precipi­ tate o f hydrated sodium perxenate. The reaction has been studied in detaiU^i a n d it is k n o w n that xenon(VI) is the immediate product o f the hydrolysis, the formation o f perxenate proceeding slowly at r o o m temperature with initial half-times ranging from 2 t o 20 hr. The reaction is catalysed by impurities o f u n k n o w n composition. The production o f perxenate in these hydrolysis experiments obeys the stoichiometry represented in the equation. 2 X e F 6 + 4 N a + + l ó O H " - ^ N a 4 X e 0 6 + X e + 0 2 + I2F- + 8H2O

Evidently, the disproportionation o f pure XeOa solution in base is m u c h slowerisi, and it is claimed^^o that only - ^ 3 3 % yields o f perxenate are obtained over a N a O H range of 0.25-4.2 Μ and a K O H range o f 2-3.6 M . It seems that high perxenate concentrations suppress the disproportionation235^ but high [ O H ] - concentrations lead t o yields o f xenon(VIII) in excess o f 50%. This implies that a x e n o n species o f oxidation state lower than xenon(VI) is contributing to the oxidation o f XeOa t o perxenate (see section 3.4.5 for a fuller discussion o f XeOs disproportionation). The most efficient synthesis o f perxenate is providedisi by ozonizing a pure XeOa solution in 1 Μ N a O H . Since the solubihty o f sodium perxenate in water is only 0.025 M , the salt precipitates out nearly quantitatively. Washing with a httle cold water readily removes excess base. The salt is a white crystalline p o w d e r which m a y contain from 0.6 to 9 molecules H2O per xenon atom, depending o n the drying procedure. The preparation of the potassium salt, K 4 X e 0 6 * H 2 0 , requires greater care279 since a mixed valence product K 4 X e 0 6 ' 2 X e 0 3 readily precipitates2. 2 5 4 . Physical properties. The perxenates are colourless, thermally stable sohds. The hydrated sodium salt becomes anhydrous at 100° and d e c o m p o s e s abruptly at 360°, and the barium salt decomposes at '^300°. The latter is almost insoluble in water^si, a saturated solution 279 A. Zalkin, J. D . Forrester, D . H. Templeton, S. M. Williamson and C. W. Koch, / . Am. Chem. Soc. 86 (1964) 3569.

324 NOBLE-GAS CHEMISTRY: NEIL BARTLETT A N D F. O.

SLADKY

XENON(VIII) COMPOUNDS

oo

OrO-OiO

Oo ÍS-1

o-

(b)

(c)

FIG. 23. The crystal structure of Na4Xe06 · 6H2O. (Reproduced with permission from A. Zalkin, J. D. Forrester and D. H. Templeton, Inorg, Chem. 3 (1964) 1417.)

325

326

NOBLE-GAS CHEMISTRY: NEIL BARTLETT A N D F. O. SLADKY

at -^25° being only 2,3 χ 10"5 Μ . The solubility of the perxenates decreases in the sequence Na+ > Li (1.0 X 10-3 M ) > A m 3 + (6.1 χ 10-5 Μ ) > B a 2 + (2.3 χ 10-5) 280. Solutions of

the

alkali salts are strongly basic, the p H corresponding approximately to the hberation of 1 mole of O H - per mole of the compound dissolved: N a 4 X e 0 6 + H2O

HXeOe-^ + O H " + 4Na+

The ultraviolet spectra of the perxenate solutions are p H dependent. Isobestic points at 220 and 270 m^ indicate that two principal species contribute to the spectra. Potentio­ metrie titrations suggest the following equihbria: H X e 0 6 3 - + H+ :f± H 2 X e 0 6 2 - ( P Ü T a - 10.5) H 2 X e 0 6 2 - + H+ <± H j X e O e - ( P / j T a - O )

The interpretation of the solution properties is also complicated by the reduction of xenon(VIII) to xenon(VI): HaXeOö- - > HXeO4-+0.5O2 + H2O

This decomposition is more rapid the lower the p H . TABLE 38. SOME PHYSICAL ΡΚΟΡΕΚΉΕ8 OF PERXENATES

Crystallographic

data Na4Xe06-6H20"

Space group b(k) c(k) dx.,„ (g c m - 3 ) ^measured Ζ

X e - O (A)

O-Xe-O

Pbca 18.44 (1) 10.103 (7) 5.873 (5) 2.59 >2.17 4 (2) 1.86 (2) (2) 1.87 (2) (2) 1.80 (2) av. 1.840 (s) 89 (l)^' 87 (1) 88 (1)

Na4Xe06-8H20 ^'

Pbcn 11.864 (5) 10.426 (5) 10.358 (5) 2.33 (5) 2.38 4 (2) 1.88 (1) (2) 1.85 (1) (1) 1.84 (1) (1) 1.89 (1) av. 1.864 (12) 89.3 (8)° 89.0 (8) 87.4 (8)

K4Xe06-9H20

Pbcli 9.049 (4) 10.924 (4) 15.606 (6) 2.35

—

4 1.86 (1)

88.8 t o 91.2 (7)°

(See Fig. 23) Vibrational data * 1.8 Μ cesium perxenate solution: 448 (mw) Ρ 402 (w) Raman bands ( c m - i ) 685 (vs) Ρ 443, 425 (s) Infrared bands ( c m - i ) 605 (s) Na4XeO60.4H2O (solid): Raman bands ( c m - i ) 683 (vs) 655 (m) 470 (mw) 390 (w) Oxidation potentials ^ Xe(VlII)-Xe(VI), acid: 3.0 V; base: 0.9 V * A . Zalkin, J. D . Forrester and D . H . Templeton, Inorg. Chem. 3 (1964) 1417. A. Zalkin, J. D . Forrester, D . H . Templeton, S. M . W^illiamson and C. W^. K o c h , Science 143 (1963) 501. J. A . Ibers, W. C. Hamilton and D . R. MacKenzie, Inorg. Chem. 3 (1964) 1412. ^ A . Zalkin, J. D . Forrester, D . H . Templeton, S. M . Williamson and S. W. K o c h , / . Am. Chem, Soc. 86 (1964) 3569. « J. L. Peterson, H. H . Claassen and E. H . Appelman, Inorg. Chem. 9 (1970) 619. ' E . H , Appelman and J. G. Malm, J. Am. Chem. Soc. 8 6 (1964) 2141. 280 Y . Marcus and D . Cohen, Inorg. Chem. 5 (1966) 1740.

XENON(VIII) COMPOUNDS

The

crystal

structures

o f several perxenates,

including

327

a h e x a h y d r a t e z s i . 282^ o c t a -

hydrate283^ a n d a n o n a h y d r a t e 2 7 9 h a v e b e e n d e t e r m i n e d a n d t h e i n T a b l e 38. T h e s t r u c t u r e o f N a 4 X e 0 6 - 6 H 2 0

findings

are summarized

i s s h o w n i n Fig. 23. T h e p e r x e n a t e i o n i s

seen t o be octahedral. M u c h o f the water in the hydrates is coordinated t o the cations, but t h e perxenate oxygen ligands are also h y d r o g e n b o n d e d t o water molecules283, 279, 281. T h e X e - O b o n d length,

1.84-1.88 A, i s s l i g h t l y s h o r t e r t h a n t h e X e - F b o n d i n XeFe (1.89 A) (1.74 A) a n d XeOa (1.76 A). As m a y b e s e e n i n T a b l e 3

and much longer than in X e 0 4

( p . 217), t h e s i z e a n d s h a p e o f t h e p e r x e n a t e i o n i s w h a t o n e w o u l d h a v e a n t i c i p a t e d o n t h e b a s i s o f t h e d a t a f o r antimonates(V), tellurates(VI) a n d p e r i o d a t e s . T h e vibrational spectra o f a q u e o u s solutions o f t h e perxenates284 suggest that a high concentration

o f symmetrical

XeO^-

ions

occurs

in the concentrated

(1.8M cesium

perxenate) solutions, b u t certain details imply t h e presence o f other ionic forms. T h e v i b r a t i o n a l s p e c t r a a r e i n c l u d e d i n T a b l e 38. T h e i n t e n s e h i g h l y p o l a r i z e d R a m a n b a n d a t 685 c m - i i n t h e s o l u t i o n

spectrum

is very close

t o t h e 683 c m - i b a n d

in the sohd

N a 4 X e O 6 * 0 . 4 H 2 O . It i s v e r y p r o b a b l y t h e t o t a l l y s y m m e t r i c a l o c t a h e d r a l v i b a n d .

This

stretching frequency is compatible with the X e - O b o n d length a n d is suggestive o f a shghtly stronger (intrinsic) b o n d t h a n t h e X e - F b o n d i n X e F g .

Bonding and bond polarity. T h e M ö s s b a u e r s p e c t r a ^ s s o f t h e p e r x e n a t e s i m p l y c h e m i c a l shifts w h i c h a r e very c l o s e t o that o f X e 0 4 , a n d t h e d e r i v e d x e n o n p o s i t i v e c h a r g e is i n ­ distinguishable from that obtained for t h e x e n o n a t o m i n X e 0 4 . T h e b o n d i n g i n XeO^" c a n b e d e a l t w i t h i n t e r m s o f t h e R u n d l e - P i m e n t e l which in its simplest form represents t h e i o n in terms o f three three-centre b o n d s i n v o l v i n g ( f o r m a l l y ) Xe2+ ( c o n f i g u r a t i o n

theory32.33^ four-electron

5^0 5p^) a n d 6 O " c o m p o n e n t s ,

each

t h r e e - c e n t r e f o u r - e l e c t r o n b o n d a r i s i n g f r o m a h n e a r a r r a y ( s e e Fig. 2, p . 226):

-O

Xe2+

O -

T h e high b o n d polarity o f this m o d e l a c c o u n t s f o r t h e greater intrinsic b o n d strength o f

X e - O i n [XeOö]^- r e l a t i v e t o X e F i n X e - F e . Alternatively, if d orbitals are involved in t h e b o n d i n g , t h e i o n c a n b e represented a s a n 5/?3ß^2 h y b r i d s y s t e m , e a c h X e - O b o n d i n v o l v i n g o n e b o n d i n g e l e c t r o n p a i r . In m o l e c u l a r o r b i t a l t e r m s , t h i s i m p l i e s t h e i n v o l v e m e n t o f t w o X e 5d o r b i t a l s i n g e n e r a t i n g leg o r b i t a l s b y t h e a g e n c y o f a l i g a n d field effect. T h e s i m p l e s t m . o . s c h e m e w o u l d b e a s s h o w n i n Fig. 24. T h e r e i s , o f c o u r s e , n o p r o o f t h a t X e 5d o r b i t a l s a r e i n v o l v e d i n t h e b o n d i n g .

Chemical properties,

Perxenate solutions are powerful a n d rapid oxidizing agents, the

x e n o n ( V I I I ) b e i n g r e d u c e d t o x e n o n ( V I ) . I o d i d e i s o x i d i z e d t o I2, e v e n i n 1 Μ b a s e . S i m i ­ l a r l y , B r - i s o x i d i z e d t o Br2 a t p H 9 o r l e s s a n d CI- t o CI2 i n d i l u t e a c i d i s i . A l s o i n d i l u t e a c i d p e r x e n a t e i m m e d i a t e l y c o n v e r t s Mn2+ t o Μ η θ 4 . l o d a t e i s o x i d i z e d t o I O 4 a n d c o b a l t ( I I ) t o c o b a l t ( I I I ) , i n b a s e a s w e l l a s a c i d . It s h o u l d b e n o t e d t h a t t h e o x i d a t i o n s i n a c i d , t o b e eflfective, m u s t b e f a s t e n o u g h t o c o m p e t e w i t h t h e r a p i d e v o l u t i o n o f o x y g e n . Americium

perxenate,

Am4(XeO6)3"40H2O,

prepared^so

from

basic

solutions

of

a m e r i c u m ( I I I ) i s o f l o w s o l u b i l i t y i n w a t e r (4.6 χ 10-5 Μ ) b u t i t d i s s o l v e s i n a c i d t o f o r m americum(VI) a n d americum(V) solutions. Up t o 80% americum(VI) h a s b e e n b y this technique. T h e formal o x i d a t i o n potentials f o r americum(III)-(VI)

obtained

a n d (V) a r e ,

r e s p e c t i v e l y , 1.75 a n d 1.83 V. 281 A . Zalkin, J. D . Forrester and D . H . Templeton, Inorg. Chem. 3 (1964) 1417. 282 A . Zalkin, J. D . Forrester, D . H . Templeton, S. M . Williamson and C. W. K o c h , Science 143 (1963) 501. 283 J. A . Ibers, W. C. Hamilton and D . R. MacKenzie, Inorg. Chem. 3 (1964) 1412.

328

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY 5d.

60"s XeO; F I G . 24. A simple molecular orbital scheme for XeOö"^- involving X e 5d orbitals.

Analysis. It is evidently characteristic of perxenate that rapid decomposition to xenon(VI) occurs in acid media. Thus if a solution of perxenate is first acidified, then iodide is added, only six equivalents of oxidizing power per mole of xenon are measured as iodine liberated. If iodide is added before the acid all eight oxidizing equivalents are measured. This can be useful in assessing the xenon(VI) impurity in a sample of perxenate. The iodine liberated is determined using thiosulphate according to standard proceduresisi.

4. RADON CHEMISTRY 4.1.

INTRODUCTION

Prior t o 1962, w h e n Fields, Stein a n d Zirin r e p o r t e d a a t r u e c o m p o u n d o f r a d o n , a l t h o u g h t h e l o w first h i n t e d a t c h e m i c a l a c t i v i t y . In t h e i r p i o n e e r i n g w o r k for

fluoride285^ t h e r e w a s n o e v i d e n c e ionization potential on

radón,

Ramsay

(10.75 eV) Soddy

and

demonstrated286 that it d i d n o t react w i t h m e t a l s a n d a large n u m b e r o f o t h e r s u b s t a n c e s .

Nikitin

w a s able287 t o p r e p a r e c l a t h r a t e s o f r a d o n , i n c l u d i n g h y d r o g e n c h l o r i d e , h y d r o g e n

bromide, a n d hydrogen sulphide clathrates, a n d he w a s thus able to separate r a d o n f r o m hehum and neon.

Unfortunately, i s o t o p e , 222Rn, h a s

there are n o stable isotopes o f r a d o n , a n d the longest h v e d "natural"

a

half-hfe o f only

226Ra (usually i n c h l o r i d e s o l u t i o n s ) .

3.83 d a y s . This Experimental

i s o t o p e is derived f r o m the d e c a y o f difficulties arise n o t o n l y f r o m

the

radiation hazard, but also from radiation decomposition o f the reagents employed in the s t u d i e s . The

latter factor rules o u t the utihty o f large-scale e x p e r i m e n t a t i o n (even a l l o w i n g

that large quantities o f 222Rn c o u l d b e collected). 284 285 286 287

J. L. Peterson, H . H. Claassen and E. H . Appelman, Inorg. Chem. 9 (1970) 619. p . R. Fields, L. Stein and M. H. Zirin, / . Am. Chem. Soc. 8 4 (1962) 4164. Ψ, Ramsay and F . Soddy, Proc. Roy. Soc. 7 2 (1903) 204. B. A . Nikitin, Comp. Rend. Acad. Sei. URSS N o . 6 (1939) 562.

COMPOUNDS OF RADON

329

4.2. C O M P O U N D S O F R A D O N

Since the X e - F bond energy (32 k c a l in all three fluorides) is much greater than that of K r - F in KrF2 (^^12 kcal), and this follows the trend observed in the fluorides of the other groups (see section 1.2.3), we anticipate the R n - F bond to be at least as energetic as that of X e - F . It also appears hkely that chlorides and oxides (RnCh, RnOa, RnOa, RnO^-, etc.) would be thermodynamically more stable than in the case of their xenon counterparts. Evidently288, however, radon, like astatine, is markedly more metallic than the element above it in its group. So far there is evidence only for a fluoride, and attempts to prepare chlorides and oxides, directly and from the fluoride, have failed. A c l a i m 2 8 9 for the oxidation of radon by strong aqueous oxidizers has been refuted29o, and effOrts to oxidize radon with ozone and with sodium perxenate in aqueous media have failed to " ñ x " the gas. 4.2.1. Radon Fluorides (s)—RnFz (?) In their initial r e p o r t s 2 8 5 , 2 9 i stein and his coworkers established that radon combines with fluorine at 400° to give a compound of low volatility which is reduced by hydrogen only at temperatures above 200°. More recent w o r k 2 8 8 , 2 9 2 has shown that 222Rn interacts spontaneously with fluorine and all of the stable interhalogen fluorides (e.g. 2 B r F 3 + 3 R n Br2H-3RnF2(?)) except I F 5 , and is even oxidized by the [NiFö]^- ion. In these experiments, gaseous 222Rn from a 5-curie radium chloride solution, dried by passing through a column of calcium sulphate and purified by distillation at —78° (to remove radiolytic hydrogen and oxygen), was condensed on to the various reagents held in Kel-F tubes. The mixtures were agitated at room temperature, for an hour or so, after which the 222Rn was found to be in the liquid phase. Removal of the excess reagent (BrFs or BrFa or C I F 3 ) in a vacuum at room temperature yielded, in all cases, a wh^ite solid containing all of the 222Rn activity. All of the solids appear to be the same material and to be identical to the fluoride prepared earlier. Since the fluoride interacts with water without generating radon oxides and leaves httle radon activity in the aqueous phase, it is probable that the compound is RnF2: RnF2 + H 2 0 R n - f - i 0 2 + 2 H F . From the observation that AsFs will not oxidize radon, whereas BrFs will, the standard free energy of formation of RnF2 has been set between the limits of - 2 9 and - 5 1 kcal m o l c T i . Stein and his coworkers have s h o w n 2 8 8 that the fluoride decomposes above 250° in a vacuum. Evidently, the radon fluoride does not vaporize as a molecular species. Electromigration studies of the radon fluoride dissolved in bromine trifluoride or anhydrous hydrogen fluoride have established that the radon is present as a cation. The species RnF+ and R n 2 F j would seem more reasonable than Rn2+. Presumably the fluoride is ionic— both Rn2+(F-)2 and RnF+(F-) are consistent with the known properties. Although complex salts containing RnF+, R n 2 F 3 , or even Rn2+ appear feasible, there is no evidence for such compounds at this stage. So far no eflOrt has been made to oxidize radon with the more powerfully oxidizing hexafluorides. Since I (Rn) = 10.75, whereas 288 L. Stein, Science 168 (1970) 362. 289 M . W^. Haseltine and H . C. Moser, / . Am. Chem. Soc. 89 (1967) 2497. 290 K. Flohr and E. H . Appelman, / . Am. Chem. Soc. 9 0 (1968) 3584. 291 See ref. 14, p. 113. 292 L. Stein, / . Am. Chem. Soc. 91 (1969) 5396.

330

NOBLE-GAS CHEMISTRY: NEIL BARTLETT AND F. O. SLADKY

I (Xe) = 12.1 eV, i t is obvious that PtFg should oxidize radon (see section 3.1.2), and it seems likely that IrF^ and possibly O S F Ö could also oxidize the gas spontaneously at ordinary temperatures. Some practical applications. The formation of radon compounds provides for the meter­ ing and location of the gamma radiation source, radon, at a specific site by bringing it into chemical interaction with fluorine, a halogen fluoride, or platinum hexafluoride. Thus the radon activity can be readily transferred to some ideal location, then "fixed". Removal of the "fixed" radon from the chosen site may be achieved by reduction (or hydrolysis). Encapsulated, involatile radon compounds may replace the radon "seeds" or needles presently in medical use. Radon occasionally produces hazardous radiation levels in uranium mines. It is feasible288 that this problem can be overcome by circulating the air through bubblers or packed towers containing oxidizing agents.