"Noble-Gas Chemistry and The Periodic System of Mendeleev" By Neil Bartlett Materials and Molecular Research Division, Lawrence Berkeley Laboratory and Department of Chemistry, University of California Berkeley, California U.S.A. The Stable Noble-Gas Electron Configuration When, in 1962, chemists were informed that a compound of a noble- gas had been prepared’, there was much expression of surprise and ini- tially even disbelief. Faith in the chemical inertness of the noble gases had been fostered in part by previous failures to prepare compounds. The greatest prejudice, however, derived from the electronic theories of the chemical bond, which stressed the noble-gas electron arrangement as the ideal to which all other atoms tended. When the noble gases were discovered’, in the last years of the 19th Century, they were quickly recognized as a new Group of elements of Mende- leev's Table of The Elements. This new Group of elements fitted naturally into the "Table", each noble-gas being Located between a halogen and an alkali metal. Since the Halogens included the most strongly oxidizing elements, whereas the Alkali Metals were the most strongly reducing ele- ments of the Periodic Table, it was appropriate, for the intervening group of elements, to exhibit neither oxidizing nor reducing properties, i.e. to be chemically unreactive. All efforts to oxidize or reduce helium and argon (i.e. to bring them into chemical combination with other elements) failed’, perhaps the most significant failure being Moissan's attempt in 1895 to prepare an argon fluoride>. The rarer noble gases were not sub- jected to the same intensive chemical investigation, and no claim for chemical activity of the gases was sustained prior to 1962. When the electronic theories of chemical bonding were developed it was natural that the chemical inertness of the noble gases should be expressed in the theory. In their pioneering papers of 1916, both W. Kossel* and c. N. Lewis” emphasized the ideality of the noble-gas con- figuration. An atom of an element other than a noble-gas was represented as gaining or losing electrons until its electron arrangement resembled that of a neighbouring noble-gas atom. The Kossel and Lewis theories unified and correlated much of what was then known of the bonding capabilities of the chemical elements. The theories quickly had wide appeal. Since the electron arrangements of the noble gases were evidently the ideal arrangements, to which all other atoms aspired, the chemical inertness of the gases was self evident, at least at a superficial level of inspection. Unfortunately in the inevi- table shorthand of convenient description, the noble-gas electron arrange- ments were usually represented by the group term ' ‘octet", this being (ex- cept in helium, which possesses a ''duet") the outermost set of electrons of the noble-gas atom. This "octet" concept helped to foster the illusion that all noble-gas electron configurations are essentially the same and of the same stability. They are not. Discovery of The First True Chemical Compounds of The Noble Gases From trends based upon the Periodic Table several chemists had pre- dicted chemical activity for the heavier noble gases. Thus Kossel, in harmony with his emphases of complete electron transfer as the key to chemical bonding, pointed out’ that the ionization potentials of the noble-gases decreased with atomic weight and that fluorides of xenon or even krypton might therefore be possible. On the basis of a survey of the chemical trends evident in the Periodic Table at that time, L. C. Pauling in 1933 suggested® that xenon and krypton compounds should be preparable. (He predicted that XeF., KrF, should exist and that XeF 6 6 8 might exist). Pauling obtained a sample of Xenon’ for his colleague D. M. Yost to attempt the synthesis of a xenon fluoride. That attempt failed.® This failure, along with the success of the simple electronic theories of valence (which emphasized the importance of the stability of the noble-gas electron configuration), contributed to a general acceptance of the complete chemical inertness of the gases. Of prime importance to the discovery of the chemical activity of the heavier noble gases was the discovery” by Bartlett and Lohmann of the remarkable oxidizing properties of the gaseous compound platinum hexa- fluoride. In 1962 they had established that a red solid, prepared by burning platinum or platinum compounds in fluorine in glass apparatus was the salt, dioxygenyl hexafluoroplatinate, 0, [Ptr] . This salt was es- pecially noteworthy for its cation, 0 + The salt formulation implied that 2° the free hexafluoride (which had previously been reported, in 1957, by Weinstock, et a1.?° of The Argonne National Laboratory) should be capable of spontaneously oxidizing molecular oxygen. This subsequently proved to be so: + _ 0 + PtF > o, [PtFe] (c) . 2(g) 6(g) The two gases combined immediately to provide the now familiar salt 0," [PtF,]. Although the salt formulation had seemed appropriate much earlier in the investigation, it had posed the difficulty that in order for the oxidation of molecular oxygen to proceed spontaneously, the electron affinity for the platinum hexafluoride, E = ~AH(PEFe (,) +e > Pte (gy ) needed to be much greater!” than 160 kcal/mole! (that is, approximately twice the value for atomic fluorine or atomic chlorine). With the spon- taneous oxidation of oxygen and the salt formulation proved, it was clear that platinum hexafluoride was the most powerful oxidizer that had been discovered. “At this point, Bartlett (who was not aware of Kossel's earlier related observation) noted that the ionization potentials of the noble gases decreased markedly, with increasing atomic number as shown in Table I. It was evident that the heavier gases should be more easily oxidize- able than the lighter .13 Most importantly, the ionization potentials of xenon (12.2 eV) and radon (10.7 eV) were as low as, or lower than, molecular oxygen (12.2 eV). Radon being difficult to handle as a consequence of the short life and a-particle activity of all of its isotopes, the oxidation of xenon appeared to be the easiest noble-gas oxidation to carry out. Xenon gas proved to be as easy to oxidize as molecular oxygen. An orange-yellow solid formed rapidly in the spontaneous gas-gas reaction as described in Figure 1. The product was designated xenon hexafluoroplatinate xe" [Pere] Subsequent work! 4 showed that the Xe + PtF, interaction is more 6 complicated. The compound XePtF, is obtained in high purity only when a large excess of xenon is employed.+> When xenon has an opportunity to interact with excess ,PtF, the overall reaction is: Xe + 2PtFy > XeF'PtF,” + PtF,. When the report of the oxidation of xenon by platinum hexafluoride appeared the reaction was immediately repeated at the Argonne National Laboratory where PF, and its relatives had first been prepared and studied. There, the PtFy oxidation of xenon was repeated and extended to the related hexafluorides ruthenium hexafluoride and plutonium hexa- fluoride .1® The ruthenium hexafluoride-xenon study proved to be highly revealing. It was clear from the appearance of the characteristic green color of ruthenium pentafluoride that the red hexafluoride was losing fluorine. It could only be lost to the xenon. Xenon fluorides, it was reasoned, ought to exist. The first xenon fluoride to be reported!’ was the tetrafluoride pre- pared by Claassen, Selig and Malm of the Argonne National Laboratory. Ironically, Bartlett and Jha soon discovered ® that the pyrolysis of Xe(PtF,) (1 XeF, + XePt,F (the last compound being a dia- 6 4 2°10 magnetic Pt(IV) compound: (XeF)" [Pt Fg]7)«, Independently of the work at the Argonne National Laboratory, study of the xenon-fluorine system by Hoppe and his coworkers” in Giessen, Germany, led to the isolation of a difluoride. Within a few weeks the fluorides XeF,» XeF) >» XeF and the oxyfluoride XeOF, were known, Within nine months of the first report 4 of XePtF, the first conference on Noble-Gas Chemistry was called and met at The Argonne National Laboratory. More than fifty papers were con- tributed in the two day meeting and the proceedings subsequently appeared as a 400 page volume .1® The Extent of Noble-Gas Chemistry In the twenty years which have elapsed since the first synthesis of a chemically bonded noble-gas compound the possible range of noble=gas chemistry has become well defined. The known oxidation states and repre- sentative ligands are illustrated in Table 2.° The requirements for noble- gas compound formation are: (1) the noble-gas atom must be a larger (more easily oxidizable) atom (Rn, Xe, Kr) and (2) each atom attached to a noble-gas atom must be highly electronegative, either intrinsically or as a result of electronegative groups linked to it. It is clear that chemical bonding of the noble gases depends upon a larger more oxidizable noble-gas atom yielding valence shell electrons to ligands which are themselves striving for a valence electron octet. The small highly electronegative fluorine ligand is the most effective ligand. Evidently the great stability of the Ne electron configuration induces a sharing of the Kr, Xe or Rn valence~shell electrons with the F ligand, which thus attains that configuration. The oxygen ligand is inferior in electronegativity to the F ligand, but it does have the capac- ity to accept two electrons to meet the Ne electron configuration. Thus oxygen can and does form stronger bonds to xenon than fluorine. The O ligand's inferior electronegativity and the great strength of the 0, bond (AH° atomization = 119 kcal mole +) relative to that of Fy (AH° atomization = 38 kcal mole‘), however, render oxides less stable thermodynamically than fluorides .7° Indeed it is the extraordinary weakness of the Fy bond, in combination with the small size and high electronegativity of the F ligand, which result in its being the most thermodynamically favorable ligand for binding to any atom. | The substituted 0, N and C ligands are only effective when the substituents are highly electronegative. In effect the substituted O, N and C mimic the small electronegative F ligand. The lower electronegativity and greater size of the halogens heavier than fluorine contribute to their inferiority to F as ligands. Noble-gas compounds derived from the heavier halogens, even the chlo- rides, are stable only at temperatures well below room temperature. Only the highly electronegative OTeF, ligand has so far revealed?! a range of chemistry (for Xe) comparable with that excited by the F ligand. Even it is inferior to F however and it is now apparent that the compound forming ability of F ligand (at least as far as the noble gases are con- cerned) is unlikely ever to be surpassed. On this basis it is possible to say that neutral compounds of argon or the lighter gases are unlikely to be made at ordinary temperatures and pressures. Similarly higher oxi- dation states than +2 are unlikely to be attained with krypton. Even other Kr(II) compounds will severely test synthetic expertise since the difluoride is itself thermodynamically unstable?” with the weakest bond known for any fluoride available in macroscopic quantities. A greater range of ligands can be anticipated for radon than is presently known. A fluoride described as RnF, has long been known 18 2 A possible tetrafluoride and an oxide believed to be Rn0, have recently been described.2° Ligands which are effective for xenon should also be suitable for radon. The radioactivity of all isotopes imposes great experimental difficulties however. To give a more quantitative evaluation of the range and limitations of noble~gas chemistry it is of value to consider some simple bonding models and energetics. Bonding in Noble-Gas Compounds As we have seen the phenomenological evidence in Table 2 implies that bonding of noble-gas atoms, N, to other atoms is associated with the removal of electrons from N. Méssbauer~“ and ESCA“? studies have established high bond polarity in the xenon and krypton compounds. Since nt is a pseudo-halogen aton, (N-L)* is expected to resemble its ‘isoelectronic halogen relative. Thus XeF* ought to resemble IF. Sim- ilarly XeF, ought to resemble IF, and have a generic relationship to such well known species as Icl, and I, - The simple molecular orbital representation=°’~’ for NL, uses the scheme set out in Figure 2. Since the non-bonding molecular orbital has no component from the noble-gas atom the pair of electrons in that orbital reside entirely on the ligands. The bonding of the two ligands derives from the two electrons in the bonding orbital. This model is therefore equivalent to the single-electron bond mode17® and to repre- sentation? as the resonance hybrid of the canonical forms {(L-N)*L"; L(N-L)*}. An advantage of the last bonding model is that it allows a rough estimation of the thermodynamic stability of NL, to be made. Examples of such evaluations are illustrated in Table 3, where the en- thalpy of atomization (i.e. total bond energy) of NL, species is related to the first ionization potential of N, the energetics of (N-L)* bond formation’, the energy of formation of the ion pair (L-n) tL’, the res- onance energy, and the electron affinity of L. With the known total bond energy of XeF, used to fix the value for the resonance energy for each of the fluorides, the total bond energy of KrF, is calculated to be 22 kcal mole !. This is in excellent agreement with experiment. A similar eval- uation for ArF, shows that it cannot be bound with respect to ground state atoms. The major cause of ATF, instability is the high ionization potential of argon. Since the first ionization potential of radon is 248 kcal mole + (i.e. 32 kcal mole! less than that of xenon) it is probable that the total bond energy of RnF, will be correspondingly greater than that of XeF,- From such an evaluation it is evident that large anions are less favorable than small because of the adverse impact which the greater charge separation has upon the electrostatic energy of ion-pair formation. A high electron affinity is also necessary. For such reasons the larger halogen atoms, and all but the least oxidizable complex species, are un- satisfactory ligands for the formation of bonds to noble-gas atoms. 10 The small size, high electronegativity and electron-pair accepting capability of the oxygen atom, combine to make it a satisfactory ligand for the xenon atom in high oxidation states. The same may also be true for radon.*> From the data in Table 4, a simple electrostatic evaluation provides a satisfactory accounting for the observed trends in stability of the xenon oxides. This assumes that each oxygen-atom ligand accepts a share in one xenon valence~electron pair, i.e. Xe: + 0. Indeed, ESCA studies”> indicate that the negative charge on O ligand in Xe0F, is approx- imately twice that of the F ligands. The shorter bond length?! and greater force constant -* for the XeO bonds in Xe0,, (Xe-O = 1.76 g, fr = 5.66 mdyn gly relative to those? for the Xe-F bonds in XeFy (Xe-F = 1.89 R, fr = 3.3 mdyn gly fit the representation of Xe-O as an electron pair bond, if the Xe-F bond is a single electron bond. However, the mean thermochemical bond 20,34 31 and « 21 kcal. Mean energies for XeF, and Xe0, are respectively thermochemical bond energies are defined with respect to molecules and atoms in ground states. A more appropriate valence state for an oxygen atom to accept an electron pair from xenon, (i.e. Xe: > 0} would be 1n(0). Since ty oxygen atom is 45 kcal mole + more energetic” than 3p(0), we therefore evaluate the intrinsic bond energy per Ke-O linkage, formed from ground- state Xe and 1D (0) to be < 66 kcal mole. This is in much more satis- factory accord with the value of 31 kcal mole for the Xe-F linkage. 11 The Group Relationships of Noble-Gas Compounds In the highly selective and brief review which follows, of the compounds of noble-gases and their reactions, the emphasis will be on trends and re- lationships within the Group and across Periods of the Periodic Table. For fuller information on which this sketch is based the reader is re- ferred to more detailed reviews. Work to 1971 was reviewed by Bartlett and Sladky.>° The comprehensive review by Legasov and Chaivanov?’ is more up to date. Stein?® has recently made a detailed assessment of radon chemistry to early 1981. The ease of oxidation of the noble gases follows that indicated by the fonization potentials (Table 1). The trend is the same as for the oxidative chemistry of the halogens with which, as we shall see, the chemistry of the noble-gases has much in common. Radon?® is the easiest of the noble-gases to oxidize and reacts spontaneously with gaseous fluorine at room temperature and even with liquid fluorine at -195° (activation energy coming from the intense @ - radiation). It also interacts spon- taneously at 20° with solids containing the moderate oxidizers IF,", clr”, + , + or BrF, . The reactions are assumed to form RnF salts, e.g.: Rn + IF,’SbF,. + RnF” SbF, + IF 6 6 5 Such solids do not oxidize the lighter noble gases. Indeed XeF* salts oxidize iodine pentafluoride according to the equation:°” + + XeF + IF, > IF, + Xe 12 Reagents which oxidize xenon, also oxidize radon but by using the moderate oxidizers (IF,* etc.) in an initial scrubbing of mixed gases, the radon alone can be oxidized. Stronger oxidants such as + -_ - . 0, SbF, or NF’ SbF, can then be used”? to oxidize the xenon: ton 7 ton 7 Xe + NoF SbF, > XeF SbF, +N, Thus separation of Rn, Xe and Kr ternary mixtures can be achieved by successive oxidation of radon and xenon. To oxidize krypton it is essential to have a source of atomic fluorine and efficient quenching of products to low temperatures. Because of greater ease of oxidation of radon relative to xenon, Stein has pointed out that an extensive chemistry could be expected.>® 3 4 appear to be possible. Until recently it seemed that this was not to The compounds RnF, >» RnF; > RnCl,, RnCl,; RnO; Rn0, 5 RnO, and RnOF, all be, but Avronin and his coworkers? have found that a higher fluoride than the difluoride exists. It could be either a tetrafluoride or a hexafluoride. The hydrolysis of the new fluoride yields an oxide, -which behaves as though it is Rn0,. Early attempts’! to prepare a chloride failed but other synthetic approaches should be explored be- fore the non-existence of radon chlorides is accepted. The weakness of the bonding in KrF, results in that fluoride being a more effective source of F ligands than the Fy molecule itself. It can oxidize xenon to the hexafluoride: 3 KrF, + Xe > XeF, + 3 Kr. It 2 6 should be the most effective reagent for higher fluorides of radon. In 13 contrast the fluoride of radon, which Stein has persuasively argued” is RnF, » is not reduced by hydrogen until 500°, and at a hydrogen pressure of 800 torr. The energetics of formation of the cations (Ar-F)*, (Kr-F)* and (Xe-F)*, given in Table 3, parallel the-bond energies >” for C1l-F, Br-F and I-F, as does the oxidizing power. Unfortunately salts of (Ar-F)* | are not known and there is no clear evidence to encourage us to believe that an anion will be found to stabilize the cation. Since there is effectively no bonding in ArF’, the electron affinity of (ArF)* is equal to the ionization potential of Ar, less the energy associated with the process: AX (ey + Fee) +> (AtF)(,)- This gives a value for E(ArF)* ®& 325 keal mole -. The electron affinity of (KrF)* is less (¥ 280) but, even so, salts of this cation are remarkably effective oxidizers and have pro- 42 KrF also oxidizes?” 0, to 2 whereas. 0," will -oxidize>® Xe to XeF*, The XeF cation, as befits vided efficient syntheses for BrF,* salts. 0 its lower electron affinity is much less powerful than KrF’, but as has been remarked, it is effective in the synthesis of IF". This is in spite of the oxidation of xenon by IF, at 200°: IF, + Xe > XeF, IF, Periodic Relationships Comparison of the XeF, bonding with that of (xeF)* illustrates the difference between single-electron bonding and electron-pair bonding. Similar differences are observed in the molecules C1F, and BrF, (see 14 Table 5). Indeed the formation of the trifluorides, by the attachment of two F ligands to the heavy halogen of the monofluoride, is equiva- lent to the formation of the noble-gas difluorides from the noble-gas atom. In much the same way the pentafluorides are related to XeF,- The closest relationships, however, involve isoelectronic species of the same period. The nearly octahedral hexa-oxospecies of antimony, tellurium, iodine and xenon (see Table 5) reveal the central-atom element in the highest attainable oxidation state (in which all valence electrons are involved in bonding). The general decrease in E-O bond distance from Sb to Xe correlates, as expected, with the increasing nuclear charge of E. The molecular oxide Xe0,, is slightly smaller than 10, ; as is Xe, compared with 10, . “The prefered stability of Xe0,, and Xe0, over Xe0, and Xe0 (Table 4) is strikingly similar to the iodine oxyanion system, where IO, and I0, are also favored. 3 4 The EF, species shown in Table 5 provide the most precisely described isoelectronic series for the comparison of a noble~gas compound (xeF,*) with its relatives. One observes that the more polar bonds (the E-F equatorial are 'single-electron'bonds) show a greater shortening, with increasing nuclear charge of E, than do the E-F axial bonds (in which, as a consequence of the electron pair bond, the polarity is likely to be slight). The most remarkable feature of the geometry is, however, the almost constant bond angle F axial-E-F equatorial. Such similarities have been noted elsewhere. “4 Thus PF. and SF,” have the same bond angle 15 (97.5 + 0.5°) yet there are significant differences from Period to Period, as the data for CIF, BrF, and KeF," also show. Evidently the oribtal hybridization at E does change from Period to Period but very little within a Period. Such observations imply that the s and p valence-orbital energy separations change little across a Period but significantly from Period to Period in such fluorospecies. The existence of IF, raises the possibility of the existence of the molecules XeF, and XeOF,. There is no convincing evidence for either as long lived species at ordinary temperatures and pressures. Huston has recently pointed”? to the evident instability of the latter: XeOF > XeFy + "09+ He has succeeded in synthesizing such xenon (VIII) oxyfluorides as Xe05F, and Xe0,F,. It seems that the difficulty with XeF, and XeOF. is the very unfavorable energy associated with coordi- nation numbers beyond six. The major cause of this must be the limited orbital set (s and p) available for bonding at the central atom. Ligand crowding may also have an impact. Even in IF, (as seen in its ability to oxidize Xe) the bond energy is markedly lower than in the iodine fluo- rides. 7° One notes also that XeF, when represented with a sterically active non-bonding valence electron pair is also hepta coordinate. Both XeF, and IF, are good fluoride ion donors and superior in that respect to 7 their lower fluoride relatives (er, and IF;). Indeed the molecules approach ion-pair behavior: KeF,” and IF,"F. Octahedral coordination 6 is favored by both the Xe(VI) and I(VII). It should now be evident that there is no qualitative difference be- tween the bonding in noble-gas compounds and that observed in other high 16 oxidation state compounds of the non-transition elements. More sur- prisingly perhaps there also appears to be little difference from related compounds of the transition elements. Compounds of the transition elements and the non-transition elements of the same group number are most alike in the highest oxi- dation states and least alike in the lowest oxidation states. Simi- larities of Xe0, with OsO, and XeF. with OsF, may be seen from the 4 6 6 data given in Table 6, where similarities of ReF. with IF. and TeF 7 7 6 with WF can also be seen. This suggests that in the high oxidation states the involvement of d orbitals (outer d for the non-transition elements and inner d for the transition elements) in the molecular orbitals, may be approximately the same. Presumably the orbital-con- tracting influence of the highly electronegative ligands in the high oxidation states is sufficient to bring the 5 d orbitals into an ef- fective bonding role in Xe0, or XeF, > whereas in OsF, and Os0, > those same influences are rendering the 5 d orbitals less effective in bonding than they are in lower oxidation states. Future Possibilities for Noble-Gas Chemistry As we have seen,an extension of noble-gas chemistry to other elements than Rn, Xe and Kr is only possible with Ar in Ar-F* salts. New oxidation states and a greater range of ligands than those presently known for xenon might occur for radon. It remains to be seen if the +1 oxidation state of that element is a favorable one, as has been argued by Pitzer © on the basis of relativistic effects which should be important for such a heavy atom. 17 Otherwise the extension.of the chemical compounds of the noble-gases “is likely to be confined to extension (with other electronegative li- gands) in the known oxidation states of xenon and krypton. Because of the stronger band in (n-L)* relative to NL,» salts of (XeL)* and (KrL)* might be preparable, even though the NL relatives may never be made. As usual L will need to be an electronegative ligand, although the high proton affinities of the heavier noble gases’! (Xe 2 6; Kr 2 4 eV) also raises the possibility of (XeH)* and (Kri)* salts being preparable. Compared with their NL relatives, the wit salts are not only more stable but are also more reactive (as a result of their high electrophilicity). It is in the exploitation of. noble-gas compounds as reagents that we are likely to witness the greatest extension of noble-gas chemistry. Although the very weak bond present in (NL)° radicals is of-impor- tance to the efficiency of the laser emission from excited states, ‘© for most chemical purposes this bond can be ignored. The (NL)° radical behaves essentially as the L° radical, although N atom can also serve to carry off energy (generated in the interaction of (NL)° with a sub- strate) as kinetic energy. Of course the NL, compounds (particularly for X = 2) can be used as a route to oxidative reagents which do not con- tain a noble gas atom. An instance of this is the generation of the high purity peroxide S50¢F5 (which is a source of So3F radicals)” by the sequence of reactions:>° XeF, + 2HSO,F > Xe (SOF). + 2HF; Xe (SOF), > 3 Xe + S50, Fe. An illustration of the direct application of an NL, com- pound as an L source is in the substitution of F ligands (from XeF,) 18 into aromatic hydrocarbons:>- XeF, + Couey > Xe + CHF + HF. The most valuable reagents are likely to be the (NL)* salts. In effect they act as suppliers of Lt, The synthesis of BrF + salts is 6 a spectacular instance” of the application of such a salt (KrF). No doubt interaction of wit with an electron-rich substrate will some- times result in its electron-oxidation: (NL) + Sub > (NL)” + Sub’. With salts of cations>- such as Ke,” this may be an excellent route to novel salts (Xe,” + Sub > 2Xe + Sub‘) but in other cases the (NL)* radical will pass the L° radical to the cation: NL° + Sub* > (SubL)* +N. The bonding and the energetics of some noble-gas compounds mean that in certain cases (e.g. the oxides) the ligands ean be provided (for inter- action with substrates) in what amounts to excited states... Thus the use”? of XeF, in aqueous solution as the oxidative reagent for conversion of bromate to perbromate, suggests the possibility of 1y(0) availability, per- 4 + Xe). It is possible that the xenon oxides, xenates and perxenates could be similarly haps via a [XeO] intermediate: (Br0, + [Xe0] + Bro exploited. Their application in oxidations as clean sources of oxygen is in any case assured. The greatest utility of the noble-gas compounds will surely derive from the weakness of their bonds and the near-inertness of the reduction product--the noble-gas atom. ACKNOWLEDGMENT This work was supported by the Director, Office of Energy Research, Office of Basic Energy Science, Chemical Sciences Division of the U. S. Department of Energy under Contract Number DE-ACO3-76SF00098. 10. 11. 12. 19 References N. Bartlett, Proc. Chem. Soc., 218 (1962) An 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. H. Moissan, Bull. Soc. Chim., 13, 973 (1895). W. Kossel, Ann. Phys., (Leipzig) 49, 229 (1916). G. N. Lewis, J. Am. Chem. Soc., 38, 762 (1916). L. C. Pauling, J. Am. Chem. Soc., 55, 1895 (1933). Professor Pauling obtained the necessary sample of xenon from his old friend and teacher Professor Fred Allen (then at: Purdue Univer- sity) -- private communication from LCP to NB. D. M. Yost and A. L. Kaye, J. Am. Chem. Soc., 55, 3890 (1933). N. Bartlett and D. H. Lohmann, Proc. Chem. Soc., 115 (1962). B. Weinstock, H. H. Claasen, and J. G. Malm, J. Am. Chem. Soc., 79; 5832 (1957). “Argon, Helium and The Rare Gases," G. A. Cook, ed., Interscience Publishers, New York, London, Vol. 1, p. 13 (1961). Calorimetric data, by P. Barberi and N. Bartlett (to be published) indicate that AH” (05 (9) + PtF +O + per -59(t1) kcal mole !. 6(g) ~ 2 PtFecey? = Combined with a lattice enthalpy for 05 PERE (gy: evaluated to be -135 kcal mole“, and an ionization potential for 0," of 278 kcal mole -, the electron affinity of Pt, is required to be 202 kcal mole |, 13. 14. 15. 16. 17. 18. 19. 20. al. 22. 23. 24. 20° Although the greater sizes of the more easily ionizable gas are somewhat disadvantageous to bond formation, this adverse size effect is much less significant than the lower ionization potentials. N. Bartlett, B. Zemva, and L. Graham, J. Fluorine Chen., 7; 301 (1976). XePtF, is paramagnetic unlike XePdF, which is diamagnetic. +4 The latter is prepared from XeF, and PdF, >» whereas interaction of XeF, and PtP, under comparable conditions gives KeF Pt,” (with Xe evolution). The formulation Xe ‘Ptr, is the most probable one for XePtFy whereas XePdF, is almost certainly KeF PdF (the anion being polymeric). C. L. Chernick, et al., Science, 138, 136 (1962). VN H. H. Claassen, H. Selig, and J. G. Malm, J. Am. Chem. Soc., 84, 3593 (1962). . "Noble Gas Compounds" H. H. Hyman, ed., The University of Chicago Press, Chicago and London, 1963. | R. Hoppe, W. Dahne, H. Mattauch, and K. M. Rodder, Angew. Chem., 74, 903 (1962). | . JANAF Thermochemical Data, Dow Chemical Company, Midland, Michigan, 30 Sept. 1965, 30 June 1966, and 30 June 1977. E. Jacob, D. Lentz, K. Seppelt, and A. Simon, Z. Anorg. Allgem. Chen., 472, 7 (1981). S. R. Gunn, J. Phys. Chem., 71, 2934 (1967), and J. Am. Chem. Soc., 88, 5924 (1964). Vv. V. Avrorin, R. N. Krasinova, V. D. Nefedov, and M. A. Toropova, Radiochemiia, 23, 879 (1981). G. J. Perlow and M. R. Perlow, J. Chem. Phys., 48, 955 (1968), and G. J. Perlow and H. Yoshida, J. Chem. Phys., 48, 1474 (1968). 25. 26. 27. . 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 21 S.-E. Karlsson, K. Sieghbahn, and N. Bartlett, U.C.R.L. Report 18502, 1969 (Lawrence Berkeley Laboratory, Berkeley, California, U.S.A.) and T. X. Carroll, R. W. Shaw, Jr., T. D. Thomas, C. Kindle, and N. Bartlett, J. Am. Chem. Soc., 96, 1989 (1974). G. C. Pimentel, J. Chem. Phys., 19, 446 (1951). R. E. Rundle, J. Am. Chem. Soc., 85, 112 (1963). J. Bilham, and J. W. Linnett, Nature, 201, 1323 (1964). Cc. A. Coulson, J. Chem. Soc., 1442 (1964). + + .._20 The values for AH(N +L > (N-L) ) parallel the bond energies for the isoelectronic halogen monofluorides: for which: I-F = 66, Br-F = 59, and Cl1-F = 59 kcal mole. D. H. Templeton, A. Zalkin, J. D. Forrester and S. M. Williamson, J. Am. Chem. Soc., 85, 817 (1963). H. H. Claassen, and G. Knapp, J. Am. Chem. Soc., 85, 2341 (1964). R. M. Gavin, Jr., and L. S. Bartell, J. Chem. Phys., 48, 2460 (1968) and L. S. Bartell and R. M. Gavin, J. Chem. Phys., 48, 2466 (1968). S. R. Gunn, J. Am. Chem. Soc., 87, 2290 (1965) and S. R. Gunn in ref. 18. G. Herzberg, "Atomic Spectra and Atomic Structure," Dover Publications, New York (1944). N. Bartlett and F. 0. Sladky, “The Chemistry of Krypton, Xenon, and Radon", in Comprehensive Inorganic Chemistry, A. F. Trotman-Dickenson, Editor, Pergamon Press, Oxford, 1973, Vol. 1, pp. 213-330. vV. A. Legasov and B. B. Chaivanov, Khim. Zhizhn, 28 (1976). 38. 39. 40. 4l. 42. 43. 44, 45. 46. 47. 48. 49. 50. 51. 52. 53. 22 L. Stein, Radiochimica Acta, in press. D. E. McKee, C. J. Adams, A. Zalkin, and Neil Bartlett, J.C.S. Chem. Communs., 26 (1973) F. A. Hohorst, L. Stein, and E. Gebert, Inorg. Chen., 14, 2233 (1975). P. R. Fields, L. Stein, and M. H. Zirin, in ref. is. R. J. Gillespie, and G. J. Schrodilgen, Inorg. Chem., 13, 1230 (1974). D. E. McKee, Ph.D. Thesis, University of California, Berkeley 1973, LBL Report 1814 and N. Bartlett, Endeavor, 31, 107 (1972) D. D. Gibler, C. J. Adams, M. Fischer, A. Zalkin, and N. Bartlett, Inorg. Chem., 11, 2325 (1972). J. L. Huston, Inorg. Chem., 2i, 685 (1982). K. S. Pitzer, J. Chem. Soc., Chem. Comm., 760 (1975). G. von Bunan, Fortschr. Chem. Forsch., 3» 374 (1965). J. E. Velazco, and D. W. Setser, J. Chem. Phys., 62, 1990 (1975). F. B. Dudley, and G. H. Cady, J. Am. Chem. Soc., 79, 513 (1957). M. Wechsberg, P. A. Bullimer, F. 0. Sladky, R. Mews, and N. Bartlett, inorg. Chem., il, 3063 (1972). R. Filler, Israel J. Chemistry, 17, 71 (1978). L. Stein and W. H. Henderson, J. Am. Chem. Soc., 102, 2856 (1980). vow E. H. Appelman, J. Am. Chem. Soc., 90, 1900 (1968). 23 Figure 1: The Xe + PtFeg Experiment The first experiment was carried out in dry glass and quartz apparatus. A small sample of PtFe was transferred to the quartz sickle gauge, and was allowed to vaporize in the gauge, closed by the metal valve 1, Following pressure measurement it was transferred to (b) via the break-seal by-pass which was then sealed at X. Xenon was admitted to the gauge to the same pressure as the PtFe sample. The sample cf Xe ‘from the gauge was condensed in (a) at -196°C and valves 2 and 1 closed to ensure a small volume. Both the Xe and PtFe were vaporized, then the break-seal separating them was broken with nickel balls, moved © within the system by means of an external magnet. The interaction of the gases, to produce an orange solid, was immediate and the gauge showed that the residual pressure in the system was low. 24 Figure 2. Simplified representation of the po m.o.s. for XeF,. CEO CO OC Ce — & CEO EGO CC -_—, Xep, Fp 25 Table 1: Atomic radius and first ionization potential for each Noble Gas @ Noble-Gas He Ne Ar Kr Xe Rn Radius (&) 1.3 1.6 1.92 1.98 2.18 -- First Ionization . Potential: (eV) 24.586 21.563 15.759 13.999 12.129 10.747 2G. A. Cook, ed., Argon, Helium and the Rare Gases, 2 vols., Interscience, New York and London, (1961), Vol. I, p. 237. 26 * Table 2: A selection of Noble-Gas Compounds to illustrate known oxidation states and ligands. , Noble-Gas and Oxidation State Ligands Fluorides Oxyfluorides Oxides Other* Kr +2 KrF,* Db ‘ n Xe +2 XeF, / FXeOR oO FXeN(SO,F), 5 Xe (OR) > P Xe(CF,), oO Xe(N(SO,F),], xeor* 4 c g , o> 4 +4 XeF), Xe0F, F,Xe0R d h - 1 iv +6 XeF eo XeOF, Xe0, F,Xe0R Xe0,F,t XeF,. (OR') * 2°2 6-x x (x = 1 > 6) k tn +8 Xe0,F, Xe0, k Xe0,F, Rn +2(2) RnF,~ +4 rnF,* or or £ +6 RnFy Rn0, * Available in macroscopic quantities. + -OR includes -OTeF., -OS0,F, -OC10 5 2 0,CCF and OSO,CF.,. -OR' = -OTeF 37 ~2°>°-3 2-3 5 a 27 Table 2 References J. J. Turner and G. C. Pimentel in ref. 18. A. A. Artyukhov, V. A. Legasov, G. N. Makeev, and B. B. Chaivanov, Khim. Vys. Energ., ll, 89 (1977). V. N. Beymelnitsya, V. A. Legasov, and B. B. Chaivanov, Dokl. Akad. Nauk, S.S.S.R., 235, 96 (1977). See ref. 18 for historical aspects. For preparation by thermal means see: W. E. Falconer and W. A. Sunder, J. Inorg. Nucl. Chem., 29, 1380 (1967). For preparation by photochemical means see J. H. Holloway, J. Chem. Soc., Chem. Commun., 22 (1966). Ref. 17 See Ref. 18 for various syntheses. Refs. 18 and 38 Ref. 23 J. S. Ogden and J. J. Turner, J. Chem. Soc., Chem. Commun. 693 (1966). E. J, Jacob and R. Opferkuch, Angew. Chem., Int. Ed., 15, 158 (1976). D. F. Smith, Science, 140, 899 (1963) and J. Shamir, H. Selig, D. Samuel, nN and J. Reuben, J. Am. Chem. Soc., 87, 2359 (1965). J. L. Huston, J. Phys. Chem., 71, 3339 (1967). J. L. Huston, Inorg. Chem., 21, 685 (1982). Ref. 18 and B. Jaselskis, T. M. Spitter, and J. L. Huston, J. Am. Chem. Soc.,. 88, 2149 (1966). H. Selig, H. H. Claassen, C. L. Chernick, J. G. Malm, and J. L. Huston, Science, 143, 1322 (1964). ws M. Wechsberg, P. A. Bullimer, F. 0. Sladky, R. Mews, and N. Bartlett, Inorg. Chem., 11, 3063 (1972). D. D. DesMarteau, R. D. LeBlond, S. F. Hossain, and D. Nothe, J. Am. Chem. Soc., 103, 7734 (1981). van L. J. Turbini, R. E. Aikman, and R. J. Lagow, J. Am. Chem. Soc., 101, 5834 (1979). N. Keller and G. J. Schrobilgen, Inorg. Chem., 20, 2118 (1981). “Ref. 21 TABLE 3: Estimation of the heat of atomization of a noble-gas dihalide, NL, (Values in kcal mole 1 + - (N-L)t (L-N)"L ion pair AH(electrostatic) AH(electron pair bond Resonance nt ‘N Energy { L N + L-N-L (8) AH(atomization) (8) Resonance Experimental Quantities ——_—_—_______, energy AH(atomization) _ + + a assumed Molecule I(N) AH(L + e>L ) AH(N + L~* (N-L) ) AH(electrostatic) - constant from cycle observed XeF, 280 -20 2 -47° -166 -52 65 assumed 65 KrF, 323 -80 -37° -176 -52 22 23 ArF, 365 -80 ~38° -195 -52 0 Molecule not known XeC1, 280 -83 -~40(est.) -138 -52 32 b a. The AH(electrostatic) is estimated as the attraction energy (E = ~e” /estimated or observed N-L distance). The bond stretching force constant for XeC 1, = 1.3 and that for XeF 2-8 mdyn gt, L. Y. Nelson and G. C. Pimentel, Inorg. Chem. 6, 1758 (1967). c. C. J. Berkowitz and W. A. Chupka, Chem. Phys. Lett., 7, 447 (1970). d. A reasonable assumption since this is an expression of the fact that the electron is delocalized over two F ligands and not localized on one. 8z ry é TABLE 4: An estimate of the relative stabilities of the xenon oxides (values in kcal role” +) oe . xe™"(07), AH(electrostatic) (an ion cluster) X+ Xe + xO x x Xe: O electron- / EI, f Xx =33 pair-bond energy ~xX Xe + xO €& XeO 4H (atomization) x __t . Electron-pair MT atomization oad ~bond energy . Molecule ZI, x 4H(x07) “(electrostatic)” x(Xe:0) Cycle Observed? Xe0, 2520 ~132 -2320 -152 =obs. ~84 xe0, 1507 ~ 99 -1339 “114 - 45 x -50 Xe0, 768 - 66 - 634 - 76 - 8 —_ Xe0 280 - 33 - 176 - 38 + 33 — Estimated as the point charge attraction energy using observed (or where necessary) estimated interatomic distances. The electron-pair-bond energy for the Xe:0 bond was obtained from the cycle for the Xe0,, case and the unit energy (-38 kcal mole”) then used for all other cases. 62 30 Table 4 References a ref. 34 b It is of interest that spectroscopic studies (C. D. Cooper, G. C. Cobb, and E. L. Tolnas, J. Molec. Spectrosc., 7 223 (1961)) indicate that XeO is bound with respect to an unspecified singlet oxygen species, the reported dissociation energy being 8 kcal mole‘. If the singlet species were 1 (0) this would imply that XeO should be bound by 37 kcal mole + with respect to 3p(0), thus providing remarkable agreement with the simple calculations. c J. L. Franklin, et al., NSRDS - NBS 26, National Bureau of Standards, Washington, D.C., June 1969. d ref. 20 31 Table 5a EF (Coy) XeF* IF Bond Length(&) 1.84(4)° 1.906° v(em +) 621° 610° force constant (md/&) 3.7? 3.6° a V. M. McRae, R. D. Peacock, and D. R. Russell, Chem. Comm., (1969) 62. b F. O. Sladky, P. A. Bulliner, and N. Bartlett, J. Chem. Soc., (1969) 2179. cL. G. Cole, and G. W. Elverum, Jr., J. Chem. Phys., 20 (1952) 1543. NN Ne d R. A. Durie, Proc. Roy. Soc., A207 (1951) 388. e G. R. Somayajula, J. Chem. Phys., 33 (1960) 1541. Table 5b EF, (Doh) KrF XeF r*2(g) e*2(g) E-F (3) 1.875(2)? 1.977(2)° vz (em) 449° 515° fr(mdynes 8) 2.46° 2.84> a C. Murchinson, S. Reichman, D. Anderson, J. Overend, and F. Schreiner, J. Am. Chem. Soc., 90, 5690 (1968). b S. Reichman and F. Schreiner, J. Chem. Phys., 51, 2355 (1969). c H. H. Claassen, G. L. Goodman, J. G. Malm, and F. Schreiner, J. Chem. Phys., 42, 1229 (1965). E- F a b c 32 Table 5¢ EF,(C,_) clF,° BrF,” -F equatorial ®) 1.598(2) 1.721 F cial 1.698(2) 1.810 ax” Fag 87.5° 86.2° D. F. Smith, J. Chem. Phys., 21, 609 (1953). D. W. Magnuson, J. Chem. Phys., 27, 223 (1957). XeF,” ¢ 1.83(1) 1.88(1), 1.89(1) 82, 80 D. E. McKee, A. Zalkin, and N. Bartlett, Inorg. Chem., 12, 1713 (1973). 33 Table 5d EO, (Cay symmetry) a b 10, Xe0, E-O 1.79(2) 1.76(3) 96 100 O-E-0(°) {209 {308 102 108 aH. Schulz, Acta Cryst., B29, 2285 (1973) b D. H. Templeton, A. Zalkin, J. D. Forrester, and S$. M. Williamson in "Noble Gas Compounds," H. H. Hyman, Ed., The University of Chicago Press, Chicago and London, (1963) pp. 229-237. Table 5e EO, Za symmetry) a ob 10, Ke0, E-O 1.775(7) 1.736(3) a A. Kalman, and D. W. J. Cruichshank, Acta Cryst., B26, 1782 (1970) NNN E-0 34. Table 5f EF, (C,. symmetry) b e d 2-9 - + SbF, TeF, IF, XeF, 1.916(4) 1.862(4) 1.817(10) 1.813(7) 2.075(3) 1.952(4) (1.873(5) 1.843(8) -E-F eq 79.4(1) 78.8(2) 80.9(2) 79.2(4) R. R. Ryan and D. T. Cromer, Inorg. Chem., 11, 2322 (1972). S. H. Mastin, R. R. Ryan and L. B. Asprey, ibid. 9, 2100 (1970). G. R. Jones, R. D. Burbank and N. Bartlett; ibid. 9, 2264 (1970). K. Leary, D. H. Templeton, A. Zalkin, and N. Bartlett, ibid. 12, 1726 (1973). Table 5g EO, (0, symmetry) . a an) c d SbO, TeO, 10, Xe0, 1.97 1.913(3) 1.888(2) 1.864(12) Interatomic Distances, L. E. Sutton, Ed., Chem. Soc. Special Publ. No. 11 (1958). H. Schulz, and G. Bayer, Acta Cryst., B27, 815 (1971). K. Tichy, A. Ruegg, and J. Benes, Acta. Cryst., B36, 1028 (1980). J. Ibers, W. C. Hamilton, and D. R. MacKenzie, Inorg. Chem. 3, 1412 (1964); A. Zalkin, J. D. Forrester, D. H. Templeton, S. M. Williamson, and C. W. Koch, J. Am. Chem. Soc.; 86, 3569 (1964; A. Zalkin, J. D. Forrester and D. H. Templeton, Inorg. Chem. 3, 1417 (1964). 35 Table 6 A Comparison of Some Transition and Non-Transition Element Compounds Molecule WF¢ : TeF¢ ReF, IF 7 7 030), Xe, Symmetry 0, (@) Dept 2p_,0, 009°) 7,2) oe (e) 1.83¢*) (b) (g) re) E-L (A units) 1.833 a 1.825 1.7487 7h £_(mdyn/A) 5.104) 5,0, 65) _ 3-4 s) qe) 5,754) v,(em™*) 769'*) qo () 7366©) 676) 971 906 est T.B.E.(kcal mole!) 101") — gp (n) 100 est 55/9) yo7'P) ay (9) (a) K. Nakamoto, Infrared Spectra of Inorganic and Coordination Compounds, John Wiley, 1963. (>) R. D.. Burbank, and N. Bartlett; Chem. Commn., (1968) 645. (c) E. W. Kaiser, J. S. Muenter, W. Klemperer, W. E. Falconer, and W. A. Sunder, Bell Telephone Report, 1970. _ (4) W. A. Yerarios, Bull. Soc. Chim. Belges, J (1965) Mah. (e) M. Kimura, V. Schomaker, D. W. Smith, and B. Weinstock, J. Chem. Phys., 48 (1968) 4001. . (f) Mean of 1.84 and 1,824 quoted respectively by L. Pauling, and L. 0. Brockway, Proc. Natl. Acad. Sci., 19 (1933) 68 and H. Braune, and S. Knoke, Z. Phys. Chem., (Leipzig) Bel (1933) 297. . 36 (g) TT. Ueki, A. Zalkin, and D. H. Templeton, Acta Cryst., 19 (1965) 157. (h) C. Gunderson, K. Hedberg, and J. L. Huston, Acta Cryst, A 25 $3 (1969) lek. (i) H. H. Claassen, J. Chem. Phys., 30 (1959) 968. (j) K. 0. Christy, and W. Sawodny, Inorg. Chem., 6 (1967) 1783. (k) R. K. Khanna, J. Mol. Spectrose., & (1962) 134. (1) H. H. Claassen, E. L. Casner and H. Selig, J- Chem. Phys., 49 (1968) 1803. (m) P. A. C. O'Hare, and W. N. Hubbard, J. Phys. Chem., 70 (1966) 3353. (n) P. A. C. O'Hare, J. L. Settle, and W. N. Hubbard, Trans. Faraday Soc., 62 (1969) 558. (0) JANAF Thermochemical Tables, The Dow Chemical Co., Midland Michigan, ‘supplement 32 (December 31, 1969). (p) Nat. Bur. Stand. Technical Note 270-4, U.S. Department of Commerce, Washington, D.C. (May 1969). (a) S. R. Gunn, J. Amer. Chem. Soc., 87 (1965) 2290.