COMPOUNDS AND METHODS FOR SEPARATION AND

MOLECULAR ENCAPSULATION OF METAL IONS

BACKGROUND OF THE INVENTION

Field of the Invention:

The present invention relates to novel compounds which are useful for the separation and encapsulation of metal ions, a method of separating and encapsulating metal ions using such compounds, and metal complexes with such compounds. The present invention also relates to an improved method for reprocessing spent nuclear reactor fuel and an improved method for disposing of radioactive wastes.

Discussion of the Background: Nuclear power offers the potential of an abundant and affordable source of energy. However, the safe disposal of radioactive wastes poses a major obstacle for the further development of the nuclear power industry, particularly in the United States. Specifically, the disposal of radioactive lanthanides present in spent radioactive fuels remains a problem (see: G.R. Choppin and J. Rydberg, Nuclear Chemistry, Pergamon, Oxford, pp. 502-559, 1980, incorporated herein by reference) .

Typically, spent fuel rods are first cooled in the reactor for a few weeks and then transferred to a cooling basin. After

storage in the cooling basin for 6-12 months, the spent fuel is either transferred to a special storage facility or sent for reprocessing. At the reprocessing plant, the spent fuel is chopped, and then the oxide fuel is extracted by leaching with boiling 6-11 M HN0 ₃ . This leaching process results in an aqueous solution which contains radioactive wastes, including lanthanide fission products, which must be removed from the elements which are to be recycled into reactor fuel, uranium and plutonium. Currently, the separation of plutonium and uranium is achieved by the solvent extraction process in which the spent fuels are dissolved in nitric acid and contacted with an organic solvent to selectively extract the desired elements. Alternatively, the uranium and plutonium may be separated by first forming UF ₆ and PuF ₆ in a molten fluoride salt eutectic and then distilling these compounds. In another process, the spent fuel may be dissolved in a salt melt and selectively extracted with another salt melt. Lastly, the metallic fuel elements can be melted or dissolved in a molten metal, such as a zinc alloy, and the strongly electropositive fission products removed as oxides after addition of a deficiency of oxygen, and the volatile fission products removed by distillation.

Currently, storage in stable geological formations such as salt domes is contemplated as the preferred method for long term storage of radioactive wastes from spent fuel. Thus, all of these methods require that the separated fission products be

solidified before storage. When the waste is obtained in the form of nitrates it is usually first calcined to convert the metals to oxides. Since such calcines have only a low resistance to leaching, a low heat conductivity, and can be easily dispersed in air, they cannot be used as the final storage product. Thus, the emphasis has been on the development of methods for fixing the waste in borosilicate or phosphate glass. However, such methods are costly and time consuming. Further, the great number of steps increases the risk with regard to safety. Thus, there remains a need for an improved method for reprocessing radioactive wastes and chemical agents to be used in such improved methods. There also remains a need for compounds which are useful for the separation and encapsulation of metal ions. There also remains a need for a method of separating and encapsulating metal ions. There also remains a need for stable complexes of metals, in particular lanthanides, suitable for long-term storage.

SUMMARY OF THE INVENTION Accordingly, it is one object of the present invention to provide novel compounds for the separation of metal ions.

It is another object of the present invention to provide novel methods for separating metal ions using such compounds. It is another object of the present invention to provide novel compounds for the encapsulation of metal ions.

It is another object of the present invention to provide novel compounds for the encapsulation of lanthanide metals.

It is another object of the present invention to provide compounds for the encapsulation of actinide metals. It is another object of the present invention to provide novel methods for encapsulating metal ions using such compounds

It is another object of the present invention to provide novel methods for the encapsulation of lanthanide metals.

It is another object of the present invention to provide novel methods for the encapsulation of radioactive lanthanide metals.

It is another object of the present invention to provide novel methods for the encapsulation of actinide metals.

It is another object of the present invention to provide novel methods for the encapsulation of radioactive actinide metals.

It is another object of the present invention to provide stable complexes of metals suitable for long-term storage.

It is another object of the present invention to provide stable complexes of lanthanide metals suitable for long-term storage.

It is another object of the present invention to provide stable complexes suitable for the long-term storage of radioactive lanthanide metals.

It is another object of the present invention to provide stable complexes suitable for the long-term storage of actinide metals.

It is another object of the present invention to provide stable complexes suitable for the long-term storage of radioactive actinide metals.

It is another object of the present invention to provide a novel method for reprocessing spent nuclear reactor fuel .

It is another object of the present invention to provide a novel method for disposing of radioactive wastes.

It is another object of the present invention to provide novel glasses suitable for the long-term storage of metals.

It is another object of the present invention to provide novel mixed valence tungsten "bronze" materials suitable for the long-term storage of metals.

These and other objects, which will become apparent during the following detailed description, have been achieved by the inventors\' discovery that the polyoxometallate anions of the formula (I) : [DA ₅ M ₃₀ . _x O ₁₁₀ . _x (M\'L) _x ] ^m - (I) in which D is Na ⁺ , Ca ⁺² or an ion of similar size; M is W ⁶⁺ , W ⁵⁺ , or mixtures thereof; M\' is any metallic element from groups 2 to 15 of the periodic table; L is 0 ^2~ , OH ^" , H ₂ 0 or another suitable ligand, A is P, As, Sb, Si, Ge, or combinations thereof; x is 0- 10; and m is the charge based on the valence states of the atoms,

typically 10-20, react selectively with the ions Z ⁿ⁺ to afford an anion of formula (II) :

tZA ₅ M ₃₀ - _x O ₁₁₀ - _x (M\'L) ^(m+1 - ⁿ »- (II.)

in which Z = Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu, Y, or Bi, when n = 3, and Z = Ce, U, Np, Pu, or Am, when n = 4, and that the compounds of formula II may be vitrified to form glasses which are stable and suitable for long-term storage of the metal Z.

BRIEF DESCRIPTION OF THE DRAWINGS A more complete appreciation of the invention and many of the attendant advantages thereof will be readily obtained as the same becomes better understood by reference to the following detailed description when considered in connection with the accompanying drawings, wherein:

Figure 1 is a drawing of [NaP _s W ₃₀ O ₁₁₀ ] ^14~ viewed perpendicular to the anion\' s C ₅ axis. Tungsten and oxygen atoms are represented by large and small circles, respectively, phosphorus atoms by shaded circles, and the sodium atom by a dotted circle;

Figure 2 illustrates the part of the [NaP ₅ W ₃₀ O ₁₁₀ ] ^14~ structure showing the position of the sodium atom relative to th tungsten atoms in inner and outer rings. Tungsten and oxygen atoms are shown as large and small circles, respectively, the

phosphorus atom as a shaded circle, and the sodium atom as a dotted circle;

Figures 3a-c are cyclic voltammograms of A : [NaP ₅ W ₃₀ O ₁₁₀ ] ^14~ and B : [GdP ₅ W ₃₀ O ₁₁₀ ] ¹² - , 10 ^"3 M in 1 M HC1 and C : [GdP ₅ W ₃₀ O ₁₁₀ ] ^12~ , 10" ³ M at pH 10 . 0 ( 0 . 025 M Borax/NaOH, 1 M Na ₂ S0 ₄ ) . Glassy carbon working electrode, Ag/AgCl reference electrode, sweep rate 200 mV/s;

Figure 4 is an X-Band ESR spectrum of powdered K ₁₂ tGdP ₅ W ₃₀ O ₁₁₀ ] at 298°K; Figure 5 is a visible-near IR absorption spectrum of [UP ₅ W ₃₀ O ₁₁₀ ]" ^" in 1 M H ₂ S0 ₄ ;

Figure 6 is an emission spectrum of H _n [EuP ₅ W ₃₀ O ₁₁₀ ] ^(12"n)" in a concentrated aqueous solution. Excitation wavelength 464.7 nm, 303°K. Values of J for the transitions ⁵ D ₀ - ⁷ F _j are given in the figure.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Thus, in a first embodiment, the present invention relates to a method of selectively isolating a particular metal from a complex reaction mixture by taking advantage of the selective nature of the reaction of an anion of formula (I) : [DA ₅ M ₃₀ - _x O ₁₁₀ . _x (M\'L) ^m - (I)

wherein D is Na ⁺ , Ca ⁺² or an ion of similar size; M is W ⁶⁺ , W ⁵⁺ , or mixtures thereof; M\' is any metallic element from groups 2 to 15

of the periodic table; L is O ^2" , OH ^" , H ₂ 0 or another suitable ligand, A is P, As, Sb, Si, Ge, or combinations thereof; x is 0- 10; and m is the charge based on the valence states of the atoms, typically 10-20; with Z ⁿ⁺ to afford an anion of the formula (II) : [ZA _E M ₃₀ . _x O ₁₁₀ . _x (M\'L) _x ] ^(ra+1 - ⁿ⁾ - (II) wherein Z = Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu, Y, or Bi, when n = 3 , and Z = Ce, Np, Pu, or Am, when n = 4.

In the above-given definition of formula (I) , M\' may be any metallic element from groups 2 to 15 of the periodic table. For the purposes of the present invention, such groups include the alkaline earth metals, the transition metals, and metals from groups Ilia, IVa, and Va of the main group. Preferred metals for M\' include Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, Ga, Ge, Sn, Nb, Ta, Mo, and combinations thereof. Particularly preferred metals for M\' include Mo, V, Nb, Ta and combinations thereof.

The choice of L will, in part, depend on the identity of M\' . However, the identification and selection of the appropriate choice for L is within the abilities of one having ordinary skill in the art. In addition to O ^2" , OH ^" , and H ₂ 0, other suitable ligands include CN ^" , NH ₃ , NH ₂ R, NHRR\' , NRR\'R" (wherein R, R\' , and R" are each independently selected from the group consisting of Ci- ₄ alkyl, phenyl, benzyl, and tolyl) , CO, pyridine, C^,, alkyl- substituted pyridine, C ₁ . ₄ -alkyl, C ₆ _ ₁₂ aryl (including phenyl, biphenyl and naphthyl) , C^ alkyl-substituted C ₆ . ₁₂ aryl, and C ₇ . ₂₀ aralkyl (including benzyl) .

Suitably, A is an element selected from the group consisting of P, As, Sb, Si, Ge, and combinations thereof. It is preferred that A is P, As, or combinations thereof. It is particularly preferred that A is phosphorus. Preferably, the anion of formula (I) is [NaP ₅ W ₂₉ VO ₁₁₀ ] ^16"

(equal to [NaP _E W ₂₉ O ₁₀₉ (VO) ] ^~16 ) or [NP ₅ W ₃₀ O ₁₁₀ ] ^14" . It is particularly preferred that the anion of formula (I) is [NaP ₅ W ₃₀ O ₁₁₀ ] ^14" .

The metallate anion [NaP ₅ W ₃₀ O ₁₁₀ ] ^14" is known as the Preyssler anion and will hereinafter sometimes be referred to as NaP ₅ W ₃₀ . This particular anion may be synthesized as described in Jeanin, Y. et al, Inorg. Synth., 1990, 2_1_, 115; or Preyssler, C, Bull. Soc. Chim. Fr. , 1970, 30, both of which are incorporated herein by reference.

The remaining anions of formula (I) may be prepared by a number of methods. In a first method, Preyssler\'s anion is treated with base by dropwise addition in an aqueous solution to obtain a solution having a pH of 10 to 12, preferably about 12. The addition of alkali is continued, at a rate such that the pH does not exceed 12 until about 12 equivalents of base have been added. Addition of saturated KCl yields a white precipitate which then is added to an aqueous solution containing 1 to 3 equivalents of M\'L. The anion of formula (I) may be precipitated by the addition of saturated KCl . This method is exemplified for K ₆ Na ₁₀ [NaP ₅ W ₂₉ VO ₁₁₀ ] "40^0 in Alizadeh, M.H. et al, J. Am. Chem. Soc.. 1985, 107, 2662, which is incorporated herein by reference.

Other anions of formula (I) may be prepared by substituting various sources of M\'L for the VOS0 ₄ used in the above-referenced preparation. Suitable examples of M\'L sources include TiCl ₄ , Cr(N0 ₃ ) ₃ , CoS0 ₄ , MnS0 ₄ , Fe(N0 ₃ ) ₃ , Ni(N0 ₃ ) ₂ , Cu(N0 ₃ ) ₂ , Zn(N0 ₃ ) ₂ , Ga(N0 ₃ ) ₃ , GeCl ₄ , K ₆ Ta ₆ 0 ₁₉ , Na ₂ Mo0 ₄ , KV0 ₃ , K ₇ HNb ₆ 0 ₁₉ , and C ₆ H ₅ SnCl ₃ .

Preferred M\'L sources include Na ₂ Mo0 ₄ , KV0 ₃ K ₇ HNb ₆ 0 ₁₉ and C ₆ H ₅ SnCl ₃ . Alternatively, the anions of formula (I) may be prepared by carrying out the steps for the synthesis of Preyssler\' s anion described in the Examples, while replacing a portion of the Na ₂ W0 ₄ «2H ₂ 0 used in the preparation of Preyssler\' s anion with a source of M\'L. In this case, the suitable and preferred sources of M\'L are the same as those described above.

The selective encapsulation of Z ⁿ⁺ with the anion of formula (I) may be carried out as follows. Generally, the anion of formula (I) is contacted, in an aqueous solution, with Z ⁿ⁺ . The order of addition is not critical. Thus, a solid salt containing the anion of formula (I) may be added to an aqueous solution of Z ⁿ⁺ or a solid containing Z ⁿ⁺ may be added to an aqueous solution containing the anion of formula (I) . It is also possible to mix a solid containing Z ⁿ⁺ with a solid salt containing the anion of formula (I) and then dissolve them by adding water or an aqueous solution.

The contacting of Z ⁿ⁺ with the anion formula (I) may be carried out over a wide range of conditions . Particularly good results have been obtained by carrying out the contacting step at

a temperature of 120 to 200°C, preferably 140 to 180°C, for a time of 5 to 30 hrs, preferably 10 to 15 hrs. The pressure is not critical, and the contacting step may be carried out at atmospheric pressure, slightly reduced pressure, or even elevated pressures as high as 5 atm, preferably 1-2 atm.

Although, the aqueous phase in which the anion of formula

(I) and Z ⁿ⁺ are contacted may contain a variety of additional components, it is preferred that the pH of the aqueous phase be maintained within a range of 2 to 10, preferably 4 to 8, during the contacting step.

After the contacting step is complete, the anion of formula

(II) may be obtained in the form of an isolated salt by a number of methods. For example, addition of KCl to an aqueous solution will precipitate the anion of formula (II) in the form of the potassium salt. Other methods of isolating the anion of formula (II) include: (1) precipitation of other salts, e.g. Cs ⁺ , and NR ₄ " (especially R = Me, Et, n-Pr) ; (2) Extraction into organic solvent, e.g. toluene, hexane by means of phase transfer agent NR ₄ ⁺ ; R = C ₇ or greater, [general method described by Katsoulis and Pope, J. Am. Chem. Soc. , 1984, 106 2737] ; and (3) absorption on to anion exchange material.

It should be understood that the present anions of formulae (I) and (II) may exist in the form of salts with a wide variety of one or more cations. The number of cations associated with the anions of formulae (I) or (II) will , of course, depend on

the oxidation state of both the cations and anions . It should also be understood that, for any given encapsulation reaction as described above, the identity of the cations associated with the anion of formula (I) may differ from those associated with the anion of formula (II) . Thus, it may be desired to start with a salt in which the cations are selected such that the anion of formula (I) is soluble in the aqueous phase under the contacting conditions and then add a salt containing cations that form an insoluble salt with the anion of formula (II) to precipitate the anion of formula (II) after the contacting step is complete.

Examples of cations which form soluble salts with the anions of formula (I) include H\ Li\ Na ⁺ , K ⁺ , and NH ₄ ⁺ . Cations which form insoluble salts with the anions of formula (I) include NMe ₄ ⁺ , NEt ₄ ⁺ , NPr ₄ ⁺ , NMePh ₃ \ and Cs ⁺ . Although not intended to be limiting, the selective reaction of Z ⁿ⁺ with the anion of formula (I) will now be further illustrated by way of a detailed discussion of the reaction of Z ⁿ⁺ with NaP ₅ W ₃₀ . A crystallographic investigation (Alizadeh, M.H. et al, J. Am. Chem. Soc.. 1985, 107, 2662) of the so-called Preyssler anion revealed it to be [NaP ₅ W ₃₀ O ₁₁₀ ] ^14" (NaP ₅ W ₃₀ ) and not [HP ₃ W ₁₈ 0 ₆₆ ] ^8" as first assigned by Preyssler (Preyssler, C, Bull . Soc. Chim. Fr. , 1970, 30) . The anion has approximate D _5h symmetry (see Figure 1) . It consists of five PW ₆ 0 ₂₂ units derived from the Keggin anion [PW ₁₂ O ₄₀ ] ^3\' , arranged in a crown so that the tNaP ₅ W ₃₀ O ₁₁₀ ] ^14" anion has an unusual fivefold symmetry axis. The

tungsten atoms are distributed in four parallel planes perpendicular to this axis. The two outer planes each contain five tungsten atoms and the two inner planes ten tungsten atoms. The structure leaves a cylindrical vacancy along the five-fold axis and the sodium ion is positioned in this vacancy, not in the center as might be expected, but in one of the inner oxygen planes so that it is coordinated to five oxygens in this plane and to five in an outer oxygen plane at a larger distance, see Figure 2. The structure explained previously-recorded ³¹ P NMR (Massart, R. et al, Inorg. Chem. , 1977, JL6 _. 2916) which showed a single line for the five equivalent phosphorus atoms, and ¹⁸³ W NMR (Acerete, R. et al, Inorg. Chem.. 1984, 23 _. 1478) which gave four lines in a 2:2:1:1 ratio as expected from the dissymmetry introduced by the sodium ion. Sodium-23 NMR showed separate signals for internal and free cationic Na ⁺ (Alizadeh, M.H. et al, J. Am. Che . Soc.. 1985, 107, 2662) .

The sodium ion was very tightly bound but could apparently be replaced under severe experimental conditions (heating in aqueous solution at 120°C in a bomb) by Ca ²⁺ , which has a similar radius (1.14A) (The "ionic radius" of a cation is known to depend upon its coordination number. In order to compare sizes of the different cations discussed in this application, the inventors have chosen the set of effective ionic radii of Shannon and Prewitt fActa Crystallogr. , 1976, A32, 751] appropriate for six- fold coordination by oxide. The effective coordination number of

the central atom in P ₅ W ₃₀ is larger than six, of course, but sufficient reliable data are not available for larger coordination numbers) to Na ⁺ (1.16A) (Alizadeh, M.H. et al, J. Am. Chem. Soc. , 1985, 107. 2662) . The ionic radii of the _. trivalent lanthanides (Ln) (r = 1.17 to 1.00A) are similar to those of Na ⁺ and Ca ²⁺ , and Ln ³⁺ ions are often used to replace Ca ²⁺ in complexes in biological studies . Because of their higher charge the lanthanide complexes would be expected to be even more stable than those of Na ⁺ and Ca ²⁺ . The reaction between Ca ²⁺ and NaP ₅ W ₃₀ has been shown (Alizadeh, M.H. et al, J. Am. Chem. Soc. , 1985, 107, 2662) to be accompanied by the appearance of Na ⁺ in the solution as measured with a sodium electrode, leading to the conclusion that Ca ²⁺ had replaced Na ⁺ in the cavity of the heteropolyanion. By reacting a neutral or slightly acid aqueous solution of K ₁₂₅ Na _1-5 [ [NaP ₅ W ₃₀ O ₁₁₀ ] at 145-180°C with 1-2 equivalents of a variety of lanthanide and other metal ions having radii similar to that of Na ⁺ , products were obtained which contained the reacting cations and which were characterized by cyclic voltammograms and ³¹ P-NMR spectra that differed from that of NaP ₅ W ₃₀ , see below. The yields varied with the metal ion and temperature and are summarized in Table 1.

Table 1 . Conditions and Yields for Preparation of rzP _s W _3P O ₁₁₀ ] :

Sm ³⁺

Eu ³⁺

Gd ³⁺

Tb ³⁺

Dy ³⁺

Ho ³⁺

Er ³⁺

Tm ³⁺

Yb ³⁺

Lu ³⁺

Ce ⁴⁺

Y ³⁺

Bi ³⁺

U ⁴⁺

Ca ²⁺

[NaP ₅ W ₃₀ O ₁₁₀ ] ^{14 "} + Z ⁿ⁺ > [ ZP ₅ W ₃₀ O ₁₁₀ ] ^{( 15 "n> "} + Na ⁺

C[NaP ₅ W ₃₀ ] = 8.0 mM, C _Zn , * 8-16 mM

While Gd ³⁺ and Tb ³⁺ reacted practically quantitatively at 145°C, Eu ³⁺ reacted only partly at this temperature but almost quantitatively at 165°C. The smaller lanthanide ions Yb ³⁺ and Lu ³⁺ required still higher temperatures and reacted only partly even at 180°C.

All lanthanide ions with ionic radii smaller or equal to that of Sm ³⁺ (r=1.10A) were found to react with NaP ₅ W ₃₀ , but no reaction was observed for the larger lanthanide ions Nd ³⁴

(r=1.12A) , Pr ³⁺ (r=1.13A) and Ce ³⁺ (r=1.15A) even at temperatures

up to 180°C. Assuming that the cations have replaced Na ⁺ in the central cavity, this behavior indicates an extraordinary and unprecedented size selectivity. Although Sm ³⁺ and Nd ³⁺ differ in ionic radius by only 0.02A, Sm ³⁺ readily entered the heteropolyanion, but no detectable reaction took place with the larger cations. In contrast, other lanthanide-containing heteropolyanions such as [Ln(PW _1:L 0 ₃₉ ) ₂ ] ^11_ and [Ln(SiW _xl 0 ₃₉ ) ₂ ] ^13" can be made with all lanthanide ions (Peacock, R.D. et al, J. Chem. Soc. (A) , 1971, 1836; and Fedotov, M.A. et al, Polyhedron, 1990, 9_, 1249) .

Although not intended to be limiting, a possible explanation for the dramatic size selectivity of P ₅ W ₃₀ O ₁₁₀ as compared to [Ln(PW ₁₁ 0 ₃₉ ) ₂ ] ^11" and [Ln(SiW _ι:L 0 ₃₉ ) ₂ ] ^13" may be that the latter complexes are more flexible with the lanthanide ions sandwiched between two lacunary heteropolyanion ligands, while in the P ₅ W ₃₀ O ₁₁₀ heteropolyanion the size of the central cation is restricted by the diameter of the cavity which may not have the flexibility to expand beyond a certain well defined size. The sudden cut-off between Sm ³⁺ and Nd ³⁺ is still surprising considering the capability of the lanthanide ions to adopt to a variety of coordination numbers with different effective ionic radii.

At the other end of the lanthanide series, Yb ³⁺ (r=1.01A) and Lu ³⁺ (r=1.00A) with the smallest ionic radii replaced Na ⁺ only

partly. A possible, but not limiting explanation for this result is that these ions are too small to compete effectively with Na ⁺ .

The smaller yields with Ce ⁺ and U ⁴⁺ were partly due to the low solubility of the tetravalent cations with the heteropolyanion necessitating using lower concentrations of cations ( <1 equivalent compared to two equivalents for the other metal ions) . Not only did Ce ³⁺ not react with NaP ₅ W ₃₀ as mentioned above, but no reduction could be observed of Ce ^IV P ₅ W ₃₀ in aqueous solution (see below) further illustrating the point that Ce ³⁺ might be too large for the cavity.

While ease of substitution of Na ⁺ by lanthanide ions seems to follow the size of their ionic radii strictly, substitution by other metal ions of similar size did not always proceed as expected. For example, Y ³⁺ (r=1.04A) , U ⁺ (r=1.03A) , Bi ³⁴ (r=1.17A) , and Ce ⁴⁺ (r=1.01A) did react, but other metal ions with similar radii such as Cd ²⁺ (r=1.09A) , Tl ³⁺ (r=1.03A) , Sn ²⁺ , Hg ²⁺ (r=1.16A) , and Th ⁴⁺ (r=1.08A) did not, under the same conditions. The fact that Cd ²⁺ did not react is especially surprising as the similarly sized Ca ²⁺ reacted readily. It is certainly possible that stereochemical (Hg ²⁺ , Sn ²⁺ ) and "hard- soft" preferences might also contribute to reactivity - the polyanion cavity is lined with "hard" oxide ions.

The ZP ₅ W ₃₀ ions were stable in aqueous solution for at least 3 weeks in 1 M HCl at 25°C, but a slow decomposition ( [ZP ₅ W ₃₀ O ₁₁₀ ] ^(15"n)" → W0 ₄ ^2" , HPO. ^2" , Z ⁿ⁺ ) took place at pH 9-10 with a

halflife of about 2 days in a pH 10, 0.025 M borax buffer and several weeks at pH 9, as measured voltammetrically with GdP ₅ W ₃₀ using a supporting electrolyte of 1 M Na ₂ S0 ₄ .

When a solution containing EuP ₅ W ₃₀ in 1 M NaCl was heated at 160°C for 24 h, no detectable amount of NaP ₅ W ₃₀ was observed although LuP ₅ W ₃₀ under the same conditions gave mostly NaP ₅ W ₃₀ . These observations combined with the yields in the preparations of ZP ₅ W ₃₀ (Table 1) show increased stability of LnP ₅ W ₃₀ over NaP ₅ W ₃₀ , as expected from the more highly charged Ln ³⁺ ions, and that LuP ₅ W ₃₀ is less stable than EuP ₅ W ₃₀ . This was further shown by the observation that Eu ³⁺ replaced ca 85% of Lu ³⁺ in LuP ₅ W ₃₀ when a 2:1 mixture was heated at 160°C for 20 hr.

The observation that different Ln ³⁺ ions reacted with NaP ₅ W ₃₀ with different efficiency suggested that NaP ₅ W ₃₀ might be used to separate mixtures of lanthanides. To test this notion some competition experiments were performed in which mixtures of two Ln ³⁺ ions and NaP ₅ W ₃₀ in 1:1:1 ratio were heated at 160°C for 20 hr and the products examined by ³¹ P NMR of the reaction mixture. Solutions containing Sm ³⁺ and the smaller Yb ³⁺ gave only SmP ₅ W ₃₀ as product, while the neighboring Sm ³⁺ and Eu ³⁺ reacted to give a mixture of SmP ₅ W ₃₀ and EuP ₅ W ₃₀ in a 1:4 ratio. When mixtures of Eu ³⁺ and Tb ³⁺ were used EuP ₅ W ₃₀ was the only product. It thus appears that NaP _s W ₃₀ efficiently separates smaller Ln ³⁺ ions from larger ones and preferentially reacts with Eu ³⁺ rather than Sm ³⁺ .

Once incorporated in the heteropolyanion the free Ln ³⁺ ions can be recovered by basic hydrolysis. IR Spectroscopy

The IR spectra of all the products were practically identical and differed only slightly from the parent NaP ₅ W ₃₀ heteropolyanion; the only consistent differences being in the position and size of one of the P-0 stretching bands (1073 cm ^"1 (m) for NaP ₅ W ₃₀ and 1060 cm ^"1 (s) for ZP ₅ W ₃₀ ) and a better resolved peak at 980 cm ^"1 for NaP ₅ W ₃₀ . The splitting of the 1160 cm ^"1 P-0 stretching band reported (Alizadeh, M.H. et al, J. Am. Chem.

Soc. , 1985, 107, 2662) for CaP ₅ W ₃₀ was not seen and may have been due to an impurity or artifact in the earlier spectrum. NMR Spectroscopy

In the ²³ Na NMR spectrum of NaP ₅ W ₃₀ it was found that the coordinated sodium ion gave rise to a broad line at a slightly different frequency from that of the free sodium ion (Alizadeh, M.H. et al, J. Am. Chem. Soc. , 1985, 107, 2662) . In order to test whether Na ⁺ had indeed been replaced by the reacting cations in the ZP ₅ W ₃₀ complexes ²³ Na NMR spectra of the acid (more soluble) forms of YP ₅ W ₃₀ and NaP ₅ W ₃₀ were attempted using the same conditions of concentration and instrumentation. The Y ³⁺ complex was chosen because it is diamagnetic so that no significant shift from NaP ₅ W ₃₀ or line broadening would be expected if Na ⁺ were still coordinated. No Na resonances were observed for- YP ₅ W ₃₀ supporting the notion that the observed reaction is indeed a

displacement of Na ⁺ by the reacting cations and not attachment at another site of the heteropolyanion.

The ³¹ P NMR chemical shifts of ZP _S W ₃₀ are shown in Table 2.

^a Lanthanide induced shift, δ (Ln) - δ (Lu) ^{b 2} J(P-0-Y) = 1.6 Hz

A single ³¹ P line was observed for all complexes except for YP ₅ W ₃₀ which gave a doublet with a coupling constant of 1.6 Hz . These data confirm that all the phosphorus atoms are equivalent

(as expected) in ZP ₅ W ₃₀ , the doublet in YP ₅ W ₃₀ presumably being due

to splitting caused by ⁸⁹ Y (100%, 1=1/2) . ³¹ P NMR spectra of the ZP ₅ W ₃₀ ions (Table 2) gave shifts for diamagnetic Z ³⁺ ions of -8.2 to -10.2 ppm; for NaP ₅ W ₃₀ -9.4 ppm; for CaP ₅ W ₃₀ -11.1 ppm; for CeP ₅ W ₃₀ -16 ppm; and for UP ₅ W ₃₀ -15.5 ppm. For the paramagnetic Ln ³⁺ ions the shifts varied from +18 to -68 ppm and widths varied from 5-240 Hz except for GdP ₅ W ₃₀ which gave no observable spectrum.

Some of the lanthanides induced shifts to higher frequency, others to lower frequency, as was also observed for [Ln(PW ₁₁ 0 ₃₉ ) ₂ ] ^11" (Fedotov, M.A. et al, Polyhedron, 1990, 9_, 1249) . The lanthanide induced shifts, LIS _obs , were taken as the observed shifts for LnP ₅ W ₃₀ minus the observed shift for LuP ₅ W ₃₀ . LIS _0BS are known (Sherry, A.D. et al, Lanthanide Probes in Life, Chemical and Earth Sciences, Bunzli and Choppin, Eds., Elsevier, 1989, and references therein; Bleaney, B., J. Magn. Res . , 1972, 8 _. , 91; and Golding, R.M. et al, Aust. J. Chem. , 1972, 2j5, 2577) to be expressed as a sum of the pseudo-contact (dipolar, through-space) shift, LIS _pc , and the contact (scalar, through-bonds) shift, LIS _C .

LIS _obs = LIS _pc + LIS _C (I)

_ττ - _ [β ² (r )2A ₂ °(3coε ² θ-D + β ² (r ² )2A ₂ ² sin ² θcos2 ] ,_ .

J_ιlbp _c = -,m_ ( X a. )

60(kt) ² r ³

where C _j = g ² J(J + 1) (2J - 1) (2J + 3) and r, θ, φ are spherical coordinates of the observed nucleus (P) with respect to the lanthanide, β is the Bohr magneton, A is the electron-nuclear hyperfine coupling constant, γ is the magnetogyric ratio, 2 >2A ₂ ° and A ₂ ² are ligand field terms, and the other symbols have their usual significance. Values of C _j which determine shifts to higher or lower frequency, have been calculated by Bleaney (Bleaney, B., J. Magn. Res.. 1972, 8., 91) for each lanthanide ion. The spin expectation values Z > are tabulated by Golding and Halton (Golding, R.M. et al, Aust . J. Chem.. 1972, 25., 2577) .