Article
Reference
A theoretical study of plutonium diketone complexes for solvent extraction
GAGLIARDI, Laura, et al.
Abstract
We present a relativistic density functional study on some plutonium compounds with thenoyltrifluoroacetone and similar ligands which can be used in the extraction of plutonium.
The method of effective core potentials is used on plutonium. The binding energies of the complexes of plutonium in the formal oxidation states II, IV and VI have been determined and the geometries of some of the complexes have been fully optimized. The stability of the compounds in the different oxidation states and the effect of varying the side groups in the ligands are discussed. Comparisons with analogous uranium compounds are presented.
GAGLIARDI, Laura, et al . A theoretical study of plutonium diketone complexes for solvent extraction. Chemical Physics , 2000, vol. 252, no. 1-2, p. 47-55
DOI : 10.1016/S0301-0104(99)00359-6
Available at:
http://archive-ouverte.unige.ch/unige:3737
Disclaimer: layout of this document may differ from the published version.
1 / 1
www.elsevier.nlrlocaterchemphys
A theoretical study of plutonium diketone complexes for solvent extraction
Laura Gagliardi
a,), Nicholas C. Handy
b, Chris-Kriton Skylaris
b, Andrew Willetts
baDipartimento di Chimica Fisica e Inorganica, UniÕersita di Bologna, Viale Risorgimento 4, I-40136 Bologna, Italy`
bDepartment of Chemistry, UniÕersity of Cambridge, Cambridge CB2 1EW, UK Received 12 July 1999
Abstract
We present a relativistic density functional study on some plutonium compounds with thenoyltrifluoroacetone and similar ligands which can be used in the extraction of plutonium. The method of effective core potentials is used on plutonium. The binding energies of the complexes of plutonium in the formal oxidation states II, IV and VI have been determined and the geometries of some of the complexes have been fully optimized. The stability of the compounds in the different oxidation states and the effect of varying the side groups in the ligands are discussed. Comparisons with analogous uranium compounds are presented.q2000 Elsevier Science B.V. All rights reserved.
1. Introduction
Extensive use is made of ligand extraction in the reprocessing of nuclear fuel. Although there has been much experimental work in refining the process, our theoretical understanding of the chemistry of the actinide extraction is still limited. It is also the case that much of the chemistry developed for this process has not changed much for the last forty years. A more detailed theoretical understanding would be also useful in gaining a knowledge of the interactions of the actinides with the various constituents of cells and tissues. This is important for the understanding of the mechanisms which control their specific tissues deposition pattern, the initiation of toxic effects and for the development of methods for treating people who may become contami- nated with actinides. It may then be possible to select a specific ligand appropriate for their complete removal.
The reactions of the actinides with various components of mammalian blood, cells and tissues has been the
w x
subject of a number of publications 1–3 .
Because it is the most abundant actinide produced in the fission nuclear fuel cycle, and because it is now present in minute quantities in our normal environment, the chemistry of plutonium has been more widely studied than that of any other actinide. Chelation therapy for the removal of plutonium following contamination
w x
has been the subject of a number of reviews 2–5 . Several compounds have been proposed as chelating agents.
Potentiometric pH titrations and ion-exchange and spectrophotometric studies show that stable 1:1 metal
Ž . Ž . Ž . Ž .
chelates are formed with Pu III , Pu IV , Pu VI and Pu VI O .2
)Corresponding author. E-mail: [email protected]
0301-0104r00r$ - see front matterq2000 Elsevier Science B.V. All rights reserved.
Ž .
PII: S 0 3 0 1 - 0 1 0 4 9 9 0 0 3 5 9 - 6
( ) L. Gagliardi et al.rChemical Physics 252 2000 47–55 48
Many organic reagents in aqueous solution form chelates which are often extractable into organic solvents, such as benzene and chloroform. They contain an acidic hydrogen and a donor group, forming a bidentate
Ž .
chelate with a metal ion. The acidic hydrogen is usually from an OH or SH group. Structural factors affect the acidity of the OH group, the basic strength of the donor atom and the effectiveness of the extractant through ring size. In general, a more basic chelating agent will form a more stable metal chelate. However, a balance needs to be achieved as a more acidic chelating agent will be of greater use in extracting from acidic aqueous solutions. This is often required to prevent hydrolysis of the metal ions. Adjustment of the initial aqueous solution may even be used to make the chelatermetal ion reaction elective or even specific. One of the first steps in uranium and plutonium removal from spent nuclear fuel is to dissolve the fuel in nitric acid. Hence it is necessary to use an extracting ligand which is acidic.
w x
1,3-diketones are often used for the extraction and purification of actinides 1,6 . In the enol form they have a hydrogen replaceable by a metal and a ketonic donor. The nature of the alkyl or aryl side groups affects the
Ž .
acidity of the hydrogen. In the case of Pu, a possible chelating species is 4,4,4-trifluoro-1- 2-thienyl -1,3-
Ž .
butanedione, or thenoyltrifluoroacetone TTA . TTA can be obtained in a pure state and has a high acidity, which is useful in extraction at low pH. In aqueous solution it forms a keto hydrate which has a low distribution ratio into chloroform and benzene. The chelation reaction involved is given by the following equation:
The magnitude of the equilibrium constant for this reaction is a measure both of the stability of the metal
Ž . Ž .
chelate and also of its formation at a given aqueous pH. In the case of Pu, the chelation of Pu III , Pu IV and Ž .
Pu VI involves three, four and two molecules of TTA, respectively. From a study on the equilibrium constants w x6 , it turns out that the extractability of these species is determined by the series
Pu IVŽ .4Pu IIIŽ .)Pu VIŽ .fU VIŽ .
This is in line with the observation that a common ratio of the coordination number to the metal charge is two 7 . It has also been observed that the extractability increases with increasing charge on the ion and, for thew x lanthanides and actinides, with increasing atomic number, although not uniformly.
Little information about the molecular structure and electronic properties of these complexes is available in
w x
the literature. A few experimental papers however address similar problems 8–12 . These references contain some information on the stability of the complexes and typical actinide–oxygen bond lengths, that we here will
w x Ž .
use to compare with our results. Ref. 8 discusses an X-ray absorption fine structure XAFS determination of
Ž . w x
Np VII in alkaline solution. Ref. 9 is an experimental study on a Np crown ether inclusion complex, wNpO2Žw18 Crown-6 ClO . The authors synthesized the complex, characterized it spectroscopically and showedx .x 4
q 2q
w x
that NpO2 is easier to encapsulate in aqueous solution than UO2 . Ref. 10 is an experimental study on a
Ž . w Ž . x6y
Pu IV carbonate system, Pu CO3 5 in the solid state and in solution. The authors obtained single crystals of wNa Pu CO6 Ž 3 5 2. x Na CO2 3P33H O and collected XAFS data for Pu IV in solution. The study indicates that2 Ž . wPu COŽ 3 5. x6y is the limiting species in high-carbonate solutions. Ref. 11 is a theoretical analysis of X-rayw x
Ž .
absorption near-edge structure XANES for Pu hydrates with a formal oxidation state of Pu ranging fromq3 w x
to q6, using an ab initio multipole scattering code. Finally, in Ref. 12 the crystal structures of uranyl complexes with four hydroxypyridinone ligands are compared and correlated with their chemical and biological properties.
Ž . Ž . Fig. 1. Pu II complex with thenoyltrifluoro-acetone TTA .
In this paper we present the first ab initio study on Pu complexes with TTA and some model ligands related
Ž . Ž . Ž . Ž .
to TTA Figs. 1 and 2 . Our aim is to compare the known trend in extractability of Pu IV , Pu VI and U VI with the calculated stability of their complexes with TTA. We also decided to look at the influence of the ligand structure on the stability. This was done with the simplest complexes of Pu and U in the formal oxidation state II. Our interest here was simply in the magnitude of the effect we might expect by varying the side groups compared with varying the actinide oxidation state. The calculations on the corresponding uranium complexes are intended to examine the trend in the variation of the stability of the complex with changes in the actinide element.
Ž . Ž .
Some emphasis is placed on the Pu II and U II oxidation states although these are not observed in reality.
We regard the complexes formed by these oxidation states as our simplest model systems with which to examine the magnitude of changes in the binding energy due to variations in the ligand structure and functional.
Given the fact that all of our calculated complexes are simply approximate descriptions of the real world
Fig. 2. Pu complexes with model ligands similar to TTA.
( ) L. Gagliardi et al.rChemical Physics 252 2000 47–55 50
Ž .
systems e.g., due to lack of solvent effects we feel that this is reasonable. Similarly, we fully optimize the
Ž Ž . Ž . .
structures of only a few selected complexes, Pu L12 and PuO L12 2 and regard these geometries as typical of all of our complexes.
The treatment of complexes containing heavy atoms like U or Pu implies the inclusion of relativistic effects w13,14 . Moreover, here one has to be able to assess the suitability of chelation agents for the solvent extractionx process, which implies the calculation of the energies of complexes formed between organic ligands and the various cations, like the uranyl or plutonyl ions.
Until recently the computational cost of quantum chemistry and the consequent limitations on the sizes of systems that could be treated has resulted in little industrial interest in the application of these techniques.
Ž .
Recently, however, the improvement in computer technology and the advent of density functional theory DFT have led to the re-evaluation of the benefits of quantum chemistry to industry.
In Section 2 the theoretical method and the computational details are described. In Section 3 a discussion of the results is presented. Finally, in Section 4 some conclusions are presented.
2. Theoretical approach and computational details
The calculations have been performed by using the newly developed MAGIC quantum chemistry code, directed at providing a means of performing chemically accurate calculations on systems containing many
w x
atoms, some of which are heavy. TheMAGIC code is described elsewhere 15–20 , but, in its essentials, it is a code based on the use of GAUSSIANbasis sets that allows DFT calculations within the Kohn–Sham paradigm.
Ž . w x
In most of the calculations the local density approximation exchange LDAX functional 21 was used.
Ž .
Some comparative calculations were done with LDAX plus the Vosko, Wilk and Nusair VWN correlation w x
functional 22 , commonly called LDA, and some with the generalized gradient approach functional, BLYP w23,24 .x
w x
For the radial quadrature we followed the scheme by Mura and Knowles 25 , and for the angular quadrature
w x w x
the scheme by Lebedev 26 . The Coulomb integrals were evaluated by Rys quadrature 27 , and an auxiliary
w x w x
basis set was used, according to the methods of Dunlap 28 and Eichkorn et al. 29 . The relativistic effects are
Ž .
treated through an implementation of the relativistic effective core potentials RECP of Hay and co-workers w30,31 .x
w x
The RECP employed on the U and Pu atoms are those reported by Hay and Martin 32 . The valence basis
Ž .
set used to represent the 6s, 7s, 6p, 7p 6d and 5f orbitals was a 10s8p2d4f primitive set contracted to w3s3p2d2f for U and a 12s10p2d4f primitive contracted to a 3s3p2d2f for Pu, as reported in Hay andx Ž . w x
w x w x
Martin’s paper 32 . The all-electron basis used for the light atoms was the Dunning DZ 33 . For some of the complexes we repeated the calculations by using a DZP basis on the ligand atoms bound to the heavy element to investigate the effect of polarization functions. The auxiliary basis set for the light atoms were those optimized
w x w x
by Eichkorn et al. 29 for a split valence plus polarization 34 basis. For U and Pu, since a specific auxiliary w x
basis set does not exist, we used the rubidium auxiliary basis, reported in Ref. 29 . A basis set superposition
Ž . w x
error BSSE 35 investigation was carried out for most of the complexes. The initial geometry of the
w x 2
complexes was optimized by using the force field by Rappe et al. 36 within the packageCERIUS . Subsequent full optimizations were made of some Pu complexes usingMAGIC.
Ž . Ž .
As already mentioned in Section 1, we initially studied the Pu II and U II complexes, although not observed in reality, because these are the simplest model systems, with which we can examine the ligand side
Ž .
group effect on the stability of the whole complexes. We thus considered the complex Pu TTA 2containing two TTA molecules, corresponding to Pu in its valence II . We then collapsed TTA to some smaller model ligands,Ž .
Ž .
which we indicate as L1, L2, L3 and L4, and the corresponding Pu complexes were formed see Fig. 2 . The
Ž . y
Fig. 3. Pu IV square antiprismatic complex with four L2 molecules.
ligand L1 does not contain any side group, L2 contains the electron-withdrawing CF3 group, L3 the electron-donating sulfur ring and L4 a different electron-donating group, NH . In this way we investigated the2 effect on the Pu–O bond of the different side groups present in TTA. We considered the Pu II complexes withŽ .
Ž .
these four ligands and determined their complexation energy. The geometry of the Pu L12 was then fully
Ž . Ž .
optimized using DFT. We did not optimize the geometry of the remaining Pu II compounds because: 1 the
Ž . Ž .
binding energy of the fully optimized Pu L1 2 differs from the initial one only by 3% and 2 the side group effect on the stability should not be strongly dependent on small variations in the geometry parameters.
Ž . w x
We then investigated Pu IV , which may form stable complexes 3 . The coordination number is generally Ž .
eight in Pu IV complexes. A qualitative assessment of the idealized geometry of an eight-coordinate Pu complex is often described in terms of cubic distortion which closely resemble a dodecahedron, square
w x Ž .
anti-prism or bicapped trigonal prism 3 . We chose to look at a square anti-prismatic complex of Pu IV ,
Ž . Ž .
Pu L24 see Fig. 3 .
Fig. 4. PuO2q2 complex with two L2ymolecules.
( ) L. Gagliardi et al.rChemical Physics 252 2000 47–55 52
Ž . Ž . Ž .
We finally considered two plutonyl complexes, PuO L12 2 and PuO L2 , corresponding to Pu VI . We2 2 Ž .
determined their complexation energy and we fully optimized the geometry of PuO L1 . In these compounds2 2
2q
Ž .
the PuO2 lies perpendicular to the ligand plane see Fig. 4 .
Some comparative calculations have been carried out on the analogous uranium complexes in its oxidation states II and VI. In this case we also extended the range of investigation to include different functionals. The
Ž . Ž . Ž . Ž .
dissociation energy of U TTA , U L22 2 and UO L22 2 has been computed at the LDAX level, and U TTA2 has been investigated also at LDA and BLYP level.
In this preliminary study, all the calculations on the complexes were performed for the closed-shell singlet state. As already mentioned, the main aim of this paper is the investigation of trends in energetics, and in this respect, the closed-shell state can be considered a representative model for the real system. The open-shell systems will be further studied in future work. As regards the Pu and U ions, Pu2q, Pu4q, PuO22q, U2q and UO22q, several multiplicities were investigated by simply examining states with different Sz values in our unrestricted code. In the subsequent energetic calculations, the lowest energy states were used.
A Mulliken-type population analysis was finally performed on three of the compounds mentioned above,
Ž . Ž . Ž .
Pu L1 , Pu L22 2 and PuO L2 , which were taken as models to investigate the electronic properties, effective2 2 charges and amount of d and f character of the plutonium atom.
3. Results and discussion
Ž . Ž . Ž .
In Table 1 we report the energy of Pu II , Pu IV and U II for different multiplicities, represented by different S values.z
Ž . Ž . Ž . Ž .
For Pu II the septet Szs3 is the spin state with the lowest energy, while for Pu IV and U II the quintet is the lowest energy. In the calculation of the dissociation energy of the complexes we thus considered Pu II asŽ .
Ž . Ž . Ž . Ž .
a septet and Pu IV and U II as quintets. As regards Pu VI and U VI , in plutonyl and uranyl, respectively, we investigated several spin states and it turned out that PuO22q is a triplet while UO22q a singlet.
In Table 2 the binding energies of the complexes are reported. It is seen that, among the Pu II compounds,Ž . Ž .
the most stable is Pu L1 , with no side groups. The introduction of the side group destabilizes the complexes in2
Ž . Ž .
general, even though the complexes with the electron-donating side groups, Pu L3 2 and Pu L4 , appear to be2
Ž . Ž .
more stable than the one with the electron-withdrawing side group, Pu L2 . Pu TTA2 2 which has both Ž .
electron-donating and electron-withdrawing groups has a stability similar to Pu L2 , which might indicate that2 the CF side group has a stronger effect, in the opposite direction, than the S-cycle or NH groups. We repeated3 2
Ž .
the calculations for Pu L1 , by using a DZP basis on the oxygen atoms bound to the Pu, in order to investigate2 the effect of polarization functions. The dissociation energy varied by 2%. The BSSE correction was introduced
Ž . Ž .
in the determination of the energy of Pu II complexes by computing the energy of Pu II and of the ligand in the presence of the basis set of the whole complex and it reduces the stability by at most 3%.
Table 1
Ž . Ž . Ž .
Atomic energies of Pu II , Pu IV and U II for different input S values. We report the energy difference with respect to the closed-shellz
Ž Ž . Ž . Ž ..
singlet energy y70.1265 for Pu II ,y67.9834 for Pu IV andy50.0823 for U II . All energy differences are in hartrees
Ž . Ž . Ž .
Sz Pu II Pu IV U II
0 0.0000 0.0000 0.0000
1 y0.0614 y0.0671 y0.0590
2 y0.1649 y0.1959 y0.1581
3 y0.3359 q0.1889 q0.4081
4 q0.1788
Table 2
Ž . Ž . Ž .
Pu II , Pu IV and Pu VI complexes:DHTO Tis the binding energy, i.e. the difference between the energy of the complex minus the energy
Ž 2q.
of Pu or PuO2 and the energy of the ligand molecules.DHPu – Ois the energy per Pu–O bond, namelyDHTOTdivided by the number of Pu–O bonds. For Pu II complexesŽ . DHTO T has also been computed including the BSSE correction. Next toDHPu – O, in parentheses, the number of Pu–O bonds present in the molecule is reported
System DHTO T DHTOT BSSE DHPu – O
Žhartree. Žhartree. Žkcal moly1.
2q y
Ž . Ž .
Pu L1 2 y0.5021 y0.4756 y78.77 4
2q y
Ž . Ž .
Pu L2 2 y0.4572 y0.4343 y71.72 4
2q y
Ž . Ž .
Pu L3 2 y0.4967 y0.4619 y77.92 4
2q y
Ž . Ž .
Pu L4 2 y0.4915 y0.4651 y77.11 4
2q y
Ž . Ž .
Pu TTA 2 y0.4586 y0.4383 y71.66 4
4q y
Ž . Ž .
Pu L2 4 y2.1113 y165.51 8
2q y
Ž . Ž .
PuO2 L1 2 y0.6398 y100.37 4
2q y
Ž . Ž .
PuO2 L2 2 y0.6615 y105.08 4
Ž .
The plutonyl complexes, corresponding to Pu VI , are calculated to be more stable than the corresponding
Ž . y y Ž .
Pu II ones. The ordering of the stability of L1 and L2 complexes is reversed compared to the Pu II case.
Ž . Ž .
The Pu IV complex is the most stable. In the square anti-prismatic Pu IV compound, the Pu is bound to eight
Ž . 2q
oxygen atoms, while in the Pu IIrPuO2 compounds only four Pu–O bonds occur. For this reason we report the complexation energy per Pu–O bond, obtained from the total complexation energy,DHTO T, divided by the number of Pu–O bonds between PurPuO22q and the ligand. In comparing the complexes of Pu in different
y Ž . Ž y1.
ionization states with L2 , one notices that the Pu–O bond in the Pu VI complex y105.08 kcal mol is
Ž . Ž y1. Ž . Ž y1.
;50% stronger than in the Pu II one y71.72 kcal mol , and in the Pu IV complex y165.51 kcal mol Ž .
it is ;60% stronger than in the Pu VI complex.
Ž . Ž . Ž . Ž .
Regarding uranium see Table 3 , we performed the calculations on U L2 , U TTA2 2 and UO L2 . One2 2 notices that the U compounds are more stable than the corresponding Pu compounds. The extra stabilization in
Ž . Ž .
going from Pu II to Pu VI complexes is also present in the U series. We explored the effect of different
Ž .
functionals in the U TTA2 calculations and recomputed the dissociation energy at LDA and BLYP level. The LDA result demonstrates the well-known over binding characteristic of this functional.
Ž . Ž .
We fully optimized the geometry of Pu L12 and PuO L12 2 without imposing any symmetry constraints.
The two structures are rather flexible and the potential energy surfaces are very flat. It is not unlikely that we have not converged to global minima. However, considering the flatness of the surfaces, we believe that these optimizations give a preliminary idea of the shape of the molecules. More accurate calculations, including force constants and vibrational energies, will be presented in a forthcoming publication. Table 4 reports some relevant bond distances and angles and the dihedral angle formed by the four oxygen atoms, which indicates the
Table 3
Ž . Ž . Ž 2q.
U II and U VI complexes:DHTO Tis the difference between the energy of the complex minus the energy of U or UO2 and twice the
Ž .Ž y.
energy of the ligand.DHU – O is the energy per U–O bond, namelyDHTOT divided by the number of U–O bonds. For U II TTA 2
several functionals have been used. Next toDHU – O, in parentheses, the number of U–O bonds present in the molecule is reported
System DHTO T DHU – O
y1
Žhartree. Žkcal mol .
2q y
Ž . Ž .
U L2 2 y0.6666 y104.57 4
2q y
Ž . Ž .
UO2 L2 2 y0.9421 y147.70 4
2q y
Ž . Ž .
U TTA 2LDAX y0.6754 y105.95 4
2q y
Ž . Ž .
U TTA 2LDA y0.7371 y115.63 4
2q y
Ž . Ž .
U TTA 2BLYP y0.6202 y97.30 4
( ) L. Gagliardi et al.rChemical Physics 252 2000 47–55 54
Table 4
˚
Ž . Ž . Ž . Ž .
Typical bond distances A and angles degrees for Pu L12 and PuO L1 . R2 2 Pu – O eq is the distance between Pu and the oxygen on the ligand, RPu – O axis the distance Pu–O distance in the plutonyl, RC – Oand RC – Care the C–O and C–C distances in the ligand, respectively.
aOP uO is the angle between Pu and two oxygens on the same ligand,bOPuOis the angle formed by Pu and two oxygens, each on a different ligand anddO is the dihedral angle formed by the four equatorial oxygens
System RPu – O eq RPu – O ax RC – O RC – C /aOPuO /bOPuO /dO
Ž .
Pu L12 2.506--2.512 – 1.323--1.324 1.423--1.425 73.01 96.88 12.32
Ž .
PuO L12 2 2.486--2.492 1.998 1.316--1.318 1.422--1.426 72.59 106.10 5.20
planarity of the structure. Because of the absence of symmetry constraints, the four Pu–O bonds are slightly different and what we report is merely the range of values that it assumes. The same has been done for the C–O
Ž . Ž .
and C–C bonds. The Pu–Oeqbond is slightly longer in Pu L1 2than in PuO L1 . The angle formed by Pu and2 2 the two oxygens on the same ligands is;738in both the two complexes, while the angle formed by Pu and two
Ž . Ž .
oxygens, each on a different ligand, is ;978in Pu L12 and ;1068in PuO L1 . The dihedral angle formed2 2
Ž . Ž .
by the four oxygen atoms is ;128in Pu L1 , while it is2 ;58in PuO L1 . The presence of the plutonyl2 2 Ž .
oxygens in PuO L12 2 forces the equatorial oxygens to lie on a plane and the minimum becomes more well defined.
w x
In Ref. 10 ten Pu–O distances to the carbonate ligands span a relatively narrow range of values,
˚ ˚ ˚
Ž . Ž . Ž .
2.381 6 –2.430 6 A, with an average value of 2.415 7 A. These Pu–O bonds are ;0.1 A shorter than those we have obtained. This might be due to the fact that we are considering a single molecule in the gas phase,
w x
while Ref. 10 refers to experiments in the solid and liquid phase, where the molecules may be more compact.
The results of the Mulliken analysis are reported in Table 5. The effective charge on Pu, q , is ca.Pu q1.1 in
Ž . Ž . Ž .
Pu L1 2 and Pu L2 , and ca.2 q0.9 in PuO L2 . The sum of the effective charges on the four oxygen atoms2 2
Ž . Ž . Ž .
on the ligand, q , is ca.O y1.30 in Pu L1 2 and Pu L2 . In PuO L22 2 2 the four oxygens on the ligand bear a negative charge of ca.y0.9, and the two oxygens of the plutonyl bear a charge of ca.y0.7. This shows that the oxidation states of Pu and the ligands are just formal, since the overall bonding is much more covalent.
Ž . Ž . Ž .
Pu L1 2 and Pu L22 have a more ionic character than PuO L2 , because of the strong plutonyl Pu–O bond2 2 Ž . present in the last complex. The total amount of d and f electrons on Pu has also been estimated. Pu L12 and
Ž . Ž .
Pu L2 2 have ;0.7 d and ;5.3 f electrons. PuO L22 2 has ;1.1 d and ;4.6 f electrons, in line with what one should expect considering the formal oxidation states of Pu in the three complexes.
4. Conclusions
The factors which determine the suitability of a ligand for the extraction of a particular actinide are many and varied. This study has used quantum chemistry to examine the properties of a range of plutonium complexes,
Table 5
Ž . Ž . Ž .
Mulliken population analysis for Pu L1 , Pu L22 2and PuO2 L2 . Effective charge on plutonium atom, q , sum of the effective charges2 Pu
Ž Ž . Ž . Ž . .
on the oxygen atoms bound to plutonium, q , for PuO L2O 2 2the contributions of the equatorial eq and axial ax oxygens are separated , total amount of d and f electrons on plutonium
System qPu qO d f
Ž .
Pu L12 q1.11 y1.30 0.67 5.34
Ž .
Pu L22 q1.14 y1.27 0.69 5.32
Ž . Ž .
PuO L22 2 q0.88 y0.91 eq 1.13 4.57
Ž . y0.74 ax
with the view to find a connection between these and the previous factors. To simplify the problem, we chose a standard ligand, TTA, and varied its structure. We also examined the effect of changing the actinide element.
Ž . Ž .
Varying the oxidation state of the plutonium atom, we find that Pu IV is more stable than Pu VI , which is
Ž . Ž . Ž .
in turn more stable than Pu II . This is in line with experimental observations of Pu IV and Pu VI . The effect of varying the ligand structure produces a much smaller change in the binding energy. Comparison with the corresponding uranium complexes shows analogous trends, although the uranium complexes are more stable.
We are aware of the fact that the systems investigated in the gas phase are model compounds and the inclusion of environmental effects, in particular the introduction of a solvent, is currently under investigation.
Acknowledgements
L.G. thanks MURST, and the European Union for support within the ‘‘Training Mobility Research’’, contract No. ERBFMRXCT96r0088.MAGICis jointly owned by BNFL and the University of Cambridge. L.G.
thanks Professor B.O. Roos for valuable discussions.
References
w x1 J.J. Katz, G.T. Seaborg, L.R. Morss Eds. , The Chemistry of the Actinide Elements, Chapman and Hall, London, 1986.Ž . w x2 A.J. Freeman, C. Keller Eds. , Handbook of the Physics and Chemistry of the Actinides, Elsevier, New York, 1986.Ž . w x3 N.C. O’Boyle, G.P. Nicholson, T.J. Piper, D.M. Taylor, D.R. Williams, G. Williams, Appl. Radiat. Isot. 48 2Ž . Ž1997 183.. w x4 C.W. Mays, G.N. Taylor, R.D. Lloyd, M.W. Wrenn, Actinides in Man and Animals, RD Press, Salt Lake City, UT, 1981.
w x5 K.N. Raymond, W.L. Smith, Struct. Bonding 43 1981 159.Ž . w x6 J.C. Cuninghame, G.L. Miles, J. Inorg. Nucl. Chem. 3 1956 54.Ž .
w x7 Y. Marcus, A.S. Kertes Eds. , Ion Exchange and Solvent Extraction of Metal Complexes, Wiley-Interscience, London, 1969.Ž . w x8 D.L. Clark, S.D. Conradson, M.P. Neu, P.D. Palmer, W. Runde, C.D. Tait, J. Am. Chem. Soc. 119 1997 5259.Ž .
w x9 D.L. Clark, D.W. Keogh, P.D. Palmer, B.L. Scott, C.D. Tait, Angew. Chem., Int. Ed. Engl. 37 1998 164.Ž . w10 D.L. Clark, S.D. Conradson, D.W. Keogh, P.D. Palmer, B.L. Scott, C.D. Tait, Inorg. Chem. 37 1998 2893.x Ž . w11 A.L. Ankudinov, S.D. Conradson, J.M. de Leon, J.J. Rehr, Phys. Rev. B 57 1998 7518.x Ž .
w12 J. Xu, N. Raymond, Inorg. Chem. 38 1999 308.x Ž . w13 P. Pyykko, Chem. Rev. 88 1988 563.x ¨ Ž .
w14 P. Pyykko, J. Li, N. Runeberg, J. Phys. Chem. 98 1994 4809.x ¨ Ž .
w15 L. Gagliardi, N. Handy, A. Ioannou, C.-K. Skylaris, S. Spencer, A. Willetts, A. Simper, Chem. Phys. Lett. 283 1998 187.x Ž . w16 C.-K. Skylaris, L. Gagliardi, N. Handy, A. Ioannou, S. Spencer, A. Willetts, A. Simper, Chem. Phys. Lett. 296 1998 445.x Ž . w17 L. Gagliardi, A. Willetts, C.-K. Skylaris, N. Handy, S. Spencer, A. Ioannou, A. Simper, J. Am. Chem. Soc. 120 1998 11727.x Ž . w18 A. Willetts, L. Gagliardi, A.G. Ioannou, C.-K. Skylaris, A.M. Simper, S. Spencer, N.C. Handy, Int. Rev. Phys. Chem. in press .x Ž . w19 C.-K. Skylaris, L. Gagliardi, N. Handy, A. Ioannou, S. Spencer, A. Willetts, A. Simper, J. Mol. Struct. Theochemx Ž . Žsubmitted .. w20 S. Spencer, L. Gagliardi, N. Handy, A. Ioannou, C.K. Skylaris, A. Willetts, A. Simper, J. Phys. Chem. A 103 1999 1831.x Ž . w21 P.A.M. Dirac, Proc. Cambridge Philos. Soc. 26 1930 376.x Ž .
w22 S.H. Vosko, L. Wilk, M. Nusair, Can. J. Phys. 58 1980 1200.x Ž . w23 A.D. Becke, J. Chem. Phys. 88 1988 1547.x Ž .
w24 C. Lee, W. Young, R.G. Parr, Phys. Rev. B 37 1988 1988.x Ž . w25 M.A. Mura, P. Knowles, J. Chem. Phys. 104 1996 9848.x Ž .
w26 V.I. Lebedev, in: S. Sobolev Ed. , Proceedings of the Conference on Differential Equations in Numerical Mathematics, Nauka,x Ž . Novosibirsk, 1980.
w27 J. Rys, M. Dupuis, H.F. King, J. Comput. Chem. 4 2x Ž . Ž1983 154.. w28 B.I. Dunlap, J.W.D. Connolly, J.R. Sabin, J. Chem. Phys. 71 1979 3396.x Ž .
¨
w29 K. Eichkorn, O. Treutler, H. Ohm, M. Haser, R. Ahlrichs, Chem. Phys. Lett. 240 1995 283.x ¨ Ž . w30 L.R. Kahn, P.J. Hay, R.D. Cowan, J. Chem. Phys. 68 1978 2368.x Ž .
w31 P.J. Hay, W.R. Wadt, J. Chem. Phys. 82 1985 270.x Ž . w32 P.J. Hay, R.L. Martin, J. Chem. Phys. 109 1998 3875.x Ž . w33 T.H. Dunning, J. Chem. Phys. 53 1970 2823.x Ž .
w34 A. Schafer, H. Horn, R. Ahlrichs, J. Chem. Phys. 97 1992 2571.x ¨ Ž . w35 S.F. Boys, F. Bernardi, Mol. Phys. 19 1970 553.x Ž .
w36 A.K. Rappe, C.J. Casewit, K.S. Colwell, W.A. Goddard III, W.M. Skiff, J. Am. Chem. Soc. 114 1992 10024.x Ž .
