Electron transfer in mixed-valent transition-metal complexes, clusters and materials is ubiquitous in both natural and synthetic systems. The degree to which intervalence charge transfer (IVCT) occurs, dependent on the degree of delocalization, places these within class II or III of the Robin–Day system. In contrast to the d-block, compounds of f-block elements typically exhibit class I behaviour (no IVCT) because of localization of the valence electrons and poor spatial overlap between metal and ligand orbitals. Here, we report experimental and computational evidence for delocalization of 5f electrons in the mixed-valent PuIII/PuIV solid-state compound, Pu3(DPA)5(H2O)2 (DPA = 2,6-pyridinedicarboxylate). The properties of this compound are benchmarked by the pure PuIII and PuIV dipicolinate complexes, [PuIII(DPA)(H2O)4]Br and PuIV(DPA)2(H2O)3·3H2O, as well as by a second mixed-valent compound, PuIII[PuIV(DPA)3H0.5]2, that falls into class I instead. Metal-to-ligand charge transfer is involved in both the formation of Pu3(DPA)5(H2O)2 and in the IVCT.


The challenge of achieving charge transfer in mixed-valent, f-block materials is ascribed to both poor spatial overlap between f and ligand orbitals and the localization of f-electrons1,2. In lanthanides, the 4f orbitals are buried within the [Xe] core and 4f electrons only interact weakly with the ligand field3,4,5,6. Examples of intervalence charge transfer (IVCT) in 4f systems are largely restricted to those containing CeIII/CeIV, EuII/EuIII, and YbII/YbIII because these are the only lanthanides with readily accessible redox couples7,8,9,10. In contrast, in the actinides, there is greater radial extension of the 5f orbitals as well as nearly degenerate 6d orbitals, which are firmly established as being involved in bonding in a variety of systems11,12. Here, we report on the synthesis, structure and a variety of physical properties that support delocalization of 5f electrons in the mixed-valent PuIII/PuIV framework material, Pu3(DPA)5(H2O)2 (DPA = 2,6-pyridinedicarboxylate, referred to as dipicolinic acid). This compound, while possessing distinct PuIII and PuIV sites, shows evidence of emergent class II Robin–Day behaviour13.

Results and discussion


The addition of DPA to a solution containing [Pu(H2O)9]3+ results in an immediate colour change from the purple of hydrated PuIII to emerald green. This is suggestive of oxidation of PuIII to PuIV, but in this case it is the result of a broad metal-to-ligand charge transfer (MLCT) band that encompasses the short visible wavelengths in [PuIII(DPA)(H2O)n]+ as demonstrated by DFT calculations and cyclic voltammetry measurements (see Supplementary Fig. 1). Similar colours have been reported for other PuIII complexes with soft donor ligands14. Heating this mixture under anaerobic conditions results in the formation of brown crystals of the mixed-valent PuIII/PuIV compound Pu3(DPA)5(H2O)2 (PuIII,IV-3) as the major product, with trace amounts of the starting PuIII DPA complex also crystallizing in the form of emerald green [PuIII(DPA)(H2O)4]Br (PuIII-1). An analogous reaction under aerobic conditions yields red prisms of PuIV(DPA)2(H2O)3·3H2O (PuIV-2) owing to complete oxidation of the plutonium. Elimination of bromide (a known reductant of PuIV) from the reaction used to prepare PuIII,IV-3, leads to the formation of a second mixed-valent PuIII/PuIV compound, PuIII[PuIV(DPA)3H0.5]2 (PuIV,III-4). In PuIV,III-4 the PuIII:PuIV ratio is reversed from that in PuIII,IV-3 and it becomes illustrative to re-express the formula of PuIII,IV-3 as [PuIII(DPA)(H2O)]2[PuIV(DPA)3]. In PuIV,III-4 there are two PuIV sites for each PuIII site and the compound should be thought of as further oxidation in the sequence: [PuIII(DPA)(H2O)4]Br → [PuIII(DPA)(H2O)]2[PuIV(DPA)3] → PuIII[PuIV(DPA)3H0.5]2 → PuIV(DPA)2(H2O)3·3H2O; that is, (PuIII-1) → (PuIII,IV-3) → (PuIV,III-4) → (PuIV-2).

Both PuIII,IV-3 and PuIV,III-4 represent intermediates between PuIII-1 and PuIV-2. The MLCT band that is observed when PuIII is complexed by DPA is indicative of changes in the relative thermodynamic stability of PuIII versus PuIV. While it may seem counterintuitive, the expectation is that an electron-accepting ligand should stabilize low-oxidation states. In fact, cyclic voltammetry measurements reveal stabilization of PuIII upon complexation by DPA. Electrochemical interrogation of PuIII in the solvent mixture described above yields a formal potential of E½ = 0.797 V (versus Ag/AgCl) or E½ = 1.021 (versus normal hydrogen electrode (NHE)), which agrees closely with established values of 1.015 V for the standard redox couple (see Supplementary Fig. 1)15. In the [Pu(DPA)3]n (n = 2 or 3) complex, however, the formal potential is E½ = 0.888 V (or E½ = 1.113 V versus NHE), which represents an overall positive shift of the potential by 91.5 mV. Therefore, the MLCT can be thought of as stabilizing PuIII by removing electron density from the metal centre and placing it on the ligand. This is consistent with the previous reports that trivalent actinides (An = Am, Cm, Bk, Cf) form [AnIII(HDPA)3] complexes in the presence of excess DPA (refs 16,​17,​18). Mild heating drives the oxidation of this complex to [PuIV(DPA)3]2− and creates the anionic cores found in PuIII,IV-3 and PuIV,III-4. When [PuIV(DPA)3]2− is combined with two [PuIII(DPA)]+ cationic units, charge neutrality is achieved and allows for the crystallization of PuIII,IV-3 from a polar solvent mixture. Likewise, further oxidation occurs in the absence of bromide and PuIV,III-4 is isolated instead. Finally, oxidation of all of the PuIII in the reaction to PuIV leads to the formation of charge-neutral PuIV-2.

Structure elucidation

The architecture of PuIII,IV-3 is best understood by first examining the moieties found in PuIII-1 and PuIV-2 because the fundamental building units in these latter compounds are linked together to create the structure of PuIII,IV-3. The structure of PuIII-1 is assembled from [PuIII(DPA)(H2O)4]+ moieties as shown in Fig. 1a. This unit consists of a PuIII cation bound by a tridentate DPA ligand that provides one N and two O donor atoms to the plutonium centre. Two oxygen atoms from the carboxylate moieties of the DPA ligand are directed away from the PuIII cation as shown in Fig. 1a. These oxygen atoms bridge to crystallographically equivalent plutonium centres forming a 3D framework structure. The network creates channels along {111} where the extra-framework bromide anions reside (Supplementary Fig. 2). The geometric constraints of the coordination environment allow bridging by two DPA ligands between the PuIII centres. These two oxygen atoms along with the N-donor from the pyridine moiety create the three capping positions of a tricapped trigonal prismatic [PuNO8] unit. The remaining four prismatic sites are occupied by water molecules. Owing to the presence of three different types of Pu−O bonds, there is a wide range of Pu−O bond lengths (2.434(3) to 2.551(5) Å, see Supplementary Tables 1 and 2). The Pu−N bond length is 2.584(4) Å (ref. 11).

Figure 1: Graphical representation of the structures of PuIII-1 and PuIV-2.
Figure 1

a, Partial coordination sphere of PuIII when chelated by 2,6-pyridinedicarboxylate (DPA) in PuIII-1. b, A view of PuIV-2 formed by the chelation of PuIV by two DPA ligands and three water molecules. In both cases a nine-coordination tricapped trigonal prismatic geometry is observed (highlighted in purple for PuIII and red for PuIV). Colour code for atoms: red, O; blue, N; black, C; white, H.

PuIV-2 forms a molecular structure consisting of a nine-coordinate PuIV centre bound by two, tridentate, DPA ligands and three water molecules yielding a tricapped trigonal prismatic geometry as shown in Fig. 1b and Supplementary Fig. 3. The bond distances between the PuIV centre and both the DPA ligands and water molecules are on average significantly shorter than those found in PuIII-1 as expected based on the higher oxidation state of plutonium and smaller ionic radius. The Pu−O bond lengths range from 2.321(3) to 2.495(4) Å and the two Pu−N bonds are 2.495(4) and 2.502(4) Å (see Supplementary Table 2). The expected bond length differences between PuIII and PuIV should be approximately 0.05 to 0.1 Å, and this is indeed observed19.

The two previously described structures significantly aid both rationalization and understanding of the structure of PuIII,IV-3. There are two crystallographically unique plutonium sites in this compound that possess noticeably different coordination environments and bond distances. The first of these sites is a nine-coordinate, tricapped trigonal prismatic PuIV site that has three, tridentate DPA ligands (that is, the tris-chelate complex of PuIV, see Fig. 2a). The bond distances are similar to those found in PuIV-2 (see Supplementary Table 2), consistent with this unit being [PuIV(DPA)3]2−. The oxygen atoms from carboxylate groups occupy the prismatic sites and the nitrogen atoms of the pyridine groups fill the three capping positions. The second site contains plutonium in an eight-coordinate, distorted bicapped trigonal prism with a single tridentate DPA ligand bound to the plutonium centre. The bond distances are consistent with PuIII and match up well with those found in PuIII-1, leading to the assignment of this plutonium centre as [PuIII(DPA)]+. The carboxylate anions from each [PuIV(DPA)3]2− unit bridge to four [PuIII(DPA)]+ moieties, part of which is shown in Fig. 2b. The PuIV centre resides on a twofold site, and therefore there are two [PuIII(DPA)]+ units for each [PuIV(DPA)3]2− core, yielding a neutral 3D network. The PuIII−O bond lengths range from 2.417(2) to 2.483(2) Å and the PuIII−N bond is 2.567(2) Å. The PuIV−O bond lengths range from 2.331(4) to 2.367(2) Å, and the PuIV−N bond lengths are 2.5076(17) (×2) and 2.515(3) Å (see Supplementary Table 2). A comparison of the Pu−N bond lengths of these three compounds reveals that the PuIV−N bonds are approximately 0.05 Å shorter than those found with PuIII.

Figure 2: Depiction of the structure of PuIII,IV-3.
Figure 2

a, A view of the tris-chelate complex of PuIV with three DPA ligands in PuIII,IV-3. b, A view of the bridging by the carboxylate moieties between PuIII and PuIV sites. Same colour code as for Fig. 1.

The structure of PuIV,III-4 is in many ways the reverse of that found in PuIII,IV-3. In PuIV,III-4, a PuIII site is bound only by four water molecules and four bridging carboxylates from DPA ligands that chelate neighbouring PuIV centres. The PuIII ion is not bound by a chelating DPA ligand as occurs in all of the other compounds and this has important electronic consequences (vide infra). The coordination environment around this PuIII ion is best described as a slightly distorted PuO8 square antiprism. The PuIII site is located on an inversion centre yielding crystallographically equivalent [PuIV(DPA)3] units that surround and bind the PuIII cation via the carboxylate oxygen atoms that are directed away from the PuIV centres as shown in Fig. 3a. This repeating sequence occurs along the b axis and thus also along a 21 screw creating a 1D chain as depicted in Fig. 3b.

Figure 3: Illustration of the structure of PuIV,III-4.
Figure 3

a, A depiction of the part of the substructure of PuIV,III-4 showing a PuIII cation bound in a square antiprismatic environment by four water molecules (hydrogens omitted for clarity) and four carboxylate oxygen atoms from two DPA ligands. b, A view of the one-dimensional chain formed from the linking of [PuIV(DPA)3] units (highlighted in red) by PuIII cations (highlighted in purple). Disordered enthanol molecules that fill voids in the structure have been omitted for clarity. Same colour code as for Fig. 1.

Pu‒O and Pu‒N bond distances for PuIV,III-4 are listed in the Supplementary Table 2, and conform to those already discussed with one key deviation. There are statistically significant differences of 0.014(4) and 0.013(5) Å between the PuIV−N bond lengths found in PuIV-2 and PuIV,III-4, respectively, versus those found in PuIII,IV-3. In short, the PuIV−N distances in PuIII,IV-3 are slightly longer than those measured from the other PuIV-containing compounds in the family. While these change are small, they corroborate computational results that show that charge transfer between the PuIII centre in PuIII,IV-3 and the DPA ligand is occurring via the overlap of a lobe of a 5f orbital and with a lobe of the 2p orbital of the pyridyl N-atom as shown in the singly occupied molecular orbitals (SOMOs) provided in Fig. 4a. The PuIII‒N distances are also statistically shorter in PuIII,IV-4 than in PuIII-1. The PuIII‒N bond lengths are showing partial 4+ character and conversely the PuIV‒N bonds are showing 3+ contributions. For comparison, there is no discernable difference between the average PuIV−N bond distances in PuIV-2 and PuIV,III-4 with Δ = 0.001(5) Å.

Figure 4: Rendering of the SOMOs and LUMOs in PuIII-1 and PuIV-2.
Figure 4

ad, Highest SOMO (a) and LUMO (b) orbitals in the [PuIII(DPA)(H2O)4]+ moieties of PuIII-1, and the highest SOMO (c) and LUMO (d) in PuIV-2.

Therefore, the structural evidence suggests that PuIV,III-4 falls into class I of the Robin–Day system with distinct sites for each oxidation state and no electronic communication between the PuIII and PuIV sites13. In contrast, these structural data point to PuIII,IV-3 being an incipient class II compound where IVCT is occurring. The large differences in coordination between PuIII and PuIV in PuIII,IV-3 would lead to the expectation that it must lie close to the class I/II boundary. Based on the solution chemistry of plutonium, which is well known to be unique in that four oxidation states can be simultaneously equilibrated, it might be expected that mixed-valent plutonium compounds are also common in the solid state. However, if one excludes cases of slight nonstoichiometry from this discussion (such as PuO2–x), these compounds actually prove to be exceedingly rare. In fact, we could only find two well-characterized examples20,21. Both are molecular structures that contain individual plutonium complexes with different oxidation states where there are large separations between the plutonium sites. For example, Pu2Cl7(thf)6 is not a traditional class I compound because while it does possess two distinct sites with two different oxidation states (PuIII and PuIV), it is a salt of co-crystallized [PuIIICl2(thf)5]+ and [PuIVCl5(thf)] (ref. 20). The second example of this type of system is also a salt that contains two different plutonium anions, in this case the [PuIVCl6]2‒ and [PuVIO2Cl4]2‒ anions21. These compounds do not have ligands bridging between the plutonium centres and therefore are not expected to exhibit class II behaviour.

Electronic structure and theory

The structural indications that PuIII,IV-3 falls within class II are bolstered by UV–vis–NIR absorption spectroscopy and DFT level (B3PW91) calculations (see Supplementary Figs 4–6)22,23,24. The typical Laporte-forbidden ff transitions of PuIII and PuIV measured from single crystals of PuIII-1 and PuIV-2, respectively, are shown in Fig. 5a. PuIII has a 6H5/2 ground state, and while many transitions partially overlap with those found for PuIV, the excitation to the 6H13/2 state at ~900 nm, and the 4L13/2 and 4M15/2 states near 600 nm are clear indications of PuIII (ref. 25). These characteristic ff transitions are present in PuIII-1. Characteristic transitions for PuIV, including the 5F2 and 5I6 transitions around 1,100 nm and the transitions near 700 nm from multiple J states, are exhibited by PuIV-2 (ref. 26). The spectrum of PuIV-2 is quite normal, as is the red colouration of the crystals. The spectrum of PuIII-1, however, shows a broad and intense CT band that extends through the UV and into the visible region of the spectrum. The fact that PuIV bound by the same ligand does not show this feature argues that this band is MLCT in origin where PuIII is transferring electron density to DPA. This supposition is supported by the calculations that reveal in PuIII-1 that the SOMO and lowest unoccupied orbital (LUMO) show significant mixing between the metal and ligand orbitals; whereas in PuIV-2, the SOMO is primarily ligand based, and the LUMO is primarily metal based as shown in Fig. 4. Thus, the intuitive understanding of PuIII-1 and PuIV-2 is validated by these calculations and provides a clear signature of the two different oxidation states of plutonium.

Figure 5: Absorption spectra of PuIII-1, PuIV-2, PuIII,IV-3, and PuIV,III-4.
Figure 5

a, Overlay of the absorption spectra of PuIII-1, PuIV-2, PuIII,IV-3, and PuIV,III-4. b, Overlay of the summed absorption spectra of PuIII-1 and PuIV-2 with PuIII,IV-3 demonstrating that the absorption spectrum of PuIII,IV-3 is not simply the sum of its PuIII and PuIV moieties.

The brown colouration of crystals of PuIII,IV-3 and PuIV,III-4 is indicative of more complex electronic structure as shown in Supplementary Fig. 7. An overlay of the absorption spectra of PuIII-1, PuIV-2, PuIII,IV-3 and PuIV,III-4 shows that the spectrum of PuIII,IV-3 is not simply a sum of its parts as shown in Fig. 5b. Summed spectra of PuIII-1, PuIV-2 and PuIV,III-4 also do not re-create the spectrum of PuIII,IV-3 as provided in Fig. 5b. The CT band in PuIII,IV-3 is significantly broader and extends deeper into visible wavelengths by at least ~200 nm than found in the other compounds. In addition, a vibronic progression is resolved that is absent in the other spectra. This progression has a frequency of 1,657 cm‒1 that corresponds to the C=O stretching mode of a carboxylate moiety. The carboxylate groups bridge between the PuIII and PuIV centres, suggesting that the IVCT band is within this broad transition, and that the charge transfer, facilitated by the MLCT in [PuIII(DPA)] and LMCT in [PuIV(DPA)3] is through the DPA ligands. Furthermore, even though PuIV,III-4 is also a mixed-valent compound, it lacks the PuIII‒N bonds found within [PuIII(DPA)] units that are required for MLCT. PuIV,III-4 should be a class I compound based on this alone and all of the experimental evidence supports this claim. In contrast, the electronic spectroscopy complements the structural and computational findings for PuIII,IV-3, and helps to place it early in the spectrum of class II compounds. The location of the MLCT, LMCT and IVCT bands at substantially higher energies than found for most transition-metal compounds is consistent with this model because it should require more energy to delocalize 5f electrons than d electrons in general27,28. The charge transfer in PuIII,IV-3 is impressive nevertheless because the distance between the PuIII and PuIV sites is 6.234(1) Å, and at a minimum the electrons have to travel through four bonds (see Supplementary Fig. 8)29,30.

Computational analysis of the electronic structure of PuIII,IV-3 shows that the highest SOMO is located on the central plutonium atom, whereas the LUMO is located on an adjacent plutonium site as shown in Fig. 6a,b31,32,33,34,35,36,37,38,39,40,41,42. Further analysis shows that the interior site contains tetravalent plutonium as found in the parent compound, PuIV-2. Likewise, the peripheral plutonium site parallels the electronic characteristics of the PuIII site in PuIII-1. The LUMO is constructed from orbitals with charge-transfer characteristics and is the former SOMO of PuIII-1. Thus, the dianionic form of the central PuIV moiety is created from the two external PuIII units as shown in Fig. 6c. Charge transfer is also evident in the SOMO of PuIII,IV-3 where the former SOMO of the PuIII moiety overlaps with the LUMO of the PuIV unit as shown in Fig. 6d. Therefore, PuIII,IV-3 is not only mixed-valent, but clearly falls into class II.

Figure 6: Representation of the frontier orbitals of PuIII,IV-3.
Figure 6

a,b, Highest SOMO (a) and LUMO (b) orbitals of the trimeric complex. These two orbitals are formed with the LUMO of monomeric PuIV and the SOMO of the monomeric PuIII, respectively, circled in blue. c, Molecular orbitals involved in charge transfer from PuIII to the DPA ligand. d, Molecular orbitals involved in the coordination to PuIV.

The oxidation states of the different plutonium centres were also confirmed by performing f-in-core calculations with fixed oxidation states36. The optimized geometry obtained with the latter calculation is in excellent agreement with the one obtained with small-core calculations, ensuring the mixed-valent nature of the compound. Additionally, the UV–vis–NIR spectra of PuIII-1, PuIV-2 and PuIII,IV-3 were computed (see Supplementary Figs 4–6). In line with the experimental observations, the spectrum of PuIII,IV-3 is not simply a superposition of the spectra of the parent complexes, but rather an additional, large MLCT band is found between 600 to 800 nm. This band involves transitions from the SOMO and the LUMO. All of these computational findings support the class II nature of PuIII,IV-3.

In conclusion, these data support the assignment of PuIII,IV-3 as a class II material. The designation of this compound as being closer to the border with class I is invoked primarily from data gathered from d-block complexes where large differences in coordination environments between the metal centres inhibits IVCT. Regardless of where PuIII,IV-3 lies in the class II spectrum, its existence directs us to target specific coordination polymers and simple multinuclear complexes of redox-active, transuranium elements that should also exhibit IVCT.



239Pu (t1/2 = 24,065 yr) and 240Pu (t1/2 = 6,537 yr) represent serious health risks, owing to their α emission. All studies with plutonium were conducted in a laboratory dedicated to studies on transuranium elements.


2,6-Pyridinedicarboxylic acid (99%, Sigma-Aldrich), ethanol (100%, Koptec), hydrobromic acid (ACS reagent 48%, Sigma-Aldrich), and plutonium (94% 239Pu, 6% 240Pu) obtained from Los Alamos National Laboratory in the form of PuCl3 or PuO2 were used without further purification. PTFE-lined Parr 4749 autoclaves with a 10 ml internal volume, and Millipore water were used in all of the following reactions. All solvents that were used in a glove box were sparged with argon.

[PuIII(DPA)(H2O)4]Br (PuIII-1) and Pu3(DPA)5(H2O)2 (PuIII,IV-3)

PuCl3 (9.4 mg, 0.0268 mmol) and HBr (8 M, 100 μl) were heated in an open PTFE-liner at 130 °C for one hour resulting in a purple-black residue. The liner was transferred and cooled to ambient temperature inside of an argon-filled glovebox in order to exclude oxygen. DPA (4.6 mg, 0.0254 mmol), ethanol (100 μl), and deionized water (100 μl) were added to the liner producing a bright green solution. The autoclave was sealed and placed into an oven inside the glovebox at 180 °C for two days. The same product forms at 150 °C. The autoclave was allowed to cool slowly at a rate of 5 °C per hour. Two products were isolated: a few, large, green block-shaped crystals of [PuIII(DPA)(H2O)4]Br and brown plates of Pu3(DPA)5(H2O)2, see Supplementary Fig. 7.

PuIV(DPA)2(H2O)3·3H2O (PuIV-2)

PuCl3 (9.8 mg, 0.0284 mmol) and HBr (8 M, 100 μl) were heated in an open PTFE-liner at 130 °C for one hour resulting in a purple-black residue. After the liner cooled to room temperature, DPA (5.2 mg, 0.0315 mmol), ethanol (100 μl), and water (100 μl) were added. The reaction was conducted aerobically (with no attempts to keep air out of the reaction) in an oven heated to 180 °C for three days with a slow cooling rate of 5 °C per hour. Red-pink tablet-shaped crystals of 1 were isolated from the reaction in high yield (Supplementary Fig. 7). No other products were present.

PuIII[PuIV(DPA)3H0.5]2 (PuIV,III-4)

PuCl3 (10.0 mg, 0.0290 mmol) and DPA (24.2 mg, 0.145 mmol) were dissolved in 200 μl of a 1:1 mixture of ethanol and water under an argon atmosphere. The resultant reaction mixture turned from dark blue to dark green. The mixture was heated in PTFE-lined Parr 4749 autoclave with a 10 ml internal volume for 4 h at 150 °C, and then slowly cooled to 25 °C over a 12-h period. The reaction resulted in the formation of brown block crystals (Supplementary Fig. 7). The crystals were isolated, brought out of the glovebox, and washed (3 × 1 ml of ethanol).

Crystallographic studies

Single crystal X-ray diffraction data were acquired using a Bruker D8 Quest X-ray diffractometer. Initial intensity measurements were performed using a IμS X-ray source, a 50 W microfocused sealed tube (Mo Kα, λ = 0.71073 Å) with high-brilliance and high-performance focusing multilayer optics. Standard Quest software was used for determination of the unit cells and data collection control. The intensities of reflections of a hemisphere were collected by a combination of four sets of exposures (frames). Each set had a different ϕ angle for the crystal and each exposure covered a range of 0.5° in ω. Quest software was used for data integration including Lorentz and polarization corrections. Supplementary Table 1 provides selected crystallographic information for all three compounds.

UV−vis–NIR spectroscopy

A Craic Technologies microspectrophotometer was used to obtain a room-temperature UV–vis–NIR spectrum for each compound. Low temperature data were acquired using a Linkam cryostat. Crystals were placed on a quartz slide under immersion oil and the data were collected from 200 to 1,200 nm.

Computational details

All the structures reported in this study were fully optimized with the Becke's 3-parameter hybrid functional31 combined with the non-local correlation functional provided by Perdew and Wang32 (denoted as B3PW91). The plutonium atom was represented by relativistic energy-consistent small-core pseudopotential obtained from the Stuttgart–Köln ECP library has been used in combination with its adapted segmented basis set33,34,35,36. Large-core calculations were also used to probe the mixed-valent nature of the coordination polymer33,34,35,36. For the remaining atoms the 6-31G(d,p) basis set was used37,38,39. In all computations no constraints were imposed on the geometry. All stationary points have been identified as minima (number of imaginary frequencies Nimag = 0) or transition states (Nimag = 1). The vibrational modes and the corresponding frequencies are based on a harmonic force field. Enthalpy energies were obtained at T = 298.15 K within the harmonic approximation. Gaussian09 program suite was used in all calculations40. Finally, for the 3D representation of the structures the ChemCraft41 program was used as well as for the visualization of the molecular orbitals. Time-dependent density functional theory calculations were carried out to determine the UV–visible spectrum using the methodology implemented in Gaussian and the lowest 50 states were considered42.

Data availability

Crystallographic data for the structures reported in this paper are deposited at the Cambridge Crystallographic Data Centre (CCDC), under the deposition numbers CCDC 1042947 (PuIII-1), 1042860 (PuIV-2), 1042862 (PuIII,IV-3), and 1533489 (PuIV,III-4). Copies of these data can be obtained free of charge via www.ccdc.cam.ac.uk/data_request/cif. All other data supporting the findings of this study are available within the article and its Supplementary Information files, or from the corresponding author on reasonable request.

Additional information

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.


  1. 1.

    et al. Heterometallic europium disiloxanediolates: synthesis, structural diversity, and photo-luminescence properties. Inorg. Chem. 53, 11662–11674 (2014).

  2. 2.

    et al. Uncovering f-element bonding differences and electronic structure in a series of 1:3 and 1:4 complexes with a diselenophosphinate ligand. Chem. Sci. 4, 1189–1203 (2013).

  3. 3.

    et al. Structural, spectroscopic, and theoretical comparison of traditional vs recently discovered Ln2+ Ions in the [K(2.2.2-cryptand)][(C5H4SiMe3)3Ln] complexes: the variable nature of Dy2+ and Nd2+. J. Am. Chem. Soc. 137, 369–382 (2015).

  4. 4.

    et al. Differentiating between trivalent lanthanides and actinides. J. Am. Chem. Soc. 134, 10682–10692 (2012).

  5. 5.

    , & The oxidation state of cerium in Ce2MoO6. J. Alloys Compd. 207–208, 444–448 (1994).

  6. 6.

    et al. Isolation of intermediate-valent Ce(III)/Ce(IV) hydrolysis products in the preparation of cerium iodates: electronic and structural aspects of Ce2(IO3)6(OHx) (x ≈ 0 and 0.44). Chem. Mater. 16, 1343–1349 (2004).

  7. 7.

    et al. Stabilization of cerium(IV) in the presence of an iodide ligand: remarkable effects of lewis acidity on valence state. J. Am. Chem. Soc. 121, 11255–11256 (1999).

  8. 8.

    A new mixed valent europium chloride: Na5Eu7Cl22. Z. Naturforsch. 57b, 901–907 (2002).

  9. 9.

    KEu2Cl6 und K1,6Eu1,4Cl5: Zwei neue gemischtvalente Europiumchloride Z. Anorg. Allg. Chem. 628, 1815–1820 (2002).

  10. 10.

    , , & Influence of impurities on the luminescence quantum efficiency of (La, Ce, Tb)PO4. J. Electrochem. Soc. 140, 216–221 (1993).

  11. 11.

    et al. Unusual structure, bonding and properties in a californium borate. Nat. Chem. 6, 387–392 (2014).

  12. 12.

    , & Covalency in f-element complexes. Coord. Chem. Rev. 257, 394–406 (2013).

  13. 13.

    & Mixed valence chemistry: a survey and classification. Adv. Inorg. Chem. Radiochem. 10, 248–422 (1967).

  14. 14.

    et al. Experimental and theoretical comparison of actinide and lanthanide bonding in M[N(EPR2)2]3 Complexes (M = U, Pu, La, Ce; E = S, Se, Te; R = Ph, iPr, H). Inorg. Chem. 47, 29–41 (2008).

  15. 15.

    , & Standard potentials of the (MO22+/MO2+) and (M4+/M3+) redox systems for neptunium and plutonium. Radiochim. Acta. 46, 85–94 (1989).

  16. 16.

    & Characterization of europium and americium dipicolinate complexes. Sep. Sci. Tech. 47, 2029–2037 (2012).

  17. 17.

    et al. Emergence of californium as the second transitional element in the actinide series. Nat. Commun. 6, 6827–6834 (2015).

  18. 18.

    et al. Characterization of berkelium(III) dipicolinate and borate compounds in solution and the solid state. Science 353, aaf3762 (2016).

  19. 19.

    et al. Syntheses, structures, and spectroscopic properties of plutonium and americium phosphites and the redetermination of the ionic radii of Pu(III) and Am(III). Inorg. Chem. 51, 8419–8424 (2012).

  20. 20.

    et al. Low-valent molecular plutonium halide complexes. Inorg. Chem. 47, 8412–8419 (2008).

  21. 21.

    , , & Supramolecular interactions in PuO2Cl42– and PuCl62– Complexes with protonated pyridines: synthesis, crystal structures, and raman spectroscopy. Inorg. Chem. 53, 383–392 (2014).

  22. 22.

    , & A DFT study of the reactivity of actinidocenes (U, Np and Pu) with pyridine and pyridine N-oxide derivatives. Dalton Trans. 39, 6682–6692 (2010).

  23. 23.

    , & Are 5f electrons really active in organoactinide reactivity? some insights from DFT studies. ChemPhysChem 11, 990–994 (2010).

  24. 24.

    , & A DFT study of the reactivity of Cp2AnMe2 with pyridine N-oxide: towards a predicted different reactivity of U/Pu and Np. C. R. Chim. 13, 870–875 (2010).

  25. 25.

    A systematic analysis of the spectra of trivalent actinide chlorides in D3h site symmetry. J. Chem. Phys. 96, 8713–8726 (1992).

  26. 26.

    , , & Analysis of the crystal-field spectra of the actinide tetrafluorides. I. Uranium, neptunium, and plutonium tetrafluorides (UF4, NpF4, and PuF4). J. Chem. Phys. 95, 7194–7203 (1991).

  27. 27.

    , & Optical transitions of symmetrical mixed-valence systems in the Class II–III transition regime. Chem. Soc. Rev. 31, 168–184 (2002).

  28. 28.

    , & Exploration of mixed-valence chemistry:  inventing new analogues of the Creutz-Taube ion. Acc. Chem. Res. 33, 755–763 (2000).

  29. 29.

    , & Ferric garnet matrices for immobilization of actinides. J. Nucl. Mater. 436, 1–7 (2013).

  30. 30.

    Interaction between the d-shells in the transition metals. II. Ferromagnetic compounds of manganese with perovskite structure. Phys. Rev. 82, 403–405 (1951).

  31. 31.

    Density-functional thermochemistry. III. The role of exact exchange. J. Chem. Phys. 98, 5648–5652 (1993).

  32. 32.

    & Accurate and simple analytic representation of the electron-gas correlation energy. Phys. Rev. B. 45, 13244–13249 (1992).

  33. 33.

    , , & Energy-adjusted pseudopotentials for the actinides. parameter sets and test calculations for thorium and thorium monoxide. J. Chem. Phys. 100, 7535–7542 (1994).

  34. 34.

    , & Valence basis sets for relativistic energy-consistent small-core actinide pseudopotentials. J. Chem. Phys. 118, 487–496 (2003).

  35. 35.

    & Segmented contraction scheme for small-core actinide pseudopotential basis sets. J. Molec. Struct. Theochem. 673, 203–209 (2004).

  36. 36.

    , & Quasirelativistic energy-consistent 5f-in-core pseudopotentials for divalent and tetravalent actinide elements. Theor. Chem. Acc. 118, 845–854 (2007).

  37. 37.

    , & Self-consistent molecular orbital methods. XII. Further extensions of Gaussian-type basis sets for use in molecular orbital studies of organic molecules. J. Chem. Phys. 56, 2257–2261 (1972).

  38. 38.

    & The influence of polarization functions on molecular orbital hydrogenation energies. Theor. Chim. Acta 28, 213–222 (1973).

  39. 39.

    , & Self-consistent molecular orbital methods. 21. Small split-valence basis sets for first-row elements. J. Am. Chem. Soc. 102, 939–947 (1980).

  40. 40.

    et al. Gaussian 09 v. A.02 (Gaussian, 2009).

  41. 41.

    ChemCraft Version 1.8 (2017);

  42. 42.

    , , & Molecular excitation energies to high-lying bound states from time-dependent density-functional response theory: characterization and correction of the time-dependent local density approximation ionization threshold. J. Chem. Phys. 108, 4439–4449 (1998).

Download references


This material is based upon work supported by the US Department of Energy, Office of Science, Office of Basic Energy Sciences, Heavy Elements Chemistry Program under award number DE-FG02-13ER16414. We are especially grateful for the assistance and supervision by the Office of Environmental Health and Safety at Florida State University, specifically J. A. Johnson and A. L. Gray of the Office of Radiation Safety for their facilitation of these studies. Magnetization measurements using the VSM SQUID MPMS were performed at the National High Magnetic Field Laboratory, which is supported by National Science Foundation Cooperative agreement number DMR-1157490, the State of Florida, and the US Department of Energy. We are grateful for helpful discussions with N. M. Edelstein, M. P. Jensen and G. Liu.

Author information


  1. Department of Chemistry and Biochemistry, Florida State University, 95 Chieftan Way, Tallahassee, Florida 32306, USA

    • Samantha K. Cary
    • , Shane S. Galley
    • , Matthew L. Marsh
    • , David L. Hobart
    • , Justin N. Cross
    • , Jared T. Stritzinger
    • , Matthew J. Polinski
    •  & Thomas E. Albrecht-Schmitt
  2. National High Magnetic Field Laboratory, 1800 East Paul Dirac Drive, Tallahassee, Florida, 32310, USA

    • Ryan E. Baumbach
    •  & Thomas E. Albrecht-Schmitt
  3. Laboratoire de Physique et Chimie des Nano-objets, Institut National des Sciences Appliquées, 31077 Toulouse Cedex 4, France

    • Laurent Maron


  1. Search for Samantha K. Cary in:

  2. Search for Shane S. Galley in:

  3. Search for Matthew L. Marsh in:

  4. Search for David L. Hobart in:

  5. Search for Ryan E. Baumbach in:

  6. Search for Justin N. Cross in:

  7. Search for Jared T. Stritzinger in:

  8. Search for Matthew J. Polinski in:

  9. Search for Laurent Maron in:

  10. Search for Thomas E. Albrecht-Schmitt in:


S.K.C., J.N.C., S.S.G. and T.E.A.-S. conceived, designed, and carried out the synthetic and crystallographic experiments. S.K.C. and J.T.S. carried out low-temperature spectroscopic experiments. S.K.C., S.S.G., J.N.C., and M.J.P. were involved in the crystallographic analysis. Cyclic voltammetry experiments were conducted by M.L.M. and D.L.H.; R.E.B. designed and carried out the magnetism experiments and analysed the data. L.M. carried out the computational analysis. All authors discussed and co-wrote the manuscript.

Competing interests

The authors declare no competing financial interests.

Corresponding authors

Correspondence to Laurent Maron or Thomas E. Albrecht-Schmitt.

Supplementary information

PDF files

  1. 1.

    Supplementary information

    Supplementary information

Crystallographic information files

  1. 1.

    Supplementary information

    Crystallographic data for compound PuIII-1.

  2. 2.

    Supplementary information

    Crystallographic data for compound PuIV-2.

  3. 3.

    Supplementary information

    Crystallographic data for compound PuIII,IV-3.

  4. 4.

    Supplementary information

    Crystallographic data for compound PuIV,III-4.

About this article

Publication history






Further reading