Probing Plutonium Materials with Magnetic Resonance

We use nuclear magnetic resonance (NMR) spectroscopy on solid materials to understand their chemical and physical properties. Of most interest are those containing isotopes with a nuclear spin of ½, such as hydrogen and carbon, due to the detailed information that can be obtained by studying this type of nucleus. The spin-½ nucleus of ²³⁹Pu meanwhile has resisted attempts by researchers to observe its direct signature because of the unusually large internal magnetic field at the nuclear site that arises from the strong interaction between nuclear spins and the orbital current of electrons. This effect causes an extraordinarily fast nuclear relaxation time, making it very difficult to directly observe ²³⁹Pu signals except under certain circumstances. This work has been further constrained because plutonium is a federally controlled substance and facilities which can handle such “hot” materials are limited.

In 2012, a historical observation of the ²³⁹Pu NMR signal in PuO₂ at Los Alamos National Laboratory (LANL) was made, followed by the second observation of this nucleus in putative topological insulators PuB₄/PuB₆ by the same group in 2018. Having determined the ²³⁹Pu isotope’s signature (namely, the nuclear gyromagnetic ratio ²³⁹γ_n) we are now in a favorable position to study the structure and chemical bonding of plutonium compounds by using this NMR technique as a local probe. It can also be used to study both the chemical properties and self-radiation effects in nuclear fuels, nuclear waste, and organometallic molecules, and the physical properties of magnets, superconductors, and other plutonium-based correlated electron materials. This short article provides an overview of  ²³⁹Pu-NMR spectroscopic research at LANL and an outlook for its future applications.

Basic aspects of NMR

A nucleus with a non-zero spin possesses a magnetic moment, like a magnetic compass. Quantum mechanics tells us that in the presence of a magnetic field the nuclear magnetic energy levels split to equally-spaced energy levels by the Zeeman interaction. In an NMR experiment we apply an oscillatory field with a frequency corresponding to the energy difference. The resonance condition is expressed by a linear relation between frequencies and fields with a proportional constant γ_n, the nuclear gyromagnetic ratio, which is the “fingerprint” of a specific nucleus.

In materials, nuclear spins sense an additional “internal” magnetic field arising primarily from an interaction with their electronic environment (hyperfine interaction) and with each other. The static component of the internal field results in a change of the measured resonance frequency—the relative shift of the resonance frequency is called the Knight shift. The fluctuating component of the internal field meanwhile is responsible for the relaxation of nuclear spins. In particular, the nuclear spin-lattice relaxation time T₁ (or the relaxation rate, 1/T_1,) is an important parameter related to the dynamical response of magnetic susceptibility, and its temperature dependence gives us a quantitative measure of the low-energy magnetic excitations.

The first observation of a ²³⁹Pu NMR signal

The search for plutonium NMR signals in liquid or solid phases has focused primarily on spin-½ ²³⁹Pu; nevertheless, this was not observed until 2012. The signal remained elusive for two reasons: first, an extremely strong hyperfine interaction between electron and nuclear spins, due to unpaired 5f electrons, gives rise to a large internal magnetic field (~100 T) at the nuclear site. This internal field shifts the resonance frequency by several orders of magnitude, and the nuclear spinrelaxation time becomes exceedingly fast (less than one microsecond), making any measurements very challenging. Second, there was a range of values in the literature reported for the ²³⁹Pu nuclear moment giving a large uncertainty to the nuclear gyromagnetic ratio.

To overcome the limitations imposed by the strong hyperfine interaction, a search for the ²³⁹Pu NMR signal in PuO₂ was most plausible because the Pu⁴⁺ ion in PuO₂ is in a tetravalent state and has cubic local symmetry. Although PuO₂ is known to have a variable stoichiometry of oxygen, a fully oxidized and well characterized sample should contain Pu⁴⁺ (5f⁴ , ⁵I₄) ions with a non-magnetic, singlet ground state in a cubic crystalline electric field. With the first excited magnetic state being higher in energy than the ground state, the Pu⁴⁺ ions should be non-magnetic below room temperature, which is reflected by the observed temperature-independent magnetic susceptibility. To address the lack of accurate data for the nuclear gyromagnetic ratio, the external magnetic field was swept at a constant frequency across a wide range of expected γn values. This approach was successful—by mapping the external field dependence of the measured resonance frequency, we determined the nuclear gyromagnetic ratio ²³⁹γ_n (PuO₂) to be 2π × 2.856 MHz/T (Fig. l). Assuming a free-ion value for the Pu⁴⁺ hyperfine coupling constant, we determined the isolated “bare” value of the gyromagnetic ratio to be 2π × 2.29 MHz/T, corresponding to a ²³⁹Pu nuclear magnetic moment of 0.15 μ_N (where μ_N is the nuclear magneton).

Of course, this discovery has not suddenly transformed Pu NMR spectroscopy into a trivial affair. Challenges remain, and as of now the pool of candidate compounds for future development is rather limited because most of the known Pu compounds are strongly magnetic and their short nuclear relaxation times prevent us from observing any signals. Nevertheless, the impact of this discovery is both profound and multi-faceted. Ending a decade-long search, it has provided a proof-of-concept, rendering Pu NMR spectroscopy a reality rather than an impossibility. We now have insight into the gyromagnetic ratio values and nuclear magnetic moment of  ²³⁹Pu, which provides a clear path forward. Especially in view of the complexity of these compounds and their technological importance, directly observing the consequences of plutonium’s 5f electrons at the atomic and structural unit-cell scales using NMR spectroscopy could potentially prove a particularly powerful tool for solid-state physics, chemistry, and materials science.

²³⁹Pu NMR signal in PuB₄

In 2018 we reported the second-ever observation of a ²³⁹Pu NMR signal. We detected this in bulk and single-crystal plutonium tetraboride (PuB₄), which has been investigated as a potential correlated topological insulator. The NMR results have provided unique f-electron physics and insight into the bulk gap-like behavior of the transport properties in this material. The temperature dependence of the ²³⁹Pu Knight shift combined with a relatively long nuclear spin-lattice relaxation time indicates that PuB₄ adopts a non-magnetic state with gap-like behavior consistent with the theoretical band structure calculations. These results imply that PuB₄ is a good candidate as a topological insulator. Simultaneously we observed an excess nuclear relaxation process below a characteristic temperature, typically ~100 K (Fig. 2), in which the high-temperature activated process crossed over into a low-temperature process of unknown character. This is a common feature with other topological insulators and Weyl semi-metals (e.g., SmB₆, YbB₁₂, and TaP); the mechanism is still an open question. Another exciting finding is that the Knight shift, which predominantly arises from the orbital hyperfine interaction, strongly depends on the chemical coordination and the nature of electronic states (Fig. 3). We found that ²³⁹Pu NMR spectroscopy is quite sensitive to f-electron configuration (metallic or ionic), nature of the chemical bond, and the environment around the Pu ions. This gives a new path for the exploration of self-radiation damage and accompanying aging effects using ²³⁹Pu NMR spectroscopy as a local probe.

Self-damage and isochronal annealing effects on local structure

Understanding the mechanisms of self-damage and aging effects in radioactive materials is a matter of significant concern in the actinide science community. In principle, ²³⁹Pu NMR spectroscopy could be a very effective technique to solve these problems because NMR spectra can be very sensitive to local environment changes, such as those caused by radiation self-damage. We can expect a split or broadened structure in the spectra, depending on the degree of local radiation damage. We observed such an effect on the ²³⁹Pu NMR spectrum of a PuB₄ single crystal which was stored at 3.95 K for nearly one month in order to accumulate damage (Fig. 4). We subsequently observed a satellite structure dependent on the local environment which might arise from Pu sites surrounding structurally damaged areas. Although we do not know which satellite peaks correspond to which areas, this spectrum encouraged us to perform another experiment in order to see if annealing would “heal” the damaged sites. The isochronal annealing results we have obtained so far are shown in the same figure. After annealing at elevated temperatures, the spectrum does not change and there are essentially no effects on the annealing temperature up to 250 K. This is because the melting point of PuB₄ is so high (ca. 20,430°C) therefore the annealing temperature is expected to be much higher than room temperature. Nevertheless, further investigations are warranted for δ-Pu metal or PuO₂ on which macroscopic measurements have shown the annealing effect at much lower temperatures (typically less than room temperature).

Summary and outlook

In view of the complexity of Pu compounds and their technological importance, directly observing the consequences of plutonium’s 5f electrons at the atomic and structural unit-cell scales using the ²³⁹Pu NMR spectroscopy could potentially prove a particularly powerful tool for the study of plutonium solid-state physics, chemistry, and materials science. Seven years passed between the first observation of the ²³⁹Pu NMR signal, in PuO₂, and the second, in PuB₄. During this time we have attempted to observe signals in several plausible materials, such as PuCoGa₅, PuPt₃, and [(Me)₄N]₂PuCl₆, but no signatures were observed. Nevertheless, our understanding of ²³⁹Pu NMR characteristics has greatly advanced. There are many areas from both physics and chemistry perspectives that future work could build on these advances:

1. The Knight shift strongly depends on the nature of the chemical bonding between Pu and ligand atoms (ranging from 0 to +25 %). Therefore the observation of ²³⁹Pu NMR signals in other compounds of interest is important not only to understand basic electronic states but also to elucidate the details of chemical bonding (compounds of interest include: Pu metal (Ga-stabilized δ-Pu), PuB₆ (topological insulator), PuCoGa₅ (superconductor), PuF₄, etc.).

2. Now we know NMR parameters for both ²³⁵U and ²³⁹Pu, a very interesting project is the study of the UO₂/PuO₂ mixed oxide system (MOX) to characterize the phase stability and thermoelectric properties.

3. From a pure physics perspective, the excess relaxation process observed below T* in PuB₄ should be clarified because this is a rather common feature observed in various topological insulators or semi-metals (e.g., SmB₆, YbB₁₂, TaP, etc.). From the external field and pressure dependence of 1/T₁, we could understand this phenomenon in more detail.

4. Future work studying self-radiation damage and aging effects should be continued using δ-Pu or PuO₂ rather than PuB₄. Because of its high melting point, annealing temperatures are much higher than room temperature and therefore we have concluded that PuB₄ is a poor choice of material for this type of study. A better choice of materials could allow us to observe how the ²³⁹Pu NMR spectra change as a function of annealing and consequently obtain the thermodynamic data of the damaged lattice by radiation and annealing effects.

Finally, I should emphasize that although we now have an excellent microscopic tool to study plutonium compounds which have the most fascinating and unique physical and chemical properties, we also need to have strong collaborations in the form of teamwork between synthetic scientists and theoretical physicists and chemists to support this effort. 

Acknowledgments

This project was supported by the LANL Laboratory Directed Research and Development program and the Glenn T. Seaborg Institute. I thank the following for all their contributions: A. Mounce, A.P. Dioguardi, S.B. Blackwell, S. Thomas, S.M. Thomas, S.K. Cary, S.A. Kozimor, A.T.E. Albrecht-Schmitt, J.D. Thompson, E.D. Bauer, H.C. Choi, J.-X. Zhu, F. Ronning, D.L. Clark.

Further reading:

1. G. Koutroulakis, H. Yasuoka, “Actinides: Nuclear magnetic resonance,” Encyclopedia of Inorganic and Bioinorganic Chemistry, 2018, John Wiley & Sons, Ltd.
2. H. Yasuoka, G. Koutroulakis, H. Chudo, S. Richmond, D.K. Veirs, A.I. Smith, E.D. Bauer, J.D. Thompson, G.D. Jarvinen, D.L. Clark, “Observation of  ²³⁹Pu nuclear magnetic resonance,” Science, 2012, 336, 901.
3. A. P. Dioguardi, H. Yasuoka, S. M. Thomas, H. Sakai, S. K. Cary, S. A. Kozimor, T. E. Albrecht-Schmitt, H. C. Choi, J.-X. Zhu, J. D. Thompson, E. D. Bauer, F. Ronning, “ ³⁹Pu nuclear magnetic resonance in the candidate topological insulator PuB₄,” Phys. Rev., 2019, B99, 035104.
4. M. Takigawa, H. Yasuoka, Y. Kitaoka, T. Tanaka, H. Nozaki, Y. Ishizawa, “NMR study of a valence fluctuating compound SmB₆,” J. Phys. Soc. Jpn., 1981, 50, 2525.
5. H. Yasuoka, T. Kubo, Y. Kishimoto, D. Kasinathan, M. Schmidt, B. Yan, Y. Zhang, H. Tou, C. Felser, A.P. Mackenzie, M. Baenitz, “Emergent Weyl fermion excitations in TaP explored by  ¹⁸¹Ta quadrupole resonance,” Phys. Rev. Lett., 2017, 118, 236403.