A digitally created image of a red and white magnet with a magnetic field around it.


By Craig Tyler| August 01, 2020

First-ever measurement of temperature-activated electron reconfiguration in the world’s strangest metal

Plutonium is known far and wide for its nuclear properties. But in certain circles, it’s also known for its material properties. The manmade metal comes in six distinct solid crystalline structures spanning a wide array of chemical properties and a stupendous range in volume. At one end of the spectrum, alpha plutonium (α-Pu) is brittle and extremely dense; at the other, the same mass of delta plutonium (δ-Pu) is much softer and more ductile, and it takes up an astonishing 25 percent more space. 

Equally confounding, these changes in structure and behavior, which run from α through δ and beyond as temperatures rise, reverse direction along the way. Up to about 320 °C, the onset of δ-Pu, volume increases with increasing temperature. This is a normal material behavior known as thermal expansion: higher temperature means stronger vibrations across the lattice of atoms that make up the material, and those vibrations are accommodated by a modest expansion in size. For the next 165 degrees, however, the volume decreases; this is not normal. It is a peculiarity of δ-Pu (and the next state after that as well) and somehow works in opposition to the increasing lattice vibrations. (Thereafter, plutonium enters its sixth and final solid state, in which volume returns to its initial dynamic, increasing with temperature.) 

Explaining the 25 percent volume disparity and its reversal in direction with anything more concrete than educated speculation has been all but impossible for as long as the metal has existed. Recently, however, Los Alamos physicist Neil Harrison and materials scientist Paul Tobash and their colleagues found fresh inspiration to attempt a new kind of experiment on plutonium. Magnetostriction, as it is known, measures volume changes occurring in response to a strong magnetic field as a tiny fraction of the atoms of the bulk material magnetize and align with the external field. Unlike elevated temperatures, magnetostriction almost exclusively affects the configuration of electrons in the material, not the atomic lattice vibrations. For that reason, it accesses a separate, electronic aspect of the properties of plutonium metal—one that might explain its anomalous volume behavior. The price for that access is an enormous magnetic field.

“It would be great if everyday magnetic fields could be used for magnetostriction, like ‘a little magnetism goes a long way,’” says Harrison. “But it’s actually the opposite. You need a tremendous magnetic field to see tiny volume changes, even though changes thousands of times larger are easily brought about by heating.”

The research team used magnetic fields ranging up to a colossal 15 teslas. (The earth’s magnetic field is 35–65  millionths of a tesla.) Even so, the plutonium stretched by only about one millionth of its original length on a side. An extremely powerful (but nondestructive) magnetic field source is needed to produce the expansion, and an extremely sensitive optical measurement apparatus is needed to notice it. But with access to the National High Magnetic Field Laboratory and Los Alamos capabilities in plutonium sample preparation, optical Bragg reflection, and cryogenic temperature regulation, the researchers were able to isolate and quantify the effect. And the result was every bit as illuminating as they hoped it would be.

A colored graph showing the six crystal structures of plutonium. On the X-axis is temperature and on the Y-axis is volume. From left to right, the plutonium structures are alpha, beta, gamma, delta, delta prime, and epsilon.
Solid plutonium metal exists in six distinct crystal structures. From absolute zero to somewhat above the boiling point of water, the first of these, α-plutonium, or α-Pu, exists. It, as well as the next two structures, β-Pu and γ-Pu, and later ε-Pu, obeys standard thermal expansion: higher temperatures require the metal to swell to greater volumes to accommodate increased atomic-lattice vibrations. However, anomalously for δ-Pu (and its successor, δ’-Pu), which exists naturally at temperatures several hundred degrees above α-Pu, the behavior reverses, with the graph sloping downward: the metal shrinks when further heated. Evidently, for δ-Pu, the metal is able to absorb heat energy in some manner other than increased vibrations in the atomic lattice.

Electronic entropy

As had been long suspected, the magnetostriction measurement confirmed an electronic effect that operates alongside normal thermal expansion. At just a hair above absolute zero, and therefore firmly in plutonium’s ground-state configuration, sweeping the magnetic field up to 15 teslas didn’t do anything; the electron configuration was essentially nonmagnetic. But starting 50 degrees hotter, the electronic configuration began to change; the atoms of plutonium became faintly magnetizable and the applied field caused the metal to expand. Then, nearing room temperature, the electrons began to reorganize again. The plutonium became more magnetizable but reversed its response to the applied magnetic field and got smaller—at long last demonstrating the origin, electronic in nature, of plutonium’s anomalous volume-reversing behavior. 

What exactly are the electrons doing differently? Well, they could be doing a lot of things. Unlike the first 88 elements on the periodic table (which have up to 88 electrons), ground-state plutonium, element 94, has six electrons in the 5f orbitals. Such 5f electrons are extremely complicated. The nature of the orbitals themselves and their large distance from the atomic nucleus give 5f electrons a lot of freedom to interact with other electrons while remaining localized to the same atom or becoming itinerant, roaming from one atom to the next across the metal (and interacting with other itinerant electrons). That makes plutonium’s electronic configuration far more complicated than that of an element without 5f electrons. But even among the handful of elements with 5f electrons, plutonium is arguably the most difficult to understand.

Take americium, for example, plutonium’s next-door neighbor on the periodic table. It has seven 5f electrons; any or all of these have the potential to be localized (confined within atomic orbitals) or itinerant. In each case, there is an energy-minimizing configuration that corresponds to a distinct volume, and in americium the energy is always minimized when six electrons are localized. However, for plutonium’s six (total) 5f electrons, it happens that for most of the energy-minimizing configurations, many different volume states share nearly the same energy, making it very difficult to determine which configurations are most relevant. Fortunately, the volume for each energy-minimizing configuration always gets larger when an additional 5f electron becomes localized. Therefore, Harrison and Tobash were able to pick up on all the measurable energy differences between plutonium’s electronic configurations accessed with increasing temperature, allowing them to go from the ground state to expansion and then contraction.

In thermodynamic parlance, the experiment had isolated the metal’s electronic entropy: increasing energy (whether by magnetism or by heating) brings about “disorder” in the form of a proliferation of electron states within the atoms, rather than just amplifying lattice vibrations. In effect, solid plutonium was shown to have new ways to absorb energy without getting hotter. And in the case of δ-Pu, absorbing that additional energy involves reconfiguring the electrons in such a way that the atoms pack together much more tightly, greatly overpowering the normal effect from lattice vibrations.

“Nothing like this occurs in ordinary metals,” says Harrison. “Minor temperature changes access whole new electronic states in a way that no prior theory of plutonium could explain. And by observing a much smaller effect in response to an external magnetic field, we were finally able to show what’s happening.”

Get real

From a theoretical point of view, the magnetostriction results are a godsend. And for Tobash, who prepared the plutonium samples for the experiment, this discovery is just the beginning.

An image demonstrating magnetorestriction. On the left is a cluttered group of nine round items and on the right is the nine items connected in a three-by-three group.
Magnetostriction occurs when a material is subjected to a large external magnetic field. Microscopic magnetic regions within the material undergo a reorientation, stretching or squeezing the overall material along different axes.

“We started with new, high-purity plutonium samples,” he says. “But most of the time in real life, plutonium is neither newly manufactured nor pure.” Plutonium used for real-world applications is generally doped with trace amounts of other materials. Gallium, for example, is known to stabilize δ-Pu so that it can exist at room temperature, where pure plutonium would naturally assume the α-Pu state. Of course, one might be tempted to just avoid the mood swings of δ-Pu and be content to work with α-Pu instead, but it turns out that δ-Pu is much more useful. Not only is it softer and therefore more shapeable, but it’s also less oxidizing and more stable, both chemically and mechanically. Therefore, it is standard practice to lock the δ-Pu crystal structure into place, even at room temperature. Gallium can do that because it forms bonds with the same crystal structure as δ-Pu; the α-Pu lattice, on the other hand, can’t accommodate gallium quite so easily.

So Harrison, Tobash, and their collaborators performed the magnetostriction measurement with various levels of gallium doping. They found that increasing gallium concentrations corresponded to the signature δ-Pu electronic volume reversal becoming smaller than that in pure δ-Pu. Reassuringly, this is a reasonable result; replacing plutonium atoms with gallium, which has no 5f electrons, should tend to suppress plutonium’s distinctive behavior.

Apart from being doped, the other key aspect of real-world plutonium is that it often sits for long periods of time, such as inside a warhead at a military installation. Because of its radioactivity, high-energy particles are constantly streaming through it, creating more and more defects in its atomic structure over time. Los Alamos conducts a great deal of research on this effect in order to assess the safety, effectiveness, and reliability of aging nuclear weapons. The magnetostriction measurement and its implications for plutonium’s electronic entropy provide an important basis for developing a more accurate understanding of the aging weapons in the nation’s stockpile.

A digital image showing atom arrangement in alpha plutonium and delta plutonium. The atoms in the alpha plutonium are horitontally shifted, while the atoms in the delta plutonium lack this shift. Two red particles in the delta plutonium image represent gallium.
(Left) The low-temperature state of plutonium, α-Pu, takes the form of a simple monoclinic crystal: essentially, atoms arranged in a pattern of flat squares layered on top of one another, with each layer shifted horizontally relative to the one below by the same amount and direction. (Right) By contrast, δ-Pu has a face-centered cubic structure, without the shifted layering. In order to preserve the useful properties of δ-Pu at lower temperatures, including room temperature, the metal can be doped with gallium (red)—i.e., a few percent of the plutonium atoms are replaced with gallium atoms—because gallium shares the face-centered cubic structure of δ-Pu, making it more difficult for the crystal to realign into the monoclinic structure as it cools.

“Having conducted the magnetostriction experiment on newly synthesized plutonium samples, we now have a much-needed baseline,” says Tobash. “From a practical perspective, the next step will be to try it on aged material, since that will add a bit more complexity with additional variables to sort out. We’re planning those experiments now.” 

From a less purely practical perspective, another next step might be to adapt the plutonium magnetostriction results to advance scientists’ theoretical understanding of its electronic properties.

“We all tend to think of plutonium in terms of the nuclear properties that make it dangerous—and no question, it certainly is that,” says Harrison. “But it’s also something else. From a chemistry standpoint, it is arguably the most complicated element known. And that means that when it comes to difficult, important concepts like electronic entropy, which barely even show up in the practical applications of plutonium today, it can lead us to the scientific and technological advances of tomorrow.” LDRD