One universal knob across all these systems to access different properties
At the University of Michigan, researchers have discovered that the invisible crystalline order electrons form inside metals is not as fixed as science once assumed — it melts, deforms, and accumulates defects just as physical matter does. This insight, drawn from watching tantalum sulfide under an electron microscope and confirmed across decades of prior research, suggests that the quantum world obeys principles metallurgists have long applied to ordinary metals. The discovery opens a path toward engineering quantum materials with the same deliberate craft we bring to steel and silicon, potentially transforming how superconductors are built and how computers learn to think like brains.
- Electrons in metals secretly organize into crystals — and those crystals can melt, a behavior no one had fully recognized until now.
- As temperature rises, the orderly electron patterns blur and dissolve, disrupting electrical flow in ways that could be deliberately triggered or suppressed.
- Researchers combed through 28 prior studies and found the same melting signature hiding across nearly every two-dimensional metal examined, suggesting a universal phenomenon long overlooked.
- Controlling this melting could let engineers flip materials between conducting and insulating states — mimicking the firing of neurons for energy-efficient, brain-like computing.
- The field of 'quantum metallurgy' is now taking shape: one universal control mechanism, one tunable knob, potentially unlocking a new generation of superconductors and neuromorphic devices.
Electrons in metals can arrange themselves into orderly patterns called charge density waves — structures long assumed to be stable and fixed. Researchers at the University of Michigan have now shown that these electron crystals behave far more like ordinary physical solids than anyone realized: they deform, accumulate defects, and ultimately melt.
The discovery came when associate professor Robert Hovden heated a thin sheet of tantalum sulfide and observed its electron crystal through an electron microscope. As the temperature approached 568 degrees Fahrenheit, the once-crisp diffraction pattern smeared and faded — the unmistakable signature of melting. The electron crystal was dissolving in real time.
The deeper significance lies in what this makes possible. Metallurgists have long shaped the properties of ordinary metals by engineering their defects — introducing impurities, manipulating disorder. Hovden's team realized the same logic could apply to electron crystals. Control the melting, and you control the material's quantum properties. "Quantum metallurgy could be the future," Hovden said.
The applications are concrete. Superconductivity often emerges near defects in charge density waves, suggesting that precise defect engineering could yield better superconductors. And because a melting electron crystal switches a material from conductor to insulator — mimicking the way neurons fire — these materials could form the basis of neuromorphic computers that process data with far less energy than silicon.
To test the universality of their finding, the team reviewed 28 published studies of metals with charge density waves and found evidence of melting across nearly all of them, spanning materials with vastly different electrical and magnetic properties. Co-author Jeremy Shen noted the power of the insight: one universal control mechanism exists across all these systems — a single knob capable of unlocking different quantum properties. The results, published in the journal Matter, suggest that quantum order and disorder exist on a spectrum — and that spectrum is now available to engineer.
Electrons in metals do something unexpected: they arrange themselves into orderly patterns, much like atoms in a crystal lattice. These electron crystals, formally known as charge density waves, have long been understood as fixed structures—neat, periodic, stable. But researchers at the University of Michigan have discovered they behave more like ordinary solids than anyone realized. They can deform. They can melt. And that discovery could reshape how we build superconductors and neuromorphic computers.
The finding emerged from work led by Robert Hovden, an associate professor of materials science and engineering, who heated a thin sheet of tantalum sulfide and watched what happened to its electron crystal through an electron microscope. As the temperature climbed toward 568 degrees Fahrenheit, the orderly arrangement of electron clusters began to break down. Rows that had been uniformly spaced grew irregular. Defects accumulated. The pattern that had been crisp and periodic started to blur and fade, exactly as a physical crystal does when it melts into liquid. The electron beam Hovden's team fired at the metal revealed the transformation: spots in the diffraction pattern that had been sharp and distinct smeared into ovals, then faded into a faint halo. The electron crystal was melting.
What makes this significant is not merely that it happens, but that it opens a new way of thinking about quantum materials. Metallurgists have long controlled the properties of ordinary metals by engineering defects—introducing impurities, creating grain boundaries, manipulating the disorder in the atomic structure. Hovden's team realized the same principle might apply to electron crystals. If you can control how much an electron crystal melts, you can control its properties. You can tune it. "Quantum metallurgy could be the future," Hovden said in describing the approach.
The practical implications are substantial. Superconductors—materials that carry electrical current without any resistance—often exist in states that coincide with defects in charge density waves. By learning to engineer those defects precisely, researchers might be able to design better superconductors. Equally promising is the potential for neuromorphic computing. When an electron crystal melts, it disrupts the flow of electricity through the metal, effectively switching it from conductor to insulator. That switching behavior mirrors the way neurons fire and transmit signals. Materials that can rapidly toggle between conducting and insulating states could process and move vast amounts of data while consuming far less energy than conventional silicon chips.
The discovery gains weight from its universality. Hovden's team suspected the melting signature might be hiding in older research and combed through 28 published studies of metals containing charge density waves. They found evidence of melting in nearly every two-dimensional metal they examined, and in several three-dimensional ones as well. The materials studied had wildly different electrical and magnetic properties, yet the underlying physics of their electron crystals followed the same pattern. Jeremy Shen, a master's student who co-authored the work, captured the significance: across all these different systems, there is one universal control mechanism—one knob that could be turned to access different properties. That simplicity, that universality, is what makes the finding so powerful.
The work was supported by the U.S. Department of Energy and the National Science Foundation, and the experiments were conducted at the Michigan Center for Materials Characterization. The results were published in the journal Matter. What Hovden's team has shown is that quantum structures long assumed to be rigidly ordered actually exist on a spectrum of disorder, and that spectrum can be leveraged. The question now is how quickly engineers can turn that knowledge into devices.
Notable Quotes
These quantum structures, which are often thought to have a highly ordered structure, actually span a continuum of disorder that could be leveraged to engineer and control these materials.— Robert Hovden, University of Michigan
A similar approach might help us harness the potential of quantum materials in future devices. Quantum metallurgy could be the future.— Robert Hovden, University of Michigan
The Hearth Conversation Another angle on the story
When you say an electron crystal melts, what's actually happening? Is it the same as ice turning to water?
It's analogous but not identical. In ice, atoms break free from their fixed positions and flow. In an electron crystal, the electrons are already mobile—they're not locked in place. What melts is the pattern itself, the orderly arrangement. The electrons stop clustering in uniform rows and start to scatter. The periodicity vanishes.
So the electrons themselves don't change, just their organization?
Exactly. The electrons rejoin what researchers call the background field. The structure that made them special—that made them a crystal—dissolves. But the metal itself remains solid. You're not getting a liquid in the conventional sense.
Why does this matter for superconductors? What's the connection?
Superconductivity often emerges when there are defects in the charge density wave. If you can control how many defects exist, how much melting occurs, you can potentially tune when and how superconductivity appears. Right now it's largely accidental. This could make it deliberate.
And for neuromorphic computing, you mentioned switching between conductor and insulator. How does that help a computer think?
Brain cells communicate by rapidly opening and closing channels that let electrical signals through or block them. If you can build materials that switch between conducting and insulating states just as quickly, you've created something that mirrors that biological process. You can process information the way a brain does, with far less power.
The study looked at 28 different metals. Did they all show the same melting behavior?
Nearly all of them. That's the stunning part. Materials with completely different properties—different electrical behavior, different magnetic behavior—all showed evidence of this same underlying physics. It suggests this isn't a quirk of one material. It's fundamental.
What happens next? Is this moving toward actual devices?
That's the open question. The researchers have shown the physics works. Now engineers need to figure out how to control it precisely, how to engineer the defects in ways that are reproducible and useful. It's early, but the framework is there.