At Penn State, a team of materials scientists has spent more than a decade coaxing order and disorder into coexistence within a ceramic film thinner than a human hair — and in doing so, may have found a quieter, less hungry way for machines to think and remember. By harnessing magnetic signals rather than electrical current to move and store data, their high-entropy oxide film points toward a future in which the vast power appetite of our digital infrastructure might be meaningfully tamed. The work is a reminder that foundational curiosity, pursued without a guaranteed destination, is often th
Penn State researchers develop magnetic ceramic film for ultra-efficient computing
Chemical disorder drives magnetic signals that require less energy than electrical current
Mark
So this is a ceramic film thinner than a hair. What makes it different from other magnetic materials researchers have worked with?
Mimi
The key is how they used chemical disorder. They packed multiple elements into a crystal structure, but then deliberately introduced irregularity within that order. That disorder actually drives the magnetic signals they need.
Luke
But wait—is this film already working in devices, or is it still in the lab phase?
Mimi
Still in the lab. They spent a decade just perfecting how to make it. Now they're planning to work with industry on the next steps.
Mark
And the efficiency claim—10 to 100 times more efficient in data centers. That's a huge range. What determines where you'd actually land?
Luke
That's a good question. The source attributes that range to Jon-Paul Maria, but doesn't specify what conditions or applications would achieve the lower end versus the higher end. It's a potential, not a guarantee.
Mimi
Right. The efficiency comes from using magnetism instead of relying so heavily on electrical current. But how much you actually gain depends on how the material performs when it's integrated into real systems.
Mark
Why did it take ten years just to fabricate the film?
Mimi
They were essentially inventing the process as they went. Almishal compared it to developing a paint formula and then figuring out how to apply it. They had to merge materials science and physics to create the right layered interfaces inside the film.
Luke
And those layered interfaces—are those something they can reliably reproduce at scale, or is that still an open question?
Mimi
The source doesn't address manufacturing scale. That's clearly the next frontier, but it's not discussed here.
Mark
So what's the timeline for seeing this in actual devices?
Mimi
Unknown. They're starting conversations with industry partners now. This is still foundational work, even though it's been a decade in the making.
Le Pouls
- Data centers and consumer devices are straining under their own electricity demands, and the computing industry has no clear path to efficiency without rethinking the physics beneath the hardware.
- A ceramic film thinner than a human hair now demonstrates that chemical disorder deliberately introduced into a crystal lattice can generate magnetic signals capable of transmitting and storing data with far less energy than electrical current.
- The breakthrough required over a decade of fabrication work — layer by layer, experiment by experiment — merging materials science and physics in a process one researcher likened to formulating a paint before ever touching a wall.
- The film bridges two previously separate fields, magnonics and spintronics, and a new Penn State laboratory is being built specifically to deepen that convergence with both graduate and undergraduate researchers.
- Industry partnerships are now being pursued, with projections of 10 to 100 times efficiency gains in large-scale data centers — though the distance between laboratory promise and commercial reality remains an open and honest question.
At Penn State, a team of materials scientists has spent more than a decade coaxing order and disorder into coexistence within a ceramic film thinner than a human hair — and in doing so, may have found a quieter, less hungry way for machines to think and remember. By harnessing magnetic signals rather than electrical current to move and store data, their high-entropy oxide film points toward a future in which the vast power appetite of our digital infrastructure might be meaningfully tamed. The work is a reminder that foundational curiosity, pursued without a guaranteed destination, is often the longest but most reliable road to transformation.
Researchers at Penn State have engineered a ceramic film so thin it rivals a human hair, yet its properties could fundamentally change how computers process and store information. Published in Advanced Materials, the work centers on high-entropy oxides — crystalline structures that pack multiple elements into a single lattice. The team's key insight was to deliberately introduce chemical irregularity within that ordered structure, a quality they call "anomalous crystallinity," which counterintuitively became the mechanism for generating useful magnetic phenomena.
The significance lies in what those magnetic signals can replace. Today's electronics depend on electrical current for both processing and storage. Magnetic signals require less energy to accomplish the same tasks, meaning a shift toward magnetics-driven computing could reduce power consumption across everything from smartphones to the server farms that underpin the modern internet. As assistant professor Saeed Almishal put it, the next generation of electronics will demand more efficient control of the internal signals that make devices work.
The road to this material was neither short nor straight. Almishal described the fabrication process — more than a decade in the making — as akin to formulating a paint and then learning how to apply it. The effort drew on collaborators across multiple institutions, including Oak Ridge National Laboratory, and produced layered interfaces within the film whose magnetic behavior was measured, refined, and gradually shaped toward practical utility. Together with a 2025 paper in Advanced Science, the work represents a bridge between magnonics and spintronics, two fields that had largely developed in parallel.
Almishal is now building Penn State's first magnetics laboratory within the College of Earth and Mineral Sciences, a signal of institutional commitment to a direction that was not previously central to the department. Industry collaboration is planned, with efficiency gains of 10 to 100 times projected for large-scale data centers. Jon-Paul Maria, who helped pioneer high-entropy oxides as a materials class over a decade ago, offered the broader lesson plainly: early-stage experiments pursued without a predetermined endpoint are often what make landmark advances possible.
A team of researchers at Penn State has engineered a ceramic film so thin it rivals a human hair in width, yet carries properties that could reshape how computers process information and consume power. The work, published in Advanced Materials, describes a material built from high-entropy oxides—complex crystalline structures that pack multiple elements into a single, ordered lattice. What makes this particular film novel is not just what it contains, but how its creators harnessed chemical disorder within that order to generate magnetic signals that move data more efficiently than electrical current alone.
The research emerged from more than a decade of work rooted in a 2015 breakthrough. Jon-Paul Maria, interim head of Penn State's Department of Materials Science and Engineering, helped pioneer high-entropy oxides as a materials class. Saeed S. I. Almishal, an assistant professor of materials science and engineering, built on that foundation with colleagues across multiple institutions, including Oak Ridge National Laboratory in Tennessee. The team's central insight was to exploit what they call "anomalous crystallinity"—a deliberate introduction of chemical irregularity within otherwise perfect crystal structures. This disorder, counterintuitively, became the mechanism for generating the magnetic phenomena they sought.
The distinction matters because it points toward a fundamental shift in how electronics might work. Devices today rely on electrical current to both process and store information. Magnetism has always played a role in data retention, but it has remained secondary. By elevating magnetic signals—which require less energy to transmit and store data than electrical current—engineers could reduce the power draw of everything from smartphones to massive server farms. Almishal described the goal plainly: "A new generation of advanced electronics will require more efficient control of the internal electrical and magnetic signals that make everyday devices work."
The path to this material was not direct. Creating the film itself took a decade, a process Almishal compared to formulating a paint and then applying it to a wall. The work merged materials science and physics to construct layered interfaces within the film's structure. Researchers measured magnetic behavior along these interfaces, refined the fabrication process through repeated experiments, and gradually shaped the material into something that could eventually function inside actual devices. This effort, combined with a 2025 paper the team published in Advanced Science, represents what Almishal sees as a bridge between two previously separate fields: magnonics, which studies magnetic waves, and spintronics, which focuses on the magnetic properties of spinning electrons.
Almishal is now establishing the first magnetics laboratory within Penn State's College of Earth and Mineral Sciences, housed in the Materials Research Institute. The Quantum and Magnetic Materials Measurements Lab will eventually house tools for testing magnetic properties under various conditions, and will involve both undergraduate and graduate students. The lab signals an institutional commitment to a research direction that, until recently, was not part of the department's traditional focus.
The practical implications extend well beyond the laboratory. Jon-Paul Maria suggested that materials like this one could deliver efficiency gains of 10 to 100 times in large-scale data centers—facilities that currently consume enormous amounts of electricity. Almishal and his team plan to work with industry partners to develop the film further for use in consumer electronics. The path from basic research to commercial application remains uncertain, but Maria emphasized the importance of recognizing how foundational work—in this case, the early exploration of high-entropy oxides without a predetermined endpoint—can eventually yield transformative advances. "Early-stage experiments give rise to the landmark advancements that make impactful change," Maria said.
Citations marquantes
A new generation of advanced electronics will require more efficient control of the internal electrical and magnetic signals that make everyday devices work.— Saeed S. I. Almishal, assistant professor of materials science and engineering
Such materials offer a potential approach to achieving high efficiency—10 to 100 times more efficient—in data centers.— Jon-Paul Maria, interim head of the Department of Materials Science and Engineering
Envie de l'histoire complète ?
Lire l'original sur Phys.org ↗
Culture & Arts