Faster, low-powered electronics

Material’s ‘incipient’ property could jumpstart fast, low-power electronics

Scientists at Penn State have harnessed a unique property called incipient ferroelectricity to create a new type of computer memory that could revolutionize how our devices work, such as using much less energy and being able to work in extreme environments like outer space.

Scientists at Penn State have harnessed a unique property called incipient ferroelectricity to create a new type of computer memory that could revolutionize how our devices work, such as using much less energy and being able to work in extreme environments like outer space.

The researchers noted that the societal benefits of this research could be significant. Traditional AI systems, especially those handling image recognition, consume significant energy. The ferroelectric transistors’ low power requirements present a sustainable alternative.

“AI accelerators are notoriously energy-hungry,” Harikrishnan Ravichandran, a doctoral student in engineering science and mechanics and co-author of the study, said. “Our devices switch rapidly and consume far less energy, paving the way for faster, greener computing technologies.”  

Their study, published in Nature Communications, focuses on multifunctional two-dimensional field-effect transistors (FETs) that leverage incipient ferroelectricity—a previously overlooked property—within freestanding strontium titanate nanomembranes.  

FETs are advanced electronic devices that use ultra-thin layers of materials to control electrical signals, offering multiple functions like switching, sensing or memory in a compact form. Incipient ferroelectricity refers to materials that show signs of temporary, scattered polarization, meaning parts of it can switch charges like tiny dipoles but it does not settle into a stable state under normal conditions.

Think of it like a material that is almost, but not quite, ferroelectric. It has the potential to become ferroelectric, but it needs a little push. That is kind of what incipient ferroelectricity is -- a material that is on the verge of becoming ferroelectric, meaning it can hold an electrical charge, but needs certain conditions to make that happen.  

“Incipient ferroelectricity means there’s no stable ferroelectric order at room temperature,” Dipanjan Sen, doctoral candidate in engineering science and mechanics and lead author in the study, said. “Instead, there are small, scattered clusters of polar domains. It’s a more flexible structure compared to traditional ferroelectric materials.”

While this trait is often considered a limitation, the team found ways to harness incipient ferroelectricity for possible new applications. According to Ravichandran, the devices displayed unique behaviors across temperature ranges.  

“The main goal of the project was to explore whether incipient ferroelectricity, usually seen as a disadvantage because it leads to short memory retention, could actually be useful,” said Saptarshi Das, Ackley Professor of Engineering and Professor of Engineering Science and Mechanics at Penn State and corresponding author of the study. “At cryogenic conditions, this material exhibited traditional ferroelectric-like behavior suitable for memory applications. But what we saw at room temperature, this property behaved differently. It had this relaxor nature.”  

Relaxer behavior refers to a more disordered, short-range polarization response. This type of behavior is less predictable and more fluid, which contrasts with the stable, long-range order seen in traditional ferroelectrics. This means the material's ferroelectric properties are weaker or less stable at room temperature. Instead of being a drawback, it showed potential for use in neuromorphic computing, which aims to imitate how the human brain processes information using neurons and uses much less energy than traditional computers. Like our brain, it saves energy by only using power when needed, like flipping a light switch on and off, instead of staying on all the time like traditional computers.

“These devices acted like neurons, mimicking biological neural behavior,” said Mayukh Das, doctoral candidate in engineering science and mechanics and study co-author. “To test we performed a classification task using a grid of three-by-three pixel images fed into three neurons. The devices were able to classify each image into different categories. This learning method could eventually be used for image identification and classification or pattern recognition. Importantly, it works at room temperature, reducing energy costs. These devices function similarly to the nervous system, acting like neurons and creating a low-cost, efficient computing system that uses a lot less energy.”

Strontium titanate is typically non-ferroelectric. However, freestanding nanomembranes of strontium titanate exhibit polar order, as confirmed by second harmonic generation experiments. Notably, it started to exhibit ferroelectric-like behavior, especially at very low temperatures due to reduced thermal fluctuations.

Strontium titanate thin films, along with its incipient ferroelectricity, is also a perovskite material. Perovskites, materials with a specific type of crystal structure, are valued for their exceptional electronic properties.  

“We were surprised to see that these well-known perovskite materials could exhibit exotic ferroelectric properties at the device level,” Sen said. “It wasn’t something we anticipated, but once we started fabricating the devices, we saw behaviors that could really redefine advanced electronics.”

Using molecular beam epitaxy, a technique that involves depositing a layer of atoms on a substrate to form a thin film, researchers at the University of Minnesota grew the strontium titanate films. These were then combined with molybdenum disulfide, a well-established two-dimensional material. Together, these materials created FETs with innovative functionalities, including incipient ferroelectricity.

The researchers noted that future research will include solving current challenges such as scalability and commercial viability while exploring other potential materials.  

“Right now, this is at the research and development stage,” Sen said. “Perfecting these materials and integrating them into everyday devices like smartphones or laptops will take time, so there’s so much more to explore. In addition, we’re examining other perovskite materials, like barium titanate, to uncover their potential. The opportunities for growth are immense, both in materials and device applications.”

Along with Sen, Das and Ravichandran, other authors of the study include from Penn State Saptarshi Das, corresponding author and Ackley Professor of Engineering and professor of engineering science and mechanics; Pranavram Venkatram, graduate student in engineering science and mechanics; Zhiyu Zhang, graduate student in engineering science and mechanics; Yongwen Sun, graduate student in engineering science and mechanics; Shiva Subbulakshmi Radhakrishnan, graduate student in engineering science and mechanics; Akash Saha, graduate student in materials science and engineering; Sankalpa Hazra, graduate student in materials science and engineering; Chen Chen, assistant research professor, thin films in the Two-Dimensional Crystal Consortium (2DCC-MIP); Joan Redwing, director of the 2DCC-MIP and distinguished professor of materials science and engineering and of electrical engineering; Venkat Gopalan, professor of materials science and engineering and of physics; and Yang Yang, assistant professor of engineering science and mechanics and of nuclear engineering. From the University of Minnesota, co-authors of the study include Sooho Choo, Shivasheesh Varshney, Jay Shah, K. Andre Mkhoyan and Bharat Jalan.

Magnetic semiconductor preserves 2D quantum properties in 3D material

An international team led by physicists at Penn State and Columbia University has developed a novel approach to maintain special quantum characteristics, even in three-dimensional (3D) materials.

There is a big problem with quantum technology — it’s tiny. The distinctive properties that exist at the subatomic scale usually disappear at macroscopic scales, making it difficult to harness their superior sensing and communication capabilities for real-world applications, like optical systems and advanced computing. Now, however, an international team led by physicists at Penn State and Columbia University has developed a novel approach to maintain special quantum characteristics, even in three-dimensional (3D) materials.

The researchers published their findings today (Feb. 19) in Nature Materials.

“Although the functionalities displayed by two dimensional (2D) materials are vast and their potential is revolutionary, maintaining their superior properties beyond the 2D limit remains a formidable challenge,” said first author Yinming Shao, assistant professor of physics at Penn State, explaining that such materials are typically crystals that are only one atom thick and can be applied in a variety of fashions, including for flexible electronics, energy storage and quantum technologies. “Realization, understanding and control of nanoscale confinement are, thus, crucial for both exploration of quantum physics and future quantum technologies.”

The team examined quasiparticles known as excitons, which have unique optical properties and can carry energy without an electrical charge, in a semiconductor material. Semiconductors — which are ubiquitous across computers, phones and other electronics — conduct electricity under certain conditions and inhibit it under others. Excitons are produced when light hits a semiconductor, energizing an electron to jump to the next energy level. The resulting excited electron and the hole it left are jointly referred to as an exciton. Excitons occur homogenously across typical 3D semiconductors, like silicon.

“But the binding energy for the excitons in bulk materials like silicon is usually small, meaning it’s not very stable and it’s not easy to observe,” Shao said, explaining that excitons are most stable and exhibit superior properties only in 2D monolayers.

The conventional method for preparing 2D materials was developed in 2004 and led to the discovery of graphene, the single layer of carbon that is highly conductive and stronger than steel. The process is simple, but labor intensive, as each layer must be exfoliated from a bulk crystal by applying a piece of sticky tape and peeling it off.

In this thin, 2D state, excitons can carry energy without charge, as well as emit light when its electron and hole recombine, which Shao said is useful for advanced optical applications. To preserve those properties in materials large enough for such applications, however, researchers would need to produce a huge number of layers.

To do this without peeling and stacking each layer by hand, the researchers turned to another aspect of physics: magnetism. Specifically, they focused on chromium sulfide bromide (CrSBr), a layered magnetic semiconductor that co-author Xavier Roy, professor of chemistry at Columbia University, has researched extensively and further developed since 2020.

At room temperature, CrSBr acts as a normal semiconductor just like silicon. Cooling CrSBr down, to around -223 degrees Fahrenheit, brings it to a ground state, or the state of lowest energy. This transforms it into an antiferromagnetic system, in which the magnetic moments — usually referred to as “spin” — of the system’s particles align in a regular, repeating pattern. Specifically for CrSBr, this antiferromagnetic ordering ensures that each layer alternates its magnetic alignment, effectively canceling out a magnetic moment and rendering the material insensitive to external magnetic forces. As a result, excitons tend to stay in the layer with the same spin, rather than hooping to the neighboring layers with the opposite spins. Like cars on alternating one-way streets, these established boundaries keep excitons confined to the layer with which they share the same spin directions.

“This is an effective approach to create a single layer of atomic material without exfoliating it out, while still preserving a sharp interface,” Shao said. “This means we could achieve the same behavior of confined excitons demonstrated in 2D materials in a bulk material.”

Using optical spectroscopy techniques, theoretical modeling and calculation, the researchers determined that this magnetic confinement held firm no matter how many layers were in the system and no matter which layer they confined, including surface layers.

“We did a lot of work to check that this actually holds, and it does,” Shao said.

Shao’s team’s finding was corroborated by another research group out of Germany — Florian Dirnberger and Alexey Chernikov from TUD Dresden University of Technology —who were investigating the same quirk of magnetic semiconductors. The two groups decided to compare notes, Shao said, and found that they all had come to the same conclusion.

“Our data lines up really well, which is remarkable because we used two different crystal materials in different labs,” Shao said. “Our results are in agreement with each other and align well with theoretical predictions, so we wrote this joint paper.”

The aligned result came from harnessing the behaviors of magnetism, Van der Waals interactions and excitons, according to Shao, to achieve quantum confinement with potential applications for advancing optical systems and quantum technologies.

“The marriage of these different aspects of physics was a crucial aspect of this discovery,” Shao said.

Shao completed his doctorate and a postdoctoral fellowship at Columbia University. Other contributors are Siyuan Qiu, Evan J. Telford, Brian S.Y. Kim, Francesco L. Ruta, Andrew J. Mills, Daniel G. Chica, Avalon H. Dismukes, Michael E. Ziebel, Yiping Wang, Jeongheon Choe, Youn Jue Bae, Xiaoyang Zhu, Xavier Roy and D. N. Basov, Columbia University; Florian Dirnberger, Sophia Terres and Alexey Chernikov, TUD Dresden University of Technology, Germany; Swagata Acharya and Rupert Huber, National Renewable Energy Laboratory, United States; Dimitar Pashov, King’s College London, United Kingdom; Mikhail I. Katsnelson, Radboud University, Netherlands; Kseniia Mosina and Mark van Schilfgaarde, University of Chemistry and Technology Prague, Czech Republic; and Zdenek Sofer, University of Regensburg, Germany. Dirnberger is also affiliated with the Technical University of Munich. Kim is also affiliated with the University of Arizona. Mills is also affiliated with the Flatiron Institute. A full list of the authors and their affiliations may be found in the paper.

The U.S. Department of Energy, the European Research Council, the U.S National Science Foundation, the Würzburg-Dresden Cluster of Excellence on Complexity and Topology in Quantum Matter and the Emmy Noether Program supported this work.

$3M grant targets integrated semiconductor for smarter, greener electronics

A joint project will aim to develop a novel method for integrating gallium nitride, a high-performance semiconductor material, with silicon substrates. Gallium nitride provides superior performance and faster switching speeds for power-intensive applications, while silicon offers scalability and affordability.

The microelectronics industry is nearing a tipping point. The silicon chips at the heart of everyday electronic devices are running into performance limits, raising the need for new materials and technologies to continue making faster, more efficient devices.  

To help address this challenge, researchers at Penn State will receive $3 million from the Defense Advanced Research Projects Agency (DARPA) as part of a larger grant awarded to Northrop Grumman, a defense, aerospace and technology company. The joint project will aim to develop a novel method for integrating gallium nitride, a high-performance semiconductor material, with silicon substrates. Gallium nitride provides superior performance and faster switching speeds for power-intensive applications, while silicon offers scalability and affordability. According to the researchers, this hybrid approach can lead to more efficient power electronics with lower production costs, making them ideal for high-demand applications like electric vehicles, power electronics and data centers, where efficiency and durability are critical. 

“Silicon is the common platform for microelectronics but it is challenging to combine new semiconductor materials with silicon,” said Joan Redwing, distinguished professor of materials science and engineering and director of the Penn State Materials Research Institute’s (MRI) Two-Dimensional Crystal Consortium, a U.S. National Science Foundation Materials Innovation Platform and national user facility. “To overcome this, we need new approaches to densely integrating advanced materials with silicon, and that is exactly what this project is about. Our work with Northrop Grumman is designed to explore integrating gallium nitride directly onto silicon using two-dimensional materials as interlayers.” 

To achieve this, Penn State will work with Northrop Grumman on heterogeneous integration, a process in which materials with distinct and different properties are combined to create more efficient devices. For this project, the researchers will work to integrate gallium nitride with silicon.  

Gallium nitride is a wide bandgap semiconductor, meaning it can withstand higher electric fields and sustain higher voltages and temperatures. Silicon is a lower bandgap semiconductor, but it’s cheaper and benefits from the well-established silicon manufacturing infrastructure. Combining gallium nitride’s ability to handle high voltages and high switching speeds with silicon’s wide use in digital electronics will create chips that leverage the strengths of both materials.   

“Data centers are expected to need 160% more power by 2030, largely because of the growing use of artificial intelligence,” said Joshua Robinson, professor of materials science and engineering and Penn State’s principal investigator on the DARPA project. “Our work could help reduce that energy demand and contribute to a more sustainable future.” 

The team’s work could also lead to smaller, faster and more efficient power electronics, which manage electricity flow in everything from smartphones to washing machines. For consumers, this would mean reduced energy bills and devices that generate less heat. 

A potential hurdle is traditional methods of integrating gallium nitride with silicon can be complex and costly, often requiring interlayers that introduce thermal resistance and limit device performance. With the DARPA grant, Penn State researchers aim to develop a novel solution using 2D materials that are one to a few atoms thick, such as molybdenum disulfide and gallium selenide, as "seed layers" to grow gallium nitride on industry compatible silicon (001). Silicon (001) is the preferred crystal orientation used in current semiconductor technology. A seed layer provides a template or foundation that influences the structure, orientation and quality of the material grown on top. 

“The current approach to gallium-nitride-on-silicon integration has too many drawbacks, from increased thermal resistance to device fabrication challenges on silicon (001),” Robinson said. “By using 2D materials as seed layers, we aim to eliminate these issues and develop a direct route to integrating gallium-nitride-on-silicon with improved performance compared to current technologies. This could directly impact manufacturing costs and enable market entry into energy-efficient devices.” 

According to Robinson, Penn State’s leadership in 2D materials and advanced manufacturing uniquely positions the University to tackle this challenge and makes it an ideal partner for a major company like Northrop Grumman. The project will leverage the state-of-the-art infrastructure for growing and characterizing 2D materials and wide bandgap semiconductors at Penn State 

“This program allows us to demonstrate that 2D materials could be key to enabling advances in 3D semiconductors,” Robinson said. “We’re combining our expertise in 2D research with the real-world need for improved semiconductor performance, setting the stage for years of innovation in heterogeneous integration.”  

The equipment and methodologies developed through this grant will be available to other researchers through MRI’s user facilities, Robinson said, with the goal of fostering collaboration and innovation among a variety of partners.   

“This grant strengthens Penn State’s role as a leader in semiconductor research,” Redwing said. “It also demonstrates the value of partnerships between academia, industry and government in solving complex challenges.” 

Adri van Duin, distinguished professor of mechanical engineering, of chemical engineering, of engineering science and mechanics, of chemistry and of materials science and engineering, and Rongming Chu, professor of electrical engineering, are also participating in the DARPA project.