100-year-old material key for next-generation computer chips

100-year-old material key for next-generation computer chips

Silicon-based computer chips that power modern devices It requires a huge amount of energy to operate. Although computing efficiency is constantly improving, information technology is projected to consume approximately 25% of all primary energy generated by 2030. Researchers in the microelectronics and materials science communities are looking for ways to sustainably manage the global needs of computing power.

Scientists turn 100-year-old materials into thin films for next-generation memory and logic devices

Electron micrographs show the atomic-by-atomic structure of a barium titanate (BaTiO3) thin film, sandwiched between layers of strontium ruthenate (SrRuO3) metal to form a small condenser. (Credit: Lane Martin / Berkeley Lab)

The holy grail to reduce this digital demand is to develop microelectronics that operate at much lower voltages. This is the main goal of our efforts to surpass today’s state-of-the-art CMOS (Complementary Metal Oxide Semiconductor) with less energy required. device.

There are non-silicon materials with attractive properties for memory and logic devices. However, their common bulk format still requires a large voltage to operate and is not compatible with modern electronics. Designing thin film alternatives that not only work well at low operating voltages but can also be packed into microelectronic devices remains a challenge.

Currently, a team of researchers at Lawrence Berkeley National Laboratory (Berkeley Laboratory) and the University of California, Berkeley are energy efficient by synthesizing thin-layer versions of well-known materials with the properties needed for next-generation devices. I have identified one route. ..

Barium titanate (BaTiO) first discovered over 80 years ago3) Used in electronic circuits, ultrasonic generators, transducers, and even various capacitors for sonar.

The crystal of the material reacts quickly to a small electric field, reversibly and permanently reversing the orientation of the charged atoms that make up the material, even when the applied electric field is removed. This allows you to switch between the logic and memory storage device proverbial “0” and “1” states, which requires a voltage in excess of 1,000 millivolts (mV).

Seeking to take advantage of these properties for use in microchips, a Berkeley Lab-led team has developed a pathway for making BaTiO films.3 With a thickness of only 25 nanometers (less than 1/1000 of the width of human hair), the orientation of charged atoms, or polarization, switches as quickly and efficiently as the bulk version.

“We know about BaTiO3 For most of the century, we have known for over 40 years how to make thin films of this material. But until now, we haven’t been able to make a film that can approach the structure and performance that can be achieved in large numbers, “said Lane Martin, a faculty member and professor of materials science at the Berkeley Institute’s Materials Sciences Department (MSD). I am saying. Engineering at the University of California, Berkeley, who led the work.

Historically, synthetic attempts have resulted in films containing higher concentrations of “defects” (where the structure differs from the ideal version of the material) compared to the bulk version. Such high concentrations of defects adversely affect the performance of thin films. Martin and his colleagues have developed an approach to growing films that limit these flaws.The survey results were published in the journal Nature Materials.

Understand what it takes to produce the best low defect BaTiO3 For thin films, researchers turned to a process called pulsed laser deposition. Fires a powerful beam of UV laser light onto a BaTiO ceramic target3 The material is converted to plasma, which transfers atoms from the target to the surface to grow the membrane. “It’s a versatile tool that allows you to fine-tune many knobs in film growth and see which ones are most important for controlling properties,” says Martin.

Martin and his colleagues have shown that their method can precisely control the structure, chemistry, thickness, and interface with metal electrodes of the deposited membrane. By chopping each deposited sample in half and examining its atomic-by-atom structure using tools from the National Electron Microscopy Center at Berkeley Lab’s Molecular Foundry, researchers accurately mimic very thin slices of bulk. Revealed the version that was used.

“It’s fun to think that we can take these classic materials that we thought we all knew and turn them over with a new approach to making and characterizing them,” Martin said. Told.

Finally, by placing a film of BaTiO3 Martin and his team created a small capacitor between the two metal layers, an electronic component that rapidly stores and releases energy in a circuit. When a voltage of 100 mV or less was applied and the generated current was measured, it was found that the polarization of the film switched within one billionth of a second, which could be faster. This conflicts with what today’s computers need to access memory and perform calculations.

This task follows the major goal of creating materials with low switching voltages and investigating how the interface with the metal components required for the device affects such materials. “This is a good early victory in the pursuit of low-power electronics that goes beyond what is possible with today’s silicon-based devices,” Martin said.

“Unlike our new devices, the capacitors used in today’s chips do not hold data unless they continue to apply voltage,” Martin said. Also, current technologies typically operate at 500-600 mV, while thin film versions can operate at 50-100 mV or less. Taken together, these measurements show successful optimization of voltage and polarization robustness. This tends to be a trade-off, especially for thin materials.

Next, the team plans to shrink the material even thinner to make it compatible with the actual devices in the computer and study how it works with those small dimensions. At the same time, we will work with collaborators from companies such as Intel Corp. to test the feasibility of first-generation electronic devices. “Think about how much energy you can save if you can make each logical operation on your computer a million times more efficient. That’s why we’re doing this,” Martin said. ..

This study was supported by the Department of Science (DOE) Science Department. Molecular Foundry is a user facility of the DOE Science Department at Berkeley Labs.

Berkeley Lab’s “Beyond Moore’s Law” initiative aims to pinpoint the path of memory elements to ultra-low power logic. “Since it is low-voltage operation that scales energy, we need to move to low-voltage operation,” said co-author, Senior Scientist at the Berkeley Institute, Physics and Materials Science at the University of California, Berkeley. And Lama Mosie Ramesh, a professor of engineering, said. “This study demonstrated for the first time the switching field of the model material BaTiO.3 On related platforms, at voltages below 100 mV. “

Article courtesy of Lawrence Berkeley National Laboratory (Berkeley Laboratory).

Related: Global supply shortage: no time, no chips — no problem for NREL

Featured Photo: Unsplash’s Laura Ockel Silicon Wafer Macro


 

Check out our brand new E-BikeGuide.. If you are interested in electric bikes, this is the perfect place to start your e-mobility journey!


 

Thank you for CleanTechnica’s originality and cleantech news coverage? Consider becoming a member, supporter, technician, ambassador, or Patreon patron of CleanTechnica.


 

Do you have CleanTechnica tips, want to promote, or suggest guests for the CleanTech Talk podcast? Please contact us from here.

advertisement