Unlocking the Power of Ferroelectrics in Microelectronics
Ferroelectricity was discovered one century ago in Rochelle salt by J. Valasek. Today, ferroelectrics are in commercial use for capacitors, sensors, actuators, energy harvesters, electro-optics, and nonvolatile memory. It is this last application that underpins the Center for Three Dimensional Ferroelectric Microelectronics (3DFeM). As one of the 41 active Energy Frontier Research Centers (EFRCs) funded by U.S. Department of Energy’s Office of Basic Energy Science, 3DFeM aims to exploit the 3rd dimension in microelectronics to enable closely interconnected memory and logic devices. Three dimensional integration of nonvolatile memory should enable a significantly lower energy burden for processing, while simultaneously removing a major source of latency in computation.
Just like our beloved planet Earth and the toy magnets we played with in primary school, tiny domains in ferroelectric materials have two poles, or more specifically a dipole moment—but in ferroelectric materials it is an electric dipole instead of a magnetic dipole. The ferroelectric dipole is realized via a non-centrosymmetric crystal structure with the separation of a positive charge center and a negative charge center, making each unit cell a tiny dipole moment with one side positive and the opposite side negative (as schematically shown in Figure 1a). Switching the position of ions in the lattice between two or more crystallographically defined states leads to a reorientation of the dipole moment, namely, polarization switching (Figure 1a). These dipole moments can line up to form polar clusters. In ferroelectric materials, these clusters are called ferroelectric domains, and each domain is a macroscopic polarization that is the sum of microscopic dipole moments of unit cells (Figure 1b). Ferroelectric domains can be predominately aligned in a direction by applying a strong enough electric field, as shown in Figure 1c from state I to state II. Likewise, applying a strong enough opposite electric field can reverse the ferroelectric polarization to the opposite direction. This polarization reversal is observable by measuring the polarization-electric field hysteresis loop. As shown in Figure 1c, changing the direction of the electric field will lead to the ferroelectric domain alignment in the opposite direction and the polarization reversal (state II to state III).
Since modern-day electronics function by transferring charges in controlled patterns, one can imagine that ferroelectric materials have immense potential as functional devices. For example, ferroelectric materials can be used in random-access memory (RAM). The dipoles in ferroelectric materials can be polarized up and down by applying an electric field, resulting in a power-efficient binary switch for memory application. Since the orientation of ferroelectric polarization is not affected by power disruption, this makes ferroelectric RAM a reliable nonvolatile memory, which can retain data memory when power is removed or interrupted for up to many decades. In addition, ferroelectric RAM also observes many advantages including low power consumption, fast write speed, and high endurance. In the meantime, the disadvantages of ferroelectric RAM, e.g., high cost and low storage density, can be overcome by developing new ferroelectric thin film materials with low-cost synthesis procedures.
Recently, a research team composed of scientists from the Pennsylvania State University synthesized high quality ferroelectric boron (B) substituted aluminum nitride thin films (Al1-xBxN) that exhibit fascinating ferroelectric, electronic, and optical properties, enabling applications in a broad range of functional devices. The team can grow high quality AlxB1-xN films on (110)W/(001)Al2O3 substrates by dual-cathode reactive magnetron sputtering. It was found that the AlxB1-xN films with x ranging from 0.02 to 0.15 display ferroelectricity, where x decides the ratio of aluminum and boron in the material. The remnant polarization—the polarization displayed by the material when the applied electric field is zero—of AlxB1-xN films can exceed 125 µC cm-2 in materials that simultaneously maintain a bandgap above 5.2 eV. The large band gap prevents the leakage contribution from the polarization switching at low frequencies, expanding the applications of this material. First-principles calculations, corroborating with the structural investigation by x-ray diffraction, offers an understanding of the composition-dependent ferroelectric properties of AlxB1-xN. This important research lays the groundwork of future development of electronic devices based on this material.
Hayden, John, et al. "Ferroelectricity in boron-substituted aluminum nitride thin films." Physical Review Materials 5.4 (2021): 044412. https://doi.org/10.1103/PhysRevMaterials.5.044412
The authors thank the Defense Advanced Research Projects Agency (DARPA) via Army Research Office (ARO) Grant No.W911NF-20-2-0274, and the Center for Dielectrics and Piezoelectrics (CDP) through National Science Foundation (NSF) Grants No. IIP-1841453 and No. IIP-1841466, for supporting the experimental portions of this work. The first-principles calculations were supported as part of the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences Energy Frontier Research Centers program under Award No. DE-SC0021118.
The author of this newsletter, Yongtao Liu, acknowledges the support of the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences Energy Frontier Research Centers program under Award No. DE-SC0021118. Yongtao Liu also acknowledges suggestions and comments on this newsletter from Susan E Trolier-Mckinstry.