UC Berkeley

The critical size limit of electric polarization remains a open domain in nanoscale ferroelectric research; in particular, the fundamental limit of switchable electric dipoles has extensive implications for the scaling of energy-efficient nanoelectronics due to the intrinsic ability to control electric polarization with an applied voltage. As ferroelectric materials are made thinner however, polarization is typically suppressed, as size effects in ferroelectrics have been thoroughly investigated in the prototypical perovskite oxides. Furthermore, perovskites suffer from various chemical, thermal, electrical, and interfacial incompatibilities with silicon and modern semiconductor processes. Recently, fluorite-structure binary oxides have attracted considerable interest as they overcome many of the silicon compatibility issues afflicting its perovskite ferroelectric counterparts. To address the missing void in ferroelectric research regarding ultrathin stabilization and Si compatibility, the first part of this dissertation focuses on stabilizing ferroelectric order in ultrathin HfO₂-ZrO₂ fluorite-structure oxides, grown by atomic layer deposition, on silicon. Indeed, this research demonstrates the persistence of inversion symmetry breaking and spontaneous, switchable polarization down to a thickness of one nanometer in Zr:HfO₂ films and five angstroms in ZrO₂ films, equivalent to just two and one fluorite-structure unit cell thickness, respectively; the latter result marks the thinnest demonstration of ferroelectric polarization switching in any material system, remarkably in a conventionally paraelectric material. Reversible electric dipoles at the 2D limit offers exciting possibilities for low-power high-density memory exploiting electronic order on silicon, evidenced by proof-of-principle ferroelectric tunnel junctions implementing 10 Å Zr:HfO₂ and 5 Å ZrO₂ barriers which demonstrate distinct polarization-driven resistive states. Furthermore, the emergence of atomic-scale ferroelectricity and switchable electric dipoles in fluorite-structure oxides not only cements its technological promise for silicon electronics, but also highlights its distinct nature from a fundamental ferroic perspective. These results indicate the presence of ultrathin-enhanced polar distortion in fluorite oxide thin films on silicon, which oppose conventional perovskite ferroelectric trends; notably these ‘reverse' size effects are consistent with predictions of its atypical piezoelectric origins. Therefore, simple binary oxides can serve as a model material system to explore novel 2D ferroelectricity in which ‘reverse’ size effects can counter-intuitively stabilize polar symmetry in the ultrathin regime.

Considering integrate nanoscale ferroelectricity in HfO₂-ZrO₂-based thin films can be synthesized directly on silicon, the next prong of this dissertation focuses on the atomic-scale design of negative capacitance – a novel phenomena present in ferroelectric materials which can enable energy-efficient nanoscale transistors – in ultrathin (< 2 nm) HfO₂-ZrO₂ films. The negative capacitance effect from a ferroelectric layer has a unique experimental signature: it results in a net capacitance enhancement of a dielectric-ferroelectric heterostructure over that of the constituent dielectric layer, which violates the expected equivalent capacitance of two capacitors connected in series. With the scaling of lateral dimensions in advanced transistors, an increased gate capacitance is desirable both to retain the control of the gate electrode over the channel and to reduce the operating voltage, enabling high-current and low-power operation. This led to the adoption of high-κ dielectric HfO₂ in the gate stack in 2008, which remains as the material of choice to date. Unfortunately, stabilizing negative capacitance is not as simple as stabilizing ferroelectricity, and until now, it has yet to be reported in the technologically-relevant HfO₂-ZrO₂ system. The second half of this dissertation demonstrates evidence of negative capacitance via capacitance enhancement in ultrathin 2 nm HfO₂-ZrO₂ multilayers and 1 nm ZrO₂ single layers on SiO₂-buffered Si. This atomic-scale design of negative capacitance builds off the previous materials characterization of ultrathin ferroelectricity in Zr:HfO₂ and ZrO₂; in particular, the antiferroelectric- ferroelectric phase evolution in ZrO₂ as a function of thickness enables mixed-ferroic phase competition at the ultrathin regime, the microscopic origin of negative capacitance in this system. Beyond just the stabilization of negative capacitance, this research also delves into how negative capacitance gate oxide stacks demonstrate key device-level benefits once integrated into Si transistors, providing a new path towards advanced gate oxide stacks beyond the conventional HfO₂-based high-κ dielectrics.

Overall, through the model system of HfO₂-ZrO₂ system on Si, this dissertation establishes new physical thickness limits and effective oxide thickness limits for ferroelectricity and negative capacitance, respectively, and proposes microscopic mechanisms for the observed phenomena.