EAGER: Exploring the Negative Capacitance Effect from Hf-Based Ferroelectrics and 2D Nanomaterials for Low-Voltage Transistors

Duke University, Durham NC

Investigators

Abstract

Transistors are the heartbeat of every electronic device, providing the connection between users and their data. As the size of transistors has continued to shrink based on Moore's Law, their use by the billions in computer chips has correspondingly soared. Amidst all of this scaling, the operating voltage has remained essentially constant for more than a decade, having major ramifications on the amount of electrical power used for computing. In this project, a new type of transistor will be demonstrated that brings together two-dimensional (2D) nanomaterials and ferroelectrics. The nanomaterials are superb electronic materials for very small devices and the ferroelectrics make possible a phenomenon known as negative capacitance. It is hypothesized that the marriage of these two advances will yield transistors that can operate at extremely low voltages. Such low voltage transistors could usher in a new era of "More than Moore" computing, such as reducing the power consumption in data centers, which are run by high-performance transistors, to address the national need for energy efficiency by offering a "green" data center solution. Additionally, there will be scientific learning regarding the use of ferroelectrics with nanomaterials that will drive a variety of new research directions. This project will also be impactful in promoting educational diversity, as a female graduate student along with a Latino graduate student, who is a NSF Graduate Research Fellow, will carry it out. Further, the extensive interest in nanomaterials among high school and undergraduate students makes this project ideal for attracting involvement of other underrepresented minorities around Duke through both established and new outreach programs. One promising option for overcoming the significant power problem in silicon transistors is to use the negative capacitance (NC) behavior of ferroelectric (FE) insulators to lower the operating voltage by amplifying the applied gate potential. In the last three years, a substantial uptick in research activity involving NC field-effect transistors (NC-FETs) has occurred, largely due to the demonstration that strong FE behavior can be achieved in appropriately synthesized, hafnium zirconium oxide (HfZrO2) thin films. However, NC-FETs with bulk semiconductor channels suffer from poor interface quality to the FE, a lack of device scalability due to thick FE layers, and a voltage-dependent substrate capacitance that results in drive currents that are orders of magnitude lower than traditional transistors. In this project, a possible solution to these challenges for extremely low-voltage NC-FETs will be explored by using the unique electrical and structural properties of 2D MoS2 in the gate stack to yield 2D NC-FETs. The primary goal of the project will be to demonstrate that the unique combination of a 2D channel with HfZrO2 ferroelectrics can yield sub-60 mV/decade switching (operation below the thermal limit). Several distinct approaches will be studied, including advancements in the atomic layer deposition synthesis and capping of the ferroelectric as well as transfer and contacting strategies for the 2D channel, in order to realize a 2D NC-FET that exhibits gate voltage amplification. While there is considerable work in the field on NC-FETs as well as transistors from 2D MoS2, bringing these distinct areas of research together in this device is highly distinct and potentially disruptive. Successful demonstration of a low power, high-performance 2D NC-FET will open the way for more extensive studies and programs that research the unique aspects of this device in greater depth.