Reliability of Transition Metal Dichalcogenide Field Effect Transistors

The growth of modern societies is strongly linked to the development of new electronic devices and circuits. Among them, the one that has played the most important role promoting humanity progress is the Field Effect Transistor (FET), which is the basic unit of most electronic circuits and computers. In its humble origins the performance of electronic transistors was very limited, but during the 20th century the scientific community has made them smaller, faster, cheaper and energetically more efficient. Now, the size of FETs is approaching a physical limit, as they are reaching the atomic scale, and this has accelerated the race for finding novel strategies that allow continuing the scaling trend that has prevailed during the last four decades. In this direction, a rising star for device fabrication are two dimensional (2D) materials. Due to their superior properties, these materials can provide very high performance in terms of switching speed, power consumption and durability compared to traditional silicon-based FETs. Moreover, these materials provide additional properties such as flexibility and transparency, opening a new horizon for the development of electronic devices. The first 2D material ever synthesized, which is also the most popular one, is graphene. During the last years a countless amount of electronic devices using graphene have been reported, but now the interest in graphene electronics seems to be losing its momentum. The reason is that, after many studies, academia and industry have realized that graphene presents an essential limitation: it has no bandgap, and it cannot be artificially induced without degrading graphenes genuine properties. That means graphene cannot be efficiently used in logic applications because the power consumption prohibitively increases; and that is a major problem because 90% of electronic devices we daily use require such logic circuits. Fortunately, recent discoveries have allowed the fabrication of semiconductor 2D materials, and among them the family of Transition Metal Dichalcogenides (2D/TMD) shows very promising performance. Therefore, the study of 2D/TMD FETs is going to be a key element in micro/nano electronics research for the next years. Although some seminal research papers have already demonstrated outstanding properties, the research in the field of 2D/TMDs FETs is still in its embryonic stage, and there are still many open questions that need to be addressed, with a most important one concerning their basically unknown reliability. Therefore, the aim of this project is the study of key reliability issues involved in next generation FETs made of advanced 2D/TMDs. With the help of expert partners, we will contribute to the establishment of 2D materials in the semiconductors industry, providing a realistic solution to avoid the stagnation of the scaling down of FETs, and contribute to the technology-driven development of our society.

In the era of digitalization computer chips play an increasingly important role in our daily lives. Computer chips facilitate communication, education, navigation and so many more tasks, thereby creating unimagined opportunities. To give as many people as possible access to these opportunities, the computer chips themselves need to become ever more powerful, faster, cheaper and less energy consuming. At the core computer chips consist of innumerable field effect transistors (FETs) which we studied within this research project. In particular, we investigated a promising approach to build FETs which are in themselves more powerful, cheaper and less energy consuming, namely be reducing the size of the transistors. In order to build small transistors with dimensions at the ultimate physical limits, the transistors need to consist of single atomic layers. This requires materials which are stable in their two dimensional form. A group of materials which satisfies these requirements are the transition metal dichalcogendies (TMDs) on which we focused in this project. At the beginning of the project all available prototypes for TMD based FETs showed very unstable operation and a short lifetime. To identify the physical reasons which render FETs unstable, we combined in this project three different measurement methods. First, the reliability was tested directly on FET prototypes using electrical measurements, second the microscopic reasons of instability were analyzed using conductive atomic force microscopy and third the measurement data was described using physical models and simulations. By combining these three different perspectives on the problem, the root cause of the instability could be identified. FETs often showed unstable operation, because charges become trapped at atomic defects in the insulator. The charges remain trapped for some time until the charge returns back to the channel in a random process. This insight allows to propose ideas which could improve the stability of FETs based on TMDs. The most promising approach we identified within this project aims for replacing the commonly used amorphous insulators with crystalline insulators which contain much fewer atomic defects. In this context hexagonal boron nitride and calcium fluoride are especially well suitable insulators for FETs based on TMDs as semiconducting channel. When using calcium fluoride we could show that the FETs became more stable by more than three orders of magnitude.