• PDF / 1,433,301 Bytes
• 8 Pages / 612 x 792 pts (letter) Page_size
• 48 Downloads / 241 Views

DOWNLOAD

REPORT

---

many research efforts on the basic device physics of CNTs. The status of nanotube electronics was reviewed previously,3 and since then, a large number of advances have taken place. Most notably, the demonstrations of photoluminescence4 and electroluminescence5 reported respectively in 2002 and in 2003 have further expanded CNT research into photonics and optoelectronics. In the present article, we examine some of the developments in the area of the nanotube field-effect transistors (FETs) and highlight some of the unsolved problems and new opportunities that still exist. Interested readers can find more information in recent reviews, where more detailed discussions and additional references are given.2–10

Nanotube Electronics and Photonics

Phaedon Avouris and Richard Martel

Single-Walled Carbon Nanotube Transistors

Abstract In electronics and photonics, intrinsic properties of semiconducting materials play a dominant role in achieving high-performance devices and circuits. In this respect, carbon nanotubes are prime candidates because of their exceptionally high carrier mobility, low capacitance, and strong optical response (direct bandgap). Although these properties compare very favorably with those of crystalline silicon, several issues related to their synthesis, processing, and assembly have challenged efforts for making electronic and photonic devices. Tremendous progress, nevertheless, has been achieved over the years, and much has been learned from novel photonic devices and electronic circuits. We review some of the developments in nanotube transistor performance optimization, ac operation, nanotube circuits, self-assembly, thin-film devices, and nanotube optical devices such as light emitters and detectors. We also examine the issues and opportunities that still exist.

One of the most important CNT devices, the CNT-FET, is composed of an individual SWNT covered by metal electrodes at both ends and by a gate stack made of a metal electrode and a dielectric layer (Figure 1). The first reported CNT-FETs in 1998 were based on a simple design involving the substrate (typically doped Si a

Source

Top gate High-κ insulator Nanotube SiO2

Drain

Si substrate

Introduction

306

can extract many of the electronic properties of CNTs using a simple tight-binding graphene sheet model and the proper boundary conditions.2 This model predicts a metallic or a semiconducting electronic structure depending on the CNT chirality. For the latter case, the electrical bandgap, Eg, scales with the inverse of the diameter. These predictions have been confirmed experimentally in many different ways.2 Typical sources of SWNTs found on the market provide materials having a mixture of semiconducting (Eg of the order of 1 eV) and metallic species. It was also shown that CNT materials display outstanding electrical, mechanical, and thermal properties, such as long mean free paths, high mechanical strength, flexibility, and excellent thermal conductivity.2 Because of these properties, CNTs are perceived as promising materials for f