Brings novel insights to a vibrant research area with high application potential - covering materials, physics, architecture, and integration aspects of future generation CMOS electronics technology
Over the last four decades we have seen tremendous growth in semiconductor electronics. This growth has been fueled by the matured complementary metal oxide semiconductor (CMOS) technology. This comprehensive book captures the novel device options in CMOS technology that can be realized using non-silicon semiconductors. It discusses germanium, III-V materials, carbon nanotubes and graphene as semiconducting materials for three-dimensional field-effect transistors. It also covers non-conventional materials such as nanowires and nanotubes. Additionally, nanoelectromechanical switches-based mechanical relays and wide bandgap semiconductor-based terahertz electronics are reviewed as essential add-on electronics for enhanced communication and computational capabilities.
Advanced Nanoelectronics: Post-Silicon Materials and Devices begins with a discussion of the future of CMOS. It continues with comprehensive chapter coverage of: nanowire field effect transistors; two-dimensional materials for electronic applications; the challenges and breakthroughs of the integration of germanium into modern CMOS; carbon nanotube logic technology; tunnel field effect transistors; energy efficient computing with negative capacitance; spin-based devices for logic, memory and non-Boolean architectures; and terahertz properties and applications of GaN.
-Puts forward novel approaches for future, state-of-the-art, nanoelectronic devices
-Discusses emerging materials and architectures such as alternate channel material like germanium, gallium nitride, 1D nanowires/tubes, 2D graphene, and other dichalcogenide materials and ferroelectrics
-Examines new physics such as spintronics, negative capacitance, quantum computing, and 3D-IC technology
-Brings together the latest developments in the field for easy reference
-Enables academic and R&D researchers in semiconductors to "think outside the box" and explore beyond silica
An important resource for future generation CMOS electronics technology, Advanced Nanoelectronics: Post-Silicon Materials and Devices will appeal to materials scientists, semiconductor physicists, semiconductor industry, and electrical engineers.
Muhammad Mustafa Hussain, PhD, is Professor in the Electrical Engineering program of King Abdullah University of Science and Technology (KAUST), Saudi Arabia. Before joining KAUST, he was Program Manager of Emerging Technology Program in SEMATECH, Austin. His program was funded by DARPA NEMS, CERA and STEEP programs. His research interest is in expanding the horizon of CMOS electronics and technology for future applications.
The Future of CMOS: More Moore or a New Disruptive Technology?
Nazek El-Atab and Muhammad M. Hussain
King Abdullah University of Science and Technology, Integrated Nanotechnology Lab, Thuwal, 4700, Saudi Arabia
For more than four decades, Moore's law has been driving the semiconductor industry where the number of transistors per chip roughly doubles every 18-24?months at a constant cost. Transistors have been relentlessly evolving from the first Ge transistor invented at Bell Labs in 1947 to planar Si metal-oxide semiconductor field-effect transistor (), then to strained SiGe source/drain () in the 90- and 65-nm technology nodes and high-?/metal gate stack introduced at the 45- and 32-nm nodes, then to the current 3D transistors (Fin field-effect transistors (s)) introduced at the 22-nm node in 2011 (Figure 1.1). In extremely scaled transistors, the parasitic and contact resistances greatly deteriorate the drive current and degrade the circuit speed. Thus, miniaturization of devices so far has been possible due to changes in dielectric, S/D, and contacts materials/processes, and innovations in lithography processes, in addition to changes in the device architecture , .
Figure 1.1 Intel innovation in process technology for the past decade.
The gate length of current transistors has been scaled down to 14?nm and below, with over 109 transistors in state-of-the-art microprocessors. Yet, the clock speed is limited to 3-4?GHz due to thermal constraints, and further scaling down the device dimensions is becoming extremely difficult due to lithography challenges. In addition, further scaling down the complementary metal-oxide semiconductor () technology is leading to larger interconnect delay and higher power density [. The complexity of physical design is also increasing with higher density of devices. So, what is next?
A promising More-than-Moore technology is the 3D integrated circuits (s) which can improve the performance and reduce the intra-core wire length, and thereby enable high transfer bandwidth with reduced latencies and power consumption, while maintaining compact packing densities [. Alternative technologies that could be promising for new hardware accelerators include resistive computing, neuromorphic computing, and quantum computing.
Resistive computing could lead to non-von Neumann (VN) computing and enforce reconfigurable and data-centric paradigms due to its massive parallelism and low power consumption [. Moreover, humans can easily outperform current high-performance computers in tasks like auditory and pattern recognition and sensory motor control. Thus, neuromorphic computing can be promising for emulating such tasks due to its energy and space efficiency in artificial neural network applications [. Quantum computing can solve tasks that are impossible by classical computers, with potential applications in encryptions and cryptography, quantum search, and a number of specific computing applications [.
In this chapter, four main technologies are discussed: FinFET, 3D IC, neuromorphic computing, and quantum computing. The state-of-the-art findings and current industrial state in these fields are presented; in addition, the challenges and limitations facing these technologies are discussed.
1.1 FinFET Technology
Over the past four decades, the continuous scaling of planar MOSFETs has provided an improved performance and higher transistor density. However, further scaling down planar transistors in the nanometer regime is very difficult to achieve due to the severe increase in the leakage current I off. In fact, as the channel length in planar MOSFETs is reduced, the drain potential starts to affect the electrostatics in the channel and, consequently, the gate starts to lose control over the channel, which leads to increased leakage current between the drain and source. A higher gate-channel capacitance can relieve this problem using thinner and high-? gate oxides; however, the thickness of the gate oxide is fundamentally restricted by the increased gate leakage and the gate-induced-drain leakage effect -.
An alternative to planar MOSFETs is the multiple-gate FETs (s) which demonstrate better electrostatics and better screening of the drain from the gate due to the additional gates covering the channel -. As a result, MuGFETs show better performance in terms of subthreshold slope, threshold voltage (V t) roll-off, and drain-induced barrier lowering (). Another alternative to planar bulk MOSFETs is fully depleted silicon on insulator () MOSFETs, which reduce leakage between drain and source due to the removal of the substrate right below the channel [. The performance of the FDSOI MOSFETs is comparable with the double-gate field-effect transistors (s) in terms of SS, low junction capacitance, and high I on/I off ratio. Yet, the DGFETs have better scalability and can be manufactured on bulk Si wafers instead of silicon-on-insulator () wafers, which makes them more promising [.
FinFETs or tri-gate FETs, which have three gates, have been found to be the most promising alternatives to MOSFETs due to their enhanced performance and simplicity of the fabrication process, which is compatible with and can be easily integrated into standard CMOS fabrication process (Figure 1.2) , . In fact, an additional selective etch step is required in the FinFET fabrication process in order to create the third gate on top of the channel. FinFET devices have been explored carefully in the past decade. A large number of research articles that confirmed the improved short-channel behavior using different materials and processes have been published, as is shown in the following section. Next, the industrial state of FinFETs, their challenges, and limitations are discussed.
Figure 1.2 TEM image of Intel's 14-nm transistors with sub-40-nm fin pitch.
1.1.1 State-of-the-Art FinFETs
18.104.22.168 FinFET with Si Channel
In the semiconductor industry, silicon is the main channel material. The first FinFET technology (22-nm node) was produced by Intel in 2011. The second FinFET generation (14-nm node) published by Intel used strained Si channel [. The gate length was scaled from 26 to 20?nm in the second FinFET generation, which was possible due to new sub-fin doping and fin profile optimization. With a V DD of 0.7?V, the saturation drive current is 1.04?mA?µm-1 and the off current is 10?nA?µm-1 for both nMOSFET () and pMOSFET (). The SS is ~65?mV/decade, while the DIBL for N/PMOS is ~60/75?mV?V-1. High-density static random access memory () having 0.0588?µm2 cell size are also reported and fabricated using the 14-nm node. More recently, a research group from Samsung published a 7-nm CMOS FinFET using extreme ultraviolet () lithography instead of multiple-patterning lithography. This resulted in a reduction of the needed mask steps by more than 25%, in addition to providing smaller critical dimension variability and higher fidelity. The FinFET presented in this work consumes 45% less power and provides 20% faster speed than in the previous 10-nm technology. The reported SS is 65 and 70?mV/decade, and the DIBL is 30 and 45?mV?V-1 for NMOS and PMOS, respectively. A 6T high-density and high-current SRAM memory has also been demonstrated using the 7-nm FinFET, and the results show a reduction in the bit line capacitance by 20% as a result of the reduction in the parasitic capacitance.
22.214.171.124 FinFET with High-Mobility Material Channel
The III-V materials gained growing attention for adoption as the channel material due to their promising characteristics such as high mobility, small effective mass, and, therefore, high injection velocity, in addition to near-ballistic performance. The first InGaSb pFET was demonstrated by Lu et al. [, where a fin-dry etch technique was developed to obtain 15-nm narrow fins with vertical sidewalls. An equivalent oxide thickness () of 1.8?nm of Al2O3 was used as the gate oxide. The authors also demonstrated Si-compatible ohmic contacts that yielded an ultralow contact resistivity of 3.5?×?10-8 O?cm2. Devices with L g =?100?nm and different fin widths (W f) were demonstrated. The results show that with W f =?100?nm, g m of 122?µS?µm-1 is achieved; while with W f...