Current status and challenges of aggressive equivalent-oxide-thickness (EOT) scaling of high-κ gate dielectrics via higher-κ ( > 20) materials and interfacial layer (IL) scavenging techniques are reviewed. La-based higher-κ materials show aggressive EOT scaling (0.5–0.8 nm), but with effective workfunction (EWF) values suitable only for n-type field-effect-transistor (FET). Further exploration for p-type FET-compatible higher-κ materials is needed. Meanwhile, IL scavenging is a promising approach to extend Hf-based high-κ dielectrics to future nodes. Remote IL scavenging techniques enable EOT scaling below 0.5 nm. Mobility-EOT trends in the literature suggest that short-channel performance improvement is attainable with aggressive EOT scaling via IL scavenging or La-silicate formation. However, extreme IL scaling (e.g., zero-IL) is accompanied by loss of EWF control and with severe penalty in reliability. Therefore, highly precise IL thickness control in an ultra-thin IL regime ( < 0.5 nm) will be the key technology to satisfy both performance and reliability requirements for future CMOS devices.
Dense crossbar arrays of non-volatile memory (NVM) can potentially enable massively parallel and highly energy-efficient neuromorphic computing systems. The key requirements for the NVM elements are continuous (analog-like) conductance tuning capability and switching symmetry with acceptable noise levels. However, most NVM devices show non-linear and asymmetric switching behaviors. Such non-linear behaviors render separation of signal and noise extremely difficult with conventional characterization techniques. In this study, we establish a practical methodology based on Gaussian process regression to address this issue. The methodology is agnostic to switching mechanisms and applicable to various NVM devices. We show tradeoff between switching symmetry and signal-to-noise ratio for HfO2-based resistive random access memory. Then, we characterize 1000 phase-change memory devices based on Ge2Sb2Te5 and separate total variability into device-to-device variability and inherent randomness from individual devices. These results highlight the usefulness of our methodology to realize ideal NVM devices for neuromorphic computing.
We developed an Al2O3/HfO2 bi-layer gate dielectric with an in-situ O3 treatment for interface state density (Dit) and gate leakage current density (Jg) reductions on SiGe channels. We observed Ge-content dependent equivalent oxide thickness (EOT) scaling and EOT 0.44 nm was achieved with an MOS capacitor with a Si0.05Ge0.95 substrate. The O3 treatment enabled application to non-planar device structures and we demonstrated five orders of magnitude lower off currents (Ioff), a sub-threshold slope of 68 mV/decade, and a very high hole mobility of 457 cm2V-1s-1 at an inversion carrier density (Ninv) of 1×1013 cm-2 for asymmetrically strained SiGe PMOSFETs with Ge% of 65%-70%.
Analog hardware-based training provides a promising solution to developing state-of-the-art power-hungry artificial intelligence models. Non-volatile memory hardware such as resistive random access memory (RRAM) has the potential to provide a low power alternative. The training accuracy of analog hardware depends on RRAM switching properties including the number of discrete conductance states and conductance variability. Furthermore, the overall power consumption of the system inversely correlates with the RRAM devices conductance. To study material dependence of these properties, TaOx and HfOx RRAM devices in one-transistor one-RRAM configuration (1T1R) were fabricated using a custom 65 nm CMOS fabrication process. Analog switching performance was studied with a range of initial forming compliance current (200–500 µA) and analog switching tests with ultra-short pulse width (300 ps) was carried out. We report that by utilizing low current during electroforming and high compliance current during analog switching, a large number of RRAM conductance states can be achieved while maintaining low conductance state. While both TaOx and HfOx could be switched to more than 20 distinct states, TaOx devices exhibited 10× lower conductance, which reduces total power consumption for array-level operations. Furthermore, we adopted an analog, fully in-memory training algorithm for system-level training accuracy benchmarking and showed that implementing TaOx 1T1R cells could yield an accuracy of up to 96.4% compared to 97% for the floating-point arithmetic baseline, while implementing HfOx devices would yield a maximum accuracy of 90.5%. Our experimental work and benchmarking approach paves the path for future materials engineering in analog-AI hardware for a low-power environment training.
7 September 2017 in Progress of Theoretical and Experimental Physics
Spallation reactions for the long-lived fission product $$^{93}$$Zr have been studied in order to provide basic data necessary for nuclear waste transmutation. Isotopic-production cross sections via proton- and deuteron-induced spallation reactions on $$^{93}$$Zr at 105 MeV/nucleon were measured in inverse kinematics at the RIKEN Radioactive Isotope Beam Factory. Remarkable jumps in isotopic production originating from the neutron magic number $$N=50$$ were observed in Zr and Y isotopes. The experimental results were compared to the PHITS calculations considering both the intranuclear cascade and evaporation processes, and the calculations greatly overestimated the measured production yield, corresponding to few-nucleon-removal reactions. The present data suggest that the spallation reaction is a potential candidate for the treatment of $$^{93}$$Zr in spent nuclear fuel.