28nm High-k Metal Gate Research Articles

Investigation of degradation and recovery characteristics of NBTI in 28-nm high-k metal gate process

Degradation induced by the negative bias temperature instability (NBTI) can be attributed to three mutually uncoupled physical mechanisms, i.e., the generation of interface traps (ΔV IT), hole trapping in pre-existing gate oxide defects (ΔV HT), and the generation of gate oxide defects (ΔV OT). In this work, the characteristic of NBTI for p-type MOSFET fabricated by using a 28-nm high-k metal gate (HKMG) process is thoroughly studied. The experimental results show that the degradation is enhanced at a larger stress bias and higher temperature. The effects of the three underlying subcomponents are evaluated by using the comprehensive models. It is found that the generation of interface traps dominates the NBTI degradation during long-time NBTI stress. Moreover, the NBTI parameters of the power-law time exponent and temperature activation energy as well as the gate oxide field acceleration are extracted. The dependence of operating lifetime on stress bias and temperature is also discussed. It is observed that NBTI lifetime significantly decreases as the stress increases. Furthermore, the decrease of charges related to interface traps and hole detrapping in pre-existing gate oxide defects are used to explain the recovery mechanism after stress.

Statistical Observations of Three Co-Existing NBTI Behaviors in 28 nm HKMG by On-Chip Monitor With Less Recovery Impact

An on-chip digital sensor has been demonstrated in 28nm High-k Metal Gate (HKMG) for bias temperature instability (BTI) statistical characterization with the benefits: fast statistical measurement, less recovery impact (Toff-stress@around 15ns, Fast Period Sampling (FPS) @around 300ns), and high resolution (0.1mV of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\Delta $ </tex-math></inline-formula> Vth). As far as we know, it is the first time to statistically observe the very early stage of trap recovery of individual device in practical scenario, e.g., static random-access memory (SRAM). We find that three Negative BTI (NBTI) recovery behaviors, 2/3/4-step with clear transition slope, co-exist in HKMG devices. Our further analysis ascribes the phenomena to co-existing of four types of defects in 28nm HKMG Devices Under Test (DUTs). Three types are recoverable and one unrecoverable. The transition slope instead of steep drop between steps is the aggregative effects of one certain type of recoverable defect contained across DUTs. More types of defects lead to more Vth shift. But the contribution percentage of unrecoverable defect remains quite close, while recoverable defects dominate the Vth degradation. Only when the Toff-stress is less than the starting point of 1st recover step (within 1 <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\mu \text{s}$ </tex-math></inline-formula> in our case), accurate and consistent Vth degradation data can be achieved.

Demonstration of Multiply-Accumulate Operation With 28 nm FeFET Crossbar Array

This letter reports a linear multiply-accumulate (MAC) operation conducted on a crossbar memory array based on 28nm high-k metal gate (HKMG) Complementary Metal Oxide Semiconductor (CMOS) and ferroelectric field effect transistor (FeFET). The fabrication is conducted at GlobalFoundries with their standard 28nm technology. The crossbar arrays show a 100% yield in MAC operation on a 300mm wafer. The arrays were divided into <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${8}\times {8}$ </tex-math></inline-formula> segments. The FeFET crossbar arrays were fabricated with access transistors, current-limiter transistors and a current-mode analog-to-digital converter (ADC) on the same wafer. Finally, the data retention characteristics reveal excellent data retention characteristics up to <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${5}\times {10} ^{{4}}$ </tex-math></inline-formula> seconds, which makes this memory array suitable for carrying out MAC operations in inference engine applications.

CryoMem: A 4–300-K 1.3-GHz Hybrid 2T-Gain-Cell-Based eDRAM Macro in 28-nm Logic Process for Cryogenic Applications

This letter demonstrates the first CMOS logic compatible cryogenic memory solution operating from 4 to 300 K designed in the 28-nm high-K metal gate (HKMG) CMOS. With the growing applications of cryogenic systems from quantum computing to space electronics, there is a need for memory capable of reliable functionality. While the prevailing low-temperature memories suffer from temperature scalability and integrability, the proposed test chip of 1-kb 2T hybrid gain cell-based embedded DRAM macro overcomes these issues while providing $10^{6} \times $ better retention time, 1.3-GHz peak frequency at 4 K, sub nW/kb array refresh power, and $1.7\times $ energy efficiency at 4 K compared to 300 K. This is due to the near absence of leakage, improved ON current, and subthreshold slope which leads to enhanced performance of critical path circuits at lower temperatures.

Drain–Erase Scheme in Ferroelectric Field-Effect Transistor—Part I: Device Characterization

Ferroelectric-doped HfO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> -based ferroelectric field-effect transistors (FeFETs) are being actively explored as emerging nonvolatile memory devices with the potential for in-memory computing. In this two-part article, we explore the feasibility of FeFET-based 3-D NAND architecture for both in situ training and inference. To address the challenge of erase-by-block in NAND-like structure, we propose and experimentally demonstrate the drain-erase scheme to enable the individual cell's program/erase/inhibition, which is necessary for individual weight update in in situ training. We describe the device characterization of different drain- erase conditions and results in this article. The experimental conditions are characterized on 22-nm fully depleted silicon-on-insulator (FDSOI) and 28-nm high-k metal gate (HKMG) FeFET devices from GLOBALFOUNDRIES. With appropriate biasing, up to 10 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">4</sup> ON/OFF ratio could be achieved by drain-erase. The 3-D NAND array architecture design and verification for in-memory computing will be described in Part II of this article.

Reconfigurable ultra-thin film GDNMOS device for ESD protection in 28 nm FD-SOI technology

We propose a novel ESD protection device (GDNMOS: Gated Diode merged NMOS) fabricated with 28nm UTBB FD-SOI high-k metal gate technology. By modifying the combination of the diode and transistor gate stacks, the robustness of the device is optimized, achieving a maximum breakdown voltage (VBR) of 4.9V. In addition, modifications of the gate length modulate the trigger voltage (Vt1) with a minimum value of 3.5V. Variable electrostatic doping (gate-induced) in diode and transistor body enables reconfigurable operation. A lower doping of the base enhances the bipolar gain, leading to thyristor behavior. This innovative architecture demonstrates excellent capability for high-voltage protection while maintaining a latch-up free behavior.

On the need for a new ESD verification methodology to improve the reliability of ICs in advanced 28nm UTBB FD-SOI technology

The need for a novel multi-scale ESD (ElectroStatic Discharge) network recognition and verification methodology is described in this paper. The proposed solution is used to limit the risk of ESD design errors and to enhance IC reliability, independently of the implemented ESD protection strategy and the type of package assembly technique. This method relies on a topology-aware & graph-based verification paradigm which is generic enough to be usable at every step of the design flow. Its efficiency is illustrated with examples involving custom I/O ring portions in 28nm UTBB FD-SOI High-K metal gate technology.

Effect of Pregate Carbon Implant on Narrow Width Behavior and Performance of High- Metal-Gate nMOS Transistors

This paper experimentally shows the reduction of anomalous narrow width effect (NWE) observed in gate-first high- $k$ metal-gate (HKMG) nMOS transistors by using pregate carbon implants. The experiments are performed with different carbon implant doses and energies, in collaboration with a semiconductor foundry. The 28-nm gate-first HKMG CMOS technology is used as the baseline flow. The physical mechanisms responsible for this improvement are identified and explained in detail. It is further shown that the pregate carbon implant used to suppress the NWE also increases junction leakage, improves the device electrostatics, and improves the universal curve.

Self-Rectifying Twin-Bit RRAM in 3-D Interweaved Cross-Point Array

A new self-rectifying twin-bit RRAM in a novel 3-D interweaved cross-point array has been proposed and demonstrated in 28-nm high-k metal gate CMOS back end of line (BEOL) process. This high density of array architecture with the cell size only $70 \times 100 \times 187$ nm can be manufactured without additional mask or process. The RRAM film is formed by via plug over shifting between two metal lines in back-end process with TaN/TaO x N RRAMs on both sides of a single via. The BEOL RRAM shows large read window between states. Fast switching time of 1 us for set operation and 10 us for reset was demonstrated. Excellent selectivity by its asymmetric IV characteristic enables the twin-bit 1R cells to be efficiently stacked in 3-D cross-point arrays without select transistors.

A New High-Density Twin-Gate Isolation One-Time Programmable Memory Cell in Pure 28-nm CMOS Logic Process

A new and compact high-k dielectric breakdown one-time programmable (OTP) cell in pure 28-nm high-k metal gate (HKMG) process is proposed. By adopting a self-aligned twin-gate isolation (TGI) made by merged gate spacer, the new OTP cell can operate independently with a very small cell area. Fabricated by a pure 28-nm HKMG CMOS logic process, this OTP cell successfully achieves an ultrasmall cell size of 0.0441 μm 2 on 28-nm HKMG CMOS logic platform. Using high-k dielectric breakdown as its program mechanism, the antifuse TGI OTP memory has more than three orders of ON/OFF read window with a low program voltage of 4 V in 20 μs. Furthermore, a highly density 64-kbit TGI OTP array has been fabricated and successfully demonstrates the new superior isolation and reliability performances.

New insight on negative bias temperature instability degradation with drain bias of 28 nm High-K Metal Gate p-MOSFET devices

The degradation of negative bias temperature instability (NBTI) on 28nm High-K Metal Gate (HKMG) p-MOSFET devices under non-uniform stress condition has been systematically studied. We found the asymmetry between forward and reverse Idsat shift under non-uniform stress condition is significant for long channel devices even under low drain bias stress (e.g., Vds=−0.1V and gate channel length L=1μm), and seems to be dominated by a minimally required critical length (L=0.2μm derived from the experimental data). To the authors’ best knowledge, these are new phenomena reported. We attribute these anomalous NBTI characteristics with drain bias to the local self-heating (LSH) activated NBTI degradation mechanism. One semi-empirical analytical model, which fits well with our experimental data, is then proposed in this paper.

28-nm 2T High-&lt;formula formulatype="inline"&gt;&lt;tex Notation="TeX"&gt;$K$&lt;/tex&gt;&lt;/formula&gt; Metal Gate Embedded RRAM With Fully Compatible CMOS Logic Processes

A new two-transistor embedded resistive RAM (RRAM) cell with fully Taiwan Semiconductor Manufacturing Company 28-nm CMOS logic compatible process is reported. The new 28-nm logic compatible RRAM cell consists of two logic standard high- k metal gate (HKMG) CMOS transistors with a composite resistive gate dielectric of TiN/ HfO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> /SiO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> /Si as a resistive memory storage node. Using one of the transistor gates as a source line in RRAM SET/RESET operation, the resistive memory states can be efficiently read and sensed by selecting the other transistor gate and its corresponding bitline. Therefore, the new 2T embedded RRAM cell has realized a logic nonvolatile memory (NVM) solution with cost effective, and fully compatible with 28-nm HKMG CMOS logic platforms. Besides, through adapting the existing high- K gate dielectric in the RRAM cell, the embedded memory cell does not need any additional deposition of resistive film or extra process steps and it will be very scalable and compatible with the fast progress of CMOS technologies in embedded NVM applications. Furthermore, its low voltage requirement makes this cell conveniently fit in logic intellectual properties and circuits for local data storages or level trimming devices on system-on-chip logic NVM products.

Preliminary results on low power sigmoid neuron transistor response in 28 nm high-k metal gate Fully Depleted SOI technology

The purpose of this paper is to describe a preliminary approach to achieve a sigmoid neuron transistor response using the 28nm high-k metal gate Fully Depleted SOI (FDSOI) technology. It is well known that a neural network is an ambitious way to handle signal and/or data flow. Of interest also is the ‘learning phase’ of the proposed structure. However, the major difficulty of such structures, where the elementary device is a “Neuron Design (ND)” is in their integration. The elementary ND is based upon a circuit with at least ten interconnected CMOS transistors in order to obtain a sigmoid response activation function (in this example) with multiple inputs typically as per the McCulloch and Pitts model. Given that a large number of NDs are required to build an Artificial Neural Network (ANN), the power consumption of such a structure is a key topic that is also addressed. Another open question concerns the dispersion response due to process variability. This study reports on a new single undoped Formal Neuron Transistor (NT) solution.