Search

Buried power rail (BPR) is a key scaling booster for CMOS extension beyond the 5-nm node. This work demonstrates, for the first time, the integration of tungsten (W) BPR lines with Si finFETs. BPR technology requires insertion of metal in the front-end-of-line (FEOL) stack. This poses risks of stack deformation and device degradation due to metal-induced stress and contamination. To assess the stack deformation, we demonstrate W-BPR lines which can withstand source/drain activation anneal at 1000 °C, 1.5 s, without adversely impacting the stack morphology. To address the contamination risk, we demonstrate a BPR process module with controlled W recess and void-free dielectric plug formation which keeps the W-line fully encapsulated during downstream FEOL processing. Suitable choice of BPR metal such as W with high melting point which does not diffuse into dielectrics also minimizes the risk of contamination. To assess the device degradation, simulations are carried out showing negligible stress transfer from BPR to the channel. This is experimentally validated when no systematic difference in the dc characteristics of CMOS without BPR versus those in close proximity to floating W-BPR lines is observed. Additionally, the resistance of the recessed W-BPR line is measured $\sim 120~\Omega /\mu \text{m}$ for critical dimension (CD) ~32 nm and height ~122 nm. The recessed W-BPR interface with Ru 3-nm TiN liner via contact can withstand more than 1000 h of electromigration (EM) stress at 6.6 MA/cm2 and 330 °C, making Ru a candidate for via metallization to achieve low resistance contact strategy to BPR.

We propose, for the first time, an indium gallium zinc oxide (IGZO)-based 2T1C compute cell (IGZO-cell) for analog in-memory computing. To assess the impact of an IGZO-cell-based array including the periphery on power and accuracy, a PyTorch framework was developed to analytically modeled analog components. The results are reported for a ResNet20 network on the Canadian Institute For Advanced Research-10 (CIFAR-10) benchmark. The state-of-the-art energy efficiency of 15 peta operations per second (POPS)/W including the periphery is achieved by using our proposed IGZO-cell with CMOS compatibility. Finally, it is shown that, with a properly trained neural network model, there is no degradation of test accuracy with 10% device to device variability for the IGZO devices.

3-D sequential integration requires top MOSFETs processed at a low thermal budget, which can impair the device reliability. In this paper, top junctionless (JL) devices are fabricated with a maximum processing temperature of 525 °C. The devices feature high k/metal replacement gate and low-temperature Si:P and SiGe:B 60% raised source and drain for nMOS and pMOS fabrication, respectively. Device matching, analog, and RF performance of the top tier devices are in-line with the state-of-the-art Si technology processed at high temperature (>1000 °C). JL devices operate at reduced electric field and can meet in specification reliability (10-year reliable operation at ${V}_{\textsf {G}}= {V}_{\textsf {th}}+ 0.6$ V, 125 °C), even without the use of “reliability” anneal. The top Si layer is transferred on CMOS planar bulk wafers with W metal-1 interconnects, using a SiCN to SiCN direct wafer bonding. Comparison with silicon-on-insulator devices fabricated with the same low-temperature flow shows no impact on device electrical performance from the Si layer transfer.

Magnetic tunnel junctions (MTJs) interconnected via a continuous ferromagnetic free layer were fabricated for Spin Torque Majority Gate (STMG) logic. The MTJs are biased independently and show magnetoelectric response under spin transfer torque. The electrical control of these devices paves the way to future spin logic devices based on domain wall (DW) motion. In particular, it is a significant step toward the realization of a majority gate, even though further downscaling may be required. To our knowledge, this is the first fabrication of a cross-shaped free layer shared by several perpendicular MTJs. The fabrication process can be generalized to any geometry and any number of MTJs. Thus, this framework can be applied to other spin logic concepts based on magnetic interconnect. Moreover, it allows exploration of spin dynamics for logic applications, Comment: submitted to Japanese Journal of Applied Physics

With the rapid progress of spintronic devices, spin-logic concepts hold promises of energy-delay conscious computation for efficient logic gate operations. We report on the electrical characterization of domain walls in interconnected magnetic tunnel junctions. By means of spin-transfer torque effect, domains walls are produced at the common free layer and its propagation towards the output pillar sensed by tunneling magneto-resistance. Domain pinning conditions are studied quasi-statically showing a strong dependence on pillar size, ferromagnetic free layer width and inter-pillar distance. Addressing pinning conditions are detrimental for cascading and fan-out of domain walls across nodes, enabling the realization of domain-wall-based logic technology.

In this work, Ru wires patterning by direct etch are evaluated for a potential 5 nm technology node. The characteristics of Ru etching by varying the bias voltage, total flow rate and Cl2/(O2+Cl2) gas flow ratio are studied in an inductively couple plasma etching chamber. Ru sidewalls profile with a tapering angle of 90° and Ru to SiO2 hard mask etch selectivity of 6 are achieved. The authors show the feasibility of patterning lines with an aspect ratio up to 3.5 and lines with a critical dimension down to 10.5 nm (with a 3σ line width roughness of 4.2 nm), which paves the way to further scaling of this approach. Finally, the authors present a study on Ru line roughness after patterning on 300 mm wafers. Here, they compare line roughness results of wafers where Ru is deposited with different deposition techniques, such as atomic layer deposition and plasma vapor deposition, and it is annealed after deposition at various temperatures.In this work, Ru wires patterning by direct etch are evaluated for a potential 5 nm technology node. The characteristics of Ru etching by varying the bias voltage, total flow rate and Cl2/(O2+Cl2) gas flow ratio are studied in an inductively couple plasma etching chamber. Ru sidewalls profile with a tapering angle of 90° and Ru to SiO2 hard mask etch selectivity of 6 are achieved. The authors show the feasibility of patterning lines with an aspect ratio up to 3.5 and lines with a critical dimension down to 10.5 nm (with a 3σ line width roughness of 4.2 nm), which paves the way to further scaling of this approach. Finally, the authors present a study on Ru line roughness after patterning on 300 mm wafers. Here, they compare line roughness results of wafers where Ru is deposited with different deposition techniques, such as atomic layer deposition and plasma vapor deposition, and it is annealed after deposition at various temperatures.