High-dielectric constant (high-k) gate oxides and low-dielectric constant (low-k) interlayer dielectrics (ILD) have dominated the nanoelectronic materials research scene over the past two decades, but they have recently reached a state of maturity and perhaps the limits of their scaling. Based on this, there is a need for a systematic review summarizing not only the historic research and achievements on high-k and low-k dielectrics, but also emerging device applications as well as an outlook of future challenges. We begin by first reviewing the factors that drove the emergence of low-k and high-k materials in nanoelectronics as ILD and gate dielectric materials, respectively, and the challenges and limits these materials ultimately approached in terms of permittivity scaling. We then illustrate that gate dielectric and ILD applications represent just a small fraction of the numerous dielectrics utilized in present day nanoelectronic products where permittivity scaling is now being increasingly demanded for materials such as dielectric spacers, trench isolation, and etch stopping layers. We conclude by examining the numerous new applications for dielectric materials that are emerging as the semiconductor industry transitions to novel patterning schemes, prepares for life post CMOS scaling, and explores ways to natively embed device functionality in the metal interconnect. For the former, we specifically examine the “colorful” requirements for the various enabling dielectric hardmask and spacer materials utilized in pitch division-multi-pattern processes and then discuss the role that selective area deposition of dielectrics and metals could play in reducing the complexity of such patterning processes. For the latter, we review the use of both high-k and low-k dielectrics in various metal-insulator-metal (MIM) structures as Fermi level de-pinning layers, tunnel diodes, and back-end-of-line (BEOL) compatible capacitive and resistive switching random access memory (ReRAM) elements. We further examine how dielectrics can hinder or aid new forms of computing such as quantum and neuromorphic in reaching their full potential. In conclusion, we find that while the field of dielectrics has a long history, it remains vibrant with numerous exciting new and old research vectors awaiting further exploration.

The paper reports on a novel process flow to manufacture conductive organic electrodes from highly conductive doped PEDOT:PSS polymer films that can be patterned and display a good adhesion to oxidized Si wafers as well as to flexible substrates, such as Mylar. Among other results, it is shown that multiple depositions of PEDOT:PSS increase the electrical conductivity by more than two orders of magnitude without increasing the film thickness of PEDOT:PSS significantly. An exponential dependence between sheet resistance and the number of PEDOT:PSS coatings has been found. The electrical conductivity of PEDOT:PSS can be increased by another two orders of magnitude doping with Cu nanoparticles when coated on the surface of a soft-baked PEDOT:PSS film. It is found, however, that both kinds of conductivity enhancement are not additive. Adhesion of PEDOT:PSS to oxidized Si wafers and BoPET (Mylar) has been ensured by applying an oxygen plasma cleaning step before spin coating. The manufactured high-conductivity PEDOT:PSS film can be patterned using a sacrificial metal layer with subsequent etching of PEDOT:PSS in oxygen plasma, followed by the removal of the patterned segments of the sacrificial metal layer in an aqueous acid solution.

Evidence for existence of composite copper/oxygen vacancy nanofilaments is presented. The composite conductive nanofilment (CF) is a serial resistive connection of Cu and V_o CFs forming a composite V_o/Cu CF where the Cu filament segment is formed first and V_o filament completes the composite CF. A composite CF has been also formed in reverse order as Cu/V_o. Both composite CFs, V_o/Cu and Cu/V_o, display V_form at considerable lower voltages than monolithic CFs.

Deep neural networks (DNNs), a brain-inspired learning methodology, requires tremendous data for training before performing inference tasks. The recent studies demonstrate a strong positive correlation between the inference accuracy and the size of the DNNs and datasets, which leads to an inevitable demand for large DNNs. However, conventional memory techniques are not adequate to deal with the drastic growth of dataset and neural network size. Recently, a resistive memristor has been widely considered as the next generation memory device owing to its high density and low power consumption. Nevertheless, its high switching resistance variations (cycle-to-cycle) restrict its feasibility in deep learning. In this work, a novel memristor configuration with the enhanced heat dissipation feature is fabricated and evaluated to address this challenge. Our experimental results demonstrate our memristor reduces the resistance variation by ~ 30% and the inference accuracy increases correspondingly in a similar range. The accuracy increment is evaluated by our deep delay-feed-back reservoir computing (Deep-DFR) model. The design area, power consumption, and latency are reduced by ~48%, ~42%, and ~67%, respectively, compared to the conventional static random-access memory technique (6T). The performance of our memristor is improved at various degrees (~13%–73%) compared to the state-of-the-art memristors.

A Cu/TaO_x/Pt resistive cell can transition from a high-resistance state to a low-resistance state by applying positive or negative set voltages. The set operation with positive bias is due to the electrochemical formation of a Cu conductive filament (CF), whereas the set operation with negative bias is attributed to the electroreduction of solid electrolyte and formation of an oxygen vacancy V_OCF. Both CFs can be reset by Joule heating. Since set voltages of the V_OCF have higher absolute values than those of the Cu CF, the switching based on Cu and V_O CFs can be decoupled within the same device. The device is inherently an antiparallel resistive switch circuit that can be consecutively switched between the Cu and V_OCFs.