National Academies Press: OpenBook
« Previous: 2 High-Entropy Materials
Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×

3

Nanoelectronics

BEYOND-COMPLEMENTARY METAL-OXIDE-SEMICONDUCTOR DEVICE CONCEPTS ENABLED BY TWO-DIMENSIONAL MATERIALS: THE VIEW FROM THE TOP DOWN

Tom Theis, Executive Director, Semiconductor Research Corporation (on assignment from IBM Research)

Tom Theis opened his presentation by explaining that it is impossible to have a manufacturing process that does not include both top-down and bottom-up fabrication processes—for example, within the semiconductor industry, self-assembled crystals are sliced into wafers, which are then carved into patterns and implanted with other materials. Theis outlined three topics of focus for his presentation: (1) semiconductor manufacturers’ investments in device research beyond that of complementary metal-oxide semiconductor (CMOS) research; (2) Nanoelectronics Research Initiative (NRI) and STARnet, which are private-public partnerships funding university semiconductor research; and (3) new device concepts enabled by two-dimensional (2D) materials bonded by van der Waals forces.

Theis displayed an image of a metal-oxide-semiconductor field effect transistor (MOSFET) and explained the simple principles of Dennard scaling that had guided the semiconductor industry for many years: in order to shrink the device, the operating voltage must be reduced simultaneously to keep the electric fields from becoming too large. The resulting smaller device will work at lower power with increased density, faster switching, and decreased cost. It is important to continue

Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×

to create new, smaller devices so that it is possible to do more with the MOSFET, Theis said. Figure 3.1 shows that progress in miniaturization continues as transistor count continues to increase, following Moore’s law. However, clock frequency and total power have been flat since 2003, as some limits to Dennard scaling have already been reached.

System designers would like to be able to innovate even faster. The level of architectural advancement measured by instruction-level parallelism (ILP) is lagging,

Image
FIGURE 3.1 Intel CPU trends, 1970-2010. The x-axis denotes years. The y-axis is defined by the box insert as either number of transistors in thousands (green); clock speed in MHz; power in W or the number of instructions that can be performed during a single clock cycle, which is described as instruction-level parallelism (ILP). SOURCE: Tom Theis, Semiconductor Research Corporation, presentation to the workshop; from H. Sutter, 2009, “The Free Lunch Is Over: A Fundamental Turn Toward Concurrency in Software,” http://www.gotw.ca/publications/concurrency-ddj.htm, copyright 2009 Herb Sutter.
Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×

even as device miniaturization has continued at a furious pace, Theis explained. The gate length of the devices (which determines how fast they can operate) is saturated, but the effective gate length scaling for the devices has continued to improve. This innovation is led by Intel with the development of a fin field effect transistor (FinFET; one electrode is shaped like a fin), which offers better electrostatic control. According to Theis, the next step might be to create a gate-all-around device, which could become the ultimate field effect transistor (FET). Though there have been problems with the creation of this device, there has been steady progress: the development of a 200-picosecond carbon nanotube ring oscillator has been reported. Unfortunately, this speed is still too slow to make this a competitive device for high-performance computing. Theis noted that carbon nanotubes are popular because their many attributes make them good for niche applications; they are substrate agnostic and there are applications in sensors and other flexible electronics, for example, but they are not close to replacing silicon in the near future.

Theis observed that clock speed and power stopped improving because the FET reached its voltage scaling limit. An external voltage applied to a current-gating electrode has to have a certain voltage swing to turn the current on and off. The minimum off current is set by thermodynamics, and the on current must be much greater in order to represent the digital state reliably. Vacuum tubes, bipolar transistors, and FETs are all subject to this voltage scaling limit. He explained that at room temperature, it would take 60 mV of swing to cause 10´ change in the current of a device, and typically a change of 104 or greater in current is needed. Therefore, operating much below 1 V at room temperature either kills the on current (which means reduced switching speed or computational performance) or increases the leakage current (which means excessive heat generation and power dissipation even in the absence of computation). There is no way for the designer to get around this, Theis said. In the last 10 years, in order to keep areal power density and total power relatively inexpensive, industry froze clock speed and slowed the deployment of multiple cores. One result is that the silicon chip area has been increasingly devoted to memory instead of processing. The sea of memory around the cores is designed as a heat radiator from the processing cores. This is not the best approach overall, according to Theis, because the net performance of the system slows down. In response, the industry has started to fund the exploration of devices with switching mechanisms that are fundamentally different from those of the conventional FET and computing architectures that are fundamentally different from the von Neumann architecture. Founded in 2005, NRI is a public-private partnership jointly funded by the National Institute of Standards and Technology (NIST), National Science Foundation, and industry. NRI provides money to university researchers who focus on these nonconventional, low-energy technologies for computation that work better than CMOS on critical applications in 2015 and into the future. This initiative was followed by STARnet, cofounded by the Defense Advanced Research Projects

Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×

Agency and industry in 2013. NRI and STARnet each fund several multidisciplinary multi-university research centers to study and explore these new devices.

Theis explained that there are numerous potential physical principles by which a digital switch could avoid the voltage scaling limit of the FET, and some of these new devices are enabled or optimized by the unique properties of 2D materials. Such materials provide a toolkit to make structures that did not previously exist. In terms of band gap engineering, different materials can be layered to make heterostructures. Hexagonal boron nitride (h-BN), for example, serves as an excellent insulator, while graphene has relevant properties for spin transport and spintronic devices. Theis provided an example from the Southwestern Academy of Nanoelectronics of using 2D materials to fabricate new devices; this group is trying to build a structure that contains tunneling between two layers of graphene isolated by h-BN. After the layering is complete, the structure is carved lithographically, and the compositional analysis shows two layers of graphene and four atomic layers of boron nitride. Theis pointed out that while this fabrication method is useful for exploration, it cannot be used for manufacturing without much further development.

Theis discussed the band-to-band tunneling FET that gets around the limit of the traditional FET because it gates through a tunneling junction. He noted that this device could show promise as a low-voltage device. Researchers are now exploring such vertical tunneling structures fabricated from 2D crystals. Theis described a new steep slope device designed to operate at a lower voltage and dissipate significantly less power by utilizing graphene p-n junctions to filter the energy distribution of carriers to get beyond supply voltage limits of FETs. He noted that, theoretically, there should be a steep turn-on and four orders of magnitude in current change for this device, which operates on the principle of current carriers tunneling from p- to n-doped. Conservation of momentum requires that the carrier trajectories be refracted in a certain way. Collimation of current carriers occurs between two junctions, and when a proper gating voltage is applied, the resulting transmission gap limits the flow of current, Theis continued. This is the first device where transport within the device is almost perfectly ballistic; however, the reflections create issues because they are difficult to control. Theis asserted that to combat such problems, the research team is returning to basic physics as well as simple device and modeling concepts.

Theis described the nanomagnetic (spintronic) device, which, in order to function properly, must have a long spin diffusion length. Current-controlled spin logic is also an important concept for this device and the graphene that is engineered for it. By hydrogenating graphene, Theis explained, a large spin-Hall effect results and spin polarization is used to switch a magnetic system. Theis further explained that running a current to create spin polarization results in undesirable resistive power loss and heat generation. He noted that researchers are exploring ways to simply apply a voltage to an electrode to control spin polarization. In this

Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×

case, the resistive loss would be very small. At some point, 2D materials might also be of interest for this very promising method of spin switching, but they are not currently used.

Theis then highlighted a specific device concept based on voltage-control of spin polarization—the magnetoelectric tunnel junction, which has potential applications in both memory and logic, is nonvolatile, and should operate at ultra-low power. Theis suggested that further developments in spin logic were likely to broaden the range of potential applications. Using surface effect and manipulating magnetism are just two strategies for innovation as the field moves from physics to electrical engineering, with the help of public-private partnerships.

Theis summarized his talk reiterating that while ultra-thin devices based on 2D materials may not have a big impact on high-performance digital devices, they may have important applications in flexible electronics and sensors. He also reiterated that the following are needed to enable new beyond-CMOS device concepts: scalable, self-limiting, and selective processes for growth of single crystals and aligned heterostructures; patterning processes for nanometer-scale lateral features; and vastly improved control of point defects. Theis’s takeaway was that although current devices could be revolutionary, none of these structures are useful if they cannot be manufactured and scaled. Although the rate of invention is increasing, researchers have not yet discovered the ultimate switch for digital computing. For this reason, continuing to focus attention on and provide funding to the device research community is essential, according to Theis.

Ward Plummer recalled Peter Liaw’s and Dan Miracle’s discussions about academia’s lack of understanding of industry needs and asked if companies are distancing themselves from research and development. Theis noted that companies continue to invest in research and development but often need to differentiate themselves and realize that value-added and leading-the-charge are two different things. Companies have to invest carefully in the areas that they are best suited for. Theis clarified that the lingering issue in this discussion is not whether research is needed but what the new path is for research. In many cases, the focus should be on new architecture and new devices, such as improving the energy efficiency and performance of artificial neural networks or integrating new devices simultaneously for both memory and logic. Theis and Todd Younkin, Intel Corporation, agreed that the next phase for NRI and STARnet will link device research with architecture-, application-, and system-level solutions.

Valerie Browning asked if, since the performance of current technology is temperature-dependent, there is a future for localized temperature control. Theis commented that this is a possibility, although the temperature extraction limits have not yet been reached. The Department of Defense (DoD) would likely have applications to support this, he continued, although implementation would have to happen at the chip level and the technology must be affordable. Theis acknowledged

Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×

that one area for potential exploration is active cooling. Robert Pohanka, retired from the National Nanotechnology Coordination Office, asked if there are any imminent international breakthroughs. Theis described NRI’s nucleation of research on tunnel-FET devices, which rapidly spread to Europe and Asia. This is important work. The United States must take care not to lose its competitive edge in post-FET device research because new device concepts, yet to be fully explored, might exceed the performance of current devices, Theis suggested.

NANO REGIME: CONTINUED CMOS SCALING THROUGH EXPLORATORY MATERIALS RESEARCH

Todd Younkin, Senior Staff Research Scientist and STARnet Program Manager, Intel Corporation

Todd Younkin noted that his presentation would be complementary to Tom Theis’s talk in covering the spectrum of top-down and bottom-up approaches. Younkin noted that his new position as STARnet program manager tasks him with managing years 4 and 5 of the current program as well as plotting approaches for a new program expected to commence in late 2017. He acknowledged that Intel is spending more money and energy on its work in recent years but with a talented staff, Intel plans to continue and in fact accelerate its effort and leadership. People trained in nanometer technologies often think this realm of research is coming to a close, but this is not the case, according to Younkin. For example, critical gate and metal pitches can be enabled by marrying continued materials and integration innovations alongside a wide variety of new designs.

For this workshop, Younkin noted, the 7-nm node (and beyond) is the critical area of focus, which equates to an approximately 45-nm pitch and below. At this length scale, there will be a complementary role that both the top-down and bottom-up approaches will play in fabrication. The top-down approach is best utilized at 30 nm and above (where current lithography, including extreme ultraviolet lithography [EUVL] works best), while the bottom-up approach functions best at a 60-nm pitch and below. Younkin gave a historical overview of progress in the field, noting that the first atomic scale lithography was performed in 1989. Since that time, much of the related funding and effort was put toward the implementation of lithography with scanners,1 utilizing 193 immersion as the current primary production technique. The next technique introduced was

___________________

1 Scanners or steppers are projection exposure systems that project the mask onto the wafer many times to create the complete wafer pattern, as compared to contact or proximity masks, which cover an entire wafer.

Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×

EUVL, a type of “optical” top-down technology that required significant research investments before it moved into the development stage in 2012. EUVL will soon prove to be a viable approach, according to Younkin.

Scanners are sold at different wavelengths to meet the required resolution. Younkin noted that in the past, it was possible to print a wide array of features with limited restrictions; now, however, a Manhattan Grid (2D layers fabricated by a combination of one-dimensional [1D] orthogonal patterns at regular spacing) must be used because everything is unidirectional as a result of pressing the tool’s capability for resolution at a given wavelength. Younkin revealed that Intel already does multipass patterning, which is costly. He also noted that lithography has limited scaling potential and that patterning materials are essential to achieving Intel’s resolution and productivity goals. For example, instead of using chemical vapor deposition or atomic-layer deposition for spacer-based double patterning, Intel now uses spacer-assisted quadruple patterning,2 which may get the industry down to only a few nanometers. Younkin explained that 1D lithography is used to make the fin, gate metal, or contact layers, called gratings, and these gratings are subsequently chopped into “plugs” or “cuts.” At this point, the contacts must be aligned to subsequent features, which often requires multiple patterning steps; this means it can take six to eight masks to create one subset of the ultimate pattern. To do this only once requires a minimum of a $300 million investment for the required infrastructure. Younkin noted that while resolution is still important, what is really of interest is pattern placement error and control of the interlocking, intertwining complexities required to deliver a functioning chip, all while saving money on the components, infrastructure, and process. He continued that although cost and heat management used to be afterthoughts, these factors are now included in the design from the onset of the development process. However, he acknowledged that there is still much progress to be made. Younkin explained that many companies are investing in EUVL because it allows the overall complexity to be simplified greatly, thereby reducing the overall cost. At the same time, there has also been increased interest in bottom-up assembly, especially since 2003, as dimensions of interest started to move below 60 nm. Trying to do everything with a top-down approach is prohibitively expensive, Younkin suggested. A complementary bottom-up approach, on the other hand, incorporates information from the atomic length scale and may deliver a more cost-effective solution. Younkin noted that it may take time to convince the community that a balance of the two approaches is best for future research and development.

___________________

2 Spacer-assisted quadruple patterning or self-aligned quadruple patterning is when self-aligned double patterning is used twice. Each application of the technique effectively cuts the line spacing in half.

Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×

In response to an audience question, Younkin noted that manufacturing led the way in the atomic-layer deposition space, with atomic-layer etching being the latest technology to mature in this arena. He suggested that the success of this combined top-down, bottom-up approach depends on the continued development and use of materials, and cost will have to be monitored and improved over time.

Younkin transitioned to a discussion of increased device complexity. He noted that it is important for the materials and manufacturing communities to figure out how to progress from FinFET to gate-all-around architectures and beyond. To do this, both new materials and new device structures are needed, and it is essential that these new devices have the capability both to accept and to relay information. At the atomic level, Younkin said, STARnet is currently looking for ways to reduce barriers, improve capacitance and material properties, and invent novel solutions that relax the design requirements.

Younkin noted the primary ongoing challenge of using new materials and techniques relates to scalability. Three substantial challenges that currently exist are (1) self-alignment; (2) selective deposition, growth, and removal; and (3) gapfill. These overarching themes include opportunities for novel patterning materials, such as EUVL, novel materials for atomic-layer deposition, and dielectric thin films by design. He reinforced the notion that continued public-private partnerships are essential to drive this research, to protect issues of national security, and to best incorporate the wide range of stakeholder interests.

Younkin moved to a deeper discussion of two subtopics related to bottom-up materials and integration strategies: (1) polymers for directed self-assembly (DSA), the next patterning step beyond EUVL; and (2) selective area deposition, based on atomic-layer deposition for metal-on-metal or dielectric-on-dielectric. He explained that the Generation One DSA material is based on polystyrene-block-polymethylmethacrylate, which phase separates upon annealing and creates ordered structures down to as low as approximately 20 nm pitch. The pitch of greatest interest, 60 nm or less, is defined by chemists who made this material in a test tube. The pitch is selected and qualified, Younkin explained, because it is a point at which the strength of top-down and bottom-up approaches overlap, providing reasonable anneal times for thermodynamic phase separation, and because simulations have indicated that long-term defectivity requirements can be achieved with this material/pitch combination. He emphasized that the patterning of this material results in nearly perfect alignment and uniformity and can also increase the line density (approximately three to four times) depending on the strategy employed for integration. To increase confidence in such an approach, Younkin’s team chose to focus on defectivity by performing a series of integrated DSA defect density studies on different geometries. This required the team to make “perfect wafers” such that it could proceed to the next level of metrology to make industry-relevant defect density comparisons. His team succeeded in reducing the

Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×

DSA effects down to less than one defect per square centimeter squared. At this pace, he stated that it should be possible to achieve the industry-required defect density goals within the next 5 years.

With confirmation that defect densities are not the problem, new materials that will better serve long-term needs should be invented, Younkin observed. Commercial-grade high-chi DSA materials3 with vertical morphology and good uniformity are readily available, so the next step forward is to study both smaller dimensions and more complex polymer architectures such as solvent annealing, top coats, Miktoarm copolymers, and ABC triblock copolymers. The best use of time and money, according to Younkin, is to use simulations for research in both smaller dimensions and more complex features that can serve in a complementary way to other patterning tools. CHiMaD, sponsored by NIST, is an example of an initiative that focuses greater attention to the combined power of computation, nanomaterials, and their industrial uses in a wide range of cutting-edge technologies.

Younkin remarked that selective area deposition is the enabler for a variety of new integration schemes. The key is the ability to deposit metal-on-metal or dielectric-on-dielectric selectively, without generating undesired deposition that would have the potential to create a defect or error. Younkin noted that there is much knowledge on precursor and process design for inherent selectivity. He believes that innovation will happen when this approach is combined with the use of the passivation necessary to block unwanted deposition. Selective cleanup steps may be needed to eliminate unwanted defects. There should also be an emphasis on the expansion of usable elements, Younkin explained. Metrology techniques can be used to understand fundamentals, thereby allowing the end user to control deposition and improve overlay. Early selective area deposition trials reveal that deposition is occurring in the desired areas and pattern placement is nearly perfect with the deposition aligned over the area of interest. Younkin noted that deposition rates need to be compatible with the desired film thickness. His team tested this approach with a defect density metrology tool. Total reflection X-ray fluorescence revealed that it is possible to deposit metal-on-metal at reasonable growth rates and discriminate against the nearby dielectric with the required level of perfection. Younkin observed that active research on integrated defect density assessments will continue into the future. He acknowledged that although the examples he provided began in development over 10 years ago, further progress will be made as the excitement builds and more funding arrives.

Robert Pohanka asked how much investment in these areas should be increased, as well as which topics DoD should consider as it revisits its strategy. Younkin said that a unified program similar to STARnet would be useful for DoD. He added that

___________________

3 DSA materials are materials for directed self-assembly, usually block copolymers. The high chi refers to polymers, where chi that is the Flory-Huggins interaction parameter has a high value.

Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×

such initiatives usually cost $50 million or more per year and operate best when researchers have a 5-to-10-year commitment and industry “steering” or guidance. Haydn Wadley observed that while directed self-assembly of polymers is impressive, there are already many examples of metals and ceramics in nature that self-align. He asked if there is currently any interest in exploring other material systems beyond polymers. Younkin said yes, but that the level of understanding in ceramics and metals is not currently at the same level of that of polymers, which makes this an important fundamental area of study and investment for the future. Wadley continued, and Younkin agreed, that HEAs might also be worth future study and investment. Tom Theis added that the Semiconductor Research Corporation has an active synthetic biology program that uses deoxyribonucleic acid self-assembly to build three-dimensional (3D) structures, which is an exciting space to explore.

MYTHBUSTING SCIENTIFIC KNOWLEDGE GENERATION AND TRANSFER WITH NANOHUB.ORG AND NEMO

Gerhard Klimeck, Director of the Network for Computational Nanotechnology; Reilly Director of the Center for Predictive Materials and Devices; Professor of Electrical and Computer Engineering, Purdue University

Gerhard Klimeck leads an infrastructure effort, nanoHUB,4 as well as a research group that creates the Nanoelectronics Modeling tool set, NEMO,5 at Purdue University. Klimeck explained that nanoHUB allows users to conduct simulations without needing to install any software on their own computers. He shared that nanoHUB has approximately 13,000 simulation users and 300,000 tutorial and lecture users annually across 172 countries. In all, nanoHUB has over 1,800 content authors who contribute to the site, and the site has had an uptime of 99.9 percent or experienced only approximately 20 outage hours throughout an entire year. The overarching goal of nanoHUB is to generate new knowledge and transfer existing knowledge throughout the field of nanotechnology.

Klimeck provided a bit of history, describing 1998 as the turning point when computers and science became accessible to many more people through the use of portals (now called science gateways). Beowulf clusters were also being built during the late 1990s in so-called garage industries, which ushered in a new wave of supercomputers, Klimeck explained. Circuit and process simulations were established in the semiconductor industry, and device simulation provided point-based optimization. All technology computer-aided design (TCAD) tools were

___________________

4 See the nanoHub website at http://nanohub.org/, accessed February 11, 2016.

5 See the NEMO website at https://engineering.purdue.edu/gekcogrp/software-projects/nemo3D/, accessed July 8, 2016.

Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×

agnostic of atoms. The materials world worked up from ab initio calculations and used DFT and equilibrium theory, whereas now real devices are on the scale of a countable number of atoms, and all electronic devices are operated out of equilibrium. The year 1997 also marked when the first nanoelectronics industrial development group (founded in the early 1990s at Texas Instruments) was sold to Raytheon and disbanded completely within 3 years. This was the last industrial research and development group, after others like IBM and Motorola, that disbanded open-ended device research efforts in the semiconductor industry.

Klimeck discussed the role of Moore’s law in software development. Two to four billion transistors are on a chip (and that number continues to grow exponentially) and that was achieved in part with circuit simulation. Klimeck noted that the two primary software tools that enabled this growth were SPICE and SUPREM. SPICE originated with one teacher and one student and became one of the first open-source codes. This experience suggests that small academic efforts can lead to a huge development in the semiconductor industry. Today, Klimeck said, approximately $2 billion is currently being invested in U.S. research and development in nanotechnology; data generated in publications can also be utilized to turn into software and applications that can be made useful for a whole community.

Portals and science gateways were promoted by computational scientists in the late 1990s to provide a new era of computing. Klimeck provided an overview of the traditional science gateway approach: researchers who are interested in advancing the science give the code to web developers, who know nothing about the tool or the science, to put the tool on the web. The web developers try very diligently to transcribe the tool for web deployment in an effort that takes 2 or 3 years. In the meantime, the researcher continues to update the original tool, as science evolves and software bugs get fixed. For the authentic research developer, the tool that is deployed on the web by someone else on the web is immediately outdated and irrelevant. In fact, the researcher no longer owns the tool that is on the web. The end result is that a facsimile of the outdated tool shows up for use on the web. It has never been tested or validated by the original tool author and it no longer advances science, since new versions are already available in the closed space of the original researcher. Such an approach cannot deliver authentic access to real research be scaled to larger communities because it renders deep research impossible, lending a bad reputation to the notion of gateways. This traditional portal or gateway concept is budgeted in proposals at a total cost of $500,000 per tool. This is very expensive to perform, especially in light of the fact that no new science is enabled in the process. However, nanoHUB fixes this problem in the process by eliminating the traditional role of web developer: the individual researcher instead is enabled to deploy software without a science-agnostic web developer. This process decreases the deployment time from approximately 2 years to 2 weeks. Longitudinal data of nanoHUB shows that in 4 years and without any significant

Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×

funding, nanoHUB deployed 175 tools. That is a value of $88 million that no one in federal agencies would have been willing or able to support for nanotechnology research and education.

And as of February 2016, nanoHUB had 1,400 versions of 400 different tools, which demonstrates the success of eliminating the intermediary and allowing nanoHUB’s network of over 380 developers to evolve its tools. Klimeck highlighted the importance of continued study of both user behavior and usage patterns; through further measurement and data collection, Klimeck’s team confirmed that intense collaboration is useful and often serves as a predictor of success. He introduced the incentives associated with using nanoHUB for both collaborators and the gateway itself. Because Web of Science now publishes nanoHUB tools and models, collaborators can count their work as published, which aids in promotion and tenure. Collaborators can also incorporate the tools into their classroom curricula while nanoHUB’s reputation continues to grow.

Klimeck moved to a discussion of nanoHUB’s usability, which he described as just as important as capability. Using PADRE, a tool from Bell Labs, one can simulate MOSFETs, but it is complicated to operate and does not have a modern input parameter description language for communication. However, it is possible to write a nanoHUB front that communicates with the PADRE code. Now it is possible to increase the number of users (and consequently the number of available tools). Although the MOSFET front on nanoHUB cannot simulate everything that PADRE can, it is more usable because it does not require the user to know a particular input parameter description language. Another crucial indicator of any gateway’s success is interactivity, according to Klimeck. He affirmed that the web must enable users’ work instead of interfere with it. For example, nanoHUB eliminates the arduous chore of compiling and assembling code because it is an open source tool. As usability increases, the number of downloads of source code drops significantly, indicating success of the gateway.

Klimeck shared that because of increased usability and interactivity, there are 1,360 papers that cite nanoHUB as a source. In essence, nanoHUB has created a social network linking 2,500 authors, 64 percent of whom are outside of the immediate collaborator relationships for a specific tool and 9 percent of whom are from industry (as opposed to the usual academic user). Over 50 percent of the papers support experiments, Klimeck continued. The quality of research being performed continues to improve as well because it is continually evaluated with metrics. For example, there are currently over 18,000 secondary citations of the 1,360 papers, and the h-index is already 65 after only approximately 12 years in existence.6

Another transformative application for nanoHUB is in education. While it typically takes nearly 4 years for a traditional textbook to be updated to a new

___________________

6 According to Klimeck, a typical researcher is expected to have an h-index of 12 in 12 years of work past the Ph.D. An outstanding researcher may have an h-index of 24, just 12 years after the Ph.D.

Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×

edition and 8 years for an initial textbook release, a nanoHUB tool takes only approximately 6 months between the first publication on nanoHUB and the first adoption into a structured educational setting through homework or project assignments. nanoHUB dramatically speeds up the transformation and innovation of education and links education to authentic research tools.

Klimeck’s next subject was simulation with NEMO5, the fifth edition of NEMO. He showed a classical TCAD FinFET model that is unaware of its own atoms. The device looks like a solid object, just like a macroscopic mechanical cooling fin. However, today’s real device has fins that are 8 nanometers, or about 64 atoms, thin. This is no longer a continuous solid material, and the atoms are countable in that direction. Over a decade, his team built an atomistic modeling toolkit to model real devices where electrons flow. (Devices are simply atoms connected together.) One goal is to bridge ab initio calculation, since it cannot handle as many atoms as there typically are in a real device, versus TCAD (which does not contain any real material physics), and map the model onto real devices. With such a tool, new materials and geometries can be explored. Klimeck explained that this is a nonequilibrium quantum statistical mechanics problem, which is something that ab initio models (or DFT) cannot capture. The implication here is that real devices have a real (large) voltage applied that drives electrons from one end of the transistor to another, via a gate or a valve that turns the flow on and off. DFT cannot handle such high-energy out of equilibrium behavior. NEMO5 allows multiphysics and is the first nonequilibrium Green’s function predictive tool that enables the modeling and simulation of such realistic devices.

The same NEMO simulation software powers eight nanoHUB tools, delivering results completely transparently to users without demanding computational science details. NEMO can tackle problems on the United States’ largest supercomputers within 1 hour—this would take 25 years to compute on a single desktop computer. These supercomputers enable exploration to a very high fidelity using the NEMO software, yet NEMO can also be used for tens of thousands of users that have relevant smaller problems to solve on nanoHUB.

Funders are rarely interested in supporting infrastructure, Klimeck suggested. And funders want the latest science and technology advancements but do not want to hear that the software infrastructure needed to deliver such realistic results might take a decade of serious investments. He advocates for more investment in software as well as a focus on sustainability, which often comes from leverage in multiple fields. Such investments must be driven not by computer science and standards, but by application researchers and developers who have essential computational needs.

Klimeck noted that his next goal is to have experimental data available on the gateway and to foster reproducible nanoengineering. Ultimately, Klimeck explained, he would like to develop a nanoHUB professional society that would

Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×

include a paid membership. He concluded by reiterating that nanoHUB is usable and accessible both for research and classroom use.

Daniel Gopman, NIST, asked if people have to queue up for time on NEMO. Klimeck responded that this is possible, depending on the type of user: there is no need to wait in a queue unless the task is extensive (e.g., high-end computation). Todd Younkin asked if the next step involves integrating stochastics. Klimeck replied that the current tools are XML-driven and embedded in an uncertainty quantification framework to do swaths of statistical exploration of the output from the tools. Klimeck added that nanoHUB is also starting to store simulation results instead of recomputing to save money.

SELF-ASSEMBLED NANOCOMPOSITES FOR MULTIFUNCTIONALITIES: CURRENT AND FUTURE

Haiyan Wang, Professor, Purdue University (Formerly at Texas A&M University)

Haiyan Wang noted that both Tom Theis’s and Todd Younkin’s presentations provided a helpful introduction for her discussion of self-assembled nanocomposite structures for mutlifunctionalities. She shared three examples of FET devices constructed with 2D layered materials, interfaced and integrated materials, and microelectronics beyond silicon. She clarified that her list is not exhaustive. With so many possibilities of materials and devices for future generations, her team uses a different approach: a thin film epitaxy approach to grow nanocomposite thin films. Using this approach, the atoms are either manipulated into a perfect atomic arrangement (i.e., epitaxial growth) or randomly distributed (i.e., polycrystalline or amorphous growth), based on one’s overarching goal (see Figure 3.2).

Wang noted that although materials scientists have many tools to use in this approach, they would benefit from a greater selection of materials. Relying on the concept of nanocomposites (i.e., hybrid materials), it is possible to “co-grow” two different materials and force them to self-assemble into new materials and new devices, Wang explained. This allows the development of vertically aligned nanocomposites (VANs), high-density nanowires, and nanocomposites with checkerboard matrices, for example.

Over the past 8 years, Wang and her colleagues have focused on oxide-oxide-based nanocomposite systems. In 2008, Wang’s team developed the first VAN that grew epitaxially. The concept discovered during this work led to the understanding of the low field magnetoresistance property using a vertical tunneling concept. The team was able to make lanthanum strontium manganese oxide that changes resistivity as the magnetic field changes and then added a secondary material,

Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×
Image
FIGURE 3.2 (Left) Perfect atomic arrangement and (right) random distribution of atoms. SOURCES: Haiyan Wang, Purdue University, presentation to the workshop. H. Wang, A. Sharma, A. Kvit, Q. Wei, X. Zhang, C. C. Koch, J. Narayan, 2001. Mechanical properties of nanocrystalline and epitaxial TiN films on (100) silicon, Journal of Materials Research and Technology 16(9):2733-2738.

zinc oxide, which interacted with the conductor material. According to Wang, this enhances the magnetoresistance properties significantly through tunneling effects. Wang’s team proceeded to work on thin-film solid oxide fuel cells, which have three layers: cathode, electrolyte, and anode. A VAN was built between the cathode and the electrolyte to make the two interact with one another effectively. The higher interface density that is created between the two layers, Wang explained, leads to high-density reaction sites for oxygen reduction processes that happen away from that interface. Whereas solid-oxide fuel cells are usually made very bulky and require high operating temperatures, the thin-film structure can work more efficiently at lower operating temperatures.

At present, Wang’s team works on new material systems and functionalities in VANs, including dielectricity, ferroelectricity, magnetotransport, magnetic anisotropy, and superconductors. They are also working on perpendicular exchange bias (PEB) via VAN, which couples the ferromagnetic and anti-ferromagnetic materials in vertical fashion. Typical exchange bias is demonstrated in layered structures where there is a thickness limitation: if the thickness increases beyond approximately 0.6 nanometer, the exchange bias phenomenon is lost, Wang explained. To remedy this problem, the materials can be integrated vertically, and the thickness constraints are eliminated. Epitaxial growth is also possible in the VAN microstructure, where both the ferromagnetic and antiferromagnetic materials can be grown on strontium titanium oxide single crystalline substrates. Strain coupling is also evident in the VAN microstructure, Wang continued. Depending upon the direction of the magnetic field applied, there is a shift of the magnetization

Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×

hysteresis loop to the left or to the right with PEB. If the field is pulled up or down, the magnetoresistance response can be measured.

In the future, in-plane growth control will be a critical area of study for nanocomposite thin films—in other words, can certain nanopatterns be used to make nanocomposites grow in a more ordered way using selective growth processes? Concerning functionality, Wang said that the goal is to evolve through VAN and move to the nanoscale device. To do this, it is necessary to improve the vertical pattern integration and to couple top-down and bottom-up approaches. Wang shared some of the new concepts being explored by her team at Purdue (formerly at Texas A&M University). The first concept is a vertical tunnel junction structure; the nanostructure domain will hopefully result in a functional device design. Wang noted that scalability is always an important concern with any of these new concepts. Second, instead of continuing to work only on oxide-oxide-based systems, Wang recently expanded her work on systems based on metal-oxide nanocomposites. The team created high-density gold nanopillars in oxide-dielectric media, and pure gold pillars can be grown epitaxially within the oxide matrix. Next, Wang would like to work on seed layer growth in an ordered fashion, as she realized that metal oxide can be made in an ordered fashion. Wang also discussed the novel 2D oxide system (also called supercell structures) that results when instead of growing the materials in pillar or vertical fashion, they become single phase. Wang commented that the group will continue to work on a number of projects involving nanostructured design in the coming years.

PANEL DISCUSSION: NANOELECTRONICS

Participants:

Curt Richter, National Institute of Standards and Technology

Luigi Colombo, Texas Instruments, Inc.

Joan Marie Redwing, Professor of Materials Science and Engineering and Chemical Engineering and Electrical Engineering and Associate Director, Materials Research Institute, The Pennsylvania State University

Moderator:

Robert Pohanka, Director (Retired), National Nanotechnology Coordination Office

Innovation

Curt Richter opened the panel discussion asserting that electronics will continue to evolve only if innovation continues. He also suggested that society, especially politicians, may not fully understand the logistical and financial difficulties of moving the nanoelectronics industry forward. Richter referenced Tom Theis’s

Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×

assertion that new architectures and new devices are needed in addition to more integrated device and design teams at the development stage; the field cannot advance if new technology is simply dropped into existing devices. Richter continued by stating that as more and more devices are manufactured by foundries, the design step is value-added: people do not make a single device to be used for all applications. If engineers creating hardware start to worry more about productivity than performance, as in the software industry, progress will slow, according to Richter. He also noted that foundational research and measurement science research are needed to drive industry improvements.

Joan Marie Redwing reviewed the central themes from Tom Theis, Todd Younkin, Gerhard Klimeck, and Haiyan Wang. During her summaries of the talks, Redwing asked Younkin if Intel is interested in other materials, such as III-Vs, transition metal dichalcogenides (TMDs), or graphene, and Younkin responded that Intel is most interested in exploring the TMD space because there has already been a large global commitment to graphene. With regard to Klimeck’s talk, Redwing wondered if nanoHub will have a greater focus on process simulation in the future given that emphasis has been lacking in understanding process.

Graphene

Luigi Colombo noted that his recent work includes a study of 2D graphene, which he described as simple compared to the multicomponent compound semiconductors, high-K materials, and gate dielectrics that he has worked with previously. He added that h-BN and TMD are also difficult materials with which to work. Colombo explained that given the past experiences with compound semiconductors, the fact that III-Vs are still left out of the computation, and that the bulk of interfaces are still difficult to control, the materials and manufacturing community needs resources to study the nucleation and growth of new materials (e.g., both thin films and monolayer single crystals). When the community attempts the move to heterostructures, it will be difficult to get clean and well-controlled interfaces, so it is important to know how to manipulate and control defects early in the process, he said. Colombo returned to a discussion of graphene, a material that is inert and can be grown on copper. He suggested that high-quality materials be developed and that current device development (i.e., one at a time) is far too slow for progress to be realized at a rate generally wished for in the community.

Haydn Wadley referred to James Hone’s slide from the previous day that showed a graphene monolayer that would sustain a strain of 25 percent when pulled at the sides. He asked if people are considering that there could be a stress variable with 2D materials. Colombo noted that this is not being evaluated in TMDs but it might be in graphene. Wadley asked if it is possible to lay a thin layer of graphene over a 300-millimeter-diameter wafer, for example. Colombo said that

Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×

it is possible to do this, despite current high costs. However, Colombo noted that wrinkles can emerge when the graphene is laid on top of the copper due either to the lattice mismatch or to the large difference in thermal expansion coefficient between the two, creating a “3D” graphene surface. Richter asked if this problem would be addressed by direct growth as opposed to transfer. Colombo noted that for copper in an ultra-high vacuum system, it is difficult to grow graphene with methane, so a catalyst like oxygen is also necessary. A more important question relates to the grain size, as it is difficult to grow large (of the order of 300 mm) single crystals of graphene. Colombo also asserted that it is difficult to deposit/grow large 2D materials in single crystals on any substrate. Redwing said that TMDs have an advantage over graphene because it is unnecessary to have a catalyst to promote deposition of TMDs. Colombo agreed but noted that this process may still be difficult if the desired outcome is a monolayer and single crystal. He reiterated the need for additional funding to be directed to the study of nucleation and growth to better understand such issues.

Hardware and Materials Suppliers

Todd Younkin asked what roles the hardware and materials suppliers should have with the end users, fabrication companies, and university research centers contributing to this community. He also asked how outsiders can be educated on the importance of the materials community. Colombo suggested that although equipment suppliers have much to contribute, they will not know what to do or where to invest if they do not know exactly which materials to deposit. He continued that it is the materials community’s responsibility to have both knowledge and quality materials to develop and demonstrate devices so that the equipment suppliers will feel comfortable developing tools. Colombo said that this community is also responsible for determining whether there are other DoD applications that would benefit and whether it is possible to devote resources to crystal growth, deposition of materials, and other fundamental challenges.

Erik B. Svedberg, National Academies, asked about the opportunities to link traditional electronics with photonics. Colombo said that the Strategic Advisory Council of the Graphene Flagship, of which he is a member, is studying photonics using graphene and other materials, since graphene can be used for infrared detectors. He said that a group in Barcelona is collaborating on this with a group at Cambridge. Richter added that 2D materials are easier to incorporate, so it is easier to envision integrated photonics on a silicon platform. Haydn Wadley asked if there are solutions to address the instability of TMDs in atmospheric environments if such materials are to be incorporated into devices. Colombo responded that capping, heterostructures, and integrated processes can be used to address this issue. Richter added that faster turnaround times and faster techniques

Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×

to determine material quality are possible when using in situ approaches to grow new materials. Colombo said it would be helpful for the United States to grow its own h-BN (and other high-impact materials), as there is only one worldwide distributor, which is located in Japan. Valerie Browning noted that the line between materials and devices is blurred, so the challenge should ultimately be addressed by both the materials and the manufacturing communities. Colombo agreed that both communities are essential in this discussion and in the progress of the field; they need to understand one another and work together.

Investments and Incentives

Katherine Stevens, GE Aviation, asked if the government truly understands the consequences of not investing and thus not making advancements. Richter responded that although some areas of the government might understand the issues at hand, the elected government that is in charge of setting budgetary guidelines about research does not recognize the value of research in moving the industry forward. Daniel Gopman asked if a facility that combines a manufacturing line with boutique research and development exists, as this is one method to advance the field of nanoelectronics. Colombo acknowledged that he was unaware whether such a facility currently exists but that it would make sense to have one. It would reduce duplicated efforts and duplication of materials that do not have an immediate application; it would increase the confidence in the quality of materials and the reliability of the data; and it would ultimately save money and time, Colombo explained. Gerhard Klimeck added that unreliability is also an issue for computational studies, as the provision of reliable material and toolboxes is generally not supported. Colombo noted that material variability must be eliminated from device development. Redwing added that the material and device should work together, making it easier to identify where the material problems are before the device fails. Richter added that materials and interfaces should be characterized and tied to the final performance of the device. Colombo said it is possible that the programs are subcritical and do not have enough resources to perform well. Rosario Gerhardt pointed out the mistake of only characterizing the device for a particular application instead of looking at the individual interfaces. Metrology (i.e., materials development) is one of the most important areas of study, according to Gerhardt, even if it is not currently given the credit it deserves, because it offers a method to understand how defects affect device performance.

Ward Plummer asked why the Department of Energy, which has several national laboratories, has not offered more funding. Robert Pohanka asked what level of infrastructure funding is needed from the government and whether the government is involved in public-private partnerships to move the field’s technology forward. Colombo said that it is likely that many U.S. laboratories

Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×

already have the facilities, equipment, and talent for further research; these groups simply need to be redirected in their efforts and funded. Richter said that the problem may actually be in the lack of critical mass more than the funding. He recommended that the principal investigators (PIs) work together so that funding can be optimized and better progress can be made. The process could improve if these groups could prioritize their critical needs and develop relevant teams. Younkin advised that researchers consider the IMEC research electronics initiative in Belgium as a good example of collaboration among industry, academia, and government. He encouraged participants to consider whether something similar could be done in the United States. Colombo noted that IMEC is a stellar example of collaboration; because it is not on U.S. soil and it does not have the culture of developing new materials that the United States has, however, it may be difficult for DoD to support such projects abroad. He also reiterated the point that it is the responsibility of the materials community to educate the equipment suppliers so that they have essential basic knowledge. Younkin agreed with Colombo, and Pohanka added that there could be engagement from DoD at a place like IMEC on the research side but not on the application side. Klimeck added that this is an issue of incentives; there is no incentive for the material groups to collaborate with the device groups for single-PI grants. If DoD wants progress, they should fund centers that will foster greater collaboration.

Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×
Page 43
Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×
Page 44
Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×
Page 45
Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×
Page 46
Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×
Page 47
Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×
Page 48
Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×
Page 49
Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×
Page 50
Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×
Page 51
Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×
Page 52
Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×
Page 53
Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×
Page 54
Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×
Page 55
Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×
Page 56
Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×
Page 57
Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×
Page 58
Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×
Page 59
Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×
Page 60
Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×
Page 61
Suggested Citation:"3 Nanoelectronics." National Academies of Sciences, Engineering, and Medicine. 2019. High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop. Washington, DC: The National Academies Press. doi: 10.17226/25106.
×
Page 62
Next: References »
High-Entropy Materials, Ultra-Strong Molecules, and Nanoelectronics: Emerging Capabilities and Research Objectives: Proceedings of a Workshop Get This Book
×
Buy Paperback | $65.00 Buy Ebook | $54.99
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

High-entropy materials, ultra-strong molecules, and nanoelectronics have become a focus of active research because of their unique potential and applications. Global research is rapidly accelerating and unlocking major recent breakthroughs. It is important to highlight these recent developments and explore possibilities for future research and applications.

The National Academies convened a workshop on February 10-11, 2016 to discuss issues in defense materials, manufacturing, and infrastructure. Key topics of discussion included emerging capabilities and research objectives for ultra-strong molecules, high-entropy materials, and nanoelectronics. This publication summarizes the presentations and discussions from the workshop.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    Switch between the Original Pages, where you can read the report as it appeared in print, and Text Pages for the web version, where you can highlight and search the text.

    « Back Next »
  6. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  7. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  8. ×

    View our suggested citation for this chapter.

    « Back Next »
  9. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!