SILICON-FREE SEMICONDUCTORS Nanomaterials provide alternatives for powering devices for the Internet of Things
Smartphones, tablets, laptops, TVs, drones, smartwatches – even refrigerators – all get their “intelligence” from tiny semiconductors, which carry an electric current when exposed to heat, light or electric fields. As the shrinking size of electronic devices tests the limits of silicon-based semiconductors, however, the market for conductive nanomaterials is growing.
As Internet of Things (IoT) devices decrease in size, semiconductor manufacturers are also looking for ways to make their products smaller, as well as more powerful, energy-efficient and reliable. Increasingly, that means seeking alternatives to silicon. Nanomaterials are emerging as a top contender.
Broadly defined as materials measured in billionths of a meter, nanomaterials are faster, lighter and more energy-efficient than silicon.
“Today’s silicon semiconductors are already nanotechnology – the feature size of silicon devices has reached as small as 10 nanometers,” said Aravind Vijayaraghavan, lecturer in nanomaterials at the UK’s University of Manchester. “A number of nanomaterials, like nanotubes, nanowires and nanoparticles, are being investigated for various roles in semiconductor device technology.”
Identifying alternatives is important to developing the next generation of semiconductors, said Raman Chitkara, Global Technology Industry leader at professional services network PwC.
“Emerging disruptions coming from digitalization and the Internet of Things will include smart manufacturing, autonomous cars, drones, augmented and virtual reality, robots and other new forms of artificial intelligence – and semiconductors will be an essential element of these and other major technological innovations,” Chitkara said.
Carbon nanotubes (CNTs) – hollow cylindrical tubes composed of carbon atoms that have a 1 nanometer diameter and are stronger than steel – are a promising semiconductor alternative. Although CNTs are 10,000 times thinner than a human hair, their unique structure – a large surface area relative to their ultra-small dimension – allows them to carry a current at higher speeds and detect electrical changes more precisely than silicon transistors.
“CNTs are currently in high demand,” said Andrew McWilliams, research analyst at market research firm BCC Research, based in Wellesley, Massachusetts. “The most interesting property of carbon nanotubes from a semiconductor standpoint is their extremely high electrical conductivity. Meanwhile, nanotubes’ extremely high thermal conductance helps to avoid the excessive thermal buildup associated with semiconductors.”
For example, in 2016, a team at the University of Wisconsin-Madison announced it had developed a CNT transistor that could conduct current 1.9 times higher than a comparable silicon transistor. The team predicted that CNT transistors eventually will be five times faster, or use five times less energy, than silicon transistors.
“WE CAN’T MARCH TOWARD A HYPERCONNECTED PLANET [...] WITHOUT THE CONTINUING INNOVATION IN THE SEMICONDUCTOR INDUSTRY.”GLOBAL TECHNOLOGY INDUSTRY LEADER, PWC
“This breakthrough in CNT transistor performance is a critical advance toward exploiting CNTs in logic, high-speed communications and other semiconductor electronics technologies,” said Michael Arnold, professor of materials science and engineering at the university, in a paper published in Science Advances.
Another promising material: multiferroics, which are both magnetic and ferroelectric, with reversible electric polarization. They have the potential to enhance device functionality, thanks to the special nature of “spin waves” associated with polarization in these materials.
“We will see a lot more integration of functional materials – magnetic, ferroelectric, multiferroic, 2D materials – because they can bring sufficient new capabilities to devices to make it worthwhile to deal with the manufacturing challenges of bringing in a new material,” said Caroline Ross, associate head of the Department of Materials Science and Engineering at the Massachusetts Institute of Technology (MIT) in Cambridge, Massachusetts. Ross specializes in magnetic materials and nanotechnology.
Identifying which nanomaterials could potentially be used as semiconductors takes a significant amount of research time, resources and money. However, virtual simulation and design software can provide a cost-effective tool for helping researchers to quickly model and predict the behavior of different materials at a nanoscale, pinpoint feasible options and create new design rules for engineering them into semiconductors.
Researchers at the US Department of Energy’s Argonne National Laboratory in Lemont, Illinois, for example, used a computer model to simulate the growth and electrical conductivity properties of 2D silicone, and quickly discounted it as a contender. The model has advanced since then, enabling researchers to quickly explore the semiconducting properties of other 2D materials.
“Essentially, we did virtual ‘experiments’ to optimize different variables, all at a much lower cost than in the lab,” said Badri Narayanan, Argonne materials scientist and joint lead author, when the research became public. “Now, others can avoid much of the trial and error within the lab. Instead, they can experiment using the optimized set of conditions our model predicts to best yield the structures and properties they desire.”
Additive manufacturing, also called 3D printing, holds promise for engineering the complex structures of nanomaterials. However, progress has been slow because the beneficial performance and structural properties of nanomaterials are difficult to maintain when scaling them to a workable level for use in semiconductors.
But a breakthrough in July 2016 by a team at Virginia Polytechnic Institute and State University in Blacksburg, Virginia, produced flexible, lightweight metallic nanostructures with good electrical conductivity. Digital light processing enabled the team to scale up its designs to usable size.
Industry observers have suggested that this process could, in the future, be applied to single-atom 2D materials like graphene – the thinnest and strongest known material – making it easy to mass produce. Although graphene is pliable, transparent, inexpensive to produce and has high thermal and electrical conductivity, researchers must find ways to introduce a band-gap (the distance between the valence band of electrons and the conduction band) to the material to transform it into a semiconductor.
“Graphene is not a traditional semiconductor so it won’t directly replace silicon, but new tunneling transistors that combine graphene with other 2D materials could potentially replace silicon devices,” the University of Manchester’s Vijayaraghavan said. “Other graphene-like 2D materials that have a band-gap could also be used to fabricate electronic devices. The €1 billion European Graphene Flagship – the European Union’s academic-industrial consortium for graphene research – is the biggest concerted effort on this topic. Looking beyond conventional computer chips, graphene could be used in quantum computing, but this research is still in its infancy.”
Graphene may be a relatively new material, but researchers have been quick to exploit its qualities. For example, the Norwegian University of Science and Technology has grown semiconductor nanowires on graphene to create 1 micrometer-thick hybrid material that could act as a semiconductor in solar cells, LED components, sensors and batteries.
According to MIT’s Ross, a major barrier to progress in engineering semiconductors with new nanomaterials like graphene is the difficulty of integrating them into current manufacturing processes. “Even materials with obviously good properties have to be produced in a way that is compatible with the underlying silicon, which dominates the industry,” she said.
Now, graphene is being used to produce conductive inks that can be 3D printed, enabling manufacturers to integrate semiconductors and electrical circuits for use in IoT devices. In August 2017, for example, the University of Manchester unveiled a flexible, battery-like device that can be screen-printed directly onto washable textiles with conductive graphene-oxide ink.
“The development of a graphene-based, flexible textile supercapacitor using a simple and scalable printing technique is a significant step toward realizing multifunctional, next-generation wearable e-textiles,” Nazmul Karim, who is Knowledge Exchange Fellow at the UK’s National Graphene Institute and co-author of the research, said when the technology was announced. “It will open up possibilities of making an environmental-friendly and cost-effective smart e-textile that can store energy and monitor human activity and physiological condition at the same time.”
Using graphene-based ink in radio-frequency identification (RFID) antennas is another promising area of activity for IoT applications. “If we can commercialize graphene inks that are suitable for printing the entire RFID tag, they may occupy a sweet spot of cost, conductivity and other properties that will enable them to carve out a growing share of the market for conductive inks,” BCC’s McWilliams said.
Despite early signs of success and rapid nanotechnology development, the adoption of nanomaterial semiconductors in mainstream devices may be at least a decade away.
“The most significant barrier to innovation in semiconductor materials is the cost for the integrated circuit industry to introduce new semiconductor materials and the corresponding new equipment and technology,” said Luo Jun, professor of the Integrated Circuit Advanced Process Center at the Institute of Microelectronics of Chinese Academy of Sciences.
Those challenges must be solved, PwC’s Chitkara said. “We can’t march toward a hyperconnected planet where people and devices can communicate with each other at increasingly higher speeds, with greater reliability and at rapidly declining costs, without the continuing innovation in the semiconductor industry.”Retour en haut
Learn how to create the core of next-generation electronics through virtual design