The demand for smaller devices with better performance has driven the development of carbon nanotube-based chips, which open up exciting possibilities for the semiconductor industry. A carbon nanotube is basically a sheet of graphene, which has carbon atoms arranged in a hexagonal pattern thus forming a sheet having thickness of a single atom. This sheet is rolled up in the shape of a cylinder to form a nanotube.
Carbon nanotubes, which were discovered in 1991, are the largest cylindrical nanostructures known to exist and they are ideal for designing tiny things such as chips. Similar to silicon, these nanotubes are also good semiconductors making them very useful in electronic design.
IBM’s New Carbon Nanotube Chip
Nature Nanotechnology recently reported experiments by scientists at IBM that demonstrate various methods to achieve computer chips with better performance. Though carbon nanotubes were known to have better electronic properties compared to existing silicon-based devices, barriers in manipulation of these tubes have been a stumbling block in producing chips based on carbon nanotubes.
To overcome the challenge of Integration of several billion nanotubes into a single chip, researchers "double-dipped" the nanotube chip in two solutions and created a two-part epoxy, in which nanotubes were firmly bound to hafnium regions but not to the silicon on the chip. This gave rise to several nanotubes aligned in series with every square centimetre having a billion nanotubes.
The IBM research group has designed a massive transistor array, with each array having six nanotubes at a distance of 10 nm apart. This model is said to offer a 10-fold performance increase at one-third of the power consumption of existing devices. Although this is a great improvement from existing methods, the research team feels that a lot more work is needed to find out better ways to manipulate these nanotubes of various size and shapes.
IBM researcher Hongsik Park observes different solutions of carbon nanotubes. Image credit: IBM News Room
The Road to Carbon Computing
The use of carbon nanotubes in computing has been debated and predicted at several global meetings and conferences in the past.
For instance, the 2008 Condensed Matter and Materials Physics conference held by the institute of Physics discussed about the future of computing and predicted that carbon nanotubes will replace silicon in the semiconductor sector in a few years. The conference highlighted that silicon may not sustain the demand to produce miniaturized electronic devices. Leeds University researchers also reported that carbon nanotubes conduct electricity and can be ideal for use in electronic circuitry in the future. Other talks at the conference discussed the short-comings of silicon computers and the need for a better technology to improve computing speed.
Similarly, in 2011, IBM scientists demonstrated their nanotube and graphene prototypes that help integration of computing and electronic devices at the IEEE International Electron Devices Meeting. The The research team has developed a memory device featuring CMOS technology on a 200 mm wafer, leading to development of sophisticated data-centric computing device, which can store high volumes of data that can be processed in less than a fraction of a second. The team displayed its carbon nanotube transistor and had also predicted that it will revolutionize computing technology in the coming years.
IBM SEM image of carbon nanotubes deposited on a trench coated in hafnium oxide (HfO2) showing extremely high density and excellent selectivity (scale bar: 2 μm). Credit: IBM News Room
Other Nanotechnologies for Computing
Thanks to the several discoveries in the microchip industry, the CMOS technology continues to rule this industry. However, the need for new materials and designs to improve the performance of electronic devices is becoming increasingly evident.
Apart from carbon nanotubes, experts list out the following as the likely successors to the silicon-based CMOS technology in the coming years.
Graphene has been in the news for its unique electronic properties and its possible use as transistors in next generation nanoelectronic devices. As new, feasible methods for separation and fabrication evolve, graphene is expected to play a key role in computing devices very soon.
Photonics is another emerging technology that uses photons instead of electrons and is again facing problems with commercial production due to the huge expenses in producing exotic optical components. This technology, once in the market, is claimed to have the potential to offer great advantages in terms of computing speeds and power consumption.
The aim of molecular electronics is to produce transistors and logic gates using single molecules. As with many other latest technological innovations, there are several barriers in mass producing these molecular components though these are said to have extraordinary benefits such as excellent power savings and good data density.
Ferromagnetic materials have nanoscale domains, which are being studied extensively to assess the possibility of their use in microprocessors and data storage, with the help of their magnetization properties. Their intrinsic properties render high stability to data stored in them as their domains are radiation-resistant and non-volatile.
Although these technologies are at their infancy, the huge demand for smaller and faster devices in the computing industry will hopefully help drive more research and development, which will lead to commercialization of these techniques in the near future.
Meanwhile, the CMOS technology is being fully exploited by microprocessor manufacturers. In early 2012, Intel introduced its Ivy Bridge series featuring a 22-nm architecture having a 3D "fin" design. They also intend to introduce processors based on 14nm architecture in a couple of years time.
Regardless of the technology adopted, the future is bright for nano- and microprocessors and it is just a matter of time before these find their way to the mainstream semiconductor market.