Linear Arrays
A growing number of electronics researchers are fascinated by the potential of single-walled carbon nanotubes—one-atom thick sheets of graphite rolled into seamless cylinders. These tiny objects have attractive electrical properties and physical features. But incorporating single-walled carbon nanotubes into scalable ICs has proven to be a challenge, largely because of difficulties in manipulating and positioning the tiny objects and in achieving sufficient current outputs.
Engineers at the University of Illinois, Lehigh University, and Purdue University studied this problem and developed an approach that uses dense arrays of aligned and linear nanotubes as a thin-film semiconductor material. According to John Rogers, a University of Illinois professor of materials science and engineering, the nanotube arrays can be transferred to plastic and other unusual substrates.
To create the arrays, the researchers begin with a wafer of single-crystal quartz, upon which they deposit thin strips of iron nanoparticles. The iron acts as a catalyst for the growth of carbon nanotubes by chemical vapor deposition. As the nanotubes grow past the iron strips, they lock onto the quartz crystal, which then aligns their growth (Fig. 3).
According to William Dichtel, a project researcher, the new memory is based on a series of perpendicular nanowires, similar to a tic-tac-toe board, with 400 bottom wires and another 400 top wires crossing the bottom (Fig. 1). Sitting at each crossing of the tic-tac-toe structure, and serving as the storage elements, are approximately 300 bistable rotaxane molecules. These molecules can be switched between two different states. Each crossbar junction can be addressed individually by controlling the voltages applied to the appropriate top and bottom wires, forming a bit at each nanowire crossing.
A rotaxane is a molecule with a dumbbell-shaped component, consisting of a rod section and two stoppers, encircled by a ring. It has the potential to be a molecular abacus. The molecule can act as a switch by inducing the ring to slide from one side of the rod to the other. "It's not just a simple storage of charge, like most of today's memory devices use," Dichtel said. "There are actually molecules which respond to applied voltage and the molecules switch from one state to another."
The bistable rotaxane molecules used in the crossbar memory can be switched at very modest voltages from an "off" (low conductivity) state to an "on" (high conductivity) state. The stoppers for the rotaxane molecules are designed to allow the molecules to be organized into layers that are a single molecule thick, after which they're incorporated into the memory device. The 160-kbit molecular memory was fabricated at a density of 100 billion bits per square centimeter—a density predicted by Ditchel for commercial memory devices by approximately 2020.
"One of the most exciting features of this research is that it moves beyond the testing of molecular electronic components in individual, nonscalable device formats and demonstrates a large, integrated array of working molecular devices," Dichtel said.
The resulting linear arrays consist of hundreds of thousands of nanotubes, each approximately 1 nm in diameter, and up to 300 µm long. The nanotubes are spaced approximately 100 nm apart. The arrays function as an effective thin-film semiconductor material in which a charge moves independently through each of the nanotubes. In this configuration, the nanotubes can be integrated into electronic devices in a straightforward fashion by conventional chip-processing techniques.
"Our approach has been not to figure out how to grow electronically identical tubes everywhere, or to position or control the location of any individual tube, Rogers said, "but to instead make a thin form that consists of very well aligned arrays of individual tubes." The researchers used the technique to build and test a variety of transistors and logic gates, as well as to compare the properties of nanotube arrays versus individual nanotubes.
Rogers notes that nanotube arrays aren't likely to replace silicon, but could be added to a silicon chip and exploited for special purposes. Higher-speed operation, higher power capacity, and linear behavior are a few options for enhanced functionality, and they can also be used for applications that silicon can't easily support, such as flexible devices.
Nano Now
Mamikunian believes that nanotechnology, and the new and enhanced products the field promises to generate, will open an almost endless array of fresh opportunities for the electronics industry. Furthermore, the field's benefits won't only be limited to major industrialized nations. "A lot of developing countries are looking at nanotechnology as a way to leapfrog from where they are today to become more of a player in the larger state of commercialization and manufacturing," he said.
According to Mamikunian, enough progress is being made in nanotechnology research that the field is on the verge of becoming mainstream. "Over the next five years or so you'll begin to see this transformation where the terms ‘nano' and ‘nanotechnology' will fade away and it will be just ‘technology,'" he said.
Maynard agrees with Mamikunian, adding, "One thing is certain, and that is that the next generation of electronics has got to rely on our ability to engineer at the nanoscale."
Reprints
