Researchers up reliability of nanoscale circuit measurements
By R. Colin Johnson
EE Times
(10/29/01 12:00 p.m. EST)
TEMPE, Ariz. — Researchers at Arizona State University report that they have created a method of "soldering" metal electrodes to each end of a single molecule, creating "through-bond" electrical contacts that enable reproducible measurements to be made of molecular wires and other nanoscale devices. Initial tests on the conductivity of 1-nanometer-long single-molecule "wires" are currently being expanded to characterize the electrical properties of potential single-molecule circuit elements including diodes, transistors and logic gates.
"This is work that is pointed at a fundamental problem in using single molecules as components in electronic circuits," said Devens Gust, chairman of the chemistry department at Arizona State University. "Quite a few people around the world are trying to design molecules that might act as wires, transistors and logic gates . . . but one of the problems that faced them was how do you go about wiring those molecules into the circuit so you can see how they behave electrically."
Gust collaborated with ASU professor of physics Stuart Lindsay; ASU physicists Xiadong Cui, John Tomfohr and Otto Sankey; ASU chemists Alex Primak, Xristo Zarate, Ana Moore and Thomas Moore; and Motorola Inc. scientist Gari Harris.
Elusive characterization
Nanotechnology researchers have designed a wide variety of single-molecule devices to replace the macroscopic transistors and logic gates used in today's chips, but so far have been unable to accurately characterize their electrical behavior.
Because of the small size of those devices, even the most advanced methods have proved to be unreliable, resulting in molecular circuits that work sometimes, but the results often are not reproducible by other researchers. Even the measurements of the same devices have often not been reproducible, because, according to Gust, measurements cannot be narrowed down to an individual molecule.
"People have tried different approaches to this, but what we have done is work out a way to reliably measure single-molecule conductivity and show that we are in fact dealing with one molecule," Gust said. "People have been measuring the properties of large ensembles of molecules as circuit elements, but there is no guarantee that when the molecule is isolated that it will function on its own in the same way that it does when it's packed together with a whole bunch of other molecules."
Progress in the development of nanoscale devices has thus been hampered by researchers' inability to reproduce measurements, plus an inability to guarantee that measurements are being made on individual molecules. In order to test out designer molecules, characterize their electrical properties and insert them into nanoscale circuits, the industry has until now sought in vain to create a reliable testbed.
"When you look at macroscopic circuits, it is pretty clear to everybody that the best way to make contact is to solder all your components at each end, and not to just stick them into a breadboard where you only have mechanical contacts," Gust explained. "What we have been able to do is essentially the same thing at the molecular level. We basically take a molecular component and solder it to a metal electrode at each end, [thereby] creating a more robust and more highly conductive circuit than when you just push on the end of the molecule with a piece of metal."
Past attempts to electrically characterize single-molecule devices have been "all over the map," according to Gust. And worse, even individual measurements have not been reliably reproducible, making it impossible to know whether a device is working properly.
In addition to being reproducible, Gust's soldering method enables researchers to determine exactly how many molecules have been connected to each individual electrode.
"What we have done in our work is really two things: We've been able to take a molecule and attach gold electrodes to both ends, and we have been able to electrically address that single molecule and show that it's not some large, undefined number of molecules. Then we were able to unambiguously measure the properties of one molecule chemically bonded to the metal electrodes at each end," said Gust.
To accomplish their goal, the researchers deposited gold on an atomically flat surface of mica to form one electrode, after which they deposited a layer of long octanethiol insulator molecules to it through chemical bonds. Using a solvent, they then removed just a few of the insulator molecules and replaced them with 1-nm-long electrically conductive hydrocarbon molecules (1,8-octanedithiol) to act as the molecular wire. The wires were then "soldered" to the underlying gold with a chemical bond.
At the free end of the wires, the team then soldered 2-nm round gold particles to act as the second electrodes to the wires. A gold-coated atomic-force microscope probe was then run across the surface of the free ends of the molecular wires and their conductivity was measured. The team took 4,000 individual measurements to create a histogram of the conductivity of the wires.
Molecular proof
"One of the main advantages here is that we have been able to show by our histogram of conductivity that we really are looking at one molecule, whereas in previous work it has been hard to know how many molecules are actually bridging the contact," Gust said.
Since it is impossible to determine by inspection whether the end of the probe is touching an individual molecule or several adjacent ones, Gust had to infer this from analysis of the histograms.
Analysis proved that all 4,000 measurements fell into five distinct conductivity curves, each of which was an integer multiple of a single fundamental curve. Gust inferred from this that the fundamental curve was a measurement of an individual molecule, and the other four integer-multiple curves resulted when the probe was touching two, three, four or five molecules, respectively.
The relationship between the curves was to provide progressively more current-carrying capability, in precise integer multiples, each time an additional wire was added to the parallel circuits, resulting from the probe touching more than one wire end at a time. Using this method, Gust claims to now be able to characterize the electrical properties not only of wires, but also of any single-molecule electronic component.
"What we have now is a way to reliably measure the electrical properties of single molecules, so next we want to do two things: We want to look at different kinds of molecules now — not just conductors and molecular wires, but also transistors, diodes and the like. Then, once we have measured their electrical properties, we want to do parallel calculations and try to understand theoretically why they are behaving as they are. Once we can do that, then we can design much more complicated circuits that will be useful someday," Gust said.
Gust predicts that it will take his team two years to characterize the electrical properties of the molecules that have already been designed to act as nanoscale electronic devices, after which he plans to begin building useful circuits.
Meanwhile, the publication of the results of his molecular soldering method enables other research groups worldwide to pursue parallel development efforts.
Test time
"There are a number of molecules that have been designed by other people that should function as diodes and transistors, but the problem is figuring out how they work on the one-molecule scale. That's the thing that people haven't had — a testbed to reliably evaluate their [molecules'] performance.
"So what we will now do is take those ideas and make molecules that we can insert into our apparatus and find out how they perform at the one-molecule level," Gust said. "We hope within the next two years to be able to expand this to the point where we have evaluated various molecular-scale circuit components and can start building useful circuits."
An audio recording of reporter R. Colin Johnson's full interview with Arizona State's Devens Gust can be found online at AmpCast.com/RColinJohnson. |
