Nanoscale Engineering: Major Advances

Nanoscale tools are getting more and more functional and general-purpose. Researchers at Northwestern University have just accomplished the following:

- Designed a tiny sensitive system for applying and sensing force.
- Had samples welded to the device using a new and very powerful nanoscale manufacturing system.
- Put the device in a tunneling electron microscope (TEM), and watched the tube while they pulled it apart.

This is really impressive stuff. There are apparently three different systems involved: the sensor, the manufacturing system, and the final sensing configuration.

The sensor is a two-chip MEMS (micro electro mechanical system) device that can sense 10 nanoNewtons -- about the strength of a single strong inter-atomic bond -- using capacitive sensing, while using thermomechanical components to apply the force. It's small enough to fit into a TEM.

The manufacturing system (which was at the University of Illinois) is an integrated focused ion beam (FIB) and scanning electron microscope (SEM). This is a very flexible and powerful configuration. FIB systems shoot ionized atoms at the sample, while electron microscopes shoot electrons. The FIB atoms erode the sample. Sometimes, this is what you want: leave the beam in one spot, and you can drill a deep hole; scan it, and you can drill other shapes. You can scan the beam to perform scanning microscopy similar to SEM, but the FIB still erodes the sample -- rather quickly, for sub-micron samples.

By including an SEM in the same system, you can scan the sample non-destructively and use the FIB only where you want to mill the sample. FIB systems can mill ~8 nanometer spots with sub-nanometer precision. Both FIB and SEM can deposit material onto the sample, though I think SEM is more delicate and less disruptive to the sample. They can build quite intricate shapes, including spirals and arches, perhaps 25 nm wide. I'm guessing that the system also has a tilting and rotating stage, so the milling and deposition can take place at a wide range of angles, not just vertical.

The system also includes a nano-manipulator. Putting together microscopy, material deposition, material removal, and manipulation, this means that sub-micron pieces can be cut to a desired shape, moved into position, and welded together. It should even be possible to build devices out of heterogeneous materials. One other thing -- the FIB-deposited material can be conductive or insulating, so it's possible to use this to wire up electrical devices.

Back to what Northwestern did: Once they got back a stress-tester with a cut-to-size and carefully-positioned sample welded to it, they stuck it into a transmission electron microscope. This is a very high-resolution kind of electron microscope that works by shooting electrons through thin samples. So they could pull on the sample, measure the force, and watch it stretch and break, all at the same time. They did this for carbon nanotubes, metallic nanowires, and free-standing polysilicon films. They say it can also be used on organic materials such as DNA, proteins, and nanofibers.

Although nothing in this work is atomically precise (with the possible exception of the TEM microscopy), it's getting close. The ability to integrate MEMS, nano-manipulation, FIB, and SEM in a single manufacturing system opens a vast new array of experiments. And the ability to watch and measure a nano-shaped sample during mechanical tests will speed all sorts of nanomaterial and nanostructure research.

There are several possible paths to molecular manufacturing. Drexler favors starting with biopolymer-based systems and improving them incrementally. Freitas and Merkle propose using scanning probe microscopes to do direct mechanosynthesis of diamondoid systems. I have changed my mind about once a year as to which of these was likely to be quicker. But another researcher, Josh Hall, favors a third path: using a more traditional machining approach to build small systems that can perform increasingly precise operations. (This is similar to what was originally proposed by Feynman.) Although I have not previously seen that as a major contender, this work has gone a long way toward changing my mind.

Chris Phoenix

Tags: nanotechnology nanotech nano science technology ethics weblog blog

Comments

[Dave Howorth]

You say: "FIB systems can mill ~8 nanometer spots with sub-nanometer precision".

Could you provide a reference to systems that provide this kind of accuracy?

Thanks, Dave