Assembly of Laterally Coupled Molecular Nanostructures

Organic molecule-based compounds have tremendous potential for use in sensors, lowdielectric films, light-emitting devices, polymer coatings, and electronics. Many of these applications will benefit from revolutionary enhancements if the fundamental components can be reduced to nanometer lateral scales. It is unlikely that any fabrication method requiring each molecule to be positioned individually will be practical. Therefore, organic molecular nanostructures will require a high degree of self-assembly, both in bonding to the substrate and in intermolecular bonding between nanostructures. The development and use of atomic or nanometer-scale molecular structures where the function derives from a characteristic of an isolated molecule or a small collection of coupled molecules (in zero or one dimension) will require the following advances:

1. The ability to anchor a structure in a defined location, preferably without anchoring every molecule; and
2. The ability to laterally assemble and chemically couple (if desired) molecules between anchor points.

In this program, we are establishing new methods and chemistries to anchor and laterally couple organic molecules into nanostructures (illustrated in the cartoon).

The assembly of laterally coupled molecular nanostructures will require the development of methods for anchoring the structures and for coupling them to each other and to other structures (such as contact pads) with as much self-assembly as possible. Our work is proceeding on a number of fronts. In addition to developing new techniques to create the nanoscale patterns that define the anchor points, we are also establishing the chemistry required to selectively anchor, assemble, and couple molecules of interest on substrates of interest. This research involves an interdisciplinary combination of organic chemistry, gas-surface reactions, semiconductor surface chemistry, scanning probe microscopy, scanning probe lithography, nanostructure fabrication, and nanoelectronics.

To create templates for growth, we are using both active patterning and controlled nanoparticles growth. Active patterning is being achieved using a technique known as "dip-pen" nanolithography (DPN, developed by Mirkin and collaborators at Northwestern University). In DPN, an atomic force microscope (AFM) tip that has been precoated with loosely bound molecules is used to "write" those molecules in nanoscale patterns on a substrate. As we learn more about the fundamentals underlying the transfer mechanism, we are discovering new combinations of materials that can be written. For example, by writing a pattern of reactive molecules and then surrounding that pattern with inert molecules, we find we can simply dip the sample into a beaker containing a solution of another molecule, and it will self-assemble onto the reactive pattern. In the 3x3 µm2 image shown, a 100-nm-wide, 2-nm-thick spiral of a biocompatible polymer (poly(ethyl)amine) has been deposited this way.

Nanoparticles of relatively uniform size and shape can also be deposited without patterning if the substrate surface chemistry is well understood. We have found that reacting a silicon surface with hydrogen after first depositing a small amount of copper creates uniform defects that can act as nucleation sites. As shown in the scanning tunneling microscopy image (right), when additional metal is deposited, quite uniform nanocrystals will selfassemble over the defects, about 10 nm in diameter in this case. In the next step, these nanocrystals will serve as anchor points for organic molecules deposited from solution.

Both active and passive methods of laterally assembling and chemically coupling molecules are being pursued. Organic molecules capable of electrical conduction when polymerized are being deposited on semiconductor crystal surfaces that act as natural templates for wire formation. For example, the surface atoms on the basal plane of silicon selfassemble into rows of dimers (left). By adsorbing monomers that are sized to adsorb across the rows and also have reactive end groups, (diiodophenanthrenequinone, in the example shown), molecular wires will then be "zipped" together by a photoreaction. An alternate approach we are also pursuing uses conventional lithography to create shallow trenches in quartz (not shown). Conducting polymer films are then selectively grown only on the side walls. This approach already shows promise for chemical sensing applications.

Fascinating and useful structures can be passively coupled by taking advantage of the unique properties of single-wall carbon nanotubes, long cylinders of graphite one atomic layer thick and about 1 nm in diameter. We have discovered that a randomly assembled network of nanotubes behaves like a semiconducting thin film that is compatible with conventional microprocessing technology. Within the network the individual nanotubes are interwoven into an electrically continuous film that can be processed into devices and circuits. As shown in the figure below, electrical contacts can be deposited on top of the random network to easily create a practical device (an AFM image of the network is shown along with a schematic of the structure of a single nanotube). In this way, these networks have already been incorporated into transistors and chemoresistive chemical sensors with properties that are superior to commercial electronic materials.

Contact the Principal Investigator, Lloyd Whitman, for more information