Last year, Erez Braun and colleagues developed a “sequence-specific molecular lithography” technique. The researchers harnessed a basic biological process - homologous recombination that mixes genes in cells - to manipulate DNA. This enabled them to create sequence-specific DNA junctions and networks, to coat DNA with metal in a sequence-specific manner and to localize molecular objects on any address on a DNA molecule.

Now, the scientists have built on this work to assemble a carbon nanotube field-effect transistor. They used a three-strand homologous recombination reaction between a long double-stranded DNA molecule and a short auxiliary single-stranded DNA. These DNA molecules encode the information to guide the assembly process: the short molecule has a sequence identical to the long one at the desired location of the transistor.

First, the team polymerized RecA - a major protein responsible for genetic recombination in bacteria - onto the short molecules to form nucleoprotein filaments. These nucleoprotein filaments then bound to the long molecules at the designated location, according to the sequence matching between both molecules.

Next, the scientists functionalized single-walled carbon nanotubes with the protein streptavidin, which served to locate the nanotube at the correct address. The team stretched the DNA/nanotube assembly on a passivated oxidized silicon wafer prior to a metallization process, which coated the DNA molecules with gold. The RecA doubled as a sequence-specific resist, so that the active area of the transistor remained uncoated. Moreover, because the nanotube was longer than the gap caused by the RecA, gold covered the ends of the nanotube and created contacts to the transistor.

Braun and colleagues made 45 devices in total. Fourteen acted as field-effect transistors with partial or full gating, and ten conducted but could not be gated - probably because they contained metallic rather than semiconducting nanotubes.

“Carbon nanotubes will be one of the major future building blocks of molecular electronics due to their small dimensions and excellent electronic properties,” said Braun. “However, you cannot self-assemble a circuit directly with nanotubes because they lack recognition. Our research demonstrates that you can harness biology to self-assemble nanoelectronics.”

Although it is too early to predict applications, the researchers now plan to construct a device on a DNA junction. This will allow for more complex logic circuits.