Tools and devices

Molecular electronics and nanowires

For half a century the electronics industry has made its circuits faster and cheaper by making transistors smaller. Molecular electronics asks what happens when that shrinking reaches the size of individual molecules, and whether molecules, nanotubes and nanowires could themselves become the components.

From bulk materials to molecules

The Wikipedia article on molecular electronics describes the field as the study and application of molecular building blocks for making electronic components, an interdisciplinary area spanning physics, chemistry and materials science. It distinguishes molecular-scale electronics, sometimes called single-molecule electronics, which uses individual molecules or small collections of them as components. Because single molecules are the smallest stable structures available, the article notes, using them would be the ultimate miniaturisation of electrical circuits.

The idea is older than most people assume. The same article records that the German physicist Arthur von Hippel mentioned it in 1956, proposing to build electronics from atoms and molecules rather than prefabricated materials, and that in 1974 Aviram and Ratner described a theoretical molecular rectifier: a single molecule with electron-donating and electron-accepting parts that would let current flow in only one direction, like a diode. That proposal is still cited as the conceptual starting point of the field.

Nanoelectronics and the silicon roadmap

Molecular electronics sits inside a wider field. The Wikipedia article on nanoelectronics defines it as the use of nanotechnology in electronic components, covering devices so small that inter-atomic interactions and quantum properties need to be studied closely. Among the candidate technologies it lists hybrid molecular and semiconductor electronics, one-dimensional nanotubes and nanowires, and advanced molecular electronics. It also points out that ordinary silicon transistors have already entered the nanoscale: recent manufacturing generations have critical dimensions within the 1 to 100 nm range, although the article cautions that the name of a technology node does not directly represent its minimum feature size.

That context matters when reading the literature. Many papers on molecular or nanowire devices justify themselves by reference to the limits of conventional scaling, sometimes called Moore's law after the 1965 observation, recorded in the nanoelectronics article, that silicon transistors were steadily shrinking. Whether a given alternative can compete with silicon is a question about manufacturing as much as physics, and papers that report a single working device rarely settle it.

Nanowires

Between single molecules and conventional circuits lie nanowires. According to the Wikipedia article on nanowires, they are structures with a diameter constrained to tens of nanometres or less and an essentially unconstrained length, typically with length-to-width ratios of 1000 or more, which is why they are often called one-dimensional materials. At such small diameters quantum effects become important, hence the older name quantum wires.

The article lists many kinds: superconducting, metallic (nickel, platinum, gold, silver), semiconducting (silicon and III–V compounds) and insulating (silica, titania), as well as molecular wires built from repeating units, including DNA. It adds that their conductance and mechanical strength can differ greatly from those of the bulk material. Production routes named in the article include suspension, electrochemical deposition, vapour deposition and vapour–liquid–solid growth, in which a small catalyst droplet guides the growth of a crystalline wire. A peptide-templated route, in which silver was cast inside self-assembled peptide nanotubes, is described on the self-assembly page.

Nanotubes as wires and switches

Carbon nanotubes are the best-known one-dimensional conductors, and depending on their structure they behave as metals or semiconductors. Semiconducting tubes can act as transistor channels, a line of work described on the carbon nanotubes page. Around 2001 and 2002 that work was among the most visible in the physics and materials journals, alongside semiconductor nanowire transistors and simple logic circuits assembled from them.

The measurement problem

Much of the difficulty in molecular electronics is not making the molecule but connecting to it. A single molecule is far smaller than any wire a lithography tool can draw, so researchers use tricks: probe tips that touch a molecule on a surface, gaps formed by breaking a thin metal wire, or self-assembled monolayers sandwiched between electrodes. Each method raises the same question: is the measured current really passing through one well-defined molecule, or through contacts, defects or many molecules at once? The best papers address it directly, with statistics over many junctions and control experiments with molecules that should not conduct in the same way.

For the spin-based alternative to charge electronics, see magnetic nanostructures and spintronics. For how the smallest conventional features are patterned, see making things small.