An international team of researchers has developed a new way to pattern complex 3D arrays of wires and interconnects on a microchip with the help of self-assembling block co-polymers. The technique could be used to pack more electronic components onto a memory chip – an important advance as device feature sizes continue to shrink.

The team, led by Caroline Ross and Karl Berggren of the Massachusetts Institute of Technology (MIT) in the US, has shown that a block co-polymer – polystyrene-polydimethylsiloxane – can be "forced" to form a complex set of 3D patterns on a substrate surface. Block co-polymers are made of blocks of different polymerized monomers. The technique, detailed in the journal Advanced Materials, involves using a simple template comprising an array of small pillars made of silica, explains team member Amir Tavakkoli. "We have found that we can create a rich variety of microdomain morphologies on a single substrate by modifying the layout of the pillars," he says. "One such morphology is a high-resolution square lattice of dots, but cylinders, spheres, ellipsoids and double cylinders can easily be produced too."

Smaller the better

The spaces between features in the final patterns can be smaller than the original periodicity in the block co-polymer, Tavakkoli says. This means that the number of components that can be packed onto the sample substrate is increased. Indeed, the feature sizes that can be made in this way are very small, at about 10–20 nm. In contrast, those produced by conventional photolithography are at least 10 nm bigger. "Being able to fabricate such small structures will be important in the future because feature sizes are continuing to shrink, in accordance with Moore's law," he adds.

Shapes in a box

Another big advantage of the new technique is that it can produce square and rectangular structures. These shapes are the basis of most microchip layouts but are quite difficult to produce through conventional self-assembly processes. "When molecules self-assemble, they have a natural tendency to create hexagonal shapes – as in a honeycomb or an array of soap bubbles between sheets of glass. They do not naturally form squares or rectangles," says Ross.

The MIT researchers' fabrication technique starts with the construction of a precisely controlled pattern of nanopillars on a silicon substrate surface using high-resolution electron-beam lithography. Next, the pillars are chemically coated with a thin polystyrene "brush" layer that subsequently interacts with the block co-polymer when it is applied to the substrate surface. The co-polymer then self-assembles into a pattern that is guided by the pillars.

Repellent shapes

The process works because the template coating is arranged in such a way as to repel one of the components in the polymer. This produces a significant amount of strain in the polymer, forcing it to twist and turn. "In doing so, the polymer rearranges itself on the substrate surface into more interesting patterns," explains Berggren.

The team says that it now plans to investigate how to remove the physical post template from the final pattern, and then transfer the pattern to the substrate. "We would then like to make some functional devices and we also want to understand and model the self-assembly process so that we can generalize this work to other block co-polymers and feature geometries," Tavakkoli told