Block copolymers (BCPs) can be used to make well-defined periodic nanopatterns with feature sizes ranging from 5-50 nanometres. Directed self-assembly of these BCPs by pre-pattern templates could extend patterning capabilities beyond the resolution limits of conventional optical lithography and so produce components with feature sizes well below the sub-20 nm regime.

A team of researchers led by Tamar Segal-Peretz has investigated the three dimensional structure of P2VP-b-PS-b-P2VP BCP assembled on pre-pattern templates using solvent vapour annealing. The researchers used the three dimensional data to analyse the interfacial fluctuations between the BCP domains and found that these fluctuations are not constant, as previously thought, but vary through the depth of the film. Such an analysis is important, explains Segal-Peretz, since the interfacial fluctuations between the BCP domains can create edge roughness, which will affect the quality of the produced nanopatterns.

3D metrology is crucial

"We also found that when the combined chemical and topographical pre-pattern templates were incommensurate with the BCP spacing, underneath the 'perfect' BCP surface, there were lots of defects and curved features," she adds. "Most metrology tools in semiconductor lithography examine only the surface of films and they thus cannot detect defects underneath the surface. Our research shows that 3D metrology is crucial for correctly probing BCP domains deeper within the polymer film, something that will ultimately lead to better design and fabrication of nanometric patterns.”

The team obtained its results using STEM tomography, which is a technique to obtain 3D information from a series of 2D images. The sample is imaged using a TEM at a series of tilt angles and a series of images is collected. This series is then used to reconstruct the 3D structure of the sample through an iterative algorithm.

Combining TEM tomography with Monte Carlo-based computer simulations

“We next performed coarse-grain Monte Carlo-based computer simulations in which each polymer is represented as a bead-spring chain,” says Segal-Peretz. “The energy of the system is defined by theoretical equations and the simulations are allowed to come to equilibrium over several thousand Monte Carlo turns until persistent structures have formed. These simulations complement the experimental data and provide further insights on the 3D structure and the interactions between the BCP and the pre-pattern template.

“Our work is extremely relevant to directed self-assembly of BCP, which is considered as the leading technology in advanced lithography for semiconductor nanopatterning,” she tells nanotechweb.org. “BCP directed self-assembly can produce feature sizes as small as 5 nm and overcome limitations in current patterning technology based on photolithography. Our research expands our knowledge on how to design pre-patterned templates to direct the self-assembly of BCPs.”

"Nice step forward"

Directed self-assembly of block copolymers has a great deal of potential for patterning surfaces, but there is still a long way to go in understanding how to reign in the fuzziness that polymers tend toward, comments Robert Grubbs of Stony Brook University in New York, who was not involved in this study. “This new work presents a nice step forward in using a combination of cutting-edge techniques in a way that will allow the many factors that affect patterning to be better understood and controlled.”

Segal-Peretz says that she and her colleagues are now busy trying to find ways to further decrease the feature sizes possible with BCP-directed self-assembly and how to reduce fluctuations and roughness in BCP films. “We are also looking into how to design and fabricate 3D structures for patterning, and how to probe these structures using a combination of TEM tomography, molecular simulations and X-ray characterization."

The experiments are described in ACS Nano DOI: 10.1021/acsnano.6b05657.