Large-scale parallel arrays of silicon nanowires via block copolymer directed self-assembly.

Extending the resolution and spatial proximity of lithographic patterning below critical dimensions of 20 nm remains a key challenge with very-large-scale integration, especially if the persistent scaling of silicon electronic devices is sustained. One approach, which relies upon the directed self-assembly of block copolymers by chemical-epitaxy, is capable of achieving high density 1 : 1 patterning with critical dimensions approaching 5 nm. Herein, we outline an integration-favourable strategy for fabricating high areal density arrays of aligned silicon nanowires by directed self-assembly of a PS-b-PMMA block copolymer nanopatterns with a L(0) (pitch) of 42 nm, on chemically pre-patterned surfaces. Parallel arrays (5 × 10(6) wires per cm) of uni-directional and isolated silicon nanowires on insulator substrates with critical dimension ranging from 15 to 19 nm were fabricated by using precision plasma etch processes; with each stage monitored by electron microscopy. This step-by-step approach provides detailed information on interfacial oxide formation at the device silicon layer, the polystyrene profile during plasma etching, final critical dimension uniformity and line edge roughness variation nanowire during processing. The resulting silicon-nanowire array devices exhibit Schottky-type behaviour and a clear field-effect. The measured values for resistivity and specific contact resistance were ((2.6 ± 1.2) × 10(5)Ωcm) and ((240 ± 80) Ωcm(2)) respectively. These values are typical for intrinsic (un-doped) silicon when contacted by high work function metal albeit counterintuitive as the resistivity of the starting wafer (∼10 Ωcm) is 4 orders of magnitude lower. In essence, the nanowires are so small and consist of so few atoms, that statistically, at the original doping level each nanowire contains less than a single dopant atom and consequently exhibits the electrical behaviour of the un-doped host material. Moreover this indicates that the processing successfully avoided unintentional doping. Therefore our approach permits tuning of the device steps to contact the nanowires functionality through careful selection of the initial bulk starting material and/or by means of post processing steps e.g. thermal annealing of metal contacts to produce high performance devices. We envision that such a controllable process, combined with the precision patterning of the aligned block copolymer nanopatterns, could prolong the scaling of nanoelectronics and potentially enable the fabrication of dense, parallel arrays of multi-gate field effect transistors.

Nanofabrication of broad-band antireflective surfaces using self-assembly of block copolymers.

We present a simple and cost-effective method for the fabrication of antireflective surfaces by self-assembly of block copolymers and subsequent plasma etching. The block copolymers create randomly oriented periodic patterns, which are further transferred into fused silica substrates. The reflection on the patterned fused silica surface is reduced to well below 1% in the ultraviolet, visible, and near-infrared ranges by exploiting subwavelength nanostructures with periodicities down to 48 nm. We show that by choosing the appropriate block copolymers and pattern transfer parameters the optical properties of the antireflective surface can be easily tuned, and the spectral measurements verify a significant reduction of the reflectivity by a factor of 10. The experiments, confirmed with simulations based on rigorous diffraction theory, also show that the tapered shape of the nanostructures gives rise to a graded index surface, resulting in a broad-band antireflective behavior.

Graphoepitaxial assembly of asymmetric ternary blends of block copolymers and homopolymers.

Ternary blends of cylinder-forming polystyrene-block-poly(methyl methacrylate) block copolymers and polystyrene and poly(methyl methacrylate) homopolymers were assembled in trench features of constant width. Increasing the fraction of homopolymer in the blend increased the spacing and size of block copolymer domains, which were oriented perpendicular to the substrate to form a hexagonal lattice within the trench. The number of rows of cylinders within the trench was controlled by the blend composition. Depending on the domain size and spacing, the hexagonal lattice was stretched or compressed perpendicular to the trench walls but not perturbed parallel to the walls, indicating a decoupling of the perturbation in the perpendicular and parallel directions. The row spacing was uniform across the trench as a function of position from the trench wall. The results are compared with an analytical model and with Monte Carlo simulations.