Templated self-assembly of block copolymer, based on topographic templates defined by electron-beam lithography (EBL), is an attractive candidate for next generation high-resolution lithography. Templated self-assembly has two advantages compared with other lithography methods: first, the resolution can be scaled down to 5 nm, which cannot be achieved by optical lithography; second, the throughput can be increased by several folds compared with EBL. In our previous study, complex sub-20-nm patterns were fabricated with 45.5 kg/mol poly(styrene-block-dimethylsiloxane) (PS-b-PDMS) block copolymer ^[1] .

Here, we demonstrate high throughput sub-10-nm fabrication by using templated self-assembly of block copolymer. To achieve 10-nm resolution, the dimensions of a block copolymer and a topographic template were scaled down to 10-nm-length scale. We used 16 kg/mol PS-b-PDMS block copolymer, which yields 9-nm half-pitch PDMS cylinders. To control the orientation of 9-nm half-pitch PDMS cylinders, rectangular lattices of posts with height of 19 nm, diameter of 8 nm, and various periods were fabricated and annealed with the block copolymer. As a result, PDMS cylinders formed a long-range ordered region when the post array satisfied the commensurate condition. By varying the periods of posts, a broad range of block copolymer lattice orientation angles was achieved (Figure 1).

On a lattice with the period larger than 72 nm, PDMS cylinders lost long-range order. To further decrease the density of the posts and therefore increase the throughput without losing long-range order, a sparse lattice of dashes was tested. As a result, a region of well-aligned PDMS cylinders with width of 708 nm was achieved (Figure 2d). The dashes occupy only 1/66 of the final PDMS line pattern. This result suggests that if instead of writing the complete pattern, EBL is used to create template arrays and the pattern is then completed by a block copolymer, the throughput of EBL could be increased dramatically.

1. J. K. Yang, Y. S. Jung, J. Chang, R. A. Mickiewicz, A. Alexander-Katz, C. A. Ross, and K. K. Berggren, “Complex self-assembled patterns using sparse commensurate templates with locally varying motifs,” Nature Nanotechnology, vol. 5, pp. 256-260, Mar. 2010.

Triblock terpolymers are interesting because they can form a much wider diversity of 3D structures than diblock copolymers, including rings and square-symmetry patterns, which may be useful in nanolithography. Two examples of pattern formation in triblock terpolymers have been investigated, showing square arrays of dots from PI-PS-PFS ^[1] and close-packed arrays of rings from and PS-PFS-P2VP ^[2] . Square patterns are of particular interest for structures such as arrays of vias. In the PI-PS-PFS triblock terpolymer, the minority blocks (PI and PFS) form cylinders alternating with square symmetry. Oxygen etching removes the PI and PS, leaving oxidized PFS arrays with a period of 44 nm. On a smooth substrate, the correlation length of the square pattern is increased dramatically to several microns by the use of brush layers and specific solvent annealing conditions. The interaction between the square pattern and nanoscale topographical trenches and posts can be controlled by substrate functionalization, templating the structure.

For the PS-PFS-P2VP, the bulk structure consists of P2VP core-PFS shell cylinders in a PS matrix. Removing the P2VP and PS leaves ring-shaped PFS features. The cylinders can be oriented perpendicular to the top surface of the film by controlling the film thickness and annealing conditions. However, the cylinders typically lie in plane at the substrate-film interface, and the ring pattern therefore cannot be transferred directly by etching an underlying film. Pattern transfer was achieved instead by imprinting using the PFS rings as an imprint stamp. These examples show some of the diversity of geometries that can be achieved using triblock terpolymers. A wide range of geometries remains to be investigated, including lines and spaces of unequal widths, lines with width modulation, or tiling patterns, and self-assembly of these patterns may facilitate the fabrication of new types of devices.

1. V. P. Chuang, J. Gwyther, R. A. Mickiewicz, I. Manners, and C. A. Ross,  “Templated self-assembly of square symmetry arrays from an ABC triblock terpolymer,” Nano Lett., vol. 9, pp. 4364-9, 2009.
2. V. P. Chuang, C. A. Ross, J. Gwyther, and I. Manners, “Self-assembled nanoscale ring arrays from a polystyrene-b-polyferrocenylsilane-b-poly(2-vinylpyridine) triblock terpolymer thin film,” Adv.Mater., vol,  21, pp. 3789-93, 2009.

Block copolymers can be used to make a variety of functional electronic or magnetic devices. We have developed pattern transfer processes to produce metal, silicon, oxide, or polymer patterns using a mask made from a PS-PDMS (polystyrene-polydimethylsiloxane) block copolymer.

Metal films are patterned using a “damacene” process in which a metal film is deposited over a pattern made from the oxidized PDMS microdomains. Reactive ion etching planarizes and slowly removes the metal until the oxidized PDMS is exposed. The oxidized PDMS etches much faster than the metal, and termination of the etch process at this point leaves a metal pattern that is the reverse contrast of the block copolymer pattern. This process has been used to make magnetic nanostructures such as dot, line, and antidot arrays, as well as narrow metallization lines.

We have also deposited films of block copolymers on top of materials such as conductive polymers or graphene, and then used the block copolymer pattern as an etch mask to form structures such as arrays of conducting polymer nanowires or graphene nanoribbons. Other pattern transfer includes the formation of high aspect silicon nanowires by catalytic etching of a Si wafer using a BCP-patterned gold catalyst and the formation of inverted pyramid arrays by anisotropic etching of silicon through a block copolymer mask.

1. Y. S. Jung and C. A. Ross, “Fabrication of diverse metallic nanowire arrays based on block copolymer self-assembly,”Nano Lett.  vol. 10, pp. 3722-3726, 2010.
2. S.-W. Chang, V. P. Chuang, S. T. Boles, C. A. Ross, and C. V. Thompson, “Densely-packed arrays of ultrahigh-aspect-ratio silicon nanowire fabricated using block copolymer lithography and metal-assisted etching,” Adv. Functional Materials vol. 19, pp. 2495-2500, 2009.
3. Y. Sik Jung and C. A. Ross, “Well-ordered thin film nanopore arrays formed using a block copolymer template,” Small, vol. 5, pp. 1654-1659, 2009.
4. Y. S. Jung, W. Jung, H. L. Tuller, and C. A. Ross, “Nanowire conductive polymer gas sensor patterned using self-assembled block copolymer lithography,” Nano Lett., vol. 8, pp. 3776-3780, 2008.

Magnetic metal nanoparticle arrays have attracted considerable interest for applications in patterned magnetic recording media as well as catalyst arrays for growing carbon or semiconductor nanotubes. These applications ideally require large-area, low-cost nanoparticle arrays with controlled long-range order. Well-ordered nanoparticle arrays are typically made by “top-down” lithographic planar processing methods, but the formation of sub-50-nm features becomes increasingly difficult as the dimensions of the particles are reduced. As an alternative, chemical synthesis can be used to obtain magnetic nanoparticles with dimensions of a few nm and above, but the long-range order of arrays formed from such particles is limited.

Nanoparticle arrays may also be made by the spontaneous dewetting (or agglomeration) of a poorly wetting metal film on a substrate such as silica upon annealing at temperatures of several hundred ˚C. These arrays consist of islands with a range of sizes and spacings, and the islands lack long-range order. Giermann et al. reported the fabrication of regular Au nanoparticle arrays by dewetting a thin metal film on a topographically patterned Si substrate at elevated temperatures ^[1] . The templating effect was attributed to the local curvature of the film over the topography, which leads to grooving of the film at the edges of the pits, and its eventual separation into nanoparticles located within the pits. We applied this technique to thin Co films on oxidized silicon substrates that were topographically prepatterned with an array of 200-nm period pits. The Co nanoparticle size and uniformity are related to the initial film thickness, annealing temperature, and template geometry. One particle per 200-nm period pit is formed from a 15-nm film annealed at 850˚C; on a smooth substrate, the same annealing process forms particles with average interparticle distance of 200 nm. Laser annealing enables templated dewetting of 5-nm thick films to form one particle per pit. Although the as-deposited films exhibit a mixture of hcp and fcc phases, the ordered cobalt particles are predominantly twinned fcc crystals with weak magnetic anisotropy. Templated dewetting is shown to provide a method for forming arrays of nanoparticles with well-controlled sizes and positions. Recent work includes the use of block copolymers to fabricate inverted pyramid arrays with a 40-nm period and analysis of the dewetting behavior on such substrates. We are also examining dewetting of films patterned with small holes made using a block copolymer process. ^[2]

1. A. Giermann and C.V. Thompson, “Solid-state dewetting for ordered arrays of crystallographically oriented metal particles,” Appl. Phys. Lett., vol. 86, p. 121903, 2005.
2. Y.-J. Oh, C. A. Ross, Y. S. Jung, Y. Wang, and C. V. Thompson, “Cobalt nanoparticle arrays made by templated solid-state dewetting,” Small, vol. 5, no. 7, pp. 860–865, 2009.

We are investigating the fabrication and magnetic properties of rings (Figure 1) for magnetic logic and memory devices. The magnetic multilayer rings show giant magnetoresistance, in which the resistance is a function of the relative orientation of the magnetization directions in the magnetic layers. These small structures have potential uses in magnetic-random-access memories (MRAM), magnetic logic devices, and other magneto-electronic applications. Unlike that of conventional MRAM devices, the ring-shaped geometry of these devices allows for a complex response with multiple stable resistance states. This capability can be used for multi-bit memory and for programmable, non-volatile memory.

These devices are programmed using either a magnetic field or a current. We have shown that spin-polarized currents can reverse the devices by domain wall motion. We are currently investigating concatenation of multiple devices through stray field interactions and control of the behavior of individual devices through interlayer magnetostatic pinning interactions and manipulation of 360° domain walls with a spin-polarized current.

We are also investigating the effects of repeated cycling on rhombic ring devices. Stray field interactions between layers give rise to complex reversal paths, which depend on the field history. These complex reversal behaviors can be attributed to the formation of 360° domain walls and mirror domains in the structure. Understanding these interactions will facilitate the development of multilayer domain wall devices.

1. C. Nam, M. D. Mascaro, and C. A. Ross, “Magnetostatic control of vortex chirality in Co thin film rings,” Appl. Phys. Lett. vol. 96, pp. 012505:1-3, 2010.
2. Y. Ren, S.  Jain, A. O. Adeyeye, and C. A. Ross, “Magnetization states in coupled Ni80Fe20 bi-ring nanostructures,” New J. Phys. vol. 12, pp. 093003:1-11, 2010.­­­

Many domain wall (DW) devices require the presence of multiple DWs, each of which may need to be independently controlled and moved. As a result, we are investigating the interaction of domain walls in a single ferromagnetic nanowire. In particular, we are focusing on 360^o domain walls (360 DWs) in which the magnetization makes a full in-plane 360° turn in a localized region of the stripe, while the rest of the stripe is magnetized parallel to its edges.

Applied current, utilizing the spin-torque effect, is the most promising method for driving domain wall devices. Our simulations indicate that 360 DWs have a response to a current that is qualitatively different from the behavior of the constituent 180 DWs. The 360 DWs move at a velocity independent of applied magnetic fields and can be destroyed by a burst of applied current. The stability of the domain wall can be controlled by an externally applied field, as shown in Figure 1. These features make the 360 DW a potential candidate as a data token in novel domain wall logic devices. Additionally, we have observed and are currently investigating resonant behavior in the 360 DW, which can be tuned by an applied field, as in Figure 1(b)). Simultaneously, we experimentally demonstrate the controlled generation of 360 DWs in rings and in a nanostructure consisting of a circular pad attached to a curved wire, as shown in Figure 2. The response of a 360 DW to current is being characterized using anisotropic magnetoresistance.

1. M. D. Mascaro, C. Nam, and C. A. Ross, “Interactions between 180 and 360 degree domain walls in magnetic multilayer stripes,” Appl. Phys. Letts. Vol. 96, no. 162501, pp. 1-3, 2010.
2. C. Nam, M. D. Mascaro, and C. A. Ross, “Magnetostatic control of vortex chirality in Co thin film rings,” Appl. Phys. Lett., vol. 96, no. 012505, pp. 1-3, 2010.
3. M. D. Mascaro and C.A. Ross, “AC and DC Current-Induced Motion of a 360˚ Domain Wall,” Phys. Rev. B vol. 82, p. 214411, 2010.

We maintain a thin-film laboratory that includes a pulsed-laser deposition (PLD) system and an ultra-high vacuum sputter system. The PLD is particularly useful for making complex materials such as oxides because it can preserve the stoichiometry of the target material.

We have been using PLD to deposit a variety of oxide films for magneto-optical devices such as isolators. The ideal material for an isolator combines high Faraday rotation with high optical transparency. Garnets, such as bismuth iron garnet (BIG, Bi₃Fe₅O₁₂), have excellent properties but do not grow well on silicon substrates, making it difficult to integrate these materials. One way to solve this problem is to develop new magneto-optical active materials, which can grow epitaxially on Si by using buffer layers. Through being doped with transitional metal ions, these materials can exhibit strong Faraday rotation as well as low optical loss. We have examined Fe and Co-doped SrTiO₃ thin films (Figure 1) ^{[1] [2]} , which show high magneto-elastic anisotropy ^[3] , strong magneto-optical properties, and lower optical absorption compared with iron oxide. We also demonstrated epitaxial integration of these films on silicon using a CeO₂/YSZ buffer layer. When Sr(Ti_0.6Fe_0.4)O₃ films are doped with Ga ions on the Ti site up to 50 at.%, the optical absorption of the material decreases by more than one order of magnitude at 1550 nm (Figure 2), while it still shows a Faraday rotation of ~400 deg/cm ^[4] . A high material figure of merit of 3~4 dB/cm was achieved in Sr(Ti_0.6Ga_0.4Fe_0.2)O₃. These films could be useful for waveguide isolators and other magnetoelectronic devices in which optical absorption losses are critical.

1. H.-S. Kim, L. Bi, G. F. Dionne, and C. A. Ross, “Magnetic and magneto-optical properties of Fe doped SrTiO₃ films,” Appl. Phys. Lett., vol. 93, p. 092506, 2008.
2. L. Bi, H.-S. Kim, G. F. Dionne, and C. A. Ross, “Structure, magnetic properties and magnetoelastic anisotropy in epitaxial Sr(Ti_1-xCo_x)O₃ films,”  New J. Phys., vol. 12, p. 043044, 2010.
3. D. H. Kim, L. Bi, P. Jiang, G. F. Dionne, and C. A. Ross, “Magnetoelastic effects in transition metal-substituted Sr(Ti _1-x(Fe,Co or Cr)_x)O₃ epitaxial thin films,” submitted for publication.
4. P. Jiang, L. Bi, D. H. Kim, G. F. Dionne, and C. A. Ross, “Enhancement of the magneto-optical performance of Sr (Ti_0.6-xGa_xFe_0.4)O₃ perovskite films by Ga substitution,” submitted for publication.

Block copolymer self-assembly on nanolithographically-defined templates has great potential in fabricating patterned media and devices at the nanometer scale.  Current experimental work requires first that the templates be written lithographically and then that the block copolymer morphology around the template be observed and categorized.  An alternative approach to the problem of finding the correct template to get the desired morphology of block copolymers can be achieved through self-consistent field theory modeling (SCFT) ^[1] .  By reducing the problem of polymer interactions with the substrate to density and field interactions, the theory allows for quick computational exploration of phase diagrams for two-dimensional unit cell templates.  Large cell templates and three-dimensional unit cell templates can be modeled as well but at a cost of larger computation time.  However, these simulations can be used to predict or confirm proposed three-dimensional structures that are hard to observe using conventional planar microscopy methods, as seen in Figure 1 ^[2] .  In the theory, topographical constraints are modeled as hard potential barriers to the polymer, and chemical surface affinity effects are modeled as attractive potential fields, as schematically represented in Figure 2.  The traditional approach with the theory is to impose an initially randomly configured system of copolymers to a predefined template and have the system evolve through pseudo-dynamical relaxation schemes to reach saddle points in the system energy space, thus allowing for both metastable states and the thermodynamic equilibrium state to be observed.  Combining SCFT with surface energy analysis calculations of the observed equilibrium morphologies, one can develop a simple theory of where ordered structure transitions occur based on geometric commensurability conditions. Eventually, the theory should allow for taking desired block copolymer morphologies as the simulation input and determining what the minimum required template is to achieve that morphology.

1. R. A. Mickiewicz, J. K. W. Yang, A. F. Hannon, Y. S. Jung, A. Alexander-Katz, K. K. Berggren, and C. A. Ross, “Enhancing the potential of block copolymer lithography with polymer self-consistent field theory simulations,” Macromolecules, vol. 43, no. 19, pp. 8290–8295, 2010.
2. J. G. Son, A. F. Hannon, K. W. Gotrik, A. Alexander-Katz, and C. A. Ross, “Hierarchical nanostructures by sequential self-assembly of styrene-dimethylsiloxane block copolymers of different periods,” Adv. Mater. vol. 23, pp. 634-639, 2011.

Self-organized macromolecular materials can provide an alternative pathway to conventional lithography for the fabrication of devices on the nanometer scale. In particular, the self-assembly of the microdomains of diblock copolymers within lithographically-defined templates to create patterns with long range order has attracted considerable attention, with the advantages of cost-effectiveness, large-area coverage, and compatibility with preestablished top-down patterning technologies. Previously, we showed that spherical morphology poly(styrene-b-dimethylsiloxane) (PS-PDMS) block copolymers, which have a large interaction parameter and a high etch-contrast between two blocks, can be templated using an array of nanoscale topographical elements that act as surrogates for the minority domains of the block copolymer ^[1] . Recently, we showed that complex nanoscale patterns can be generated by combining the self-assembly of block-copolymer thin films with minimal top-down templating. A sparse array of nanoscale HSQ posts was used to accurately dictate the assembly of a cylindrical PS-PDMS diblock copolymer into a wide assortment of complex, unsymmetrical features, as shown in Figure 1 ^[2] . To extend the feature sizes to the sub-10-nm range, we demonstrated the formation of highly ordered grating patterns with a line width of 8 nm and period of 17 nm from a self-assembled PS-PDMS diblock copolymer and fabricated sub-10-nm-wide tungsten nanowires from the self-assembled patterns using a reactive ion etching process, as shown in Figure 2.

1. I. Bita, J. K. W. Yang, Y. S. Jung, C. A. Ross, E. L. Thomas, and K. K. Berggren, “Graphoepitaxy of self-assembled block copolymers on two-dimensional periodic patterned templates,” Science, vol. 321, pp. 939-943, 2008.
2. J. K. W. Yang, Y. S. Jung, J.-B. Chang, C. A. Ross, and K. K. Berggren, “Complex self-assembled patterns using sparse commensurate templates with locally varying motifs,” Nature Nanotechnology, vol. 5, pp. 256-260, 2010.