Fabrication and Applications of Self-Assembled Nanopillars

In this mini-review, we summarize fabrication methods, formation mechanisms, factors controlling the characteristics, and applications of selfassembled nanopillars. Nanopillars prepared both in the gas phase and in solutions are discussed.

At the nanoscale, materials possess characteristics that are dramatically different from their corresponding macroscale ones.This is one of the main driving forces for the recent explosive increase in research on nanomaterials such as nanowires, nanoparticles, and quantum dots.However, a key to realizing their potential applications for devices is the ability to assemble them into desirable patterned nanostructures.The two major design strategies in coupling functionalities with materials on the micro-or nanoscale are via multilayering or by forming vertical heteroepitaxial nanocomposites.In recent years, fabrication and applications of nanopillars have attracted more and more attention.There are two principle approaches to nanopillar fabrication: top-down approaches such as lithography and etching, and bottom-up growth techniques.Bottom-up fabrication methods of nanopillars include chemical vapor deposition (CVD), 1 physical vapor deposition (PVD), 2 template synthesis, 3,4 and self-assembled methods.In this mini-review, we intend to summarize fabrication methods, formation mechanisms, factors controlling characteristics, and applications of nanopillars based on the self-assembly approach, with more attention being focused on the factors that affect the growth of the nanopillars.It is noteworthy that although, to a certain extent, CVD and PVD belong to the self-assembled class of approaches and have been widely used to make nanopillars, such as nanopillars of ZnO, CdS, etc., we will only briefly summarize the reports which mentioned the concept of self-assembly.The purpose of this review is to provide new researchers in this area with a quick overview on the body of work containing self-assembled nanopillars.For reviews on nanopillars from CVD or PVD, readers are suggested to read other recent reviews. 5,6,7,8

Fabrication methods and formation mechanisms.
The self-assembly approach offers a simple and efficient method for growth aligned nanostructures.This includes: a.
self-assembled inorganic nanopillars in the gas phase by deposition of aerosol, and pulsed laser deposition (PLD).b.
self-assembled inorganic nanopillars in solutions by hydrothermal + polymer assisted template.c.
Self-assembled organic compounds in solution on surfaces through liquid evaporation.
2a. Self-assembled inorganic nanopillars in the gas phase by deposition of aerosol or pulsed laser deposition (PLD).In a film-on-substrate geometry, epitaxial composite films can be created in two forms, horizontal and vertical.Nanocomposite films with a vertical architecture, i.e. the nanopillar geometry, offer numerous advantages over the conventional horizontal multilayers, such as a larger interfacial area and intrinsic heteroepitaxy in three dimensions. 9,10bedev et al discovered 11 that (La 0.67 Ca 0.33 MnO 3 ) 1-x :(MgO) x (LCMO) composite films on a (100) MgO substrate have microstructure and magneto-transport properties depending on the MgO concentration.The LCMO films were prepared by metal organic aerosol deposition on a (100) MgO substrate for different concentrations of the (MgO) phase (0<x<0.8)(Figure 1 http://ccaasmag.org/CHEMleft).At x≈0.3 a percolation threshold in conductivity is reached, at which point an infinite insulating MgO cluster forms around the La 0.67 Ca 0.33 MnO 3 grains.This yields a drastic increase in the electrical resistance for films with x>0.3.The films exhibit a remarkable structure consisting of domains of LCMO material surrounded by an epitaxially intergrown MgO thin layer.This MgO layer creates a wall around the LCMO domains and, as a result, a uniform 3D stress is built up.They concluded that the structural phase transition is coupled with the percolation threshold.The reason seems to be a drastic increase of the 3D stress contribution when all the LCMO domains are surrounded by epitaxially grown MgO layers.The formation mechanics were proposed as shown in Figure 1 right.In a later work, Moshnyaga et al. confirmed the formation of vertical nanopillar films with a composition of (La 0.7 Ca 0.3 MnO 3 ) 1-x :(MgO) x . 12The structural and magneto-transport properties of the La 0.7 Ca 0.3 MnO 3 nanoclusters were tuned through the tensile stress originating from the MgO second phase.
In later reports, the pulsed laser deposition (PLD) method becomes the most popular method in this area and it has been successfully used in the fabrication of ceramic materials with self-assembled nanopillars embedded in films.PLD is a thin film deposition technique where a high power pulsed laser beam is focused inside a vacuum chamber to strike a target of the material that is to be deposited.The material is vaporized from the target in a plasma plume which deposits it as a thin film on a substrate.Ceramic pulsedlaser deposition target compositions were fabricated by milling appropriate starting chemicals, such as La 2 O 3 , MnCO 3 , SrCO It has been generally accepted that the formation of vertical nanostructures in films is related to the minimization of the relative interfacial and elastic strain energies between two materials and between each material and the substrate.Both the morphology and the scale of self-assembled nanostructures are controlled by varying the lattice misfit with a substrate. 16By selecting two appropriate materials and their phases, it is possible to grow vertically aligned nanopillars in an epitaxial thin-film form.The properties of such nanostructures are dependent on the nanostructure morphologies including domain patterns and shapes as well as structures and properties of the interfaces.The two requirements for self-assembled two-phase nanocomposites are 1) the intrinsic similarity in crystal chemistry between two materials that leads to crystal lattice parameters that are reasonably commensurate; 2) the compounds have little solid solubility into each other.The two phases in the nanostructure can be epitaxial with respect to each other as well as with respect to a common substrate. 13In the above mentioned papers, Li et al 14 demonstrate that varying the stress conditions via differently oriented single-crystal substrates yields different phase morphologies in the epitaxial two-phase films.For the [001] orientation the platelet habits were parallel to the [110] planes, whereas for the [110] orientation the platelets were parallel to the [111] planes (Figure 2).They also demonstrated that the morphologies of self-assembled multiferroic nanostructures can be varied by modifying the epitaxial stress state in the film.Their results confirm that the inplane elastic anisotropy of the film, as determined by the substrate orientation, has a profound effect on the morphologies of the resulting two phase nanostructures, and that the major morphological differences can be explained using a thermodynamic theory.Ji et al. have recently developed a novel, surfaceassisted self-assembly approach to harvest vertically aligned single-crystalline organic nanopillar semiconductors from a solution of organic chemicals. 18n these experiments, when cyanuric acid (CA) and melamine (M) are mixed at lower concentrations on a gold surface, an array of nearly 100% hexagonal nanopillars of the CA•M complex was obtained (scanning electron microscopy (SEM) images in Figure 4).The nanopillar is composed of a cyclic CA 3 M 3 rosette assembly 19 (Figure 4 left).CA 3 M 3 rosette molecular assemblies are hierarchically stacked or coordinated into insoluble nano-or micro-structures in 3D space due to hydrogen bonding, π-π stacking, hydrophobic interactions, and van der waals interactions.They showed that the CA•M complexes nucleate as a thin, hexagonal plate on the gold surface at the onset stage, and grow wider and taller after more solvent evaporates.The unique aspect of this result is that this assembly is the first of such well-defined selfassembled nanopillars made of small organic molecules.It is expected that arrays of pillars can be developed from several other complexes.The nanopillars prepared by this method can be made in large quantities at a low cost due to the facile self-assembling method.

Factors affecting the characteristics of the selfassembled nanopillars
The property of surfaces is the main factor that controls the formation and characteristics of nanopillars.For the nanopillars from the PSD method, it has been believed that the in-plane compressive strain caused by the misfit between evaporated materials and the substrate restricts the lateral size of the nanopillars and drives them to grow along the c-axis (the direction perpendicular to the substrate surface) forming a nanopillar structure. 20The evaporated compounds and substrate should have a similar structure and chemistry at their interface to facilitate the nucleation. 20So it is possible to control relative lattice orientation of the constituent materials and the subsequent vertical phase boundaries in a systematic way to achieve desired and tunable physical property of the nanostructures, i.e. the physical property of the nanostructures depends on the substrate-film lattice misfit and thus can be adjusted by the choice of the substrate.This leads to the possibility of designing new types of the structures with predictable physical properties.
Zheng et al, 21 reported that the morphology of selfassembled perovskite-spinel nanostructures can be controlled simply by selecting single-crystal substrates with different orientations; a (001) substrate results in rectangular-shaped CoFe 2 O 4 nanopillars in a BiFeO 3 matrix; in contrast, a (111) substrate leads to triangularshaped BiFeO 3 nanopillars in a CoFe 2 O 4 matrix, irrespective of the volume fraction of the two phases (Figure 5).This reversal can be understood as a primary consequence of different nucleation modes due to the large differences in surface energy anisotropy.Control of crystallographic orientations and physical properties of self-assembled nanostructures via rational selections of substrates has also been demonstrated by Liao et al. 22 They showed that in the perovskitespinel BiFeO 3 /CoFe 2 O 4 model system the crystal orientation of self-assembled CoFe 2 O 4 nanopillars can be tuned among (001), (011), and (111), while that of the BiFeO 3 matrix is fixed in (001).Moreover, the resultant CoFe 2 O 4 nanopillars appear in various shapes: pyramid, roof, and triangular platform, respectively (Figure 6).The tunable nanostructures through this approach enable the control of material functionality such as the magnetic anisotropy of CoFe 2 O 4 .
The substrates play a major role in the growth of the CA•M nanopillars. 18The density of the CA•M hexagonal nanopillars was approximately 30 /m 2 on a glass or silicon surface, which was much greater than the density of 0.1 / m 2 found on the gold surface.The CA•M nanopillars on a glass or silicon surface were shorter and smaller in diameter than those made on the gold surface. http://ccaasmag.org/CHEM

Applications of the self-assembled nanopillars.
Self-assembled vertical nanostructures can be used to design new functionalities.Some applications of these nanopillars are briefly summarized as follows: 4a.Magnetoelectric.The nanopillar-embedded films hold potential as an excellent magnetoelectric (ME) materials.ME materials are materials that exhibit an induction of magnetization by an electric field or an induction of electric polarization by a magnetic field.They offer increased functionality and entirely new applications for electronic devices.The creation of nanostructures with uniform size and spacing in epitaxial heterostructures through self-assembly offers great potential to tailor the physical properties and explore the fabrication of new functional devices.Strong ME responses have been realized in artificial two-phase vertically aligned nanocomposite films with well-ordered vertical phase boundaries.Laminated ceramic composites, fabricated from piezoelectric and magnetostrictive components, have been found to exhibit strain-mediated ME responses an order-of-magnitude larger than those observed in single-phase materials.
Zheng et al. 9,13 prepared self-assembled epitaxial CoFe 2 O 4 -BaTiO 3 ferroelectromagnetic nanostructures on single-crystal SrTiO 3 (001) substrates (Figure 8).The CoFe 2 O 4 -BaTiO 3 nanocomposites were formed from a 0.65BaTiO 3 -0.35CoFe 2 O 4 or 0.62BaTiO 3 -0.38CoFe 2 O 4 target by PLD and magnetic spinel CoFe 2 O 4 pillars were epitaxially embedded into a ferroelectric BaTiO 3 matrix.The films are epitaxial in-plane as well as out-of-plane with self-assembled hexagonal arrays of CoFe 2 O 4 nanopillars with average spacing of 20 to 30 nanometers.SrRuO 3 was chosen as the lattice-matched bottom electrode to enable heteroepitaxy as well as to facilitate electric measurements.This nanocomposite exhibited strong magnetoelectric coupling due to the strong elastic interactions between the two phases.Temperature dependent magnetic measurements illustrate the coupling between ferroelectric and magnetic order parameters, which is manifested as a change in magnetization at the ferroelectric Curie temperature.
http://ccaasmag.org/CHEM 4b.Ferroelectric.These nanopillar-embedded films are also used as ferroelectric materials, which are used in applications ranging from energy harvesting to highpower electronic transducers.Strain control of oxide films can achieve dramatically enhanced ferroelectricity properties, and huge changes in the ferroelectric Curie temperature Tc.
In BiFeO 3 :Sm 2 O 3 , 23 high compressive vertical strain levels of ~1.5% have been demonstrated, with a strong improvement in dielectric properties and reduced leakage. 24Harrington et al 30 reported practical ferroelectric materials based on micrometer-thick films of BaTiO 3 , which were formed by self-assembled vertical columns of Sm 2 O 3 throughout the film thickness (Figure 9).Sm 2 O 3 (SmO) is the ideal strain controlling second phase because it substitutes only minimally into the BTO, it has a large elastic modulus (125 GPa, compared to 67 GPa for BTO) and is insulating.The Curie temperature of the film is at least 330 o C, and the tetragonal-to-cubic structural transition temperature to beyond 800 o C. Such enhancements in thick films are a consequence of uniform, vertical strain coupling between stiff nanocolumns of SmO and a surrounding BTO nanomaterial and have opened a new approach to using BTO in high-temperature ferroelectric applications.This method provides strain control in much thicker films via the growth of vertical nanocomposites. 31,32,33c.Vortex pinning and high conductivity.The introduction of nanopillars in membranes is a promising approach for optimizing high-temperature superconductors (HTSs) by increasing their critical current density Jc and reducing their anisotropy.These goals require the introduction of nanostructures able to strongly pin magnetic flux vortices so as to oppose vortex motion induced by the Lorentz force.
Wee et al reported 34    http://ccaasmag.org/CHEM4d.FET spintronics.GaN-based quantum wells (QWs) are a promising candidate for optoelectronic devices, such as spin-polarized field effect transistors (spinpolarized FETs). 37When light-polarized M-plane GaN (Figure 11) and spin-polarized C-plane GaN are fabricated on the same substrate (e.g., -LiAlO 2 ), twophase GaN provides an opportunity to combine optoelectronics and spintronics to an optically controlled spin-polarized FET by the coupling of polarized light and the polarized spin of 2DEG. 38Hsieh et al showed that the formation of self-assembled c-plane GaN nanopillars is through nucleation on hexagonal anionic bases of LiAlO 2 .This heterostructure might be used in optically controlled spintronic devices.

Conclusion and Further perspectives.
The self-assembled nanopillars hold great promise for novel device applications and functional property control that goes substantially beyond what is possible in established systems.The nanostructures spontaneously form over large areas which has important ramifications for many areas of functional device materials.Compared with the lateral interface, the effect of vertical interfaces on the physical properties of metal oxide films is profound.
A functional material that spontaneously forms nanopillars embedded in a matrix of another material during the thin-film growth offers attractive new possibilities for device applications.In addition to the individual functionalities of each constituent phase, materials can display coupling between the order parameters.Investigations which are ongoing in this field include the creation of ordered structures (and their formation on a large scale) and different combinations of materials and control of strain coupling between the phases.More applications of the these structures are expected-for instance, recently, a large photon-induced strain of ∼0.5% to 1.5% was demonstrated in SRO/SrTiO 3 39 and SRO/Pb(Zr,Ti)O 3 superlattices; 40 these can be used as photosensitive materials.Furthermore, similar two phase systems have not been achieved in organic based materials since selfassembled nanopillars are still in their infancy.Realization of organic based two-phase, vertical nanocomposites may lead to new forms of ordered nanostructures for multifunctional applications and will open up a new level of control in films so that electric or optical properties of materials can be tuned by the appropriate choice of materials.This would also enable more straightforward basic research studies of physical property measurements in strained systems to be undertaken.O. I. Lebedev, J. Verbeeck, G. Van

Figure 1 .
Figure 1.Left: Cross-section electron microscopy image of the interface (La 0.67 Ca 0.33 MnO 3 ) 1-x -(MgO) x /MgO(100) for x=0.5.Right: Schematic representation of the growth mechanism of the composite film at different stages: (a) Nucleation of the LCMO film on the MgO substrate.(b) Enlargement of one LCMO island on the MgO substrate.Close to the interface the LCMO region is under a two-dimensional tensile stress; the upper part of the LCMO island is under a two-dimensional compressive stress.The shape of LCMO becomes pyramidal.(c) Nucleation of the (LCMO) 1-x :(MgO) x composite film on the MgO substrate.(d) The columnar structure of the LCMO film on the MgO substrate.(e) Structure of the (LCMO) 1-x :(MgO) x film for x=0.33.Intergrowth of the MgO islands in between LCMO domains occurs.(f) Structure of the (LCMO) 1-x :(MgO) x film for x=0.5.Every LCMO grain is surrounded by MgO through the complete film thickness.Reprinted with permission from the American Physics Society.

Figure 2 .
Figure 2. TEM images of the PbTiO 3 -CoFe 2 O 4 nanostructures grown on (001) SrTiO 3 (a) cross section, (b) plane view of the structures on (110) SrTiO 3 .(c) cross section, (d) plane view the structures on (110) SrTiO 3 .The image shown in (b) is a bright field image recorded near the [001] orientation parallel to the surface normal.In all the images, the CoFe 2 O 4 phase appears as bright regions.Reprinted with permission from the American Physics Society.

17
Figure 3. (a) Typical XRD pattern of the as-prepared CFO-BTO samples, (b) TEM image of the CFO-BTO nanostructures with the EDX spectrum in the upper right inset, (c) TEM image and (d) High resolution TEM image of an individual CFO-BTO nanostructure, (e) High resolution TEM image of a CFO nanopillar embedded in the BTO matrix.Reprinted with permission from the Elsevier B. V.

Figure 4 .
Figure 4. Molecular structures (left) and SEM images of 1:1 CA•M hexagonal nanopillars on a gold surface.Reprinted with permission from the American Chemical Society.

Figure 5 .
Figure 5. Morphologies of the BiFeO 3 -CoFe 2 O 4 nanostructures grown on a (001)-oriented (left) and a (111)oriented (right) SrTiO 3 substrate.(a) Z-contrast image from a TEM sample.(b) A TEM image of a single CoFe 2 O 4 pillar embedded in a BiFeO 3 matrix.(c) A high-resolution TEM image from the interface region marked by the rectangle in b.(d) Structural model of the interface between CoFe 2 O 4 and BiFeO 3 showing that the interfaces are 110planes (left) or 112planes (right), both along 110directions.(e) Cross-sectional TEM image of a single CoFe 2 O 4 pillar.(f) SEM image of the CoFe 2 O 4 pillars.(g) A schematic of a CoFe 2 O 4 pillar.(h) A schematic of a CoFe 2 O 4 pillar showing three crystal facets.(left) or A schematic of a BiFeO 3 pillar showing (100), (010), and (100) facets (right).Reprinted with permission from the American Chemical Society.

Figure 8 .
Figure 8. Left: Schematic illustration of a self-assembled nanostructured thin film formed on the substrate.Middle: AFM topography image of the film showing a quasi-hexagonal arrangement of the CoFe 2 O 4 nanopillars.Right: TEM planar view image showing the CoFe 2 O 4 nanostructures in the BaTiO 3 matrix.Reprinted with permission from the AAAS.

Figure 9 .
Figure 9. Self-assembled vertical nanostructure.a-c, TEM images of a 600-nm-thick BaTiO 3 :Sm 2 O 3 film on SrTiO3: cross-section (a); plan view (b).Sm 2 O 3 shows a darker contrast compared with BaTiO 3 .Yellow arrow represents c-axis direction.d, Crystallographic model of interface matching.Reprinted with permission from the Nature Publishing Group.
formation of columnar defects comprised of self-assembled Ba 2 YNbO 6 (BYNO) nanopillars in YBa 2 Cu 3 O 7-δ (YBCO) films via simultaneous laser ablation of a YBCO target and a Nb http://ccaasmag.org/CHEMmetal foil.Compared to pure YBCO, YBCO+BYNO films exhibited no Tc reduction as well as superior Jc performance with higher self-and in-field Jc by a factor of 1.5-7 and also exhibited a strong Jc peak for H ‖ c indicative of strong flux-pinning.Wee et al. also reported 35 rare-earth barium tantalates, Ba 2 RETaO 6 (BRETO, RE=rare earth elements) as promising pinning additives for superior flux pinning in YBa 2 Cu 3 O 7−δ (YBCO) films.BRETO compounds have excellent chemical inertness and large lattice mismatch with YBCO.Zhang et al showed20 the structure and chemistry of the self-assembled oxide nanopillars that form in superconducting Co-doped BaFe 2 As 2 thin film grown by PLD.The oxide nanopillars consist of a BaFeO 2 phase, form epitaxially on the SrTiO 3 template, and grow coherently with the BaFe 2 As 2 film.The high density of self-assembled non-superconducting oxide nanopillars provides very effective flux pinning centers in the Ba-122.The nanopillars provide exceptionally strong vortex pinning and high critical current density due to the very close correlation of pillar and vortex core diameters (Figure10).The use of self-assembled nanopillars as correlated defects for strong vortex pinning provides an efficient method for increasing Jc without any apparent film thickness dependence or need for post-processing.Several other pinning materials including Y 2 BaCuO 5 , BaZrO 3 , BaIrO 3 , Nd 2 O 3 , YSZ, Y 2 O 3 , and BaSnO 3 have been investigated to improve the flux pinning properties of YBCO films.For example, Kang et al36 demonstrated short segments of a superconducting wire in substantially thicker films (3-micrometers) that meets or exceeds performance requirements for many large-scale applications of high-temperature superconducting materials, especially those requiring a high supercurrent and/or a high engineering critical current density in applied magnetic fields.Enhancements of the critical current were achieved in the thick films via incorporation of a periodic array of extended nanopillars extending through the entire thickness of the film.These nanopillars are highly effective in pinning the superconducting vortices or flux lines, thereby resulting in the substantially enhanced performance of this wire.

Figure 10 .
Figure 10.A high angle annular dark field (Z-contrast) image showing the atomic structure of an epitaxial nanopillar in the Ba-122 film.Reprinted with permission from the American Chemical Society.

Figure 11 .
Figure 11.(left) (a) Lattice structure of LAO (100).(b) High quality LAO wafer.(c) Schematic diagram of twophase growth mode.A 3D c-plane GaN nanopillar and a 2D M-plane GaN film are simultaneously grown on a -LiAlO 2 substate, forming a 120 o angle between them.Reprinted with permission from the Japan Society of Applied Physics.