Microphase separation by block copolymers in thin films has been investigated from several perspectives. First, the physics of self-assembly in confined soft materials can be studied using model block copolymer materials for which reliable mean-field statistical mechanical theories have been developed.[13–16] Second, interest has expanded as a result of the potentially exciting applications that exploit self-organization to (i) fabricate high-density data storage media,[17,18] (ii) lithographically pattern semiconductors exhibiting ultra-small feature sizes,[19,20] and
(iii) prepare ultra-fine filters or membranes.[21] Although research in this field is growing at a rapid pace, the field has not been reviewed since 1998,[2,22] even though many new developments have occurred.
Block copolymer films can be prepared using spin-coating techniques, where drops of a solution of the polymer in a volatile organic solvent are deposited on a spinning solid substrate (silicon wafers are often used because of their uniform flatness). The polymer film spreads as a result of centrifugal forces and the volatile solvent is rapidly driven off. With care, the method can provide films exhibiting low surface roughnesses over areas of several square millimeters. The film thickness can be controlled through modification of the spin speed, the concentration of the block copolymer solution, and the volatility of the solvent, which also influences the surface roughness.[23] Dip coating is another reliable method for fabricating uniform thin films.[24] Whichever deposition technique is used, if the surface energy of the block copolymer is much greater than that of the substrate, dewetting will occur. The mechanism of dewetting has been investigated in some depth.[25–27]
In thin films, the lamellae formed by symmetric block copolymers can be oriented either parallel or perpendicular to the substrate. A number of possible arrangements of the lamellae are possible (Figure 1-5), depending on the surface energies of the blocks and the substrate and whether the film is confined at one or both surfaces. In the case that a different block preferentially wets the interface with the substrate or air, wetting is asymmetric and a uniform film has a thickness (n + 1/2)d. If the initial film thickness is not equal to (n + 1/2)d, then islands or holes (quantized steps of height d) form to conserve volume.[28] As well as leading to distinct orientations, confinement of block copolymers can change the thermodynamics of ordering; in particular, surface-induced ordering persists above the bulk order–disorder transition.[29]
Asymmetric block copolymers that form hexagonal or cubic-packed spherical morphologies in the bulk, form stripe or circular domain patterns in two dimensions, as illustrated in Figure 1-6. The stripe pattern results from cylinders lying parallel to the substrate; a circular domain surface pattern occurs when cylinders are oriented perpendicular to the substrate or when spheres are present at the surface. Because bicontinuous structures cannot exist in two dimensions, the gyroid phase is suppressed in thin films. Nanostructures in block copolymer films can be oriented using electric fields (if the difference in dielectric permittivity is sufficient)—an important property for applications in which parallel stripes[30] or perpendicular cylinder configurations[31]
are desired.
1-1-3 Applications
Until recently, most block copolymers have been applied industrially as adhesives or for their mechanical properties (e.g., as thermoplastic elastomers).
Only in the past 10 years have researchers taken block copolymers into the area of
“high technology,” particularly into the realm of the so-called “nanotechnologies”.
Many attempts have been made to use block copolymers in nanotechnology.
Self-assembled block copolymer microstructures having dimensions ranging from 10 to 100 nm are useful as nanometer-scale membranes, templates for the fabrication of nano-objects (e.g., metal or ceramic nanodots and wires), as 1-, 2-, and 3D photonic crystals, and as nanopattern masks for the fabrication of high density information storage media. The majority of applications proposed to date rely on the use of thin film structures; this feature is a major focus of the current review, although bulk nanoporous materials and photonic crystals are also considered. Figure 1-7 summarizes the applications of various block copolymer-enabled nanotechnologies.[32]
1-1-3-1 Nanolithography with Self-Assembled Block Copolymer Patterns
Although photolithography has played a dominant role in producing feature sizes smaller than 100 nm, feature sizes of less then 50 nm are not readily obtained using conventional lithography techniques. The minimum size that can be achieved through photolithography is determined by the wavelength of light used in the exposure. Electron beam lithography is commonly used to access feature sizes between 30 and 150 nm. Nevertheless, sizes less than 30 nm are not easy to obtain using standard lithography. One way to overcome this problem is through the use of self-assembled block copolymers.
In a pioneering paper, Park et al. demonstrated the use of block copolymer films as masks to transfer dotted and striped patterns into semiconductors.[33] They achieved a feature density of holes of ca. 1011 cm–2. The method they developed relies on the selective ozonation of polyisoprene (PI) or polybutadiene (PB) in block copolymers with polystyrene (PS) as the other (majority) component. Ozone cleaves the olefinic bonds in unsaturated polymers, such that they can be etched away. This process leaves holes or stripes in the PS matrix. This pattern is then transferred from the block copolymer into silicon nitride through reactive ion etching (RIE) using CF4
or CF4/O2 gases. The quality of the pattern transfer is excellent, producing nanoscale arrays of pits or channels. Park et al. also described how to prepare nanoscale arrays of posts by using an “inverse” mask, relative to that employed to produce the array of pits. If the PI is fixed by staining with osmium tetroxide, then etching of the matrix will occur preferentially. The regions under the PI domains will be left as an array of posts. Details of the ozone etching method for preparing block copolymer film masks were elaborated on in a subsequent paper,[21] which also contains data illustrating pattern transfer into other semiconductors, including silicon and germanium. A patent was awarded for this technology in 1999.[34] In a further
extension of the technique, arrays of nano-sized metal posts have been fabricated using the modified method illustrated in Figure 1-8.[35] A lithographic procedure in which the ozone etching step can be omitted has also been developed;[36] here, the reactive ion etch rate is sufficiently different between the two blocks in the PS-b-PFS (PFS = polyferrocenyldimethylsilane) diblock copolymer, allowing selective etching of PS to occur directly. The technology has been applied to the development of a self-assembly route to produce a high-density magnetic storage medium (in this case, cobalt nanodots). Oxygen plasma RIE leaves PFS spheres; the pattern is then transferred into silicon oxide (which improves pattern transfer) and then into tungsten through RIE. The multiplayer structure is necessary because magnetic materials, such as cobalt, nickel, and iron, are not amenable to RIE. In the next step, the polymer and silica are removed. Finally, the pattern is transferred from the tungsten hard mask into the magnetic cobalt layer using ion-beam etching. The result is the array of cobalt nanospots illustrated in Figure 1-9.[36] Ultrahigh-density metal nano-column arrays can be fabricated using block copolymer templates.
1-1-3-2 Nanoparticle Templates
Block copolymers have been utilized not only as surfactants to inhibit coalescence and aid in the dispersion of nanoscale particles (e.g., metals, metal oxides, inorganic nanostructures, molecular chromophores, and quantum dots) but also to spatially pattern them. Our group has previously reported[37] the formation (Figure 1-10) of ordered clusters of surfactant-modified TiO2 nanoparticles (NPs) in the selective block of lamellar assemblies of the diblock copolymer poly(styrene-b-methyl methacrylate) (PS-b-PMMA). Instead of using a water or alcohol phase, TiCl4 or titanium tetraisopropoxide precursors were used to synthesize the TiO2 NPs in tetrahydrofuran, which is a good solvent for block copolymers.
3-(Methacryloyloxypropyl)trimethoxy silane and cetyltrimethylammonium chloride
amphiphilic surfactants were used to modify the TiO2 NPs. This process provides a new approach toward the selective dispersal of quantum-confined NP clusters in a PS-b-PMMA diblock copolymer exhibiting an ordered lamellar phase. 3D and 2D nanostructure of CdS clusters have been prepared in the bulk and as thin films, respectively, by selectively dispersing pre-synthesized CdS NPs, containing hydroxyl groups on their surfaces, in the PEO block of templating poly(styrene-b-ethylene oxide) (PS-b-PEO) diblock copolymers.[38–42] Russell et al. reported that mixtures of diblock copolymers and either cadmium selenide- or ferritin-based NPs exhibit cooperative, coupled self-assembly on the nanoscale.[43] In thin films, the copolymers assemble into cylindrical domains, which dictate the spatial distribution of the NPs; segregation of the particles to the interfaces mediates interfacial interactions and orients the copolymer domains normal to the surface, even when one of the blocks is strongly attracted to the substrate. Figure 1-11 presents a cross-sectional TEM image of 100% PS thiol-coated Au NPs dispersed on a symmetric poly(styrene-b-2-vinylpyridine) (PS-b-P2VP) block copolymer. [44] We observe that the PS-coated Au NPs were located near the center of the PS block phase of the lamellar structure, whereas P2VP-coated Au NPs were located in the P2VP domain, as expected. Particles coated with a homopolymer similar to one of those in the copolymer can result in lower enthalpy when segregating into the corresponding domain of the block copolymer. Positioning the particle near the center of the corresponding polymer domain leads to better embedding of the NP because the polymer chains can accommodate particles by moving apart, rather than by stretching.
Particles coated with a mixture of PS and P2VP thiols become localized exactly at the interface between the PS and P2VP block phases (Figure 1-11b).
1-1-3-3 Nanoreactors for Nanostructure Production
Block copolymer domains can be used as “nanoreactors” for the synthesis of
inorganic NPs. Reviews of the subject are available.[45,46] Two basic approaches have been developed. The first involves the binding of inorganic species to the monomer prior to polymerization or to one of the blocks of the copolymer prior to micellization (which may be induced by the ion binding event).[45] The second, more-important approach, however, involves the loading of preformed micelles, whether in solution or in bulk.
Bronstein and coworkers employed various types of block copolymers to prepare micellar nanoreactors for the fabrication of metallic NPs. In many cases, they took advantage of the fact that the N atoms of poly(vinylpyridine), P2VP, and P4VP allow their copolymers to form complexes with metal salts.[47,48] For example, they produced palladium clusters through the reduction of Pd(CH3COO)2 coordinated to the P4VP micellar cores formed from poly(styrene-b-4-vinylpyridine) (PS-b-P4VP) diblocks in toluene.[47,49] Other metal nanoclusters, including cobalt,[50] Au,[47–49]
rhodium,[47] and platinum,[47] have been prepared in a similar manner. The catalytic hydrogenation properties of these nanoclusters havebeen evaluated.[47,49] TiO2 NPs have several interesting applications in such areas as catalysis, water purification, and UV blocking. In a previous study, our group synthesized arrayed needle-like TiO2
nanostructures using a PS-b-P4VP diblock copolymer as the template.[51] These nanostructures, which exhibited the rutile phase of crystalline TiO2, were grown on a Si substrate presenting TiO2 seeds prepared using a thin layer of PS-b-P4VP. This approach allowed the needle-like rutile TiO2 nanostructures to be fabricated with variable spatial locations and densities. For example, the distance between two TiO2
needle bunches could be controlled from 120 to 160 nm when using block copolymer templates of different molecular weights (Figure 1-12).
1-1-3-4 Photonic Crystals
Photonic crystals are attracting a huge amount of attention because they can be used to control and confine light. Materials with a complete bandgap reflect light (incident from any direction) within the wavelength range of the gap. Block copolymers are interesting materials with which to construct photonic crystals because they can self-assemble into periodic structures in one, two, and three dimensions. The inherently low dielectric contrast between the polymeric domains can be overcome through selective doping and/or removal of one component.
Theoretical predictions have revealed the possibility of photonic bandgaps existing in single- and double-network bicontinuous cubic structures.[52] These calculations have indicated, however, that (for the parameter space explored) no complete bandgaps exist for bicontinuous double primitive, double gyroid (body-centered cubic), or double diamond (face-centered cubic) structures.
Bandgaps are anticipated for the single-network analogues, with the best candidate being single-diamond structures exhibiting dielectric contrast as low as 3.6. The optical reflectivity characteristics of an experimentally realized large-domain double-gyroid structure have been assessed.[40] The initial material selected was a PS-b-PI diblock of high molar mass, leading to a cubic lattice parameter a of 258 nm.
The optical properties were measured for a bulk film and for a sample in which the PI block was removed through UV/ozone treatment (Figure 1-13), creating an interpenetrating PS network structure. As anticipated theoretically, a complete bandgap was not observed, although a wavelength range with high reflectivity was identified. This range shifted to lower wavelength in the etched structure.
1-2 Nanoparticles
Colloidal nanocrystals are sometimes referred to as “artificial atoms” because the density of their electronic states—which controls many of their physical
properties—can be widely and easily tuned by adjusting the crystal’s composition, size, and shape. The combination of size- and shape-dependent physical properties and ease of fabrication and processing makes nanocrystals promising building blocks for materials possessing designed functions.[53,54] The ability to control the uniformity of the size, shape, composition, crystal structure, and surface properties of nanocrystals is not only of technological interest: access to defined nanoscale structures is essential to uncovering their intrinsic properties, unaffected by sample heterogeneity. A rigorous understanding of the properties of individual nanocrystals will enable us to exploit them, making it possible to design and build novel electronic, magnetic, and photonic devices—and other functional materials based on these nanostructures.
1-2-1 Semiconductor Nanoparticles
Semiconductor NPs possess inorganic cores that are stabilized by a layer of surface surfactants. NPs featuring semiconductors as the inorganic material—so-called quantum dots—exhibit size-tunable band gaps. Quantum dots exhibit two deining characteristics: the surface area effect and the quantum confinement effect. The surface area effect exist for particles in a small size regime, where a large percentage of the atoms are located on or near the surface; for example, 99% of the atoms are positioned on the surface for a 1-nm-sized particle. [55] Such a vast interface between the NPs and the surrounding medium can have a profound effect on the particles’ properties; for example, the imperfect surface of the NPs may act as electron and/or hole traps upon optical excitation. Thus, the presence of these trapped electrons and holes can, in turn, modify the optical properties of the particles.
In the quantum confinement effect, “confinement” and “quantization” have two closely related definitions: if a particle is “confined,” then its energy is “quantized,”
and vice versa. According to the dictionary, to “confine” mean to “restrict within
limits” to “enclose”, and even to “imprison”. Quantum confinement not only causes an increase in the energy gap (a blue shift of the absorption edge) and splitting of the electronic states but also changes the densities of state and the exciton oscillator strength.[56] Many of the differences that exist in the electronic behavior of bulk and quantum-confined low-dimensional semiconductors are due to differences in their densities of state.
1-2-1-1 CdSe Nanoparticles
In 1993, Murray, Norris, and Bawendi described the preparation of nearly monodisperse CdS, CdSe, and CdTe semiconductor NPs.[57] CdSe NPs are potential building blocks for new electronic and optical nanodevices, such as light-emitting diodes, solar cells, lasers, and biological labels.
1-2-2 Metal Nanoparticles
Physicists predicted that metal NPs in the diameter range 1–10 nm (intermediate between the sizes of small molecules and bulk metals) would display electronic structures that reflect their electronic band structures, owing to quantum-mechanical rules.[58] The resulting physical properties are those of neither the bulk metal nor molecular compounds; instead, they depend strongly on the NP’s size, shape, interparticle distance, and nature of the protecting organic shell.[59] The few “last metallic electrons” are used for tunneling processes between neighboring particles—an effect that can be detected through impedance measurements that distinguish intra- and intermolecular processes.
1-2-2-1 Au Nanoparticles
Au NPs are the most stable metal NPs. They are fascinating materials for applications in several fields; for example, materials science (because of their multiple modes of assembly), physics [because of the behavior of individual particles and their size-related electronic, magnetic, and optical properties (quantum size effect)],
catalysis, and biology. The bottom-up approach exploited in nanotechnology is making them key materials and building blocks for the 21st century.
As mentioned above, the few “last metallic electrons” of Au NPs are used for tunneling processes between neighboring particles. The quantum size effect is involved when the de Broglie wavelength of the valence electrons is of the same order as the size of the particle itself. In such a system, the particles function electronically as zero-dimensional quantum dots (or quantum boxes) that behave according to quantum-mechanical rules. Freely mobile electrons are trapped in such metal boxes and display a characteristic collective oscillation frequency of the plasma resonance, giving rise to the so-called plasmon resonance band (PRB) observed near 530 nm in the 5–20-nm-diameter range. Au NPs are potential building blocks for memory cells,[60] single-electron transistors,[61,62] biological sensors,[63] and catalysts.[64]