Tải bản đầy đủ - 0trang
7 Nanostructures by Ball Milling or Strong Plastic Deformation
Nanocrystallization of materials after deformation by high strain rates also occurs
after severe torsional deformation [3.101], as well as in tribology and wear [3.102].
In addition to the grain refinement process, alloying of conventionally immiscible components such as Fe and Cu can be achieved by ball milling due to
mechanical alloying. The dissolution of 4 nm Fe particles in a Cu matrix may
be thermodynamically favorable when the difference GFe = 2Vm σFeCu /r of the
Gibbs free enthalpy between the Fe and the Cu structures is offset, where Vm denotes
the molar volume and σFeCu = 1.37 J/m2 the interface energy [3.100], and thus the
solubility limits drastically change with grain size.
By reactive milling in gaseous or liquid atmospheres oxides, carbides, nitrides or,
e.g., metal/carbide composites [3.103] with improved ductility may be synthesized.
3.8 Carbon Nanostructures
Three novel nanostructured modifications of carbon, namely fullerenes (0D
[3.104]), carbon nanotubes (1D [3.105]), and graphene (2D [3.106, 3.107]), have
been prepared during the last two decades in addition to diamond and graphite
(Fig. 3.45). The preparation techniques for these novel nanostructures with an enormous impact in solid-state physics, chemistry, and materials science will be briefly
outlined in the following.
In 1985 the teams of H. Kroto and R. Smalley discovered highly stable clusters of
60 carbon atoms in mass spectra of laser-ablated graphite [3.104]. For this cluster
a soccer ball-shaped molecule (C60 ), a truncated icosahedron comprising 12 carbon
pentagons and 20 carbon hexagons, was suggested (Fig. 3.45).
Because of its similarity to the geodesic domes of the architect Buckminster
Fuller this cage-shaped carbon cluster was named buckminsterfullerene and a
number of larger clusters was detected. In 1990 the synthesis of macroscopic quantities of fullerenes was announced [3.113] (see Fig. 3.46) yielding crystallized C60
fullerenes, higher fullerenes as C70 , C76 , C84 , C92 , C96 and giant fullerenes with
hundreds of carbon atoms [3.114].
For producing fullerene (C60 ) molecules graphite rods are vaporized by an arc
discharge, by electron beam heating, or by sputtering in a production chamber with
a dilute atmosphere of He or an other noble gas. The C atoms are thermalized in
the gas atmosphere to form clusters which are very reactive as long as they are
small because of unsatisfied bonds at the edges. These clusters grow and finally the
totally closed, inert C60 molecules without dangling bonds survive together with
higher fullerenes. The fullerenes are extracted from the carbon soot by dissolution
via a non-polar solvent (e.g., toluene) or by sublimation. Separation of the various
fullerenes and purification can be achieved by chromatographic techniques or by
selective complexation with calixarenes [3.115].
Fig. 3.45 Crystal structures of the five forms of carbon: (a) graphite [3.108], (b) diamond [3.109],
(c) crystallized fullerene molecules (C60 ) [3.110], (d) carbon nanotube [3.111], and (e) graphene
[3.112]. (Reprinted with permission from [3.108] (a), [3.109] (b), [3.110] (c), [3.111] (d), and
[3.112] (e). © 1964 Wiley Intersience (a), © 2002 Oldenburg (b), © 1996 Institute of Physics (c),
© 1999 Springer Verlag (d), and © 2006 Nature Publishing Group (e))
The distance between two C atoms in the truncated icosahedral C60 molecule
is 0.14–0.145 nm yielding a diameter of the molecule of about 0.7 nm. The C60
molecules are about 0.29 nm apart in the solid, called fullerite, compared to the
0.34 nm distance of the graphite planes.
It has been detected that C60 [3.116] and higher fullerenes [3.117] are also synthesized in nature as deduced from the analysis of the Allende meteorite. This
indicates that fullerenes existed either in the early solar nebula or as a component of
Onion-shaped graphitic spheres were formed by high-intensity electron irradiation of carbon soot in a high-voltage electron microscope [3.118]. This structure
without dangling bonds may be the most stable form of graphite. If these graphitic
Fig. 3.46 Fullerene soot production chamber with a He atmosphere and graphite electrodes.
(Reprinted with permission from [3.110]. © 1996 Institute of Physics)
spheres are further electron irradiated at 700◦ C, their cores transform into diamonds.
From the outer layers to the more interior layers of these spheres the spacings of the
graphitic layers decreases from 0.31 to 0.22 nm in contrast to the 0.34 nm layer
separation in graphite indicating high compression [3.119] (see Fig. 3.47).
3.8.2 Single-Walled Carbon Nanotubes (SWNTs) – Synthesis
Using an arc discharge method, carbon nanotubes were initially grown on the negative end of the electrode [3.105]. The tubes consist of cylindrical graphitic layers
separated by 0.34 nm (see Fig. 3.48) and capped on the ends with fullerene-like
For the synthesis of carbon nanotubes, carbon arc, laser ablation, and chemical
vapor deposition (CVD) processes are available (see [3.120]). The CVD processes
offer the best approach to the manufacturing of larger SWNT quantities, with perhaps the most scalable being the CoMoCAT process which uses a fluidized bed
reactor (see Fig. 3.49a) similar to those in petroleum refining, albeit, on a much
smaller scale. In this CoMoCAT method, SWNTs are grown by CO disproportionation (decomposition into C and CO2 ) at 700–950◦ C in flow of pure CO at a
Fig. 3.47 Spherical graphitic
particle (onion) with a
diamond core 10 nm in size
(1.8 × 10−17 carat) and the
diamond lattice fringes
separated by 0.206 nm, after
electron irradiation at 700◦ C.
(Reprinted with permission
from [3.119]. © 1996 Nature
pressure of 1–10 atm with a selectivity toward SWNTs with a diameter between
0.72 and 0.92 nm (> 90%) and a (6, 5) chirality of 52% [3.120]. The catalyst
is effective when both metals, Co and Mo, are simultaneously present on a silica
For the characterization of SWNTs, the microscopy techniques (see Chap. 2) of
scanning tunneling microscopy (STM), atomic force microscopy (AFM), and transmission electron microscopy (TEM) are available, as well as Raman spectroscopy,
optical absorbance spectroscopy, and thermogravimetric analysis (TGA) . In Raman
spectroscopy (Fig. 3.49b), the radial breathing mode (RBM) from 120 to 300 cm−1
is unique to SWNTs and can be used to determine the tube diameter d (in nm) from
the wave number ν (in cm-1 ) according to
ν = 238/d 0.93 .
The D band at 1300–1350 cm−1 is indicative of disordered carbon, multiwall
tubes, and microcrystalline graphite, and the G band at 1500–1586 cm−1 is a result
of the tangential stretching mode from graphitic-like materials and the G/D ratio has
been used as a measure of the purity of the SWNTs [3.120]. The optical absorption
measurements in the UV–Vis–NIR region show peaks which are characteristic of
individual (n, m) species. For example the (6, 5) species absorb at 566 and 976 nm
and fluoresce at 983 nm, whereas a (7, 6) SWNT absorbs at 645 and 1024 nm and
fluoresces at 1030 nm. The main peak of the TGA trace represents the oxidation
of SWNTs, while the second peak is indicative of the presence of other forms of
Fig. 3.48 Electron micrographs of nanotubes of graphitic carbon [3.105]. A cross section of each
tubule is illustrated. (Reprinted with permission from [3.105]. © 1991 Nature Publishing Group)
The caps of carbon nanotubes can be opened by oxidation [3.121] and filled
[3.122] with various materials in order to synthesize nanowires. Large areas of
highly aligned, isolated carbon nanotubes were prepared by 700◦ C deposition of
acetylene on sol–gel synthesized mesoporous silica with iron nanoparticles in the
Crystalline ropes of metallic regularly ordered 2D arrays of SWNTs can be
synthesized [3.124] by a laser-oven method with a uniform SWNT diameter of
∼1.38 nm. They self-organize to hexagonal 2D rope-like crystallites 5–20 nm in
diameter and tens to hundreds of micrometers long with a 1.7 nm lattice constant as
evidenced by transmission electron microscopy (TEM) and x-ray diffraction (XRD)
(see Fig. 3.50). All SWNTs terminate together at the end of the rope.
Fig. 3.49 (a) Illustration of a fluidized bed reactor which is able to scale up the generation of
SWNTs using the CoMoCAT process. (b) Typical Raman spectrum of (6, 5) SWNTs, obtained
with 633 nm laser excitation. (Reprinted with permission from [3.120]. © 2009 Sigma-Aldrich)
Fig. 3.50 Scanning electron micrographs (SEM) of single-wall carbon nanotube (SWNT) bundles (ropes). (a) SWNT rope consisting of ∼100 SWNTs bending through the image plane of the
microscope, showing uniform diameter and triangular packing of the tubes within the rope. (b)
Imaging of the SWNT rope perpendicular to the tube axes. Scale bars 10 nm (a) and 20 nm (b).
(Reprinted with permission from [3.124]. © 1996 AAAS)
Graphene is a single layer of sp2 -bonded carbon which can be seen as an individual
atomic plane pulled out of bulk graphite. This material was found in its free state
only recently when individual graphene samples of a few microns in width were
isolated by micromechanical cleavage of graphite. The current intense interest in
graphene is driven by the high crystal quality and ballistic transport at sub-micron
distances [3.125, 3.126] as well as the behavior of quasi-particles as massless Dirac
fermions so that the electronic properties are governed by quantum electrodynamics
rather than the standard physics of metals based on the (non-relativistic) Schrödinger
equation [3.125, 3.126].
Preparation and stability: Graphene up to sizes of 100 μm (Fig. 3.51b) can be
obtained by micromechanical cleavage of bulk graphite [3.127] although graphene
Fig. 3.51 (a) Crystallographic structure of graphene. Atoms of the two sub-lattices (A and B) are
marked by different colors [3.128]. (b) TEM image of a suspended graphene membrane. Its central
part (arrows) is monolayer graphene. Electron diffraction shows that it is a single crystal without
domains. Scrolled top and bottom edges and a strongly folded region are visible on the right.
Scale bar, 500 nm [3.135]. (c) TEM atomic resolution image of a few-layer graphene membrane
near its edge, where the number of horizontal dark lines indicates the thickness of two to four
layers. Because the electron contrast depends strongly on the incidence angle, small (few degrees)
variations of the surface normal are visible. Scale bar, 1 nm [3.135]. (Reprinted with permission
from [3.128] (a) and [3.135] (b) (c). © 2007 Materials Research Group (b) (c))
crystallites left on a substrate are extremely rare [3.129]. Graphene crystals with a
few layers (1,2,3,4,5) can be discerned by Raman spectroscopy [3.130].
The fact that 2D atomic crystals do exist and are stable under ambient conditions is amazing by itself. Peierls [3.131] and Landau [3.132] predicted that strictly
2D crystals were thermodynamically unstable and could not exist because a divergent contribution of thermal fluctuations in low-dimensional crystals should lead
to such large displacements of atoms that they become comparable to interatomic
distances and dislocations should appear in 2D crystals [3.133] at any finite temperature. However, strong interatomic bonds can ensure that thermal fluctuations cannot
lead to the generation of dislocations [3.133] and 2D crystals are intrinsically stabilized by gentle crumpling in the third dimension [3.134] which is reported from
x-ray diffraction experiments [3.135].
For future large-scale application of graphene, two requirements are of importance as discussed in the following. Convenient synthesis of larger quantities and
control of the pattering, morphology, and crystallinity of the edges of graphene
nanoribbons, because these nanoribbons can exhibit either quasi-metallic or semiconducting behavior, depending on the atomic structure of their edges [3.136].
Nanoribbons less than 10 nm wide are expected to be semiconductors, independent
of their edge patterns (see [3.137]).
Gram-scale production of graphene (see Fig. 3.52) has been achieved [3.138] by
reacting ethanol and sodium to an intermediate solid that is then pyrolyzed, yielding a fused array of graphene sheets that are dispersed by mild sonication. By a
solution-based method for large-scale production, uniform films of single and/or
few-layer chemically converted graphene can be produced over the entire area of a
silicon/SiO2 wafer [3.139]. Epitaxial graphene layers have been grown on singlecrystal 4-inch silicon carbide wafers [3.140] and a number of additional chemical
methods are available for the production of graphene (see [3.141]).
Fig. 3.52 (a) Example of the bulk quantity (∼2 g) of the graphene product. (b) Transmission
electron micrograph of the agglomerated graphene sheets. Scale bar, 200 nm. (Reprinted with
permission from [3.138]. © 2009 Nature Publishing Group)
3.9 Nanoporous Materials
Bulk materials with a nanoporous structure can be synthesized (see below) from
oxides with widespread applications in catalysis, photocatalysis, and separation,
from semiconductors due to band gap engineering, and from metals for improved
fuel cell electrodes or optical and electronic applications. The field of nanoporous
materials is sometimes subdivided into microporosity (pore diameter less than 2 nm)
and mesoporosity (pore diameter 2–100 nm). Between these two length scales the
transition between monolayer and multilayer adsorption occurs and the assumptions
of classical adsorption theories can break down (see [3.142]). Recent developments
in zeolite synthesis could bridge this gap with the generation of ever larger pore
sizes. It may be not catalysis or separation where the nanoporous materials with
larger pore sizes will find use. They may be more important in the field of sensors
or photonics or low-dielectric constant materials.
3.9.1 Zeolites and Mesoporous Metal Oxides
Zeolites are crystalline oxide materials with a tetrahedral atomic framework structure that contains cavities in which “guest” molecules can move relatively freely.
Such molecules can rapidly be adsorbed, react and desorb, making zeolites particularly useful for catalytic and separation application (see [3.142]). The framework
is composed mainly of silica tetrahedra, but many other elements, such as Al, Ti,
or Ge, can be incorporated. One of the most useful properties of zeolites is their
ability to select molecules based on size (molecular sieving). But for some applications, such as cracking of hydrocarbons, and specifically hydrocracking, this is in
fact a limitation, as the large molecules one may wish to react do not have access
to the reaction sites located inside the crystals. A measure of pore size is how many
oxygen atoms form the ring that define a pore channel. An 18-ring zeolite (ITQ-33)
with a pore diameter of 1.3 nm (Fig. 3.53a, b) has been synthesized [3.143] making
use of an organic hexamethonium template and additional fluoride and germanium
ions, yielding a large void fraction of 0.37 cm3 /g. The additional medium-size pores
(10-ring, 0.56 nm, see Fig. 3.53b) that run perpendicular to the 18-ring pores can
bypass blocked pores. It should be pointed out that the structure of the ITQ-33
was “predicted” theoretically by algorithms that generate framework structures of
One may ask the question whether materials with even larger pores are needed
[3.142]. Some of the unique properties of zeolites arise from the large curvature
of their pores. As the pores get larger, the interaction of adsorbates increasingly
resembles the interaction with a flat surface. Yet, large pore materials will find use
as sensors or in photonics, or where the low dielectric constant (low k) of such materials, arising from their porosity, can be exploited in the manufacture of improved
microelectronic devices (see Sect. 9.8).
Zeolitic materials with extralarge pores and channels displaying only one enantiomorph have been synthesized with the ITQ-37 chiral zeolite [3.145] of the
Fig. 3.53 (a) Structure of the Si–Ti–Ge zeolite ITQ-33 [3.143]. View of the framework down the
18-member ring (a pore defined by a ring containing 18 oxygen atoms). The green and yellow
molecule is benzene, which is 0.55 nm in diameter and is included for size comparison. (b) The
10-ring pore windows that interconnect the 18-ring channels [3.142], [3.143]. (c) The framework
structure model of the 30-ring ITQ-37 zeolite, viewed along the c-axis, derived from the electron
density map which is deduced from selected area electron diffraction (SAED) and powder x-ray
diffraction (PXRD). The Si and Ge atoms are in yellow and oxygen atoms are in red [3.145].
(Reprinted with permission from [3.142] (a) (b) and [3.145] (c). © 2006 Nature Publishing Group
(a) (b) and © 2009 Nature Publishing Group (c))
formula |(C22 N2 H40 )10.5 (H2 O)| [Ge80 Si112 O400 H32 F20 ] (Fig. 3.53c). The stable
30-ring ITQ-37 zeolite with an asymmetric pore opening of 0.43 nm × 1.93 nm,
a low framework density of 10.3 Si and Ge atoms per cubic nanometer, a BET
(Brunauer–Emmett–Teller) surface area of 900 m2 g−1 , and a micropore volume of
0.38 cm3 g−1 approaches the mesoporous range [3.145].
Mesoporous transition metal oxides with ∼20-nm pore size, high crystallinity
and high thermal stability can be synthesized [3.146] by making use of transition metal oxide sols and amphiphilic diblock copolymers. Mesoporous TiO2 and
Nb2 O5 may be used in photovoltaic cells and fuel cells, respectively [3.146]. When
mesoporous TiO2 is applied in photocatalysis, a highly crystalline TiO2 matrix is
required, because amorphous regions are known to be trap sites for the recombination of photoexcited electrons and holes, limiting the device efficiency (see
[3.146]). For the synthesis of the mesoporous crystalline metal oxide, first a mixture
of the block copolymer poly(isoprene-block-ethylene oxide) (PI-b-PEO) with the
metal oxide sol leads to a film of hexagonally arranged PI cylinders embedded in a
matrix containing PEO and the metal oxide sol (Fig. 3.54a). For PEO burn-off, PI
conversion to amorphous carbon and crystallization of the oxide walls, the film is
heated under argon to 700◦ C (Fig. 3.54b). Finally, the residual carbon is removed by
heating the film in air to 450◦ C, leaving a highly crystalline mesoporous transition
metal oxide behind (Fig. 3.54c). Mesoporous TiO2 (anatase) prepared by this procedure contains pores with a size of ∼24 nm and fully crystalline walls of a thickness
of 5 nm, and exhibits a surface area of 89 m2 g−1 [3.146].
Fig. 3.54 (a) Schematic representation of the formation of mesoporous transition metal oxides.
In situ formed carbon acts as a rigid support and enables the synthesis of highly crystalline
mesoporous oxides with large uniform pores. (b) High-resolution transmission electron micrograph (HRTEM) of a TiO2 mesoporous film. Inset: Image showing the d101 lattice spacing,
d101 = 0.348 nm, of the TiO2 anatase structure consistent with literature values. (Reprinted with
permission from [3.146]. © 2008 Nature Publishing Group)
Nanoporous anodic aluminum oxide (AAO) with self-organized hexagonal arrays
of uniform parallel nanopores have been used for sensing, storage, separation, and
nanowire synthesis (see [3.147]). Self-ordered AAOs have been obtained by mild
anodization (MA) within limited processing windows, whereas hard anodization
(HA) with high current densities using sulfuric acid is a faster process but the
pores are less ordered than those in the MA process. By a combination of the
MA and HA processes in a periodic pulse sequence consisting of a low-potential
pulse followed by a high-potential pulse, a tailoring of the pore structure as well
as of the chemical composition of the AAO along the pore axis can be achieved
(see Fig. 3.55).
Nanoporous membranes can be used as templates for nanoscopic metal particles.
The optical properties of the nanometal–insulator composites are of particular interest [3.148]. The length and aspect ratios of the wires or nanotubes can be controlled
by the amount of metal deposited and the pore diameter. For obtaining nanotubes the
pore walls are chemically modified for preferential deposition of the metal ions. As
a molecular anchor for metal ions a cyanosilane is used to react with the hydroxyl