Tải bản đầy đủ - 0 (trang)
2 Superlattices of Nanocrystals in Two (2D) and Three (3D) Dimensions

2 Superlattices of Nanocrystals in Two (2D) and Three (3D) Dimensions

Tải bản đầy đủ - 0trang


Superlattices of Nanocrystals in Two (2D) and Three (3D) Dimensions


Fig. 3.9 (a) Transmission electron micrograph (TEM) of 5 nm/5 nm Fe/Fe3 O4 core/shell particles. (b) TEM of 6 nm Fe3 O4 particles (see [3.18]). (c) Scanning electron micrograph (SEM) of

α-Fe2 O3 nanodisks [3.21]. (d) TEM of 20 nm FeCo particles. (e) High-resolution TEM of SmCO5

magnet particles prepared by the polyvinyl pyrrolidone (PVP) process (see [3.18]). (Reprinted with

permission from [3.18] (a) (b) (d), [3.21] (c), and [3.22] (e). © 2008 Sigma-Aldrich (a) (b) (d),

© 2009 Chinese Society of Metal (c), and © 2008 American Institute of Physics (e))

3.2 Superlattices of Nanocrystals in Two (2D) and Three (3D)


By functionalization of nanocrystals or by choosing nanocrystals with two different compositions or sizes, long-range ordered 2D or 3D structures (superlattices) with novel properties can self-assemble (see Sect. 1.5) as discussed in the


3.2.1 Free-Standing Nanoparticle Superlattice Sheets

Free-standing monolayer superlattices of alkyl-conjugated [3.23] or DNA-ligated

[3.24] gold nanoparticles were synthesized in a drying-mediated self-assembly

process starting from solution droplets on microholes. These free-standing layers exhibit thicknesses from 9.6 to 66 nm and interparticle spacings from 1.2 to

20 nm.

Close-packed nanoparticle arrays [3.23] can be formed by letting dodecanethiolligated Au nanocrystals (6 nm in diameter) suspended in toluene spread across the

top of a water droplet resting on a Si3 N4 substrate with microholes (Fig. 3.10a). As


3 Synthesis

Fig. 3.10 Self-assembled

ultrathin nanoparticle layers.

(a) Schematic diagram of the

nanoparticle layer

configuration inside a

microhole. The layer recedes

into the hole. The projection

of nanoparticles on the

vertical wall of the hole

shows up as a dark band at

the edge in (b). (b)

Transmission electron

micrograph (TEM) of a Au

nanoparticle monolayer freely

suspended over a hole with a

0.5 μm diameter in the Si3 N4

substrate. (Reprinted with

permission from [3.23].

© 2007 Nature Publishing


the water evaporates, the monolayer drapes itself over the substrate and the holes

with an average interparticle spacing of ∼1.4 nm, high order (Fig. 3.10b), and a

thickness of ∼9.4 nm. The high Young’s modulus of E = 6 GPa is a consequence of

the confinement of the ligands to the nanoparticle surface and to the spaces between

the particles, which is consistent with molecular dynamics simulations (see [3.23]).

In addition to the high Young’s modulus, the ultrathin membranes possess a high

bending flexibility (see Fig. 3.10) which should make it well suited for a wide range

of applications.

Free-standing nanoparticle superlattice sheets controlled by DNA ligands can

have both structural and functional properties rationally controlled by adjusting the

DNA length up to 20 nm [3.24]. Monodisperse 13 nm diameter gold nanoparticles are capped with 5 -thiolated single-stranded DNA (ssDNA; Fig. 3.11a);

when a droplet of colloidal solution dries on a holey substrate, 2D free-standing

well-ordered films of the capped gold nanoparticles form with interparticle spacings depending on the DNA length (Fig. 3.11b, c) and with sheet thicknesses

between ∼37 and ∼66 nm [3.24]. The optical properties of the sheets show

color variations from blue to pink under white-light illumination (Fig. 3.11d, e)

as interparticle spacing is varied which can be understood by discrete dipole

approximation simulations (see [3.24] and Sect. 7.6), suggesting dominant dipole

plasmon coupling resonance in the sheets. This behavior could enable the integration of free-standing monolayer sheets into future optoelectronic devices



Superlattices of Nanocrystals in Two (2D) and Three (3D) Dimensions


Fig. 3.11 Free-standing Au nanoparticle superlattice sheets controlled by DNA. (a) Schematic

of a DNA-capped Au nanoparticle. (b–c) Regulation of the interparticle spacing by the DNA

length DNA-T15 (b) and DNA-T90 (c). (d–e) DNA regulated plasmon coupling of superlattice

sheets. Representative photographic images of sheets with the ligands DNA-T5 (d) and DNA-T30

(e), which were acquired in transmission mode under white-light illumination. (Reprinted with

permission from [3.24]. © 2009 Nature Publishing Group)

3.2.2 3D Superlattices of Binary Nanoparticles

A wide-structural diversity can be achieved by the assembly of nanoparticles of

two different materials into a binary nanoparticle superlattice (BNSL) [3.25], which

can provide a general path to a large variety of materials (metamaterials) with precisely controlled chemical composition and tight placement of the components. In

combinations of semiconducting (PbS, PbSe), metallic (Au, Ag, Pd), and magnetic

(Fe2 O3 ) nanoparticle building blocks, space-filling, particle charging, the relative

concentrations of A and B particles, and the nanocrystal shape can be used to

engineer the structure of the self-assembled BNSLs.

Nearly spherical nanoparticles can self-assemble into BNSLs with domains

coherently packed over up to 10 μm in lateral dimensions with different structures

(see Fig. 3.12a–c). By addition of oleic acid, PbSe nanoparticles are charged positively and dodecanethiol-capped metal particles negatively, whereas the addition

of tri-n-octylphosphine oxide (TOPO) yields negatively charged PbSe nanocrystals and neutral metal particles [3.25]. Therefore, the structure of BNSLs of the

same composition can be entirely changed by changing from oleic acid to TOPO

(Fig. 3.12d–g). The nanocrystal shape can be used as an additional tool to control

the BNSL structure as shown for a combination of LaF3 triangular nanoplates and

spherical Au nanocrystals in Fig. 3.12 h.


3 Synthesis

Fig. 3.12 (a–c) Transmission electron micrographs (TEM) of the characteristic projections of

binary superlattices, self-assembled from different nanoparticles, and modeled unit cells of the

corresponding 3D structures. (a) 13.4 nm diameter γ-Fe2 O3 and 5.0 nm Au nanoparticles, NaCl

structure, (111) scale bar 20 nm; (b) 7.2 nm PbSe and 4.2 nm Ag, Cu3 Au structure, scale bar

10 nm; (c) 6.2 nm PbSe and 3.0 nm Pd nanoparticles, cubic AB13 structure, scale bar 20 nm.

The lattice projection is labeled in each panel above the scale bar. (d–g) TEM images of binary

nanoparticle superlattices (BNSLs) self-assembled in the presence of oleic acid (left column) or

tri-n-octylphosphine oxide, TOPO (right column). (d) 6.2 nm PbSe and 3.0 nm Pd nanoparticles

self-assembled into orthorhombic AB- and AlB2 -type BNSLs and (e) into NaZn13 -type BNSL.

(f–g) 7.2 nm PbSe and 4.2 nm Ag nanoparticles self-assembled into orthorhombic AB and cuboctahedral AB13 BNSLs, respectively. (h) TEM image and unit cell of BNSL self-assembled from

triangular LaF3 (9.0 nm side) nanoplates and spherical nanoparticles (5.0 nm) of Au. (Reprinted

with permission from [3.25]. © 2006 Nature Publishing Group)

Nanocrystallites of Ag (diameter 5.1 nm) and Au (4.8 nm) functionalized with

alkane thiols to final diameters of ∼8.6 and ∼8.36 nm, respectively, self-assemble

to binary nanoparticle crystals (micrometers in diameter) with a non-close-packed

diamond-like lattice and a lattice constant a = 19 nm [3.26]. Formation of this

non-close-packed structure, which is stable versus NaCl and CsCl lattices, is a


Nanowires and Nanofibers


consequence of electrostatic effects specific of the nanoscale, where the thickness of

the screening layer is commensurate with the dimensions of the assembling objects


3.3 Nanowires and Nanofibers

In the early 1980s it was theoretically predicted that quantum wires may have applications in high-performance transport devices due to their saw tooth like density

of states (see Sect. 1.3). Novel techniques for nanowire formation were reviewed


A popular technique for producing sub-10 nm features is “top-down” nanofabrication where a pattern is transferred on a material by selective irradiation

(lithography by electrons, ions, photons) of a resist wire, subsequent development

and etching. This approach, however, is hardly capable of producing structures

with dimensions in a single-digit-nanometer range. Scanning tunneling microscopy

(STM) and atomic force microscopy (AFM) fabrication techniques can achieve

these dimensions but large area patterning produced in this fashion would be time

consuming. Therefore, new techniques are required which, e.g., involve templates

for the formation of natural-forming self-organizing 1D nanostructures which are to

be discussed in the following.

Quantum wires can be grown at surface steps as templates. Atoms and molecules

adsorbed on surfaces tend to stick to the step edges of a surface where they find

extra bonding partners (see Fig. 3.13). The coverage determines the stripe width,

the miscut angle of the substrate the stripe separation. In Fig. 3.14, 3 nm wires of

Cu with a low surface energy not alloying with the high-surface energy Mo(110)

substrate are shown.

In step-flow surface growth all the familiar equilibrium growth modes reappear

such as row-by-row growth (equivalent to layer by layer), Stranski–Krastanov (layer

and island growth) and island growth.

More universal techniques emerge for structuring the surfaces of the ubiquitous

semiconductor material Si. On the reconstructed Si(111) (7×7) surface regular step

arrays appear giving rise to natural grooves (see Fig. 3.15). On this surface epitaxial

CaF2 stripes can be deposited as masks (see Fig. 3.15). Selectively between these

Step-Flow Growth

Fig. 3.13 Fabrication of

nanowire arrays by

self-assembly of atoms at

stepped surfaces with stripe

rows formed by step-growth

flow. (Reprinted with

permission from [3.27].

© 1999 Materials Research


• Coverage

Stripe width

• Miscut angle

Stripe spacing


3 Synthesis

Fig. 3.14 STM image of Cu

wires with a 3 nm width and a

50 nm separation grown by

step flow on a Mo(110)

surface. Copper appears

bright as a result of resonant

tunneling into a surface state.

(Reprinted with permission

from [3.27]. © 1999

Materials Research Society)

Fig. 3.15 Array of straight steps on Si(111) (7×7). In (a) only a single kink is detected in 20,000

edge atoms. Epitaxial CaF2 mask deposited on a stepped Si(111) (7×7) surface. (Reprinted with

permission from [3.27]. © 1999 Materials Research Society)

CaF2 stripes, e.g., organic precursor molecules for metal–organic chemical vapor

deposition (MOCVD) of metals can be deposited [3.27].

III–V semiconductor quantum wires such as gate-controlled GaAs nanowires can

be formed by using Schottky in-plane or wrap gates (Fig. 3.16) for a selective

depletion of a 2D molecular beam epitaxy (MBE) grown AlGaAs/GaAs quantum well electron gas [3.28]. The structures are formed by standard electron beam


Nanowires and Nanofibers


Fig. 3.16 Approaches for

III–V nanowire formation. (a)

Standard split-gate approach,

(b) cross-section of a

Schottky in-plane gate

(IPG)-controlled GaAs wire,

(c) cross-section of a

Schottky wrap gate (WPG)

controlled GaAs wire, and (d)

a schematic view of an

InGaAs nanowire embedded

in InAlAs formed by selective

molecular-beam epitaxy on

an InP-patterned substrate. 2

DEG denotes a 2D electron

gas. (Reprinted with

permission from [3.28].

© 1999 Materials Research


lithography, metal deposition, and etching. These gate-controlled semiconductor

nanowires exhibit quantization of conductance.

The InGaAs/InAlAs heterostructure system on a InP substrate may be attractive

for quantum devices operating at high temperatures due to its large conduction band

discontinuity and a superb electron transport. InGaAs ridge nanowires were grown

by selective MBE making use of self-organization during crystal growth on InP

substrates [3.28] (see Fig. 3.16d).

3.3.1 Vapor–Liquid–Solid (VLS) Growth of Nanowires

Semiconductor nanowires can be grown from metal nanodroplets (e.g., Au) on a

substrate. First a gold droplet reacts with Si from the gas phase to form droplets

of liquid AuSi upon Si supersaturation, then Si precipitates beneath the droplets

(see [3.31]). The droplets remain on the tips of the Si nanowires and allow them to

continue growing fed by the Si supply from the gas phase. The diameter of the wire,

which may be much less than achieved by lithography, depends on the size of the Au

particle and the length depends on the growth time. By switching between source


3 Synthesis

Fig. 3.17 Different types of semiconductor nanowires [3.31]. (a) Axially modulated InAs/InP

nanowire (30 nm in diameter) grown by chemical beam epitaxy and imaged using high-angle

annular dark-field scanning transmission electron microscopy [3.31]. The InP segments appear

darker. (b) GaP nanowires with branches of GaP, grown by metallorganic chemical vapor deposition (MOCVD) and imaged by scanning electron microscopy (SEM) [3.33]. (c) Si nanowires

grown by CVD and imaged by transmission electron microscopy. The growth direction is changed

by the periodic addition of 10% oxygen to the disilane source gas [3.31]. (d) Periodically twinned

InP nanowire imaged by SEM [3.34]. (Reprinted with permission from [3.31] (a) (c), [3.33] (b),

and [3.34] (d). © 2009 Nature Publishing Group (a) (c), 2004 Nature Publishing Group (b), 2008

Nature Publishing Group (d))

gases, nanowires of modulated composition can be created (Fig. 3.17a). Varying the

growth conditions allows cladding layers to be deposited around the wires to create

core/shell geometries. Electronic confinement within these structures is exploited

in sensors and nanoelectronic and nanophotonic devices [3.32] (see Sect. 4.2). The

addition of fresh particles to previously grown wires allows the creation of branched

structures (Fig. 3.17b). Growth direction and changes in diameter can be controlled

through surface chemistry (Fig. 3.17c).

The dependence of the structure of InP nanowires – wurtzite or zincblende –

of growth temperature, diameter, and doping level (see [3.31]) is of interest

because the two types of crystals have different band structures. In pure zincblende

InP nanowires a periodic sequence of twin planes appears (Fig. 3.17d) which is

predicted to give rise to new electronic properties, including the formation of minibands and the opening of gaps. It has been shown that in nanowires unprecedented

control of the crystal structure can be achieved [3.31, 3.34, 3.35].

The formation of nanowires in carbon nanotubes was studied soon after the

discovery of the nanotubes (see [3.29]). For this process capillarity, wetting, and

surface tension play an important role. For the synthesis of a carbon nanotube

with encapsulated Ni (see Fig. 3.18) the following mechanism is suggested [3.30].


Nanowires and Nanofibers


Fig. 3.18 Ni-filled carbon

nanotube grown by pyrolysis

of C60 –Ni substrates [3.29];

interlayer spacing of the

carbon nanotubes about

0.34 nm. (Reprinted with

permission from [3.29].

© 1999 Materials Research


Initially the metal atoms react at high temperatures with the C particles. Then a

carbide surface layer is formed and C extruded to form a concentric covering of

C nanotubes. Further Ni3 C/Ni is drawn into the tube with a simultaneous growth

of wall and core. Finally the carbon tube may be closed at decreasing temperatures (Fig. 3.19) reaching a length of up to 2.5 μm. In the C nanotubes long Fe

nanowires can be protected from oxidation and may prove advantageous in the

design of magnetic-storage devices.

The growth process of nanowires has been studied on an atomic scale by highresolution electron microscopy and density functional theory in the case of catalyzed

growth of carbon nanofibers [3.36]. The growth process occurs by surface diffusion

of C atoms to the step edges of the Ni catalytic nanocrystals (Fig. 3.20) where the

step edges act as dynamic growth sites for the carbon nanofiber. From concomitant

density functional theory (DFT) calculations, diffusion energy barriers of 0.1 eV

for Ni and 0.5 eV for C adatoms at the graphene–Ni(111) interface are deduced.

DFT shows that Ni step edges can be induced by absorbed C atoms because the C

binding energy to a Ni step is higher than the energy required for step formation.

It, furthermore, yields an estimate of the total energy for incorporation of a C atom

from the gas phase to the growing nanofiber of 1.6 eV which is in agreement with

the measured activation energy of carbon fiber growth.


3 Synthesis






Fig. 3.19 Schematic of a growth process of a Ni-filled carbon nanotube. After the reaction of the

metal with the C source, e.g., Ni-filled C60 molecules (a) or carbon-coated Ni clusters (b) may be

formed. Then carbon extruded from the coated nanoclusters may form concentric tubes (c). Upon

further uptake of Ni3 C/Ni the nanotube and the core grow simultaneously (d) and the carbon edges

close (e) when the C supply ceases. (Reprinted with permission from [3.29]. © 1999 Materials

Research Society)

Fig. 3.20 In situ TEM image

showing a Ni nanocrystal

((111) lattice fringes) during

carbon nanofiber growth.

Arrows indicate monoatomic

step edges of the Ni surface.

The Ni catalyst is exposed to

a CH4 :H2 = 1:1 mixture of a

total pressure of 200 Pa and a

temperature of ∼800 K; scale

bar 5 nm. (Reprinted with

permission from [3.36].

© 2004 Nature Publishing


3.3.2 Pine Tree Nanowires with Eshelby Twist

Hierarchical nanostructures of lead sulfide nanowires resembling pine trees

(see Fig. 3.21) were synthesized by chemical vapor deposition [3.37]. Structural

characterization reveals a screw-like dislocation in the nanowire trunks with


Nanowires and Nanofibers


Fig. 3.21 (a) High-magnification scanning electron micrograph (SEM) of PbS pine tree nanowires

highlighting the twisting (Eshelby twisting) of the central trunk and helical rotating branches. (b)

Scheme of the magnified tip of a tree structure highlighting the combined faster dislocation-driven

trunk nanowire growth and slower vapor–liquid–solid (VLS) driven branched nanowire growth. (c)

A simplified scheme illustrating that the self-perpetuating steps of a screw dislocation spiral at the

tip of a trunk can enable 1D crystal growth of nanowires. (Reprinted with permission from [3.37].

© 2008 AAAS)

helically rotating epitaxial branch nanowires. It is suggested that the screw

component of an axial dislocation provides the self-perpetuating steps to enable

1D crystal growth, in contrast to mechanisms that require metal catalysts. Using

elasticity theory, Eshelby [3.38] has shown that in a finite cylindrical rod containing

an axial screw dislocation at the center, the stress field of the dislocation exerts a

torque at the free ends of the rod, resulting in a twist of the rod along the axial direction. The rotating trunks and branches are the consequence of the Eshelby twist of

screw dislocations with a Burgers vector along the <110> directions having an estimated magnitude of 0.6 nm for the screw component. A screw dislocation-driven

nanowire growth process may be more general for other materials or may coexist

with vapor–liquid–solid (VLS) growth.

3.3.3 Ultrathin Nanowires

Ultrathin nanowires are considered to be less than 10 nm in diameter with an

especially large interest emerging in nanowires below ∼2–3 nm [3.39] with particular emphasis to nanowires of Al, Pb, Bi, Si, Rh, Ag, Cu, and Au. The synthesis

methodologies comprise templating ligand control and oriented attachment.

In the templating strategy the isotropicity of crystal growth is constrained by the

templates which may be mesoporous materials, nanocrystals, structured surfaces, or

Tài liệu bạn tìm kiếm đã sẵn sàng tải về

2 Superlattices of Nanocrystals in Two (2D) and Three (3D) Dimensions

Tải bản đầy đủ ngay(0 tr)