TEM for Characterization of Core-Shell Nanomaterials

Abstract

Transmission electron microscope is an essential tool for characterization of nanoscale materials and devices because it can shed light on the microstructure of nanomaterials. For core-shell nanostructured materials, transmission electron microscopy (TEM) can provide much more important information: overall particle size, core size, shell thickness, uniform or nonuniform shell coating, lattice fringe, elemental distribution, etc. In this chapter we will describe the application of TEM for characterization of core-shell nanomaterials.

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Definition of the Topic

Transmission electron microscope is an essential tool for characterization of nanoscale materials and devices because it can shed light on the microstructure of nanomaterials. For core-shell nanostructured materials, transmission electron microscopy (TEM) can provide much more important information: overall particle size, core size, shell thickness, uniform or nonuniform shell coating, lattice fringe, elemental distribution, etc. In this chapter we will describe the application of TEM for characterization of core-shell nanomaterials.

2 Overview

Nanomaterials have, by definition, at least one dimension in the range of 1–100 nm and subsequently show novel properties different from their bulk materials. The synthesis, characterization, and applications of nanomaterials are the most important parts among the wide range of nanotechnology areas falling under the general “nanotechnology” umbrella. In recent years, core-shell nanomaterials have attracted much attention for their excellent physical properties and chemical stability.

However, traditional characterization tools such as scanning electron microscope (SEM) and atomic force microscope (AFM) can only reflect the surface features of core-shell nanomaterials, lacking detailed information from the core. This is due to the fact that the core is embedded in the shell which is made up of complex materials such as metal, silica, and organics.

Here we describe how to employ TEM to investigate the interface between the core and shell, shape of the core and shell, uniform or nonuniform shell coating, formation of core-shell structure, etc.

3 Introduction

Nanostructured materials have drawn significant attention as potential building blocks for nanocomposites, nanoscale electronic devices, ultrahigh-density magnetic recording systems, and optical devices. The most important characteristics, among many others, on a nanoscale are as follows. First, the small size of nanomaterials leads to an increased surface area to volume ratio and as a result the quantum confinement effects dominate. Second, the increasing surface area to volume ratio leads to an increase in the dominance of the surface atoms over those in its interior.

Initially a lot of research work focused on single-phase nanoparticles because such nanomaterials had much better properties than bulk materials. In the late 1980s, it was found that heterogeneous composite or sandwich colloidal semiconductor nanoparticles had better efficiency than their corresponding single-phase particles; in some cases they even demonstrated some new properties [1–3]. More recently during the early 1990s, researchers synthesized concentric multilayered semiconductor nanoparticles with an aim to improving their properties. Hence, the terminology “core-shell” was subsequently adopted [4–6]. Furthermore, there has been a gradual increase in research activities because of tremendous need for more and more advanced materials fueled by modern technology. Simultaneously the advancement of characterization techniques has greatly helped to establish the structures of these different core-shell nanomaterials. A statistical data analysis is presented in Fig. 6.1 to show the increasing trend of published research papers in this area. These were collected in June 2012 from “SciFinder Scholar” using the keyword “core-shell nanoparticles.”

Fig. 6.1

Publications per year for core-shell nanoparticles during the period from 1994 to June 2012 (Data collected from SciFinder Scholar Database)

In recent years, the advances in new synthesis techniques have made it possible to fabricate not only the symmetrical (spherical) nanoparticles but also a variety of other shapes such as cube [7–14], prism [15, 16], hexagon [7, 8, 17–20], octahedron [11, 12], disk [21], wire [22–29], rod [22, 30–37], tube [22, 38–41], etc. Moreover, the structure and composition of core-shell nanomaterials also become more complicated. It means that core-shell nanomaterials are no longer simple spherical particles but are completely coated by a shell of different materials. The shell may have a complex multilayer structure [42–45], and the core may move freely instead of being fixed by the shell [46, 47]. Different classes of core-shell nanoparticles are shown schematically in Fig. 6.2. These core-shell nanomaterials have aroused immense interest because of their novel properties.

Fig. 6.2

Different kinds of core-shell nanomaterials: (a) Spherical core-shell nanoparticles. (b) Hexagonal core-shell nanoparticles. (c) Double shell core-shell nanoparticles. (d) Movable core with hollow shell nanoparticles. (e) Core-shell nanowires

Current applications of different core-shell nanoparticles were summarized in a review article by Karele et al. [48] The individual report from different researchers also demonstrated the fact that core-shell nanoparticles are widely used in different applications such as biomedical [49–52] and pharmaceutical applications [53], catalysis [54, 55], electronics [4, 56, 57], enhancing photoluminescence [58–60], creating photonic crystals [61], etc. In particular, in the biomedical field, the majority of these particles were used for bioimaging [51, 62–68], controlled drug release [68, 69], targeted drug delivery [51, 65, 68–70], cell labeling [51, 71], and tissue engineering applications [69, 72].

Unfortunately, traditional characterization techniques are not good enough to demonstrate the growing complexity of core-shell nanomaterials. However, TEM is an effective method [73] to unlock the secret of the core due to its unique imaging procedure and various techniques.

At present, lots of books have focused on TEM techniques or applications, but few on TEM characterization of core-shell nanomaterials. This chapter is designed to illustrate some TEM techniques for characterizing the core-shell nanostructures, such as diffraction contrast imaging, high-resolution TEM (HRETM), high-angle annular dark-field (HAADF), and elemental mapping. We will take the TEM techniques as a clue to discuss the application of TEM for characterization of core-shell nanomaterials. The characteristic of each TEM technique will be explained and some up-to-date research work will be demonstrated.

4 Experimental and Instrumental Methodology

4.1 Synthesis of Core-Shell Nanomaterials

Approaches for the synthesis of nanomaterials can be broadly divided into two categories: “top-down” and “bottom-up.” The “top-down” approach often adopts traditional workshop or microfabrication methods where externally controlled tools are used to cut, mill, and design materials into the desired shape and order. For example, the most common techniques are lithography techniques [74, 75], laser-beam processing [76], and mechanical techniques [77–79]. “Bottom-up” approach, on the other hand, exploits the chemical properties of the molecules to let them self-assemble into some useful conformations. The most common bottom-up approaches are chemical synthesis, chemical vapor deposition, laser-induced assembly, self-assembly, colloidal aggregation, film deposition and growth [80–82], etc. Currently it is hard to say which approach is superior because each has its advantages and disadvantages. However, the bottom-up approach can produce much smaller particles and has the potential to be more cost-effective in the future due to the advantages of absolute precision, complete control over the process, and minimum energy loss compared with that of a top-down approach. As far as the synthesis of core-shell nanomaterials is concerned, the bottom-up approach has proven to be more suitable since the ultimate control is required for achieving a uniform coating of the shell materials during the particle formation. A combination of these two approaches can also be utilized. For example, core particles can be synthesized by a top-down approach and then coated by a shell fabricated by a bottom-up approach which could maintain uniform and precise shell thickness. To control the overall size and shell thickness precisely, a microemulsion instead of a bulk medium is preferable because water droplets can act as a nanoreactor template.

4.2 Transmission Electron Microscopy Techniques

TEM is a characterization technique whereby a beam of electrons transmits through an ultrathin specimen and interacts with the atoms or molecules in the specimen [73]. TEM is capable of imaging at a significantly higher resolution than light microscopes, owing to the small de Broglie wavelength of electrons. TEM has various imaging techniques, such as diffraction contrast imaging, high-resolution TEM, high-angle annular dark-field, and elemental mapping.

TEM bright field (BF) image is mainly caused by amplitude contrast. Amplitude contrast results from variations in mass or thickness or a combination of the two: the thickness variation can produce contrast because the electron interacts with more material (hence, more mass). Alternatively, diffraction can vary locally because the specimen is not a perfect, uniformly thin sheet. In order to translate the electron scatter into interpretable amplitude contrast, we use objective aperture which is placed in the back focal plane of the objective lens to select the direct beam in the selected area electron diffraction (SAED) to form BF images. Regions of no specimen show a bright background, and regions of the specimen that are thick or dense will present dark in the image.

The HAADF image is also called Z-contrast image. The HAADF image contrast is usually proportional to the Z² (Z is atomic number). Because of Bragg scattering, normal ADF detector is not suited for the study of crystalline specimens. But we can decrease the camera length with the post-specimen lenses to ensure that the Bragg electrons (including any HOLZ scattering) do not hit the detector. Thus, only the electrons scattered through very high angles contribute to the image. Bragg scattering effects are avoided if the HAADF detector only gathers electrons scattered through an angle larger than 50 mrad (∼3°).

HRTEM is an important imaging technique in TEM, from which we can obtain the atomic structure information from a specimen. HRTEM image is mainly caused by phase contrast, and it is produced by interference of the transmitted beam with at least one diffracted beam. When performing HRTEM experiments, we should select a larger objective aperture and let through more beams carrying with their amplitudes and phases to produce a phase contrast image.

Elemental mapping is one important technique in energy-filtering transmission electron microscopy. Elemental maps extracted from ionization edges can obviously show the spatial distribution of elements in samples. Usually two methods, two-window and three-window, are used to get this information. Two-window method is acquiring two images from electrons in selected energy windows, a pre-edge background image and a post-edge image, and then obtaining the ratio images of pre- and post-edge windows, which can give a qualitative elemental distribution. Three-window method is acquiring three images from electrons in selected energy windows: two pre-edge windows used to calculate the background fit and one post-edge window in which the extrapolated background is subtracted from the total intensity to leave the edge intensity. This method can give quantitative images of the distribution of specific elements.

Alternative operation modes of use allow for TEM to observe modulations in chemical composition, crystal orientation, and electronic structure.

5 Key Research Findings

5.1 Application of Diffraction Contrast Imaging in Nanomaterials

TEM image contrast arises because of the scattering of the incident beam by the specimen. For core-shell structure materials, the components of core and shell are different. As a result, it will produce a strong contrast in the BF image. Through the BF image, we can determine the formation of core-shell structure and measure the thickness of core and shell. In this section, we will discuss the application of TEM BF image for characterization of core-shell nanomaterials in detail.

5.1.1 Silica-Coated Core-Shell Nanomaterials

The silica coating has several advantages. The most basic advantages of the silica coating compared with other inorganic (metal or metal oxide) or organic coatings are as follows: It reduces the bulk conductivity and increases the suspension stability of the core particles. In addition, silica is the most chemically inert material available, and it can block the core surface without interfering the redox reaction at the core surface. Silica coatings can also be used to modulate the position and intensity of the surface plasmon absorbance band since silica is optically transparent. As a result, chemical reactions at the core surface can be studied spectroscopically. Therefore, researchers are more interested in the silica coatings on different inorganic core materials such as metals [83–94], binary inorganic composites [95–97], metal oxides [98–101], and metal salts [88, 102–106] than any other combination.

According to the literature, the shell thickness from 8 to 100 nm can be controlled by adjusting the experimental parameters such as coating time, concentration of reactants, catalyst, and other precursors [83, 84, 87]. Figure 6.3 shows Au@SiO₂ nanoparticles where the shells have different thicknesses [107]. These nanoparticles were prepared in the following steps. Gold colloids were homogeneously coated with silica using the silane coupling agent (3-aminopropyl)-trimethoxysilane as a primer to render the gold surface vitreophilic. After the formation of a thin silica layer in aqueous solution, the particles were transferred into ethanol for further growth using the Stöber method [108]. From the BF image, we can find that the core is darker compared with the shell, mainly because the gold core has a stronger scattering ability than the silica shell. The image contrast is so clear that we can easily measure the size of core and shell through the image. The gold core is ∼15 nm in diameter and a silica shell thickness ranges from 8 to 28 nm.

Fig. 6.3

TEM images of Au@SiO₂ nanoparticles. The core dimension is ∼15 nm and the silica shells are around 8, 18, and 28 nm, respectively. Scale bars are 50 nm in all images [107] (Reprinted with permission from ref 107. Copyright 2002 Wiley-VCH)

CoFe₂O₄ receives much attention in the biomedical field for its high magnetic anisotropy and saturation magnetization which give rise to suitable magnetic behavior at room temperature, but the presence of cobalt makes it potentially toxic [109, 110]. To protect magnetic nanoparticles, encapsulation both in polymeric and inorganic matrices has been proposed [111], while silica has been most often used [112]. Spherical nanoparticles of surfactant-coated CoFe₂O₄ (core) were prepared through thermal decomposition of metal acetylacetonates in the presence of a mixture of oleic acid and oleylamine and uniformly coated with silica shell by using tetraethyl orthosilicate and ammonia in a micellar solution (core/shell) [113]. TEM analysis of core-shell nanoparticles evidenced the high homogeneity of the coating process in producing single core-shell nanoparticle with a narrow size distribution. TEM images (Fig. 6.4a, b) in BF mode show the formation of spherical core-shell structures with an average overall size of 30 nm and a polydispersity of 5 % (Fig. 6.4c) with a single magnetic core in the center of the sphere. TEM image in dark-field mode (Fig. 6.4d) confirms the high degree of crystallinity of the core and the amorphous nature of the shell. The assembling of the nanoparticles appears to be in the form of hexagonal close packing. In some cases a deviation from spherical shape can be observed (Fig. 6.4b); this can be caused by a slight deformation of the particles along the close packing direction.

Fig. 6.4

TEM data of CoFe₂O₄-SiO₂ core-shell nanoparticles. (a) BF image at low magnification. (b) BF image at high magnification. (c) Particle size distribution. (d) Dark-field image [113] (Reprinted with permission from Ref. [113]. Copyright 2010 American Chemical Society)

In some cases the density/concentration of one shell is different. Since the shells with different density/concentration own different scattering ability which will show different TEM image contrast, TEM BF imaging technique is also applied to characterize such kind of materials. Ge et al. [42] prepared nanoparticles in a facile and scalable way, and the procedure was outlined in Fig. 6.5. A monolayer of the metal nanocatalyst was first immobilized on the surface of silica colloids by using coupling agents. The core-satellite structures were then coated with another layer of silica of the desired thickness to fix the position of metal nanoparticles. Finally, a “surface-protected etching” technique was applied to make the outer shell mesoporous, exposing the catalyst particles to outside chemical species [114]. To improve the recyclability, they also incorporated a superparamagnetic Fe₃O₄ core at the center of the initial silica colloids [115, 116]. The etching process can be well controlled by monitoring the transmittance of the colloidal solution. With an increase of etching time, more silica materials dissolved in the form of soluble silicate oligomers, and accordingly the transmittance increased. Figure 6.6 shows typical TEM BF images of five Fe₃O₄/SiO₂/Au/por-SiO₂ composite colloids collected after 50, 65, 85, 95, and 105 min of etching [42]. It is clear that the thickness of shell shows no apparent change as time goes on, but the contrast between core and shell tends to be sharp. This is due to the fact that the shell is composed of a porous structure which has a weak scattering ability. Consistently, the transmittance of the solution increases from 29 % to 45 % with a near-linear dependence on the reaction time (Fig. 6.6f).

Fig. 6.5

Synthetic procedure of porous silica protected Fe₃O₄/SiO₂/Au composite structures [42] (Reprinted with permission from Ref. [42]. Copyright 2008 Wiley-VCH)

Fig. 6.6

(a–e) TEM BF images of Fe₃O₄/SiO₂/Au/por-SiO₂ composite colloids collected after different etching times: (a) 50, (b) 65, (c) 85, (d) 95, and (e) 105 min. (f) Changes in transmission intensity of the solution (λ = 1,000 nm) with etching time [42] (Reprinted with permission from Ref. [42]. Copyright 2008 Wiley-VCH)

As an alloying electrode material, Si has attracted much attention because of its highest known theoretical charge capacity. One interesting behavior for an amorphous Si (a-Si) is that it reacts with lithium (Li) at slightly higher potential (∼220 mV) [117, 118] than crystalline Si (c-Si) does (∼120 mV) [119, 120], which leads to an idea of using c-a core-shell Si nanowires (NWs) as an anode material. When limiting the charging potential, it should be possible to utilize only the amorphous shell material for Li storage while preserving the crystalline core as mechanical support and efficient electron transport pathways, as indicated in Fig. 6.7a. Figure 6.7 shows c-a core-shell Si NWs grown directly on stainless steel (SS) current collectors by a simple one-step synthesis. Cui et al. [121] found that large flow, high pressure, and high temperature promote the yield of c-a core-shell Si NWs on SS substrates. As shown in Fig. 6.7b, the thickness of amorphous shell increases linearly with growth time, while the core radius does not change. This suggests that c-Si cores grew first and a-Si was subsequently coated onto the cores from SiH₄ decomposition. Figure 6.7c–h shows the TEM images, selected SAED patterns, and HRTEM images of Si NWs grown at 485 °C for different growth times. After 10 min of growth, the NWs were mostly single crystalline (Fig. 6.7c, d) with little amorphous shell. After 20 min, a thick layer of amorphous shell was observed (Fig. 6.7e, f), which became even thicker after 40 min (Fig. 6.7g, h). Consistently, the SAED pattern in Fig. 6.7g shows amorphous diffraction ring which is not found in Fig. 6.7c.

Fig. 6.7

(a) Schematic illustration of the lithiation of the c-a Si core-shell NWs grown on a stainless steel substrate. (b) Statistics of the core radius (red) and shell thickness (green) versus growth time. The growth temperature was 485 °C. (c) TEM and SAED images of NWs grown for 10 min. (d) HRTEM image of a nanowire grown for 10 min. (e) TEM and SAED images of NWs grown for 20 min. (f) HRTEM images of a nanowire grown for 20 min. (g) TEM and SAED images a nanowire grown for 40 min. (h) HRTEM images of a nanowire grown for 40 min [121] (Reprinted with permission from Ref. [121]. Copyright 2009 American Chemical Society)

5.1.2 Bimetallic Core-Shell Nanomaterials

Bimetallic core-shell and alloy nanoparticles have received intense attention, owing to their novel optical, electronic, magnetic, and catalytic properties different from those of individual metals [122–124]. Since these properties strongly depend on composition, shape, and size of the nanoparticles, extensive studies have been focused on the controlled synthesis of these nanoparticles with specific compositions and morphologies [125–140]. For the bimetallic core-shell nanostructures, a direct approach to determine their structure is TEM because a clear boundary between core and shell can be distinguished by bright or dark contrast in the TEM BF image. HAADF and HRTEM techniques can also be employed to characterize bimetallic core-shell structure. We will introduce them in detail in the following section.

Tsuji et al. [141, 142] synthesized shape-dependent Au@Ag core-shell nanocrystals successfully by using a two-step method. In order to understand growth mechanisms of these Au@Ag core-shell particles, they added Au seeds with different shapes into AgNO₃/DMF solution at [AgNO₃]/[HAuCl₄] molar ratios of 1, 9, and 18, respectively. From the TEM BF images (Fig. 6.8), we can observe a mixture of Au@Ag core-shell nanocrystals with various shapes. Obviously, truncated-triangular and hexagonal plate-like Ag shells overgrew from triangular and hexagonal Au cores, respectively, whereas decahedral and octahedral Ag shells overgrew from the decahedral and octahedral Au cores, respectively. In addition, it can be clearly seen that the thickness of shell increased over [AgNO₃]/[HAuCl₄] molar ratio. At low [AgNO₃]/[HAuCl₄] molar ratio of 1, thin triangular and hexagonal shells are epitaxially formed over the triangular and hexagonal Au plate cores (Fig. 6.8a, b). With an increase of the [AgNO₃]/[HAuCl₄] molar ratio, larger triangular, truncated-triangular, or hexagonal Ag shells are overgrown. The edge length of Ag shells enlarges with an increase of the molar ratio for the [AgNO₃]/[HAuCl₄]. At the highest [AgNO₃]/[HAuCl₄] molar ratio of 18, Ag shell edges can achieve about three times longer than those of plate-like Au cores, but Au cores are still observed easily in a constant contrast. This indicates that the Au@Ag core-shell particles have a plate-like shape. In order to further confirm the crystal structure of triangular and hexagonal particles, TEM imaging has been carried out from different view angles within ±16° (Fig. 6.9). No significant change can be found in the bright and dark contrast of these triangular and hexagonal particles.

Fig. 6.8

TEM BF images of the Au cores synthesized by a microwave heating for 3 min and the Au@Ag nanocrystals prepared at different [AgNO₃]/[H₂AuCl₄] molar ratios by an oil-bath heating [142] (Reprinted with permission from Ref. [142]. Copyright 2008 American Chemical Society)

Fig. 6.9

TEM BF images of hexagonal Au@Ag nanocrystal observed from different view angles [142] (Reprinted with permission from Ref. [142]. Copyright 2009 American Chemical Society)

5.1.3 Hollow Core-Shell Nanomaterials

As a unique class of structured materials, hollow colloidal particles have attracted growing research efforts owing to their technological importance in a wide range of applications [53, 143–148]. Templating against colloidal particles is probably the most effective and general method for preparation of hollow particles, especially for studies in which a narrow size distribution is required, i.e., self-assembly of photonic crystals. Monodisperse latex and silica spheres are commonly used as colloidal templates because they are readily available in a wide range of sizes [149–156]. In this section we will introduce some kinds of hollow core-shell nanomaterials which are characterized by TEM.

Figure 6.10 shows hollow core-shell nanoparticles which were designed for a double-electrode nanomaterial composed of a V₂O₅ matrix containing a low weight ratio of SnO₂ nanocrystals (10 % or 15 %) [45]. In this nanostructured composite electrode material, SnO₂ nanocrystals are homogenously distributed in a double-shelled V₂O₅ hollow nanocapsule. The V₂O₅-SnO₂ double-shelled nanocapsules were synthesized by a solvothermal treatment and final heat treatment in air. The SEM image (Fig. 6.10a) of the V₂O₅-SnO₂ nanocapsules indicates that these nanocomposites can be produced in large scale with an average diameter of 550 nm without aggregation. The inset of Fig. 6.10a shows a schematic structure of one individual double-shelled nanocomposite capsule. The red spheres represent SnO₂ nanocrystals, and the green double shells represent the V₂O₅ matrix. The microstructure and components of these nanocapsules were further studied by means of TEM and SAED. Figure 6.10b shows a TEM BF image of double-shelled V₂O₅-SnO₂ nanocapsules consisting of nanocrystals. A typical double-shelled nanocapsule is shown in Fig. 6.10c, which clearly confirms that these hollow nanocapsules have two thin shells. The diameter of the inner hollow nanocapsules is about 430 nm, and the inner cavity is around 250 nm. The thickness of the inner and outer walls can be determined to be ∼90 nm through TEM BF image in Fig. 6.10d. To investigate the distribution of SnO₂ in the shell, HRTEM characterization was carried out. The micrographs in Fig. 6.10e and g are HRTEM images taken from the wall edge of the nanocapsules shown in Fig. 6.10d at different locations. Figure 6.10e shows a HRTEM image of a single nanocrystal that reveals the (310) lattice planes of V₂O₅. Figure 6.9f and g reveals the (110) and (101) lattice planes of SnO₂, respectively. These HRTEM images confirm that SnO₂ nanocrystals are homogeneously distributed in the V₂O₅ matrix (double shell). The polycrystalline nature of these nanocapsules was also confirmed by the SAED measurements (Fig. 6.10h). The formation mechanism of the double-shelled hollow nanocapsules is a combination of two types of Ostwald ripening processes (both inward and outward ripening cases).

Fig. 6.10

(a) SEM image of V₂O₅-SnO₂ double-shelled nanocapsules. The inset shows a schematic structure of a double-shelled nanocapsule. The red spheres represent SnO₂ nanocrystals, and the green double shells represent the V₂O₅ matrix. (b) Low-magnification TEM image. (c, d) High-magnification TEM images showing that the porous shells consist of a great deal of nanocrystals and have thicknesses of R1 ≈ R2 90 nm. (e, g) HRTEM images revealing lattice planes of the V₂O₅ matrix and SnO₂ nanocrystals. (h) SAED pattern taken from individual nanocapsules, which shows that these nanocapsules are polycrystalline [45] (Reprinted with permission from Ref. [45]. Copyright 2009 American Chemical Society)

In order to understand the growth mechanism of double-shelled V₂O₅-SnO₂ hollow nanocapsules, Liu et al. [45] investigated the morphology evolution of the intermediates involved in the formation process. Two intermediates obtained at 5 and 10 h are shown in Fig. 6.11. With a short reaction time (5 h, Fig. 6.11a), the crystallite aggregates gave a spherical morphology. When the reaction time was extended to 10 h, these solid spheres were converted into solid core-shell particles (Fig. 6.11b, c), and finally this solid core became hollow to form double-shelled hollow nanocapsules (Fig. 6.11d). They concluded that the formation mechanism of the double-shelled hollow nanocapsules is a combination of two types of Ostwald ripening processes (both inward and outward ripening cases). Ostwald ripening firstly took place at the surface of solid spheres, which differed from the previous simpler outward ripening process. Following this inward ripening process, the solid core of core-shell spheres ripened outward furthermore, and finally the double-shelled nanocapsules were achieved.

Fig. 6.11

TEM images show the formation process of V₂O₃-SnO₂ composite as double-shelled hollow nanocapsules. (a) 5 h, solid V₂O₃-SnO₂ composite nanospheres. (b, c) 10 h, core-shell intermediates. (d) 24 h, double-shelled V₂O₃-SnO₂ hollow nanocapsules [45] (Reprinted with permission from Ref. [45]. Copyright 2009 American Chemical Society)

The rattle-type nanoarchitectures, a special class of core-shell particles, have been extensively studied because of their unique structural properties and potential applications. These architectures possess spherical shells and solid cores having a variable space between them. Some rattle-type particles such as Au-polymer, SiO₂-Fe₂O₃ nanoball, and Cu-silica have been synthesized [157, 158]. Zhou et al. [159] fabricated rattle-type carbon-alumina core-shell spheres with large cavities and proposed a formation mechanism for them.

The time-dependent evolution of morphology was elucidated by TEM and it is shown in Fig. 6.12. The alumina-carbon composite microspheres were obtained via hydrothermal treatment before calcination. After calcination at 450 °C, the carbon transforms into carbon dioxide, meanwhile the loosely adsorbed Al³⁺ ions turn into dense Al₂O₃ network forming the shells of the rattle-type spheres [160]. A closer observation of the TEM images reveals that small cavities exist between the carbon cores and alumina shells resulting from the shrinkage during the calcination process. The formation mechanism of the rattle-type carbon-alumina particles was described as a two-step process. First, the carbohydrate used as a carbon precursor is subjected to dehydration, condensation, polymerization, and aromatization [161] and finally carbon spheres are formed. The surface of these carbon spheres is hydrophilic because it contains a considerable amount of reactive oxygen-containing group. Therefore, Al³⁺ ions are easily attached to the surface of the carbon spheres. Second, 2 h calcinations result in the partial removal of carbon cores, and simultaneous densification and cross-linking of the incorporated aluminum ions in the shells, which leads to the formation of rattle-type structures.

Fig. 6.12

Schematic illustration of the formation of rattle-type carbon-alumina particles; TEM images represent the samples obtained by calcination at 450 °C for 0, 0.5, and 2 h, respectively [159] (Reprinted with permission from Ref. [159]. Copyright 2012 American Chemical Society)

Most work in this area was focused on spherical shape, and the resulting hollow spheres are generally single shelled. However, Lou et al. [46] reported a simple synthesis of double-walled SnO₂ ellipsoidal hollow nanoparticles with movable α-Fe₂O₃ cores. This method is based on hydrothermal shell-by-shell deposition of polycrystalline SnO₂ on ellipsoidal α-Fe₂O₃/SiO₂ nanotemplates. Firstly, α-Fe₂O₃ spindles were coated with a SiO₂ layer to produce ellipsoidal α-Fe₂O₃/SiO₂ core-shell particles (Fig. 6.13a). Then, two polycrystalline SnO₂ layers were deposited on the surface of the α-Fe₂O₃/SiO₂ core-shell particles through hydrothermal method. From Fig. 6.13c, a black core and double shell can be clearly distinguished. After annealing at 550 °C, the sandwiched silica layer is dissolved in sodium hydroxide solution to produce double-walled SnO₂ nano-cocoons (Fig. 6.13d). As can be seen from the image, most nano-cocoons encapsulate only one α-Fe₂O₃ spindle which is usually not located in the center of the cocoon. It is therefore believed that the encapsulated α-Fe₂O₃ spindle is free to move within each cocoon at least when filled with liquid.

Fig. 6.13

TEM BF images of (a) α-Fe₂O₃/SiO₂ nanotemplates, (b) α-Fe₂O₃/SiO₂/SnO₂, (c) α-Fe₂O₃/SiO₂/(SnO₂)₂, (d) double-walled SnO₂ nano-cocoons with movable α-Fe₂O₃ spindles [46] (Reprinted with permission from Ref. [46]. Copyright 2007 Wiley-VCH)

5.2 Application of High-Angle Annular Dark-Field Imaging in Nanoparticles

The HAADF approach can detect the variation in chemical composition of the multicomponent sample with an atom-level resolution due to the enhanced contrast difference of various elements (Z-contrast imaging). The contrast of HAADF images is strongly dependent on the average atomic number of the scatterer encountered by the incident probe, not strongly affected by dynamical diffraction effects and defocus. Spatial resolution is limited by the size of the focused incident probe. So HAADF is suitable for characterization of core-shell nanomaterials. In the following we will discuss the application of HAADF images for characterization of core-shell nanomaterials, especially core-shell structured bimetallic nanoparticles [43, 162–170].

5.2.1 Au@Cu₂O Nanoparticles

Figure 6.14a and b shows typical HAADF images of the Au/Cu₂O nanocube heterostructures formed after heating copper grid in ambient environment. TEM observations show that nearly all the gold nanoparticles (>95 %) near the bars of copper grid have transformed into core-shell heterostructures. It can be clearly seen from Fig. 6.14a, b that the core is much brighter than the shell, which indicates that the core has a higher atomic number than that of the shell. Combined with energy-dispersive X-ray spectroscopy (EDS), the chemical compositions of the core and shell are determined to be Au and Cu₂O, respectively. In addition, there are two kinds of morphologies: one being a nearly perfect core-shell nanocube heterostructure and the other being formed through coalescence of two or more small particles (examples of the latter are indicated by white arrows in Fig. 6.14a). These two morphologies have nearly equal volume fractions. The edge dimensions of these heterostructures range from 15 to 45 nm, and the sizes of the cores range from 3.2 to 7.5 nm. TEM examinations of more than 200 core-shell heterostructures show that the edge length of a nanocube is proportional to the diameter of the particles at the core. The linear relationship between the core size and the edge length of the nanocubes is plotted in Fig. 6.14c. The linear relationship indicates that the gold core controls the growth of Cu₂O shell, acting as a template and catalyst. Figure 6.14d shows the SAED patterns taken from the pure Cu₂O nanocubes (the left half) and the Au-Cu₂O core-shell nanocube heterostructures (the right half). Due to the lattice parameter difference between Cu₂O (a = 4.269 Å) and gold (a = 4.09 Å), it can be seen that some rings (i.e., 111) in the right half are a little broadened.

Fig. 6.14

(a, b) Typical HAADF images of core-shell nanocube heterostructures, showing two types of morphologies. (c) The relationship between the core sizes and the edge lengths of the nanocubes. (d) SAED patterns taken from the pure Cu₂O nanocubes and the Au-Cu₂O nanocube heterostructures [162] (Reprinted with permission from Ref. [162]. Copyright 2008 Elsevier Ltd.)

5.2.2 Metal@Metal Nanoparticles

Wu et al. [169] used HAADF techniques to observe the morphology of Au@Ag core-shell nanoparticles. As the atomic number difference between Ag(47) and Au(79) is sufficient, the Z-contrast imaging should be capable of distinguishing Au and Ag within the Au@Ag nanoparticles [171, 172]. The enlarged HAADF image in Fig. 6.15 shows that the Au core is brighter than Ag shell, and the Au@Ag core-shell nanoparticles could have various shapes such as cube, triangle, decahedron, and nanorod. Through careful analysis of Fig. 6.15, they found that the cubic Ag shells can form on Au cores with different shapes such as octahedral (Fig. 6.15a), truncated octahedral (Fig. 6.15b), and cubic (Fig. 6.15c). Single-twinned bi-triangular or bi-hexagonal Au cores predominated by {111}-type facets can epitaxially evolve into the single-twinned inverted bi-triangular Ag shells predominated by {100}-type facets (Fig. 6.15d, e). Decahedral Ag shells (Fig. 6.15f) are epitaxially overgrown from the decahedral Au cores and Ag shell nanorods (Fig. 6.15g) with a five-twinned cross section from the Au nanorod cores with the five-twinned cross section.

Fig. 6.15

The HAADF-STEM images of the individual Au@Ag nanoparticles with various shapes and their evolution models from different shaped Au core [169] (Reprinted with permission from Ref. [169]. Copyright 2009 IOP Publishing Ltd.)

Serpell et al. [170] presented a new proof-of-concept method to synthesize core-shell nanoparticles in which ligand-based supramolecular forces are used to ensure surface segregation of the shell metal onto the preformed core before its reduction. They demonstrated the principle through the synthesis of Au@Pd, Pd@Au, Pt@Pd, and Pd@Pt nanoparticles using an anion coordination protocol. Conventional TEM instruments provide insufficient atomic number sensitivity for the determination of core-shell structure within the nanoparticles. Figure 6.16a, c is HRTEM images of Au@Pd and Pd@Au nanoparticles in which the lattice fringe images and the boundary between core and shell are not very clear. Therefore, aberration-corrected STEM with a HAADF detector was used to image the precise architectures of individual nanoparticles [173, 174]. The examination of Fig. 6.16b and d using a JEOL-2100F TEM/STEM with a probe correction and ∼0.1 nm point resolution clearly reveals core-shell morphology. The intensity is directly related to the square of the atomic number of the elements, making Au atoms appear brighter relative to Pd. The image of Fig. 6.16b strikingly shows the Au core surrounded by a Pd coating at atomic resolution, with a regular crystalline structure. The Pd@Au nanoparticles appear to have a less regular crystalline structure than the Au@Pd nanoparticles, suggesting a modestly defined core-shell structure for this sample (Fig. 6.16d). This clearly illustrates that the structure within the nanoparticles can be modified by varying the reaction conditions.

Fig. 6.16

HRTEM and aberration-corrected HAADF-STEM images of Au@Pd nanoparticles (a, b) and Pd@Au nanoparticles (c, d) [170] (Reprinted with permission from Ref. [170]. Copyright 2011 Nature Publishing Group)

Using aberration-corrected scanning transmission electron microscopy (STEM) and electron energy-loss spectroscopy (EELS) line profiles with Ångstrom resolution, Lin et al. [175] studied the structural changes of individual nanoparticles. After electrochemical dealloying, all of the dealloyed Pt-Ni nanoparticles revealed Pt-rich shells surrounding Pt-Ni alloy cores, as shown in Fig. 6.17. The EELS data evidence a distinct difference in the Ni distribution across the alloy cores. Figure 6.17a–d shows a typical high-resolution HAADF-STEM image and line scan EELS spectra across several nanoparticles of the D-PtNi catalyst. The Ni composition shows a monotonic decrease from the particle center to the particle surface. Contrary to that, the D-PtNi₃ catalyst (Fig. 6.17e–h) revealed an unusual Ni composition profile across the core, showing a previously undiscovered spherical enrichment of Ni at the near surface. Figure 6.17f, g presents two perpendicular EELS line scan profiles across the nanoparticle shown in Fig. 6.17e. Two off-center maxima of Ni intensity are clearly discernible in both directions, which coincide with the inflection points in the Pt intensity profiles. In other words, a Ni-enriched inner shell is formed near the surface and sandwiched between a Ni-poorer core and a Pt outer shell. This Ni-enriched inner shell is found to be universal in the D-PtNi₃ catalyst (Fig. 6.17h). In D-PtNi₅ catalyst (Fig. 6.17i–l), it is interesting to note that the Ni-enriched inner shell is located closer to the surface compared with D-PtNi₃. Figure 6.17j, k again shows two perpendicular EELS line scans from the nanoparticle shown in Fig. 6.17i, which display the Ni compositional maxima located closer to the surface compared with D-PtNi₃ catalyst.

Fig. 6.17

High-resolution HAADF-STEM images and EELS compositional line profiles of individual D-PtNi (a–d), D-PtNi₃ (e–h), and D-PtNi₅ (i–l) nanoparticles. The directions of the line profiles were plotted as dashed arrows in the corresponding nanoparticles. The EELS profiles are normalized with the elemental scattering factors, and hence the intensity ratios represent thickness-projected compositions [175] (Reprinted with permission from Ref. [175]. Copyright 2012 American Chemical Society)

Kim et al. [176] also used HAADF-STEM and EELS line profiles to study the structural changes of individual core-shell nanoparticles. Figure 6.18 shows the HAADF-STEM images and the cross-sectional compositional line profiles measured at the central (left panels) and edge parts (right panels) of Pt_0.97Ag_0.03, Pt_0.95Ag_0.05, Pt_0.90Ag_0.10, Pt_0.70Ag_0.30, and Pt_0.0Ag_1.0 nanoparticles. It can be seen that all the Pt@Ag particles were composed of Pt and Ag atoms. A closer examination reveals that the line profiles for Pt_0.97Ag_0.03, Pt_0.95Ag_0.05, Pt_0.90Ag_0.10, and Pt_0.70Ag_0.30 particles (Fig. 6.18a–d) are of the Pt core-Pt/Ag alloy shell types, while that of Pt_0.00Ag_1.00 particles are of a Pt core-Ag shell type (Fig. 6.18e). It is obvious that not all, but a substantial amount of seed Pt, has taken part in the formation of Pt/Ag alloys, except in the case of Pt_0.00Ag_1.00 nanoparticles.

Fig. 6.18

HAADF-STEM images and cross-sectional compositional line profiles measured at the central (left panels) and edge parts (right panels) of (a) Pt_0.97Ag_0.03, (b) Pt_0.95Ag_0.05, (c) Pt_0.90Ag_0.10, (d) Pt_0.70Ag_0.30, and (e) Pt_0.0Ag_1.0 nanoparticles, each of which was grown on Pt seed particles [176] (Reprinted with permission from Ref. [176]. Copyright 2011 American Chemical Society)

5.3 Application of High-Resolution Transmission Electron Microscopy Imaging in Nanoparticles

HRTEM could provide a lot of useful information about the sample, such as crystallographic orientation, defects, and interfaces at an atomic scale. For core-shell structured nanomaterials, HRTEM images can illustrate the interface between core and shell. In the following, we will discuss the application of HRTEM image for characterization of core-shell nanomaterials [162, 177–184].

5.3.1 Au@Cu₂O Nanoparticles

In order to investigate the microstructure of Au@Cu₂O core-shell nanocubes, especially the interfaces between the gold nanoparticles and Cu₂O, Wang et al. [162] carried out systematic characterization of these nanoparticles by HRTEM. Most of the core-shell nanocube heterostructures demonstrate specific orientations of Au particle in Cu₂O cube. Figure 6.19 shows two major orientation relationships normally observed in these heterostructures: (a) [001]Au//[001]Cu₂O, {100}Au//{100}Cu₂O and (b) [011]Au//[011]Cu₂O, {111}Au//{111}Cu₂O. Contrary to the HAADF images (Fig. 6. 14 in Sect. 5.2.1), the gold nanoparticle core has a dark contrast, while the Cu₂O shell has a light contrast. Figure 6.19a shows an example of a heterostructure in which gold nanoparticle and Cu₂O have an orientation relationship of [001]Au//[001]Cu₂O, {100}Au//{100}Cu₂O. However, the lattice is a little distorted around the interface due to the lattice misfit of about 4 % between Au and Cu₂O. Figure 6.19b shows an example of a heterostructure in which gold nanoparticle and Cu₂O have an orientation relationship of [011]Au//[011]Cu₂O, {111}Au//{111}Cu₂O. The {111} lattice misfit between Au and Cu₂O is about 2.3 %, which resulted in the lattice distortion around the interface. As mentioned earlier, some heterostructures are formed through coalescence of two or more small particles. Two examples are shown in Fig. 6.19c and d. Figure 6.19c shows an HRTEM image of a nanocube heterostructure formed through coalescence of two Cu₂O particles with a triangular shape. The boundary between two small particles is still evident. The final shape of this heterostructure is close to cubic. Figure 6.19d shows an HRTEM image of two coalesced particles with a trapezoid shape. The two particles coalesce through twinning, and the twinning configuration is indicated in Fig. 6.19d. Due to irregular shapes of the small particles, the final shape of the heterostructures is no longer cubic. The twinning configuration is thought to reduce the boundary energy and make the final structure more stable. It is believed that the presence of oxygen in the environment is crucial for the formation of Au-Cu₂O core-shell nanocube heterostructures since it can oxidize the copper into cuprous oxide at 300 °C.

Fig. 6.19

(a) [001] zone-axis HRTEM image of a nearly perfect Au-Cu₂O nanocube heterostructure. (b) [011] zone-axis HRTEM image of a nearly perfect Au-Cu₂O nanocube heterostructure. (c) and (d) HRTEM images of coalesced Au-Cu₂O nanocube heterostructures [162] (Reprinted with permission from Ref. [162]. Copyright 2008 Elsevier Ltd.)

5.3.2 PbTe/CdTe Core-Shell Nanoparticles

Core-shell quantum dots (QDs) are heterogeneous nanoparticles composed of an inorganic core enveloped by at least one inorganic shell of a second material. PbS/CdSe core-shell QDs can be prepared by cation exchange method [185]. Lambert et al. [183] demonstrated that the combination of the PbTe rock salt structure and the CdTe zinc blende structure allows for the direct observation of the core and the shell with HRTEM. This enables a direct visualization of the crystallographic properties of the PbTe/CdTe QDs and an evaluation of the cation exchange reaction. They observed a seamless match between the PbTe and CdTe crystal lattices and found that the formation of {111} terminated PbTe cores was favored. This intrinsic anisotropy of the exchange process leads to a strong increase in the heterogeneity of the cores formed, not only in terms of core size and shell thickness but also at the level of shape and position of the core.

Both PbTe and CdTe have a cubic structure [186] with almost no lattice mismatch (Fig. 6.20a, b). Since the appearance of the crystal lattice in HRTEM not only depends on the crystal orientation but also on the defocus, the same crystal plane may yield a different lattice image for both materials. Therefore, it is necessary to carry out a systematic HRTEM simulation. The simulated HRTEM images are shown in Fig. 6.20. When viewed along the <100> direction, PbTe and CdTe may yield two types of lattice images with a square symmetry (Fig. 6.20). The first has a lattice constant of 3.23 Å for PbTe (Fig. 6.20c) or 3.24 Å for CdTe (Fig. 6.20d); the second is tilted by 45° and has a lattice constant of 2.29 Å for both PbTe (Fig. 6.20e) and CdTe (Fig. 6.20f).

Fig. 6.20

Viewed along the <100> direction (a, b) structural models of PbTe and CdTe. (c, e) HRTEM simulation images of PbTe under different defocus. (d, f) HRTEM simulation images of CdTe under different defocus [183] (Reprinted with permission from Ref. [183]. Copyright 2009 American Chemical Society)

Figure 6.21 shows that resolved core-shell lattice images can be obtained in HRTEM for specific orientations of the particles. Viewed along the <100> direction, Fig. 6.21a represents an image where a core exhibits a 3.21 Å square pattern and a shell demonstrates a 2.34 Å square pattern. Both patterns are tilted by 45° and match seamlessly. By comparison with the simulated images, the core is determined to be PbTe and the shell is CdTe.

Fig. 6.21

HRTEM images of PbTe/CdTe core-shell particles in the (a) <100>, (b) <211>, and (c) <111> direction. A core-shell structure is observed only in the <100> and <211> direction [183] (Reprinted with permission from Ref. [183]. Copyright 2009 American Chemical Society)

Along the <211> direction, PbTe and CdTe yield a rectangular lattice image with 2.29/1.94 Å and 3.89/2.29 Å unit cells, respectively. Figure 6.21b shows an image where both lattice images can be seen.

Along the <111> direction, the lattice images of both PbTe and CdTe show a hexagonal pattern with an almost identical lattice constant of 2.80 and 2.81 Å, respectively. Figure 6.21c gives an example of a PbTe/CdTe core-shell QD that exhibits this pattern. It appears as a simple, uniform particle with a continuously resolved lattice and no indication of any core-shell structure. This result is typical for all particles viewed along the <111> direction. It shows that the <111> axes of both core and shell point in the same direction, with a coherent alignment of {111} planes.

5.3.3 Zn/ZnO Core-Shell Nanobelts

ZnO nanomaterials can be used for fabricating nanolasers [187], field-effect transistors [188], gas sensors [189], nanocantilevers [190], and nanoresonators [191]. Wang et al. [184] synthesized heterostructured metal-semiconductor Zn-ZnO core-shell nanobelts successfully by a solid-vapor decomposition process [192]. The microstructure of epitaxial Zn-ZnO nanobelt has been studied by TEM.

Figure 6.22a is a low-magnification TEM image of the nanobelt, displaying a Zn-ZnO core-shell structure. The SAED pattern (Fig. 6.22c) indicates two sets of single-crystal diffraction spots, which are indexed to be [0001] Zn and [0001] ZnO with an epitaxial orientation. The weak reflection spots were produced by double diffractions from the core and the shell. Structurally, both Zn and ZnO have hexagonal crystal structure [193] with lattice constants of a = 2.665 Å, c = 4.947 Å and a = 3.249 Å, c = 5.206 Å, respectively, and the mismatch between the two in $\left( {10\overline{1}0} \right)$ is about 21.9 %. Therefore, the interference between the Bragg reflections from the two crystals produces Moiré fringes in the image, which are apparent in the HRTEM image in Fig. 6.22b at the region where the Zn core and the ZnO shell overlaps. However, in the region where there is only a ZnO shell, the HRTEM image shows clear lattice structure. The boundary between the Zn core and the ZnO shell is fairly sharp.

Fig. 6.22

(a) Low-magnification TEM image of a Zn-ZnO core-shell heteronanobelt. (b) [0001] HRTEM image recorded near the right-hand edge of a nanobelt, showing the Zn-ZnO overlapping region and the ZnO wall. (c) The corresponding SAED pattern [184] (Reprinted with permission from Ref. [184]. Copyright 2009 American Chemical Society)

5.3.4 Pt@Pd Core-Shell Nanoparticles

Nguyen et al. [194] synthesized Pt@Pd core-shell nanoparticles and studied the structure of individual core-shell nanoparticles by HRTEM. Figure 6.23 shows the HRTEM images of Pt@Pd core-shell nanoparticles with the most characteristic polyhedral morphology and shape. The thin Pd shells grown over the Pt cores have led to form the core-shell configuration with the well-controlled size in the range of about 15–25 nm. The thickness of the coated shell was well controlled in the range of 1–3 nm. The Pt@Pd core-shell nanoparticles also show characteristic polyhedral morphology and shape, typically such as tetrahedral, octahedral, and cubes. Most of the Pt@Pd core-shell nanoparticles exhibit the low-index facets of {111}, {110}, and {100} planes.

Fig. 6.23

(a–e) HRTEM images of Pt@Pd core-shell nanoparticles with the polyhedral morphology [194] (Reprinted with permission from Ref. [194]. Copyright 2009 American Chemical Society)

5.4 Application of Elemental Mapping in Nanowires

Elemental mapping is based on inner-shell ionization of elements present in the sample giving rise to characteristic signals in well-defined energy-loss regions [195]. It is a valuable tool for core-shell materials characterization. The applications of elemental mapping have spanned the range of research from biology to polymer materials [44, 196–199]. Elemental mapping, formed by imaging with electrons that have lost energies corresponding to inner-shell ionization edges for particular elements, can give the elemental distribution images in a relatively large area with high spatial resolution. In the following, we will discuss the application of elemental mapping for characterization of core-shell nanomaterials.

5.4.1 Boron@Boron Oxides Nanowires

It has been found that crystalline silicon [200] and germanium NWs [201] were sheathed with an amorphous oxide coating. Cao et al. [202] reported the successful synthesis of well-aligned straight amorphous boron NWs. From the HRTEM observation, it is difficult to detect whether there is an oxide coating layer of BO_x because the phase contrast of amorphous boron and amorphous BO_x coating cannot be easily distinguished. Therefore, it would be helpful to use the EFTEM. Wang et al. [197, 198] carried out a comprehensive characterization of boron NWs through EFTEM.

The aligned boron NWs were prepared by a radio-frequency magnetron sputtering method. A Philips CM200-FEG TEM equipped with a Gatan Imaging Filter (model 678) was used for elemental mapping and EELS examinations. A three-window method was used to study elemental distribution of boron and oxygen in order to clarify the existence of a boron oxide outer layer coating. The ionization edges selected for elemental mapping are listed as follows: B-K edge (188.5 eV) and O-K edge (532 eV). The exposure time for the elemental mapping of B and O was 10 and 20 s, and the width of the energy windows ΔE was set to be 10 and 20 eV, respectively. The EELS spectrum was acquired in the image mode with a half collection angle of ∼13 mrad.

The EELS spectrum of single boron NW is shown in Fig. 6.24, revealing the characteristic boron K-shell ionization edges (∼188 eV). Careful examination of the EELS spectrum shows that a small peak is located at 532 eV, which corresponds to the K-shell excitation of oxygen. The magnified oxygen peak is shown in the inset of Fig. 6.24. This demonstrates that a small amount of oxygen exists in the boron NWs.

Fig. 6.24

EELS spectrum acquired from single boron nanowire (NW) showing strong boron K-shell peaks and a small peak at 532 eV corresponding to the oxygen K-edge [198] (Reprinted with permission from Ref. [198]. Copyright 2008 Elsevier Ltd.)

In order to further investigate the spatial distribution of boron and oxygen in the boron NWs, elemental mappings of boron and oxygen were achieved for the straight boron NWs (Fig. 6.25). It can be clearly seen that the boron is mainly distributed in the core (Fig. 6.25b), while oxygen is mainly located in the outer layer (Fig. 6.25c) of the boron NWs. The thickness of the outer oxidized layer is about 1–2 nm. The importance of oxide or oxygen for the both nucleation and growth of the boron NWs has been confirmed in an experiment (under the same conditions) using two targets (one is a mixture of B and B₂O₃, the other B only). The experiment showed that the quantity of NW stopped increasing after they switched the magnetron sputtering from the mixed target to the pure B target. Moreover, the diameter of the straight boron NWs (about 100 nm) using a pure B target is larger than that of the straight NWs (40–50 nm) using the mixed target.

Fig. 6.25

(a) TEM image of single straight boron NW. (b) Elemental mapping of boron in the straight boron NW. (c) Elemental mapping of oxygen in the straight boron NW [198] (Reprinted with permission from Ref. [198]. Copyright 2008 Elsevier Ltd.)

In order to preclude the possibility of the oxidization layer after growth, some precautions were adopted. Before the boron NWs were put into TEM, they were kept in an argon atmosphere, not exposed to the air. The boron NWs were only exposed to air for several minutes during the TEM sample preparation. In addition, they carried out the oxygen mapping of the straight boron NWs (about 100 nm in diameter) produced by using a pure B target (without the oxygen source). The experimental results showed that there is no outer oxidized layer. So it is assured that the observed outer oxidized layer is not produced after growth. The oxide (B₂O₃) or oxygen is crucial for the nucleation and growth of the boron NWs.

5.4.2 Core-Multishell Semiconductor Nanowires

n-GaN/In_xGa_1-xN/GaN/p-AlGaN/p-GaN core-multishell NWs were synthesized by metal-organic chemical vapor deposition (MOCVD) [203], using a strategy involving axial elongation by nanocluster-catalyzed growth followed by controlled shell deposition onto the NW core [203]. To characterize the chemical composition and thickness of individual shells in the core-multishell heterostructures, Qian et al. [189] exploited cross-sectional imaging with the electron beam parallel (vs. perpendicular) to the NW axis since this allows for direct visualization of the spatial distribution of elements.

A cross-sectional BF TEM image (Fig. 6.26a) of a GaN/In_xGa_1-xN/GaN/AlGaN core-multishell NW taken along the $\left[ {11\overline{2}0} \right]$ zone axis shows that the core-multishell wire has a triangular cross section with smooth facets. No dislocations or boundaries were observed in the NW, indicating an epitaxial deposition of the shells on the cores. Electron diffraction data (inset, Fig. 6.26a) further demonstrates that the core-multishell NW is single crystalline and that the three lateral facets can be indexed as (0001) and two $\left\{ {1\overline{1}0\overline{1}} \right\}$ crystallographic planes. This result is consistent with the previous report on core/shell/shell NWs [203].

Fig. 6.26

TEM BF cross-section image of a GaN/In_xGa_1-x/GaN/AlGaN core-multishell NW. Scale bar is 100 nm. Inset: $\left[ {11\overline{2}0} \right]$ zone-axis SAED pattern. (b) STEM image recorded at a NW corner with (0001) and $\left( {1\overline{1}0\overline{1}} \right)$ facets. Scale bar is 100 nm. (c–e) Elemental mapping of the same NW region, indicating spatial distribution of Ga (blue), In (red), and Al (green), respectively [44] (Reprinted with permission from Ref. [44]. Copyright 2008 American Chemical Society)

Additional analysis using STEM (Fig. 6.26b) revealed contrast indicative of variations in the radial chemical composition as expected for the core-multishell structure. STEM EDS mapping of the same NW region (Fig. 6.26c and e) confirmed the STEM results and defined clearly the spatial distributions of Ga, In, and Al in an individual shell that are consistent with targeted core-multishell structure. Interestingly, the thickness of InGaN shell was larger on the $\left( {1\overline{1}0\overline{1}} \right)$ versus (0001) facet, indicating that shell deposition rate depends on the specific crystal planes. This can be understood in terms of different surface energies and polarities on nonequivalent facets [204] and suggests that these NWs could also serve as a model system to study growth kinetics. In addition, localized indium (In)-rich clusters on a scale of 10–50 nm were observed in the thicker InGaN layer grown on the $\left( {1\overline{1}0\overline{1}} \right)$ facet. However, In segregation was not observed in the thinner layer grown on the (0001) facet. Similar In inhomogeneity has been reported in InGaN-based planar structures and is dependent on several factors, including InGaN layer thickness [205].

6 Conclusions and Future Perspective

This chapter concentrates on the introduction of various TEM techniques for characterization of core-shell nanomaterials. The objective was to review the versatility of TEM and the complimentary applications of the techniques. We take the TEM techniques as the clue to discuss the application of TEM for characterization of core-shell nanomaterials.

For TEM BF images, mass-thickness and diffraction contrast contribute to image formation: thick and crystalline areas appear with dark contrast. Since core-shell nanomaterials usually consist of different components, TEM BF image is applicable to most of them. We focus on three kinds of core-shell nanomaterials: silica coating core-shell nanomaterials, bimetallic core-shell nanomaterials, and hollow core-shell nanostructure. HAADF image is strongly dependent on the atomic number of the scatterer encountered by the incident probe, not strongly affected by dynamical diffraction effects or defocus conditions. So it has advantages in characterizing core-shell nanomaterials.

HRETM imaging process is very complicated and sensitive to the sample thickness and defocus conditions. HRTEM requires very thin TEM specimens free of preparation artifacts. Additionally, correct interpretation of HRTEM images requires systematic image simulations. Because of these limitations, only a small part of the core-shell nanomaterials can be characterized by HRTEM. But we can get lots of important information of core-shell nanomaterials through HRTEM image, such as orientation relationship and interface between the core and shell. So HRTEM is an effective characterization method, which cannot be ignored.

Elemental mapping combined with EELS can investigate the spatial distribution of elements in the nanomaterials. Especially when the core and shell has a same structure and similar atomic numbers, the three methods mentioned above cannot distinguish.

BF image, HADDF image, HRTEM image, and elemental mapping are the most popular and effective methods for the characterization of core-shell nanomaterials. In fact, other TEM techniques can also be used to investigate core-shell nanomaterials, such as TEM dark-field image and SAED. With the development of TEM techniques, the microstructure of core-shell nanomaterials can be studied in more detail.