Various surface science studies have been performed to investigate the interactions between the different surface facets of metals or semiconductors, and the reactant molecules. Platinum, being one of the most versatile metal catalysts, has an fcc crystal structure and is one of the most studied surfaces. On the flat surface, such as Pt{100} and Pt{111}, each platinum atom is surrounded by four and six adjacent neighbors, respectively. These two low‐index facets have highest atomic density. High‐index facets, such as Pt{530} and Pt{730}, contain many terrace structure and steps with low‐coordinated atoms [34]. The atomic arrangement of Pt{730} facet is periodically overlapped with one Pt{310} subfacet and two Pt{210} subfacets, leading to a multiple‐height terrace structure. Furthermore, the faceted metal catalysts have more atoms located at the edges and corners, where they have much lower atomic density and coordination, which can have a profound impact on surface adsorption.
Titanium dioxide (TiO2) has been the most intensively investigated binary transition metal oxide as a photocatalyst, where many theoretical calculations have been conducted to investigate the interactions between water molecules and the TiO2 surface [35]. Because the water molecule is one of the most important adsorbates present in many photocatalytic reactions, such as water splitting, photodegradation of organic pollutants, and CO2 reduction, many experimental and theoretical studies have focused on this aspect [36–40]. Anatase {101}, {001}, and {100} surfaces are three typical low‐index surfaces of anatase TiO2. Anatase {101} contains 50 % sixfold coordinated (Ti6c) Ti atoms (saturated) and 50 % fivefold coordinated (Ti5c) Ti atoms (unsaturated), whereas anatase {001} and {100} surface contains 100% Ti5c atoms. Most theoretical and experimental results suggest molecular adsorption of water on the defect‐free {101} surface [41] but dissociative adsorption of water molecules on the {001} surface at low coverages (shown in Figure 2.3) [36]. The dissociative adsorption of water is also favorable on the anatase {100} surface [12].
Figure 2.3 Side view of anatase TiO2 {101} and {001} facets. Top view for adsorbed water molecules on anatase {101} surface and side view of adsorbed water molecules on anatase {001} surface.
Source: Vittadini et al. 1998 [36]. Reproduced with permission of American Physical Society.
2.3.2 Surface Electronic Structure
Anisotropic electrical properties of surfaces are logically attributed to the anisotropy of the crystal lattice, which determines the different atomic arrangements and configurations depending on exposed facets. Work function is the minimum energy for valence electrons in the solid to overcome in order to exit into the vacuum [42], which can be defined as
where for elemental metals, EF is the potential energy of the electrons at the top of the valence band (VB) called the Fermi level, −e is the charge of an electron, and ϕ is the electrostatic potential in vacuum. For elemental metals, work function highly depends on the crystallographic orientation of the surface, as atomic density and electron charge density vary across different facets. The difference of work function between two surfaces of the same metal in different orientation can reach up to 1 eV [43]. For example, work functions of tungsten W{001} facet, W{112} facet, and W{111} facet are 4.56 eV, 4.69 eV, and 4.39 eV, respectively [44]. Work function of polycrystalline Ag is 4.26 eV, but work functions of Ag{100} facet, Ag{110} facet, and Ag{111} facet are 4.64 eV, 4.52 eV, and 4.74 eV, respectively [45]. Furthermore, surface roughness and particle size also have a profound impact on work function.
The electronic structure of semiconductors is different from that of metals because of a bandgap between the conduction band (CB) and the VB. According to the band theory, when a large number of identical atoms assemble to form a solid, the atomic orbitals with discrete energy levels will overlap. Each atomic orbital will split into discrete molecular orbitals with different energies, due to the Pauli exclusion principles stating that it is impossible for two electrons in the solid to have the same values of the four quantum numbers. For example, the VB of TiO2 is composed of O 2p orbitals, while the CB is composed of Ti 3d orbitals [46]. In the bulk of TiO2 crystals, no matter in anatase or rutile, there are numerous TiO6 octahedron units connected to their neighbors by sharing corners and edges in different ways. But at the surface, this periodic arrangement terminates, leading to variation in the coordination of Ti and O atoms. It is reasonable to deduce that the band structure at the surface is more or less different from the band structure in the bulk. A blueshift of light absorption edge was found when comparing the nanosized anatase TiO2 crystals with 82% {101} and 18% {001} facets with the micrometer‐sized anatase TiO2 crystals with 28% {101} and 72% {001} facets [47]. The 9 nm blueshift of absorption edge means a larger bandgap, which is attributed to the different dominant facets exposed. The dependencies of bandgap and exposed surface were also found in other materials [48–50].
2.3.3 Surface Electric Field
For a metal, the surface electric field is oscillating when the light strikes the surface. Light, an electromagnetic wave, oscillates the electric field in a plane perpendicular magnetic field. The electric field's oscillatory patterns would cause a rippling wave pattern in the distribution of electrons, where the resonant oscillation of conduction electrons is called surface plasmon resonance (SPR). The SPR only exists in metals or other electrically conductive materials containing conduction electrons. When the size of the metal crystals shrinks to the nanoscale, which is smaller than the wavelength of the incident light, the surface plasmon is confined to a very small surface rather than the bulk material, known as localized surface plasmon resonance (LSPR). The LSPR frequency affects the light absorption and scattering of metal nanoparticles. How the LSPR frequency is affected by facets (shapes) of a nanoparticle is explained in a later section, but as a consequence, the color of metal nanoparticles will be changed, and it is sensitive to the shape of nanoparticles (with different facets exposed). Based on the Mie theory, it is possible to tune the LSPR spectra of Ag nanocrystals of different shapes, as shown in Figure 2.4 [51].
Figure 2.4 Calculated UV–visible extinction (black), absorption (red), and scattering spectra (blue) of Ag nanocrystals, illustrating the effect of shape on its spectral characteristics, including isotropic sphere (a), anisotropic cubes (b), tetrahedra (c), and octahedra (d), triangular plate (e) and circular disc (f).
Source: Wiley et al. 2006 [51]. Reproduced with permission of American Chemical Society.
(See online version for color figure).
For