Heterogeneous Catalysts. Группа авторов

a faceted semiconductor, the surface electric field is a disparate situation. Besides the variation of the bandgap, the band edge position shifts as a function of different facets due to surface band bending. The different band edge positions provide varying redox potentials of the photogenerated electrons and holes, resulting in spatial separation of charge carriers and the built‐in electric field. In addition, selectively depositing a noble metal as cocatalysts on the surface facets can further enhance the strength of the built‐in electric field. When a single BiVO₄ crystal enclosed by {010} and {011} facets was characterized by spatially resolved surface photovoltage spectroscopy (SRSPS), {011} facets exhibited a much higher signal intensity of surface photovoltage than {010} facet [52]. This phenomenon indicated a significant difference in surface band bending between BiVO₄{011} and {010} facets. As a consequence, the different band bending will lead to the variation in the spatial distribution of the charge carriers and build an electric field between different facets. By changing the area ratio of (011)/(010) facets of BiVO₄ crystal, the surface built‐in electric field varied as well. Such an intrinsic difference in the surface photovoltage between different facets can be further enhanced by selectively depositing cocatalysts, such as MnO_x and Pt deposited on faceted BiVO₄ crystal, as shown in Figure 2.5 [53].

Figure 2.5 (a) Scanning electron microscopy (SEM) image of a BiVO₄ single crystal with Pt photodeposited on {010} facet and MnO_x photodeposited on {011} facet. (b) Spatial distribution of the surface photovoltage signals. Pink and green colors correspond to holes and electrons separated toward the external surface, respectively. Schematic band diagrams across the border between the {011} and {010} facets of (c) a bare single BiVO₄ photocatalyst particle and of (d) a single BiVO₄ photocatalyst particle with MnO_x cocatalyst selectively deposited at {011} facets (green line) and with MnO_x and Pt nanoparticles selectively deposited at {011} and {010} facets, respectively (dashed pink line).

      Source: Zhu et al. 2017 [53]. Reproduced with permission of American Chemical Society. (See online version for color figure.)

2.4 Effects of Facets Engineering

      Each facet in a single crystal has different properties. However, combining anisotropic surface properties could dramatically alter the properties of the crystal, especially when the particle size is reduced to the nanoscale and the ratio of surface atoms/bulk atoms is no longer negligible.

      2.4.1 Optical Properties

      As mentioned previously, the surface electric field of a metal oscillates when the light strikes the surface. The oscillating electric field causes a rippling wave pattern in the spatial distribution of electrons. According to Lenz's law, the wave created by the surface plasmon opposes the electromagnetic wave of the incident light. The oscillating electrons absorb the energy of light and reemit the energy as the reflected light, due to which metals have shiny and reflective surfaces. However, when the particle size becomes very small, the surface plasmon is confined to a very small surface (i.e. LSPR). When the electron cloud is excited at one of the resonance frequencies, light absorption will become stronger. This is how LSPR frequency affects the light absorption of metal nanoparticles. The plasmon frequency is determined by electron density, dielectric constant, and effective mass of an electron. The well‐defined facets of a crystal form different shapes with more symmetries compared with spherical particles [54]. The surface charges tend to accumulate at edges and corners, which further promote surface polarization, i.e. the charge separation between mobile electrons and immobile atoms. Surface polarization determines the frequency and intensity of LSPR as it provides the main restoring force for electron oscillation. Large surface polarization reduces the restoring force, resulting in a redshift of resonance peak, and multiple distinct symmetries may induce several light absorption peaks [55]. Therefore, the same metal nanoparticles with different size and shape may exhibit different colors, indicating diverse light absorption.

The light absorption of semiconductors is quite different from that in metals due to the electronic band structure in semiconductors. Between the VB and CB of semiconductors, no electron states exist in this energy range called the bandgap. In some semiconductors, the minimal energy state of the CB (conduction band minimum, CBM) and the maximal energy state of the VB (valence band maximum, VBM) are situated in the same crystal momentum in the Brillouin zone (direct gap); in other semiconductors, they are not (indirect gap). There is a slight difference in light absorption between these two types of bandgap structure. But it is not necessary to discuss in this chapter. In general, light absorption of a semiconductor is associated with its bandgap. Semiconductors only absorb photons with the energy equal to or greater than the bandgap. As a result of facets‐dependent anisotropic surface electronic properties that in turn influence the band positions, semiconductor crystals with different dominant facets show shifting in light absorption edges. In applications such as photoelectrochemical catalysis, faceted semiconductors can enhance the light harvesting of photoelectrodes [56–58].

      The combination of plasmonic metals and semiconductors with facets engineering has the great potential to adjust the light harvesting for photoelectrodes. For example, the LSPR absorption of the faceted plasmonic metal nanoparticles, such as Au and Pd, can be tuned by embedding them in Cu₂O to form core–shell heterostructure [59–61].

      2.4.2 Activity and Selectivity

      Numerous studies have shown that catalysts with facets engineering exhibit greater catalytic performance. This exploits one or more unique properties of the well‐defined crystal facets to tune the overall catalytic activity and/or selectivity.

      For metal catalysts, activity and selectivity are related to the surface atomic structure of the catalysts, where they can be tuned to enhance effective adsorption and/or promote favorable coordination of adsorbates. These processes are strongly influenced by the arrangement and coordination of surface atoms as well as by the corresponding surface density of states of the different facets. For example, there are two types of flat surfaces of Pt, namely, the Pt{111} facet (hexagonal surface) where each surface atom has six nearest neighbors, and the Pt{100} facet (square surface) where each surface atom has four nearest neighbors. The hexagonal surface is up to seven times more active than the square surface in the aromatization reaction of n‐heptane to toluene, but the square surface is seven times more active than the hexagonal surface in the alkane isomerization reaction of isobutane to n‐butane [62]. Also in electrocatalysis, the Pt{210} facet has high activity for electrooxidation of formic acid and electroreduction of CO₂ [63]; the Pt{410} facet exhibits high performance in NO decomposition [64]; the Pt{730} shows superior activity in electrooxidation of formic acid and methanol [34].

For the semiconductors in photo‐related catalytic processes, the performance of the faceted semiconductor crystals is affected by the synergetic combination of the intrinsic properties of the bulk and surfaces. A typical photocatalytic process includes the following steps: light absorption of the semiconductor catalyst, excited charges (electrons and holes) generation, excited electrons and holes recombination (bulk and surface), charge migration to the surface, charge trapping at the surface, and transfer to reactants. As such, an ideal semiconductor with good reactivity and selectivity