Plasmonic Nano-Structures of d10 Metals - the Future of Photovoltaics

Plasmonic nano-structures of d10 metals are considered to be the future of photovoltaics and photo-catalysis under solar irradiation. This article examines the effect of continuous plasmon excitation on the density of states of gold nanoparticles.

Visible Light Absorption of TiO2

Honda and Fujishima used TiO2 to photo-assist the electrochemical splitting of water. Inoue et al. showed that CO2 could also be photo-reduced to a number of hydrocarbons, with the help of powder semiconductors, such as TiO2. These discoveries were hindered by the large band gap (>3 eV) of the semiconductors. UV-A (300–400 nm) irradiation is needed for inducing charge separation, representing only a small fraction of the solar spectrum. However, TiO2 is still the best performing photo-catalyst. Researchers have made many efforts to enhance visible light absorption through direct manipulation of the TiO2 band gap by doping with C, N, and S.

The visible light absorption deficiency of TiO2 can be overcome by using sensitizers as they can harvest solar light and inject hot electrons into the TiO2 conduction band (CB). They can have discrete absorption levels (narrowband), low photo-stability, and small optical cross-sections.

Metallic nanoparticles (NPs) are intriguing sensitizers due to their localized surface plasmons (LSPs) with large optical cross-sections. Gold group metals have plasmonic resonances that can be tuned by modifying their size, shape, and composition. It has been recently shown that excitation of Au and Ag LSP nanostructures enhances solar cell charge transfer from the sensitizer to the semiconductor, increases the photocurrents under solar irradiation, and improves photo-initiated catalytic oxidations.

Furube et al. proposed that the excited plasmon band overlaps with an interband transition in gold, resulting in the excitation of electrons in the filled d-band to electronic states above the Fermi level. The collective excitation of electrons offers adequate energy to some electrons to help them overcome the Au–TiO2 Schottky barrier. This article intends to show that LSP excitation modifies the gold d-band occupancy and that some of the produced hot electrons have sufficient energy to overcome the Schottky barrier and be injected into the TiO2 CB, thereby validating the proposed mechanism.

Experimental Analysis

High-resolution X-ray absorption spectroscopy (HR-XAS) was used at the Au LIII-edge to analyze the production of hot electrons (electron-hole pairs) as a result of LSP excitation. As shown in Figure 1, the experiment was carried out on Au NPs supported on a passivized-Si substrate with the help of a von Hamos spectrometer at grazing incidence.

HR-XAS experimental procedure used to determine the changes in the Au LIII-edge induced by continuous wave laser excitation of LSP at 532 nm, with 100 mW power.

Figure 1. HR-XAS experimental procedure used to determine the changes in the Au LIII-edge induced by continuous wave laser excitation of LSP at 532 nm, with 100 mW power.

Figure 2A and 2B respectively illustrate the unexcited Au LIII-edge spectrum (open circles black trace) and the spectral difference following the laser excitation (Laser ON–Laser OFF (open circles black trace)). The resonance threshold of the Au NP spectrum falls at 11,923 eV (whiteline) related to the 2p3/2 → 5d dipole transition, representing the unoccupied d density of states above the Fermi level. The s–d hybridization causes the unoccupied states above the Fermi level in Au (5d106s1).

FDMNES calculations that show the absorption cross-sections of photons surrounding the ionization edge were performed to support the hypotheses. Figure 2A illustrates the excited spectra due to the excitation from the ground state Au 5d106s1 (black trace) electron to the unoccupied 7p orbital upon exciting the 6s1 ([Au 5d106s07p1] red trace) or 5d10 ([Au 5d96s17p1] dashed blue trace) electrons. Figure 2B illustrates the calculated spectral differences between the ground and excited states assuming 2% excitation (red trace [Au 5d106s17p0]/[Au 5d106s07p1]; dashed blue trace [Au 5d106s17p0]/[Au 5d96s07p1]). It is anticipated that the excitation of 6s1 and 5d10 electrons will be similar due to the 6s–5d hybridization. As observed experimentally, the excited spectrum demonstrates a shift in the ionization threshold and an increase in the whiteline intensity.

HR-XAS experiments depicting changes in the Au LIII-edge induced by continuous wave laser excitation of LSP at 532 nm, with 100 mW power. (A) HR-XAS ground state spectrum of Au NPs (open circles black trace), FDMNES calculated spectra of the ground state (black trace Au 5d106s1), and excited states (red trace 5d106s07p1 and blue trace 5d96s17p1); (B) difference spectra between the excited and ground state: experimental (open circles black trace), and calculated assuming 2% excitation Au 5d106s17p0/Au 5d106s07p1 (red trace); Au 5d106s17p0/Au 5d96s07p1 (dashed blue trace).

Figure 2. HR-XAS experiments depicting changes in the Au LIII-edge induced by continuous wave laser excitation of LSP at 532 nm, with 100 mW power. (A) HR-XAS ground state spectrum of Au NPs (open circles black trace), FDMNES calculated spectra of the ground state (black trace Au 5d106s1), and excited states (red trace 5d106s07p1 and blue trace 5d96s17p1); (B) difference spectra between the excited and ground state: experimental (open circles black trace), and calculated assuming 2% excitation Au 5d106s17p0/Au 5d106s07p1 (red trace); Au 5d106s17p0/Au 5d96s07p1 (dashed blue trace).

Plasmon Excitation

Plasmon excitation is a collective phenomenon that involves the distribution of the excess energy among various atoms over an extended period; yet, it is likely that a single electron carries the energy at a given instant. Hence, this electron can possess the adequate energy to be excited to a higher level (hot electron formation) and form a hole in the valence states (d-band).

As electron thermalization is the main decay channel of the excited state, spectral variations due to thermal contributions, which were estimated by expanding the Au lattice parameters, are expected. The temperatures of the Au NP surface excited at 532 nm can be more than 500 °C for sub-microsecond times.

Transient broadband mid-IR (infrared) spectroscopy was carried out to confirm that the photo-generated hot electrons have adequate energy to overcome the Schottky barrier and be injected into the TiO2 CB. Trapped and free electrons in a semiconductor CB result in the occurrence of a unique broad mid-IR absorption band. The integrated mid-IR transient signal of Au–TiO2 with an Au average particle size of 30–40 nm supported on TiO2 is illustrated in Figure 3.

Transient mid-IR measurements of TiO2 coated with a light absorber irradiated at 532 nm with 100 ps pump pulses and 33 mW energy. The spectral changes were probed with 100 ps broad mid-IR pulses at a spectral resolution of 32 cm−1. Time dependence of the integrated mid-IR signal of Au–TiO2 (Ο) and N719–TiO2 (Δ).

Figure 3. Transient mid-IR measurements of TiO2 coated with a light absorber irradiated at 532 nm with 100 ps pump pulses and 33 mW energy. The spectral changes were probed with 100 ps broad mid-IR pulses at a spectral resolution of 32 cm−1. Time dependence of the integrated mid-IR signal of Au–TiO2 (Ο) and N719–TiO2 (Δ).

Latest studies indicate that electron injection from graphene quantum dots happens in <15 fs and from the N719-dye system in <10 fs. The detection of electrons in the TiO2 CB indicates that some of the generated hot electrons have adequate energy to overcome the Schottky barrier and be injected into the TiO2 CB.

Due to the incomplete recovery of the IR signal during the measurement time (80 ns), at least one-third of the component must be involved, which is possibly associated with the electron-hole recombination of electrons that jumped between TiO2 particles. Full relaxation takes place in the millisecond and beyond time scales. The differences observed for the faster time decay are considerable and suggest a different decay mechanism.

The relative electron injection yield shows that compared to N719, Au NPs injected approximately 20%–25% of the electrons into TiO2. Efficiencies of around 11.1% have been reported for the N719 dye-system. The Au NP injection yield is quite promising since it can be enhanced by, for example, optimizing sensitizer–semiconductor contact points and increasing metal loading. The low Au loading was chosen to avoid inter-particle communication.

Conclusions

In summary, the investigation of the generation of electron-hole pairs (hot electrons) showed that the electrons have adequate energy to be injected into TiO2. The intrinsic characteristics of gold group plasmonic nanostructures render them highly suitable for harvesting visible light and can be further tapped by combining them with photo-catalysts such as TiO2. Moreover, it is possible to use these materials as composite materials (for instance, with the LSP NPs at the core and TiO2 at the surface) for a number of sunlight-driven photo-reactions. This approach allows photochemical reaction on metal-doped and pristine TiO2 under visible irradiation without compromising the surface area availability, which is a huge problem with the dye-systems.

This information has been sourced, reviewed and adapted from materials provided by Peafowl Solar Power.

For more information on this source, please visit Peafowl Solar Power.

Tell Us What You Think

Do you have a review, update or anything you would like to add to this article?

Leave your feedback
Submit