Dichalcogenide and Metal Oxide Semiconductor-Based Composite to Support Plasmonic Catalysis

Nanocomposites comprising plasmon active metal nanostructures and semiconductors have been used to control the charge states in the metal to support catalytic activity. In this context dichalcogenides when combined with metal oxides offer the potential to control charge states in plasmonic nanomaterials. Using a model plasmonic mediated oxidation reaction p-amino thiophenol ↔ p-nitrophenol, we show that through the introduction of transition metal dichalcogenide nanomaterial, reaction outcomes can be influenced, achieved through controlling the occurrence of the reaction intermediate dimercaptoazobenzene by opening new electron transfer routes in a semiconductor-plasmonic system. This study demonstrates the ability to control plasmonic reactions by carefully controlling the choice of semiconductors.


■ INTRODUCTION
Advanced materials and new techniques offer significant opportunities to advance control over surface catalytic reactions. 1−7 Such surface reactions can be potentially applied in a wide range of areas such as chemical production or for the removal of contaminants. Currently, used catalysts have wellknown limitations regarding reactivity, selectivity, and/or stability. 8−10 For example, for many current industrial catalytic processes, catalysts require high temperatures and/or pressures to operate efficiently. 11,12 The use of plasmon resonances in metal nanostructures to control the rate and selectivity of photocatalytic reactions offers significant potential. 13−17 The localized surface plasmon resonance (LSPR) excitation of metal nanostructures produces enhanced light-matter interaction, resulting in a strongly enhanced plasmonic electromagnetic field on the surface of a plasmon active nanostructure. 8−12 The oscillation of free electrons quickly decays via the excitation of energetic electron−hole pairs. 7−10 These initially excited electrons rapidly thermalize and equilibrate via electron−electron scattering, creating a "hot" Fermi−Dirac distribution. 2−6 Then the hot distribution cools via the coupling between the "hot electrons" and the phonons of the metal lattice. These generation "hot" electrons are seen as essential in catalysis through their interaction with the target chemical. However, hot electrons possess short subpicosecond lifetimes which presents a significant challenge for efficient surface plasmon-induced hot electron transfer catalytic reactions. 11−14 The development of transition metal dichalcogenide (TMDC) materials has opened up new opportunities in optoelectronics applications. Molybdenum disulfide (MoS 2 ) is emerging as a unique material with a range of optical and electrical properties. It demonstrates a layer-dependent band gap and excitons in both the visible and infrared regions of the electromagnetic spectrum. Among the unique characteristics of MoS 2 18−22 is that photogenerated excitons remain stable at room temperature because of their high binding energy. In contrast, devices made completely of MoS 2 have a significant limitation because of their limited capacity to absorb light. MoS 2 -based heterostructures formed by combining MoS 2 with materials such as zinc oxide (ZnO) offer the potential to enhance the optical qualities of the TMDC. ZnO possesses a wide direct bandgap of 3.37 eV at room temperature and a work function of 5.2 eV. 23,24 In addition, ZnO is low-cost and exhibits high resistance to defects, high stability, environmentally friendly characteristics, and biosafety. 25−28 When these plasmonic nanoparticles attach to semiconductors such as ZnO, a Schottky barrier will form at the interface between the metal and the semiconductor. The formation of this metal−semiconductor heterojunction is an effective way to enhance charge carrier separation and improve photocatalytic efficiency. 29 Combining TMDCs nanostructures with ZnO/ plasmonic metal nanomaterials offers a potential route to control surface catalytic reactions. It has been demonstrated that the addition of MoS 2 to ZnO concentrates charge onto MoS 2 through photoabsorption-based processes. 30,31 This can potentially enhance plasmonic catalysis rates through a strengthened exciton−plasmon interaction between silver nanoparticles (AgNPs) and MoS 2 , creating a stronger electric field at the AgNP/MoS 2 interface resulting in longer-lived hot electrons resulting in enhanced plasmonic catalysis properties. Additionally, MoS 2 can protect plasmonic metals, for example, by preventing the oxidation of Ag. MoS 2 can bind strongly to plasmonic metals such as Ag due to the favorable bonding between S (sulfide atom) and the Ag metal atoms. Simulations of AgNPs when attached to MoS 2 showed that strong excitation-plasmon coupling of the silver lattice with MoS 2 layers can occur, 32,33 resulting in the altering of the density of states (DOS) and increasing the hot electrons' lifetimes which can potentially improve plasmonic catalysis reaction rates. 34,35 Moreover, ZnO and MoS 2 have lattice constants that are very well matched, which makes them ideal for interfacial carrier transfer in ZnO/MoS 2 heterostructures. 36 Here we study the effect of combining TMDCs and metal oxides on a model plasmonic catalysis reaction. We undertake this study using the oxidation of p-amino thiophenol (PATP) to p-nitrophenol (PNTP). We show that in a model oxidation reaction PATP ↔ PNTP, the reaction is controlled by the introduction of the TMDC nanomaterial MoS 2 to a ZnO/ AgNP system by opening up new electron transfer routes. This study demonstrates the ability to control plasmonic reactions by carefully controlling the choice of semiconductor used to support the plasmon active nanomaterial.

■ RESULTS AND DISCUSSION
Photoluminescence (PL) emission spectra of ZnO and MoS 2 / ZnO mix (Figure 1a) revealed the predicted wide peak located at around 750 nm. When ZnO is introduced to MoS 2 , a significant quenching effect is noticed. This reduced PL intensity suggests that the rate of electron−hole recombination has decreased. This indicates that the TMDC and semiconductor have a strong interaction. Compared to MoS 2 or ZnO alone, the optical characteristics of ZnO coupled with MoS 2 exhibit no variation in absorption peak. As shown in Figure 1a, MoS 2 exhibits an absorption peak at ca. 600 nm (arising from exciton A and B transitions), while for ZnO, the absorption peaks are found at 450 nm. After mixing the composite, we found that the MOS 2 /ZnO mix exhibits stronger absorption ability than pristine MoS 2 , suggesting that the heterostructure has more intensive light-matter interaction. 37,26,29,58 In Figure 1b, a graph depicting the relationship between the photon energy hν (eV) and (ν*hν) 1/n was plotted, where n is a constant that relates to various electronic transition types (n = 3 for indirect forbidden transitions, n = 2 for indirect allowed, n = 3/2 for direct forbidden, and n = 1/2 for direct allowed).
where α stands for the absorbance coefficient, A represents the absorbance, and T represents the sample's thickness. 26 (Figure 1c). The characteristic stretching mode of the ZnO bond is assigned a large vibration band in the FTIR spectra ranging from 400 to 550 cm −1 . The presence of hydroxyl is shown by a broad peak at 3430 cm −1 stretching mode and 1330 cm −1 to 1670 cm −1 bending mode. 59,60 In addition, a band at 490 cm −1 is observed corresponding to a Mo−S vibration. Raman spectra of ZnO in (Figure 1d) showed the strongest peak at 440 cm −1 which is attributed to the phonon mode wurtzite hexagonal phase E 2 of ZnO. In addition, two peaks are seen at 330 and 380 cm −1 , which are allocated to the multiphoton process 2E 2 and A1-TO modes, respectively. In addition to the Raman spectra of MoS 2 excited at 532 nm, we noted the two main phonon peaks located at 370 cm −1 arising from the E 1 2g in-plane and 410 cm −1 assigned to A 1g out-of-plane. 51,37 The Raman spectra for MoS 2 and ZnO combined showed spectral features arising from combining the Raman spectral features from each component. Scanning electron microscopy (SEM) images of the pure MoS 2 , ZnO and MoS 2 /ZnO combined are shown in (Figure 1e) respectively. The pristine MoS 2 nanoparticles have a size range between 50 and 1000 nm and are clustered. MoS 2 :ZnO nanocomposites, the MoS 2 layers whose crystalline sizes are ca. 300 nm, are self-restacked and form thick layers. The small panels of ZnO are decorated on the surface, and edge of the large, in a small number of layered MoS 2 .
An investigation into the effect that MoS 2 /ZnO had on the plasmonic catalytic conversion of PATP to PNTP was undertaken. The Raman spectra of PATP and PNTP powder were first acquired using a dielectric substrate (Figure 2a). The Raman spectra shows A 1 modes with peaks at 1080 and 1595 cm −1 in agreement with literature values. 2  location and relative intensities. 8−14 It has been determined that a photocatalytic process takes place on the metal substrate, which is responsible for this difference in spectra when comparing PATP on a dielectric substrate to AgNPs. This photocatalytic process may result from the hot electrons formed when the Raman excitation laser excites the localized surface plasmon resonance (LSPR) of AgNPs. 1−3,8−14 There are two possible mechanisms for this plasmon-driven oxidation of PATP to p,p′-dimercaptoazobenzene (DMAB). First, the hot electrons that are produced as a result of plasmon decay are transferred to adsorbed singlet oxygen molecules from the surrounding air. This produces reactive triplet 3 O 2 , which is then engaged in the oxidation of PATP to DMAB. The second mechanism is that plasmonic hot electrons leap off the surface of the metal, and the hot holes that are left behind on the metal oxidize PATP to DMAB. It has been found that if a sufficiently enough external stimulus was introduced into the system, PATP would be oxidized to PNTP rather than oxidized to DMAB. 52,8−14 PATP on AgNPs/ZnO (Figure 2b) produces a Raman spectrum that replicates for PATP on only AgNPs with the spectrum assigned to DMAB. In contrast, when PATP is present on AgNPs/MoS 2 , the substrate prevents the formation of DMAB. The SERS spectrum possesses a peak at 1350 cm −1 arising from the presence of PNTP formed from the oxidation of PATP. When PATP is added to AgNPs/MoS 2 /ZnO, the spectra show features arising from a combination of DMAB and PNTP in comparison to when AgNPs/ZnO is studied where DMAB only is observed. Examining placing PNTP on the semiconductor-plasmonic substrate was then undertaken to assess how adding MoS 2 to ZnO/AgNPs affects this molecule. For PNTP on AgNP/ZnO, the Raman spectra of PNTP is preserved, with the Raman spectra (Figure 2c) showing the same features as recorded for PNTP (Figure 2a). In contrast, when MoS 2 is added forming ZnO/MoS 2 /AgNP, PNTP partially converts to DMAB with the Raman spectra showing peaks assigned to PNTP and DMAB. [9][10][11][12][13][14]52 A band diagram (Figure 3a) shows a band diagram of ZnO and MoS 2 . The semiconductor ZnO has an electron affinity (χ) = 3.87 eV, work function (φ) = 5.28 eV, and bandgap (E g ) = 3.26 eV. 29,53,54 While, MoS 2 has an estimated E g = 1.4 eV, φ = 5.15 eV, and χ = 4.3 eV. 25,38,55 When adding AgNPs to MoS 2 and ZnO, a heterojunction is formed (Figure 3b). Following the application of the Raman excitation wavelength (532 nm) electron−hole pairs are formed in MoS 2 . Electrons from MoS 2 conduction band can transfer the Ag Fermi level. As mentioned earlier, the transfer of electrons from MoS 2 conduction band to the Ag Fermi level will suppress PL (Figure 1a), as the Schottky contact will reduce the rate of radiative recombination. When ZnO is introduced, this semiconductor forms a ZnO/MoS 2 /AgNP system. Photogenerated electrons from MoS 2 could transfer to ZnO, and as the conduction band edge potential of ZnO is lower in energy than MoS 2 , the electrons in the conduction band of MoS 2 could transfer into the conduction band of ZnO. 29,38,51,53,54,56 This results in reduced efficiency of forming PNTP from PATP for ZnO/MoS 2 /AgNPs relative to MoS 2 /AgNPs (as observed in Figure 2).
Overall, we demonstrate that dichalcogenides when combined with metal oxides offer the potential to control charge states in plasmonic nanomaterials. This is demonstrated in a model oxidation reaction PATP ↔ PNTP. 57  Raman Spectroscopy. SERS spectra are collected using a monochromatic light green laser (HeNe, ThorLabs). Excitation wavelength is 532 (nm). Laser power and energy meter: a microscopy slide power meter sensor head (SN: 09113026, S121 C, 400−1100 (nm), 500 (mW), LMR1/M, Thorlabs) and energy meter are used to measure the incident laser power. The energy of the laser power is focused by an attenuator at 5 mW to control the laser power at a distance of ca. 2 cm for the entire experiment. Briefly, the beam passes through an interference filter and is directed by a mirror to angle prism, which drives the beam at 90 towered the sample. Then, it passes through a lens, which can be focused onto the samples to obtain the best signals. Here, the sample is excited and scatters light, which is collected by the lens and passes through a notch filter. This lowers the impact from the laser line before it enters the spectrograph. Raman spectra are collected with an exposure time of 1 s and 10 accumulation modes. Calibration of the Raman spectrographic windows is conducted by acquiring a Raman spectrum from the toluene and using it as a standard spectrum. The mean and standard deviation of 10 measurements is recorded.
Optical Spectroscopy UV−vis Absorption. Optical absorbance (UV−vis) measurements were accomplished with the use of an absorbance spectrometer (V-650, JASCO, Inc.), with the following settings: 1 nm step size, UV−vis bandwidth of 2 nm, and 200 nm/min scan speed across a range of 200− 800 nm. For the purpose of performing out the measurements, a coverslip substrate was used.
Fourier Transform Infrared Spectroscopy. Setup for Fourier transform infrared spectroscopy (FTIR) measurement parameters included a resolution of 4 cm −1 , a sample scan time of 8 scans, a measurement period of more than 10 s, data stored between 400 and 4000 cm −1 , result spectrum transmission mode, and accessory ATR platinum diamond. As a solid state, we recorded the FTIR spectra of both ZnO and MoS 2 as well as the composite of ZnO and MoS 2 . The Alpha Platinum Bruker system was used in order to acquire data from the FTIR instrument.
Transmission Electron Microscopy. TEM is used to examine the thin sample ultrastructure (limited by the penetration of electron beam). The transmission electron microscope utilizes an electromagnetic lens to concentrate electrons into a very tiny beam. The electrons then either scatter or strike a fluorescent screen at the bottom of the microscope after passing through a very thin object. An picture of the specimen with its many components shown in various hues based on its density shows on the screen. This picture may then be examined or photographed immediately inside the TEM.

■ ASSOCIATED CONTENT Data Availability Statement
The data that supports the findings of this study are available within the article.