Plasmon Effect of Ag Nanoparticles on TiO2/rGO Nanostructures for Enhanced Energy Harvesting and Environmental Remediation

We report Ag nanoparticles infused with mesosphere TiO2/reduced graphene oxide (rGO) nanosheet (TiO2/rGO/Ag) hybrid nanostructures have been successfully fabricated using a series of solution process synthesis routes and an in-situ growth method. The prepared hybrid nanostructure is utilized for the fabrication of photovoltaic cells and the photocatalytic degradation of pollutants. The photovoltaic characteristics of a dye-sensitized solar cell (DSSC) device with plasmonic hybrid nanostructure (TiO2/rGO/Ag) photoanode achieved a highest short-circuit current density (JSC) of 16.05 mA/cm2, an open circuit voltage (VOC) of 0.74 V and a fill factor (FF) of 62.5%. The fabricated plasmonic DSSC device exhibited a maximum power conversion efficiency (PCE) of 7.27%, which is almost 1.7 times higher than the TiO2-based DSSC (4.10%). For the photocatalytic degradation of pollutants, the prepared TiO2/rGO/Ag photocatalyst exhibited superior photodegradation of methylene blue (MB) dye molecules at around 93% and the mineralization of total organic compounds (TOC) by 80% in aqueous solution after 160 min under continuous irradiation with natural sunlight. Moreover, the enhanced performance of the DSSC device and the MB dye degradation exhibited by the hybrid nanostructures are more associated with their high surface area. Therefore, the proposed plasmonic hybrid nanostructure system is a further development for photovoltaics and environmental remediation applications.


Introduction
The significant increase in energy requirements and the depletion of fossil fuels have caused researchers to develop energy harvesting from renewable energy resources [1,2]. The use of solar-driven photovoltaic technologies and heterogeneous photocatalysis offers an appealing solution to the current global energy crisis and environmental remediation [3][4][5]. In terms of solar energy conversion technologies, dye-sensitized solar cells (DSSC) and heterogeneous photocatalytic dye degradation are the most attractive and promising areas of research to address energy and environmental concerns [6,7]. Semiconducting titanium dioxide (TiO 2 ) nanostructures are typically potential candidates as photoanode materials in DSSC devices and photocatalysts for the heterogenous photodegradation of toxic dyes due to their unique physicochemical properties [8][9][10]. However, the wide bandgap of the TiO 2 nanostructures (3.2-3.3 eV) has the absorption region below 4% in the entire solar spectrum compared to that of the visible region (43% solar energy) [11]. As a result, inefficient use of the visible light spectral portion and the fast recombination of electron-hole pairs (TiO 2 ) are significant constraints in the large-scale development of efficient photoanodes for DSSC devices and photocatalysts [12][13][14]. To overcome these constraints, different TiO 2 nanostructure composites with two-dimensional (2D) carbon materials have been employed as an effective strategy in recent decades. In nature, 2D-single molecular layered structures with sp 2 hybrid carbon atoms, i.e., graphene, possesses high surface area (ca. 2600 m 2 /g), electron mobility (ca. 15,000 m 2 /V.s at room temperature) and unrestricted movement of electrons in the crystal lattice [2,15,16]. During the formation of the nanocomposite, TiO 2 nanostructures are bonded with the graphene surface due to the presence of intermolecular forces. This increases the number of electron spots and electron bridges, which promotes electron transport with a suppressed recombination rate at the interface [17]. Moreover, reduced graphene oxide (rGO) nanosheet composites with TiO 2 nanostructures offer enhanced active surface area, good electrical conductivity and a lower recombination rate of photon-induced-charge carriers for photovoltaic and photocatalytic performance [18]. Manikandan et al. reported TiO 2 along with rGO not only enhanced the surface area but also influenced the short-circuit current (J SC ) in a device due to the high carrier mobility behavior of the rGO [19].
In recent decades, the localized surface plasmon resonance (LSPR) phenomenon of noble metal nanoparticles has played a dual role as potential visible sensitizers and electron sinks in the degradation of pollutant dyes and in DSSC device performance. It is a wellknown strategy to trap electrons and harvest maximum light for the development of high-efficiency photovoltaic devices. Amine-functionalized TiO 2 composites with GO and Ag nanoparticles exhibited high current density due to the improved electron transfer at the photoanode/electrolyte interface, as reported by Kandasamy et al. [20]. The incorporation of Ag nanoparticles (NPs) into TiO 2 -carbon nanotube (CNT) nanocomposites exhibited a photocatalytic activity of 66% for methylene blue (MB) degradation under visible light due to the presence of the CNTs and Ag NPs, as reported by Zhao et al. [21]. MB is a well-known, highly carcinogenic thiazine pollutant that has been manufactured and used in a variety of industries for various purposes. Therefore, it is strongly recommended to remove such a persistent contaminant from any given aqueous solution [22].
Moreover, the introduction of Ag NPs onto a TiO 2 nanocomposite improved the absorption-coefficient of the organic dye and eventually enhanced the optical absorption in the visible-light region [23,24]. It is similarly promising to achieve superior electrical conductivity with a prolonged lifetime of the photogenerated charge carriers for rGO/TiO 2 nanocomposites [25]. For example, Huan et al. prepared a flower-shaped nanosheet rGO/TiO 2 composite material that exhibited a photocatalytic efficiency of 92.3% under UV-visible light irradiation for the degradation of a rhodamine B (RhB) solution [26]. Duygu et al. reported the development of rGO-TiO 2 -CdO-ZnO-Ag based composites that exhibited an excellent degradation rate of methylene blue (MB) dye (15 min) with 91% photocatalytic efficiency under UV light irradiation [27]. Similarly, Zohreh et al. reported the influence of surface plasmon resonance on the photovoltaic characteristics of Ag/TiO 2 in a photoanode-based DSSC device with a power conversion efficiency (PCE) of 6.5% under 1 sun simulated solar irradiation [28]. As a result, the combination of the LSPR influence of Ag NPs and the conductivity of rGO nanosheets with the reduced recombination rate of the TiO 2 /rGO/Ag hybrid nanostructures is advantageous for long-term dye degradation photocatalysts and photoanodes for DSSC devices.
In the present work, we report the in situ growth synthesis of mesosphere TiO 2 /rGO nanosheets/Ag NPs as a plasmonic hybrid nanostructures for visible-light-responsive DSSCs and photocatalytic applications. The prepared hybrid nanostructures are further examined with respect to their structural, morphological and optical properties using various advanced characterization techniques. The influence of the LSPR properties on a constructed DSSC device with the visible-light-driven photoanode and photocatalyst dye (MB) degradation is systematically investigated with an appropriate mechanism.

Preparation of TiO 2 Mesospheres
TiO 2 mesospheres were prepared through a combined route of sol-gel and solvothermal synthesis. In the first step (sol-gel), 3 mL of TTIP was slowly added into 150 mL of ethylene glycol, and the solution was continuously stirred for 12 h at room temperature. After a 12 h stirring process, 300 mL of acetone with 2 mL of de-ionized water (DI) were added into the solution and stirring was continued for 2 h to obtain a white suspension. Subsequently, the white suspension was collected and subjected to several centrifugations with ethanol and DI water, respectively. The obtained product was dried at 60 • C for 10 h to remove the impurities and form titanium glycolate. In the second step of the procedure (solvothermal route), 1 g of titanium glycolate was dispersed into 60 mL of a mixed solvent of ethanol and DI water under stirring. The prepared white solution was then transferred into a 100 mL autoclave and kept at 180 • C for 12 h. Finally, the TiO 2 mesospheres were obtained after annealing at 350 • C for 1 h.

Preparation of Graphene Oxide (GO)
Graphene oxide (GO) was prepared under room-temperature conditions by the oxidation of natural graphite powder using a modified Hummer's method [29]. Briefly, 1 g of natural graphite powder and 0.5 g of sodium nitrate were blended in 23 mL of concentrated sulfuric acid under vigorous stirring for 30 min. A total of 3 g of well-ground potassium permanganate was then slowly added into the mixed solution and stirred for 30 min under ice-bath conditions at 7 • C. The mixed solution was then stirred at 35 • C for 30 min. To quench the vigorous oxidation process in the solution, 3 mL of hydrogen peroxide was added to 60 mL of DI water, and the solid GO powder was obtained after washing several times with 5% of HCL and DI water, then dried overnight at 60 • C.

Preparation of Mesosphere TiO 2 /rGO Sample
Prepared TiO 2 mesospheres (0.2 g) were added to a mixed solution of an equal portion of DI water (15 mL) and ethanol (15 mL) (as solution A). Then, 30 mg of GO powder was dispersed in the same ratio of solvent (DI and ethanol) under stirring for 1 h (as Solution B). Solutions A and B were then mixed with continuous stirring for 1 h. The resultant mixed solution was transferred into a 100 mL autoclave and kept in hot air at 180 • C for 12 h. The final product of mesosphere TiO 2 /rGO was obtained by washing the product with DI water and drying overnight at 60 • C.

Preparation of TiO 2 /rGO/Ag Hybrid Nanostructure by In Situ Growth
The TiO 2 /rGO/Ag hybrid nanostructures were synthesized by an in situ hydrothermal process. In this procedure, 10 mg of silver nitrate was dissolved in 60 mL of DI water (30 mL) and ethanol (30 mL), and then 0.2 g of the prepared TiO 2 /rGO was added to the above solution under continuous stirring for 1 h. The net solution was transferred into an autoclave and kept at 180 • C for 4 h. The final product was washed with DI water and dried at 60 • C.

Characterization
Structural analysis was conducted using X-ray diffraction (XRD; PANalytical, Malvern, UK) with Cu Kα radiation (λ = 1.5406 Å) in the 2-theta range between 10 • and 80 • with a scanning rate of 0.02 • /min. Raman spectra of the prepared sample were obtained using a micro-Raman spectrometer (LABRAM HR Evolution, Horiba, Longjumeau, France) with an excitation wavelength of 532 nm. The surface morphology of the prepared samples was analyzed using high-resolution scanning electron microscopy (HR-SEM; Apreo S, Thermo Fisher Scientific, Hillsboro, OR, USA) with an acceleration voltage of 15 kV. Further analysis of the surface morphology was conducted using high-resolution transmission electron microscopy (HR-TEM; JEM-2100, JEOL, Tokyo, Japan) with an acceleration voltage of 200 kV to reveal the atomic interplanar morphology and the elemental composition of the prepared samples. Ultraviolet-visible diffuse reflectance spectroscopy (UV-DRS, V-750, JASCO, Tokyo, Japan) measurements were conducted in the range of 200 nm to 800 nm. To analyze the emission properties of the prepared sample, photoluminescence (PL; FP8600, JASCO, Tokyo, Japan) spectra were measured at room temperature. The surface area and pore size distribution of the samples were characterized by the BET (Brunauer-Emmett-Teller) and BJH (Barrett-Joner-Halenda) methods (Autosorb IQ series, Quantachrome Instruments, Boynton Beach, FL, USA). X-ray photoelectron spectroscopy (XPS) was performed via a Kratos analytical instrument (ESCA 3400, Shimadzu Corporation, Kyoto, Japan). The percentage of mineralization efficiency was determined from total organic carbon (TOC) measurements (TOC-L, Shimadzu, Kyoto, Japan). Photovoltaic characterization of the fabricated devices was performed using a solar simulator (Sciencetech, Class A, Lamp: 300 W, London, ON, Canada). The I-V measurement and incident photocurrent efficiency (IPCE) of fabricated devices was measured using the same solar simulator over the wavelength range of 200 nm to 800 nm.

Photocatalytic Experiments
The photocatalysts measurement were carried out in our laboratory, SRM Institute of Science and Technology, Chennai (28 • 4 N; 82 • 25 E), in April 2020 (8 April to 10 May). Daylight from 9 am to 12 pm was utilized to perform the photocatalytic experiment with an average light intensity of 68.2~89.4 mW/cm 2 . The photocatalytic properties of the as-synthesized samples were performed using MB dye as a model pollutant. In a typical photocatalytic reaction, 10 ppm of MB dye was added to 50 mL of DI water and stirred for 5 min. The solution was maintained in the dark for 20 min under stirring to achieve an adsorption-desorption equilibrium. At regular time intervals of the photocatalytic dye degradation reaction solution (20 min), 3 mL aliquots of the solution were sampled and UV-Vis spectra were measured.

DSSC Device Fabrication
Prepared samples, such as TiO 2 , TiO 2 /rGO and TiO 2 /rGO/Ag (0.25 g each), were dispersed in 2 mL of stock solution (prepared by mixing an equal amount of DI water and ethanol) and ground for 15 min. To obtain a paste-like formation, 2 mL of the stock solution was mixed with 200 µL of acetic acid and 100 µL of surfactant (Triton 100-X) during the grinding process. The resultant colloidal solution was then uniformly coated on a fluorine-doped tin oxide (FTO) substrate with optimized conditions using a doctor blade technique. After drying at 100 • C for 10 min, the coated substrate was annealed at 450 • C for 30 min. The coated FTO was then immersed in dye solution for 12 h, which consisted of 0.03 M dis-tetrabutylammonium cis-bis(isothiocyanato)bis(2,2 -bipyridyl-4,4dicarboxylato) ruthenium (II) (N719). The prepared photoanode was subsequently clamped with a Pt-coated counter electrode to form a sandwich type of device. An electrolyte solution consisting of 0.6 M dimethylpropylimidazolium iodide, 0.1 M lithium iodide, 0.01 M iodine and 0.5 M 4-tert-butylpyridine in acetonitrile was then filled in between the layers of the sandwich-type device.  Figure 1a shows the XRD patterns of prepared samples of pristine GO, mesosphere TiO 2 , TiO 2 /rGO and TiO 2 /rGO/Ag. A strong and sharp diffraction peak is observed at 9.8 • , which corresponds to the (0 0 1) crystal plane of GO. Using Bragg's equation, the interlayer distance of the as-prepared GO sheet was estimated to be 0.9 nm, whereas graphite powder shows an interlayer distance around 0.33 nm [30]. The XRD pattern obtained for GO indicates that the bulk graphite is successfully reduced as rGO nanosheets [31]. with a Pt-coated counter electrode to form a sandwich type of device. An electrolyte solution consisting of 0.6 M dimethylpropylimidazolium iodide, 0.1 M lithium iodide, 0.01 M iodine and 0.5 M 4-tert-butylpyridine in acetonitrile was then filled in between the layers of the sandwich-type device. Figure 1a shows the XRD patterns of prepared samples of pristine GO, mesosphere TiO2, TiO2/rGO and TiO2/rGO/Ag. A strong and sharp diffraction peak is observed at 9.8°, which corresponds to the (0 0 1) crystal plane of GO. Using Bragg's equation, the interlayer distance of the as-prepared GO sheet was estimated to be 0.9 nm, whereas graphite powder shows an interlayer distance around 0.33 nm [30]. The XRD pattern obtained for GO indicates that the bulk graphite is successfully reduced as rGO nanosheets [31].   To explore the structural characteristics, Raman spectroscopy was performed for GO, TiO 2 , TiO 2 /rGO and TiO 2 /rGO/Ag samples, as shown in Figure 1b. The obtained Raman spectrum of pristine GO revealed the D band at 1350 cm −1 due to the disordered nature of the graphene structure with sp 3 defects (as given in Figure S1). In addition, the G band (at 1587 cm −1 ) was observed in the Raman spectrum of pristine GO, which was attributed to the in-plane vibrations of C-C stretching in graphitic materials, as well as the doubly degenerate phonon mode in the Brillouin zone [32]. As shown in the inset image of Figure 1b, the G band shifted from 1587 to 1599 cm −1 , which is direct evidence of the effective reduction of GO to rGO with a constant D band for TiO 2 /rGO [33]. In comparison with pristine GO, the I D /I G intensity ratio of the TiO 2 /rGO and TiO 2 /rGO/Ag samples was slightly reduced to 0.83 from 0.87 (as shown in Supporting Information Figure S1). However, the reduction in the I D /I G ratio indicates a considerable reduction in the sp 2 domain size of carbon atoms, as well as the reduction of sp 3 to sp 2 [34]. In the case of TiO 2 /rGO/Ag, the Raman vibrational modes were observed at 147, 199, 396, 515 and 639 cm −1 , which corresponds to the E g , E g , B 1g , A 1g and E g modes of anatase TiO 2 [35]. On the other hand, the broadening of the E g mode (147 cm −1 ) with a considerable peak shift revealed the presence of TiO 2 mesospheres, as well as Ag NPs, on multilayer rGO. Furthermore, there were no significant changes in the observed Raman scattering modes of the plasmonic hybrid nanostructures compared with the hybrid sample (TiO 2 /rGO). Nevertheless, the single Raman characteristic peak of the anatase TiO 2 phase was considerably enhanced due to the LSPR properties of Ag NPs on the TiO 2 /rGO/Ag hybrid surface [36]. The obtained Raman characteristic modes further confirm the successful formation of plasmonic hybrid nanostructures.

Structural and Compositional Analysis
XPS measurements were performed to evaluate the chemical state and interaction of the prepared hybrid nanostructures and pristine GO samples. The surveyed spectra of TiO 2 , TiO 2 /rGO and TiO 2 /rGO/Ag indicate the presence of Ti, O, Ag and C elements without any impurities in the prepared hybrid nanostructures (as given in Supporting Information Figure S2). The deconvoluted C 1s spectrum of the as-prepared GO sample is shown in Figure 1c as follows: (i) At 282.55 eV, C bonds with sp 1 carbon atoms [37], (ii) C=C bonds denote the aromatic sp 2 structure groups present in the GO at 284.66 eV, (iii) the peak at 286.30 eV represents the carboxyl groups (C-O bonds) including epoxy and hydroxyl group [38]. In the case of the TiO 2 /rGO sample, the deconvoluted XPS peak of the C 1s spectrum shows the four diverse carbon bonds with various binding energies of 283.18 eV, 284.74 eV, 286.07 eV and 288.72 eV. Thus, the peak that appears at 283.18 eV indicates the chemical bonding between the C atom and the Ti atom (Ti-C) [34]. Furthermore, the peaks with reduced intensity emerges at 284.74 eV, 286.07 eV and 288.72 eV with extensively reduced intensity purely attributed to C=C, C-O, O=C-OH bonds as oxygenated functional groups, which highlights the formation of rGO from GO [39]. Figure S3 shows the deconvoluted O 1s spectrum with two bands at 529 eV and 530 eV that correspond to the Ti-O-Ti and Ti-O-C groups. There is a slight peak shift in the TiO 2 /rGO sample from 530.95 eV to 531.12 eV, which indicates the strong binding of rGO with TiO 2 mesosphere during the solvothermal process [40,41] . The chemical states of Ti species were thoroughly investigated using the XPS spectra, as shown in Figure 1e. The core level peak of Ti is split into a doublet peak observed at 458.40 eV and 464.10 eV, which corresponds to Ti 2p 3/2 and Ti 2p 1/2 with a splitting energy of 5.7 eV, mainly attributed to the Ti 4+ state for the TiO 2 /rGO/Ag samples [42,43]. After infusing the Ag NPs into the hybrid nanostructure, the doublet peaks (Ti 2p 3/2 and Ti 2p 1/2 ) of the Ti 2p state have observed notable shifts to higher order, such as 458.40 eV to 458.51 eV (Ti 2p 3/2 ) and 464.10 eV to 464.27 eV (Ti 2p 1/2 ). As a result, a positive shift in the doublet peak of the Ti species directly reflects the alteration of the Fermi level, whereas the Ag species holds a lower Fermi level offset than that of TiO 2. Therefore, the resultant Fermi level alignment in the plasmonic hybrid nanostructures are more favorable for fast electron transfer processes at the rGO/TiO 2 /Ag interface [44]. Figure 1f shows the doublet peaks of the Ag element at binding energies of 367.23 eV and 373.31 eV, which correspond to Ag 3d 5/2 and Ag 3d 3/2 with a splitting energy of 6.12 eV, which confirms the presence of the Ag NPs in the prepared TiO 2 /rGO/Ag hybrid nanostructures [45].

Morphological Analysis
The surface morphology features of GO nanosheets, TiO 2 mesospheres, TiO 2 /rGO and TiO 2 /rGO/Ag hybrid nanostructures were investigated using HR-SEM, TEM and HR-TEM measurements, as shown in Figure 2. The transparent ultrathin GO nanosheets are visible in the HR-SEM and HR-TEM micrographs, as shown in Figure 2(a1-a3). As-prepared TiO 2 mesosphere samples are agglomerate-free, uniformly distributed and spherical shaped, with an average size of~570 nm (see Figure 2(b1)).

Morphological Analysis
The surface morphology features of GO nanosheets, TiO2 mesospheres, TiO2/rGO and TiO2/rGO/Ag hybrid nanostructures were investigated using HR-SEM, TEM and HR-TEM measurements, as shown in Figure 2. The transparent ultrathin GO nanosheets are visible in the HR-SEM and HR-TEM micrographs, as shown in Figure 2(a1-a3). As-prepared TiO2 mesosphere samples are agglomerate-free, uniformly distributed and spherical shaped, with an average size of ~570 nm (see Figure 2(b1)). The surface morphology and atomic interplanar distance of prepared samples were investigated using TEM and HR-TEM, as shown in Figure 2(b2,b3). Figure 2(c1-c3) shows the uniform TiO2 mesospheres with an interplanar distance of 0.35 nm for the pure anatase TiO2 phase over the rGO nanosheet surfaces. The presence of Ag NPs embedded in the TiO2 mesosphere-coated rGO nanosheet surfaces is revealed by the TEM and HR-TEM images of the hybrid nanostructures. The interplanar distance of the prepared hybrid The surface morphology and atomic interplanar distance of prepared samples were investigated using TEM and HR-TEM, as shown in Figure 2(b2,b3). Figure 2(c1-c3) shows the uniform TiO 2 mesospheres with an interplanar distance of 0.35 nm for the pure anatase TiO 2 phase over the rGO nanosheet surfaces . The presence of Ag NPs embedded in the TiO 2 mesosphere-coated rGO nanosheet surfaces is revealed by the TEM and HR-TEM images of the hybrid nanostructures. The interplanar distance of the prepared hybrid nanostructures was estimated from the HR-TEM image using a line profile analysis, as shown in Figure 2(d3). The calculated interplanar distance values were 0.35 nm and 0.23 nm for the anatase TiO 2 crystal plane of (1 0 1) and the (1 1 1) crystal plane of the Ag NPs, respectively, which is well matched with the standard JCPDS cards (Nos. 21-1272 and 04-0783) [46]. Moreover, the elemental mapping and energy dispersive X-ray spectroscopy (EDX) spectrum of the prepared hybrid nanostructures indicate the coexistence of C, Ti, O and Ag elements without the presence of any impurities in the resultant hybrid sample (as shown in Supporting Information Figure S4). Therefore, the resultant surface morphology of the prepared hybrid ternary nanostructures indicates that the plasmonic Ag NPs are uniformly embedded into the TiO 2 mesospheres anchored in the rGO nanosheet surfaces. Moreover, the thickness of the fabricated photoanode was estimated to be 14.2 µm (TiO 2 ) and 13.8 µm (TiO 2 /rGO/Ag) and presented Figure S5a,b.

Optical and Surface Area Analysis
UV-DRS was used to investigate the optical absorption of the prepared TiO 2 , TiO 2 /rGO and TiO 2 /rGO/Ag samples and is shown in Figure 3a. Due to the electronic transition of O 2p to Ti 3d, all of the prepared samples exhibit a typical optical absorption edge located at 392 nm of TiO 2 , which corresponds to a bandgap energy of 3.3 eV for TiO 2 [47]. After the introduction of Ag NPs over the TiO 2 /rGO surface, the optical absorption of the hybrid nanostructure is significantly higher in the visible region compared to the other samples, due to the LSPR of Ag NPs on the TiO 2 /rGO surface [48]. Overall, the improved visible-light-region absorption observed for the prepared hybrid nanostructures indicates a greater number of photons harvested from the visible-light region, which is favorable for efficient photocatalysts and photoanodes for solar energy conversion devices. PL emission spectroscopy measurements are widely used to understand the desired photon-induced charge separation properties of the prepared samples. PL spectra of the prepared samples with an excitation wavelength of 300 nm and in the range from 350 to 500 nm are shown in Figure 3b. A strong band edge emission peak is observed at around 389 nm for all the prepared nanostructure samples [49]. Another emission peak centered in the visible region of 480 nm indicates the presence of oxygen vacancies on the TiO2 PL emission spectroscopy measurements are widely used to understand the desired photon-induced charge separation properties of the prepared samples. PL spectra of the prepared samples with an excitation wavelength of 300 nm and in the range from 350 to 500 nm are shown in Figure 3b. A strong band edge emission peak is observed at around 389 nm for all the prepared nanostructure samples [49]. Another emission peak centered in the visible region of 480 nm indicates the presence of oxygen vacancies on the TiO 2 surface. For the plasmonic hybrid nanostructures, the resultant emission peaks are significantly quenched, which is a direct indication of the reduction of the photon-induced electron-hole pair recombination rate that facilitates electron transfer at the TiO 2 /rGO/Ag interface [50,51]. The function group of GO and prepared materials were analyzed by FTIR spectrum [52,53] (Figure S6).
The surface area and pore size distribution of the prepared samples were evaluated using the BET and BJH methods shown in Figure 3c,d. In comparison, the specific surface area of Ag NP-based hybrid nanostructures have a surface area of 297.71 m 2 /g due to TiO 2 mesosphere formation with the Ag NP-coated rGO nanosheets. The estimated surface areas of the TiO 2 mesosphere and TiO 2 /rGO samples were 234.16 m 2 /g and 281.31 m 2 /g, respectively. As shown in Figure 3d, the estimated pore sizes were 2.15 nm, 1.52 nm and 1.91 nm for the TiO 2 , TiO 2 /rGO and TiO 2 /rGO/Ag hybrid nanostructures, respectively. The enhanced specific surface areas and pore size distributions of the hybrid nanostructures are more beneficial for the photoanodes in DSSC assemblies and photocatalysts for photocatalytic reactions.

Performance of DSSC Device
The influence of the LSPR properties of Ag NPs on the photovoltaic characteristics of a DSSC based on a TiO 2 /rGO/Ag photoanode under simulated sun irradiation with an AM 1.5 G filter was investigated. The current density vs. voltage characteristics of DSSC devices assembled with various photoanodes (TiO 2 , TiO 2 /rGO and TiO 2 /rGO/Ag) are given with error bars in Figure 4a. Compared with the TiO 2 and TiO 2 /rGO photoanodes, the TiO 2 /rGO/Ag photoanode show superior photovoltaic performance with a maximal J SC of 16.05 mA/cm 2 and an improved open circuit voltage (Voc) of 0.74 V, with a fill factor (FF) of around 62.50%. In addition, the hybrid-nanostructure-based plasmonic DSSC device achieved a higher power conversion efficiency of 7.27%, which is 1.7 times higher than that of a pristine TiO 2 photoanode (4.01%). An overall comparison of a fabricated DSSC device's photovoltaic characteristics are given in Table 1 and SI. Figure 4b shows the IPCE of all of the fabricated DSSC devices in the range of 400 nm to 800 nm. The Ag-based DSSC device exhibited the maximum IPCE owing to the presence of N719 dye adsorption. The IPCE of the TiO 2 /rGO/Ag photoanode-based plasmonic DSSC device achieved a maximum value of ca. 77.82% at an incident wavelength of 550 nm due to improved photon harvesting efficiency and a higher number of electron extraction from the N719 dye molecule than the other devices [27,28,52]. The photovoltaic performance of the fabricated TiO 2 /rGO/Ag photoanode-based plasmonic DSSC cell has been compared with a recent report, as illustrated in Figure 4c [54][55][56][57][58][59]. A tentative operating mechanism of the proposed plasmonic DSSC device is illustrated to understand its enhanced photovoltaic performance. As shown in Figure 4d,e, upon 1 sun irradiation the electrons are excited from higher occupied molecular orbitals (HOMO) to lower occupied molecular orbitals (LUMO) levels of the dye molecule (N719). Under this irradiation, if the incident light coincides with the LSPR effect of Ag NPs (work function (WF) = −4.2 eV), then the electrons close to the Fermi level are excited into a higher energy state by receiving the energy from plasmon resonance, known as a hot electron, via non-radioactive process [60]. These electrons have sufficient energy to break the Schottky barrier formed between the TiO 2 mesosphere (WF = −4.4 eV) and Ag NPs and can flow through the conduction band of the TiO 2 mesosphere. Moreover, it is possible that this hot electron can be accepted and shuttled via rGO nanosheets (WF = −4.4 eV) [61][62][63]. The superior conductivity of rGO nanosheets could assist in accepting or transmitting electrons generated by TiO 2 or Ag NPs to the appropriate photoanode. The hybrid nanostructure attained higher efficiency than the TiO 2 mesospheres for the following reasons. (i) The large surface area provides high dye loading and efficiently scatters the incoming light within the device. (ii) The rGO nanosheets promote extraordinary charge transport where electrons come from TiO 2 mesospheres or Ag NPs and limit the electron-hole recombination rate. (iii) The LSPR effect of the Ag NPs extends the light absorption range from the UV to the visible region.   Figure S7 shows the transient photocurrent response of the TiO2/rGO/Ag composite under 1 sun light irradiation with the highest photocurrent density, which is approximately one-fold greater than that of TiO2/rGO. This confirms that the incorporation of Ag into the TiO2/rGO/Ag nanocomposite not only enhances the light harvesting but also significantly expedites the photon-induced charge carrier separation and transport properties of Ag NP-coupled hybrid composites. The transient photocurrent response of the TiO2/rGO/Ag composite was also performed under a UV filter, which confirmed the plasmonic Ag response photocurrent in the visible region. These results confirm that Ag could act as an acceptor of the photogenerated electrons by TiO2/rGO and encourage fast charge transportation due to the high metallic conductivity, which effectively suppresses the charge recombination in the composites.

Photocatalytic Performance
To demonstrate the photocatalytic MB dye degradation process, TiO2-, TiO2/rGOand TiO2/rGO/Ag-based photocatalysts were employed under natural sunlight irradiation for three repetitions as shown in Figure S8.    Figure S7 shows the transient photocurrent response of the TiO 2 /rGO/Ag composite under 1 sun light irradiation with the highest photocurrent density, which is approximately one-fold greater than that of TiO 2 /rGO. This confirms that the incorporation of Ag into the TiO 2 /rGO/Ag nanocomposite not only enhances the light harvesting but also significantly expedites the photon-induced charge carrier separation and transport properties of Ag NP-coupled hybrid composites. The transient photocurrent response of the TiO 2 /rGO/Ag composite was also performed under a UV filter, which confirmed the plasmonic Ag response photocurrent in the visible region. These results confirm that Ag could act as an acceptor of the photogenerated electrons by TiO 2 /rGO and encourage fast charge transportation due to the high metallic conductivity, which effectively suppresses the charge recombination in the composites.

Photocatalytic Performance
To demonstrate the photocatalytic MB dye degradation process, TiO 2 -, TiO 2 /rGO-and TiO 2 /rGO/Ag-based photocatalysts were employed under natural sunlight irradiation for three repetitions as shown in Figure S8. Figure 5a-c shows dye degradation profiles with MB as the pollutant with different irradiation times under natural sunlight for all of the prepared photocatalysts. The maximum absorption peak for the photocatalytic dye degradation process in the presence of a photocatalyst is displayed at around 664 nm due to the absorbance characteristics of MB dye. Under dark conditions, there were no significant changes in the UV-vis absorption spectra of all the prepared photocatalysts. The UV-vis absorption spectra of all the prepared photocatalysts were successfully recorded with a time interval of 20 min under continuous natural sun irradiation. The hybrid nanostructures of the TiO 2 /rGO/Ag photocatalyst exhibited a higher dye degradation efficiency of 93% under continuous natural sunlight irradiation for 160 min. Compared with pristine TiO 2 , the enhanced dye degradation efficiency of the plasmonic hybrid photocatalyst was almost 1.3 times higher due to the LSPR properties of the Ag NPs. On the other hand, the presence of rGO/Ag on the TiO 2 surface creates a more favorable environment for fast photon-induced charge separation and transfer with an extended lifetime of charge carriers, whereas Ag NPs act as electron sinks for improved photocatalytic dye degradation [64].  To further understand the effect of dye degradation in the form of mineralization efficiency, TOC analysis of the TiO2/rGO/Ag photocatalyst was performed with respect to various time intervals, as shown in Figure 5e. After 160 min of natural solar irradiation, the mineralization efficiency of the TiO2/rGO/Ag photocatalyst in terms of carbon content removal in MB dye was approximately 80.35% (Figure 5e). The improved TOC of the plasmonic hybrid TiO2/rGO/Ag photocatalyst suggests that it efficiently mineralized MB dye into residual organic molecules, such as CO2 and H2O, in the MB dye solution. Furthermore, the photocatalytic MB dye degradation activity of TiO2/rGO/Ag was evaluated by the addition of various radical scavengers (Ag, EDTA, IPA and BQ) to determine the more active species e − , h + , OH − and O 2− in the degradation system [65,66]. Figure 5f shows various scavenger photocatalytic MB dye degradation (MB) experiments with the TiO2/rGO/Ag system as a photocatalyst. For this study, 2 mg of various scavengers were To further understand the effect of dye degradation in the form of mineralization efficiency, TOC analysis of the TiO 2 /rGO/Ag photocatalyst was performed with respect to various time intervals, as shown in Figure 5e. After 160 min of natural solar irradiation, the mineralization efficiency of the TiO 2 /rGO/Ag photocatalyst in terms of carbon content removal in MB dye was approximately 80.35% (Figure 5e). The improved TOC of the plasmonic hybrid TiO 2 /rGO/Ag photocatalyst suggests that it efficiently mineralized MB dye into residual organic molecules, such as CO 2 and H 2 O, in the MB dye solution. Furthermore, the photocatalytic MB dye degradation activity of TiO 2 /rGO/Ag was evaluated by the addition of various radical scavengers (Ag, EDTA, IPA and BQ) to determine the more active species e − , h + , OH − and O 2− in the degradation system [65,66]. Figure 5f shows various scavenger photocatalytic MB dye degradation (MB) experiments with the TiO 2 /rGO/Ag system as a photocatalyst. For this study, 2 mg of various scavengers were mixed with MB dye solution in the presence of a prepared photocatalyst (15 mg) and exposed to natural sunlight for 160 min. The photocatalytic MB dye degradation efficiency of TiO 2 /rGO/Ag was significantly inhibited by 58.89% and 23.00% in the presence of IPA and BQ as scavengers, respectively. This reduced the photocatalytic dye degradation efficiency, and it reveals that OH − and O 2− radicals are the main active species during the degradation of MB by TiO 2 /rGO/Ag under natural sunlight [67].
A possible mechanism for the photon-induced charge carrier separation and transfer in the prepared photocatalyst surface is depicted in Figure 6. The valence band (E VB ) and conduction band (E CB ) potentials of TiO 2 , TiO 2 /rGO and TiO 2 /rGO/Ag were estimated using the Mulliken electronegativity theory [53] and the following equations: where x is the electronegativity (5.81 eV for TiO 2 ), E CB is the conduction edge potential, E VB is the valence band edge potential, E e is the free energy of the electrons in the reversible hydrogen scale (4.5 eV), and E g is the bandgap of the material. The E VB and E CB potentials of the TiO 2 /rGO/Ag system were estimated to be −0.34 eV and 2.96 eV, respectively, while E VB and E CB for pure TiO 2 were estimated to be −0.32 and 2.94 eV, respectively.
of TiO2/rGO/Ag was significantly inhibited by 58.89% and 23.00% in the presence of IPA and BQ as scavengers, respectively. This reduced the photocatalytic dye degradation efficiency, and it reveals that OH − and O 2− radicals are the main active species during the degradation of MB by TiO2/rGO/Ag under natural sunlight [67]. A possible mechanism for the photon-induced charge carrier separation and transfer in the prepared photocatalyst surface is depicted in Figure 6. The valence band (EVB) and conduction band (ECB) potentials of TiO2, TiO2/rGO and TiO2/rGO/Ag were estimated using the Mulliken electronegativity theory [53] and the following equations: where x is the electronegativity (5.81 eV for TiO2), ECB is the conduction edge potential, EVB is the valence band edge potential, E e is the free energy of the electrons in the reversible hydrogen scale (4.5 eV), and Eg is the bandgap of the material. The EVB and ECB potentials of the TiO2/rGO/Ag system were estimated to be −0.34 eV and 2.96 eV, respectively, while EVB and ECB for pure TiO2 were estimated to be −0.32 and 2.94 eV, respectively. Figure 6 shows the proposed photocatalytic mechanism for the natural-sunlight-responsive TiO2-RGO-Ag nanostructure. After natural sunlight irradiation, the Ag NPs' dipolar characteristics of the SPR effect enable electrons generated from the Ag NPs to migrate to the CB of TiO2. In contrast, electrons in the CB of TiO2 could be transferred to the surface of the RGO nanosheets due to the favorable electron affinity of TiO2 and the lower WF of RGO. Further, it is postulated that the photogenerated electron from Ag particles can also be transferred into the RGO sheets. TiO2and plasmon-excited Ag particles serve as electron transfer channels, which ensure charge separation efficiency. The oxidized Ag + species accept electrons from water molecules (H2O) or OHadsorbed on the TiO2 surface or from the dye molecules present in the solution and are regenerated. The reaction with H2O or hydroxide ions produces hydroxyl radicals (OH). These radicals (O2 and OH) are influential oxidizing agents for the degradation of MB dye molecules. Therefore, the prepared ternary nanostructures offer superior photocatalytic dye degradation (MB) due to efficient photoinduced charge carrier separation [52,53].

Conclusions
In this study, TiO2/rGO/Ag hybrid nanostructures were successfully synthesized using a combination of solution processes and in situ growth and were then employed as photoanodes for DSSCs and as catalysts for photodegradation applications. The plasmon-  Figure 6 shows the proposed photocatalytic mechanism for the natural-sunlightresponsive TiO 2 -RGO-Ag nanostructure. After natural sunlight irradiation, the Ag NPs' dipolar characteristics of the SPR effect enable electrons generated from the Ag NPs to migrate to the CB of TiO 2 . In contrast, electrons in the CB of TiO 2 could be transferred to the surface of the RGO nanosheets due to the favorable electron affinity of TiO 2 and the lower WF of RGO. Further, it is postulated that the photogenerated electron from Ag particles can also be transferred into the RGO sheets. TiO 2 -and plasmon-excited Ag particles serve as electron transfer channels, which ensure charge separation efficiency. The oxidized Ag + species accept electrons from water molecules (H 2 O) or OH − adsorbed on the TiO 2 surface or from the dye molecules present in the solution and are regenerated. The reaction with H 2 O or hydroxide ions produces hydroxyl radicals (OH). These radicals (O 2 and OH) are influential oxidizing agents for the degradation of MB dye molecules. Therefore, the prepared ternary nanostructures offer superior photocatalytic dye degradation (MB) due to efficient photoinduced charge carrier separation [52,53].

Conclusions
In this study, TiO 2 /rGO/Ag hybrid nanostructures were successfully synthesized using a combination of solution processes and in situ growth and were then employed as photoanodes for DSSCs and as catalysts for photodegradation applications. The plasmonenhanced DSSC devices demonstrate enhanced photovoltaic performance of 7.27% along with a higher short-circuit current of 16.05 mA/cm 2 and an IPCE efficiency of 77.82% at 550 nm. The results suggest that the high photovoltaic performance of the plasmon-based TiO 2 /rGO/Ag device can be attributed to (i) the large specific area of TiO 2 /rGO/Ag, which leads to high dye loading; (ii) TiO 2 mesospheres enhancing the light scattering effect of incoming light; and (iii) the incorporation of Ag NPs facilitating more induced photons and fast electron transport in the device. Upon natural sunlight irradiation, the prepared hybrid nanostructure shows an improved photocatalytic degradation of MB by 93% within 160 min, and the effects of different scavengers on the obtained photocatalytic activity were systematically investigated. The effects of optimum active surface area, the LSPR properties of Ag NPs and the enhanced electrical conductivity of the prepared ternary nanostructures combine to provide an enhanced visible-light-driven plasmonic DSSC device and photocatalyst for dye degradation (MB). The proposed plasmonic and hybrid-based nanostructures demonstrate an emerging strategy to establish large-scale applications of solar energy conversion technologies.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.

Conflicts of Interest:
The authors declare no conflict of interest.