Incorporating Solution-Processed Mesoporous WO3 as an Interfacial Cathode Buffer Layer for Photovoltaic Applications

Dextran-templating hydrothermal synthesis of monoclinic WO3 exhibits excellent specific surface area of ∼110 m/g and a monomodal pore distribution with an average pore diameter of ∼20 nm. Dextran plays a crucial role in generating porosity on WO3. The role of supporting dextran has been investigated and found to be crucial to tune the surface area, porosity, and morphology. The photoluminescence and X-ray photoelectron spectroscopy studies reveal the existence of oxygen vacancies in substoichiometric WO3, which creates localized defect states in WO3 as synthesized through this templating method. The highly mesoporous WO3 has been further explored as an interfacial cathode buffer layer (CBL) in dye-sensitized solar cells (DSSCs) and perovskite solar cells (PSCs). A significantly enhanced photoconversion efficiency has boosted up the performance of the counter electrode used in traditional DSSC (as platinum) and PSC (as carbon) devices by ∼48 and ∼29%, respectively. The electrochemical impedance and incident photon to current conversion efficiency (IPCE) studies were also analyzed in order to understand the catalytic behavior of the WO3 interfacial CBL for both DSSCs and PSCs, respectively. The much higher surface area of WO3 enables rapid electron hopping mechanism, which further benefits for higher electron mobility, resulting in higher short circuit current. Through this study, we were able to unequivocally establish the importance of buffer layer incorporation, which can further help to integrate the DSSC and PSC devices toward more stable, reliable, and enhanced efficiency-generating devices. In spite of this, using WO3 constitutes an important step toward the efficiency improvement of the devices for futuristic photoelectrochromic or self-powered switchable glazing for low-energy adaptive building integration. ■ INTRODUCTION In the recent past, there has been a great impetus to develop and design functional materials for energy-harvesting applications in the field of dye-sensitized solar cells (DSSCs) and perovskite solar cells (PSCs). While researching new ways to increase the efficiency of utilizing solar energy, the technology is constantly being developed into new and better-advanced products. DSSCs stand out as one of the front runners in view of the basic novelty of the concept derived from nature’s principles, which allows facile and costeffective processing alternatives. PSCs are thus considered as an advancement over DSSCs, where the photoanode thickness is reduced to a few microns and the dye is replaced by a superior light-absorbing organometal halide perovskite supported by a hole-transporting layer, which can be deposited directly over this cell architecture. Among the research and development versatility of the third-generation solar cell photovoltaic (PV) area, PSCs are the best technology developed during the last few years that promise a cheaper and more efficient alternative to the existing technologies for converting light to electric power. Implementation of various structures and fabrication strategies such as nanostructured materials (core−shell, mesoporous, one-dimensional, composite), advanced light-harvesting materials (quantum dots, NIR sensitizers), solid-state electrolytes, elimination of hole-transport materials (HTM), flexible substrates, scattering layers, up-conversion materials, and encapsulation methods, therefore, is envisaged as a potential effort to achieve prolonged stability and high efficiency in DSSC and PSC devices. The motive of all these strategies is efficient collection and extraction of charges. Additionally, introducing a buffer layer can provide enormous support, depending on where it has been used in the device. In general, a cathode buffer layer (CBL) excels as a supportive or optional layer for the cathode or back electrode material, which protects the cell from air and moisture and hence induces enhanced performReceived: April 1, 2020 Revised: June 10, 2020 Published: June 11, 2020 Article pubs.acs.org/JPCA © 2020 American Chemical Society 5709 https://dx.doi.org/10.1021/acs.jpca.0c02912 J. Phys. Chem. A 2020, 124, 5709−5719 This is an open access article published under a Creative Commons Attribution (CC-BY) License, which permits unrestricted use, distribution and reproduction in any medium, provided the author and source are cited. D ow nl oa de d vi a U N IV O F E X E T E R o n A ug us t 2 0, 2 02 0 at 1 2: 47 :1 5 (U T C ). Se e ht tp s: //p ub s. ac s. or g/ sh ar in gg ui de lin es f or o pt io ns o n ho w to le gi tim at el y sh ar e pu bl is he d ar tic le s. ance and stability of the device. Besides, as a protective layer, CBL employment further catalyzes the electron−hole separation too. Similarly, the primary function of an anode buffer layer in a heterojunction is to form a junction with the absorber layer while admitting a maximum amount of light to the junction region and absorber layer. Several commonly used CBLs were first established for organic light-emitting diodes and then transferred to organic solar cells (OSCs). Traditionally, CBL imparts as an inert, solution-processable substitute of calcium oxide; it is a hot spot in the field of OSCs and has achieved a significant efficiency of 9.2% in this field. In this regard, CBL can no more be considered as an optional layer but is an essential layer for achieving maximum performance in OSCs. In OSCs, there are also several studies on the application of fullerene derivatives as buffer layers. Liang et al. (2015) showed an enhanced efficiency of 15.44% using the fullerene buffer layer for CH3NH3PbI3-based PSCs, 23 whereas using TiO2 buffer layer, there is a significant and exceeds the enhancements of ∼42% efficiency has been achieved by Lungu et al. (2019), for the DSSC device. The effective hole-blocking role of the TiOx layer in the poly(3-hexylthiophene) (P3HT) and phenyl-C61butyric acid methyl ester (PC61BM) system was investigated, and significant improvements in the efficiency were obtained: from 2.3 to 5.0% and from 1.4 to 4.0%, respectively. The implementation of CBL for DSSCor PSC-based devices has been overlooked because of other predominatant alternative strategies, as a result, realization of CBL’s performance further opens up ample scope for further research. Lithium fluoride (LiF)-based CBL in PSCs has been demonstrated to enhance the device performance through the formation of an interfacial dipole at the cathode interface. However, ZnO is the most extensively investigated one for CBLs among the metal oxidebased CBL for the inverted PSCs, mainly because of its suitable energy levels, high electron mobility, good transparency, environmental stability, and low cost. Recently, Rajbhandari et al. (2020) reported the effect of three different oxide CBLs, such as Al2O3, ZnO, and TiO2, deposited by atomic layer deposition method in PSCs. Tran et al. (2016) demonstrated about the implementation of SnO2 nanoparticles as a CBL for an inverted OSC, which revealed excellent longterm device stability, and the photoconversion efficiency (PCE) was retained as ∼95% (2.89%) of its initial value after 10 weeks in ambient air. Using MoOx as a CBL could also be able to improve the efficiency from 1.3 to 5.52% for an all-inorganic CsPbIBr2-based PSC. 31 Aside from rapid progress in the field of PSCs, originated from a variety of device configurations, metal-oxide buffer layer implementation is still a less-explored research interest. Along with this, mesoporous characteristics in the devices stand out as a skillful technique, leading to heightened efficiency because of greater availability of active sites in the substrate. The effect of mesoporosity has been demonstrated not only for the metal oxide scaffold but also for the counter electrode. The redox functionality of the devices is intimately related to the dynamics of charge transfer (CT)/electron transport across the mesoporous layer, which, in turn, is strongly influenced by the nanostructured morphology of the film. In this work, a biocompatible, water-soluble dextran has been employed as a soft templating agent for porosity improvement in the WO3 structure. Dextran has been found to be a promising soft templating agent for metallic sponge and high-surface-area developing and structure-directing synthesis. WO3 is an extensively studied potential candidate for PV devices. With a view to enhance the PCE, research focused on the exploration of WO3 as a photoanode material retarding the efficiency because of high band gap and less electron mobility than that of DSSCs based on the other semiconducting metal oxides, such as TiO2 and ZnO, without employing any surface modification. On the other hand, efforts to replace organic polymers with WO3, as a hole injection layer (HTL) in PV cells leads to exhibit higher device performance, which are currently receiving increased attention. However, the methods such as sputtering, thermal evaporation, and pulsed laser deposition, involve depositing the WO3 layer, which is incompatible with a low-cost and solution-processed layer for future scalable manufacturing. In this study, our target is to integrate the mesoporous WO3 as an n-type metal oxide-based CBL for DSSC and PSC devices. To achieve a high surface area and considerable mesoporosity features of WO3, a dextran-templating hydrothermal synthesis process has been attempted. The synthesized WO3 has been further characterized by various physiochemical techniques and finally introduced as a CBL for DSSC and PSC devices. Further, our subsequent agenda is to imply this WO3-based PV devices for electrochromic (EC) switchable glazing applications. The glazing has the ability to tune its color in the presence of applied direct current electric power. By changing the transparency, an EC switchable window attenuates glare and undesired solar gains, which provides improved indoor comfort. ■ MATERIALS AND METHODS Materials. Sodium tungstate (Na2WO4·2H2O), HNO3, and HCl were purchased from Merck, and dextran (molecular weight ca. 75,000) was acquired from Alfa Aesar, Ward Hill, MA. Isopropanol and propylene glycol were purchased from Sigma-Aldrich, UK. All the chemicals were used without any further purification. Synthesis of Porous WO3. Hierarchically porous WO3 assemblies were synthesized by the hydrothermal method. In a typical experiment, 1.98 g of sodium tungstate (Na2WO4· 2H2O) was first dissolved in 40 mL of distilled water, and then, 1.5 mL of HNO3 aqueous solution was added dropwise into the above solution under constant stirring to prepare tungstic acid H2WO4. H2O (8 mol L −1, pH 2). To this solution, 10 g of dextran has been added to further prepare an aqueous viscous solution to control the growth, leading to porous textures at room temperature. To allow the dextran-formatting network throughout the aqueous solution of sodium tungstate for texture formation and further porosity development, the entire viscous solution was further kept completely under an undisturbed condition overnight. After this, the solution was transferred into a 50 mL Teflon-lined autoclave for the hydrothermal treatment at 160 °C for 24 h. The yellowish product as obtained from the hydrothermal reaction was accordingly centrifuged at 10,000 rpm and successively washed with 0.1 M HCl to neutralize the solution pH to 7. Besides, HCl treatment further helps to remove any unreacted Na ion also. The collected product was dried under an infrared lamp at ∼80 °C for 12 h followed by calcination at 300 °C for a period of 1 h to get the final product. The same synthesis process has been carried out without dextran to understand the role of dextran in controlling the morphology and porosity. The overall synthesis scheme has been represented in Figure 1. In this synthesis, dextran excels as a porosity developer and The Journal of Physical Chemistry A pubs.acs.org/JPCA Article https://dx.doi.org/10.1021/acs.jpca.0c02912 J. Phys. Chem. A 2020, 124, 5709−5719 5710 morphology-texturing agent, which is required to obtain a high surface area of WO3. It is anticipated that the dextranformatting network behaves like a mould where the WO3 precursor was attached. During the hydrothermal treatment, the removal of dextran further leads to the creation of pores on the particles. Fabrication of DSSCs. Pieces with an area of 2 × 2 cm of transparent conducting oxide [fluorine-doped tin oxide (FTO)] glass substrates were first ultrasonicated in deionized water and ethanol. Next, they were subjected to surface cleaning by using an UV-Ozone cleaner (Ossila UV Ozone Cleaner, UK) for 15 min to remove any dirt, patches on the surface of the FTO glasses. Next, a layer of transparent TiO2 paste (18NR-T, GreatCell Solar) was allowed for screen printing (120 T mesh/inch) on the UV-ozone-treated FTO glasses and dried for 6 min on a hotplate at 125 °C. After this, another layer of a similar transparent TiO2 was deposited by the same method. The screen-printed active area of the working electrode was selected as 0.28 cm. Then, a final layer of opaque TiO2 paste (18NR-O, GreatCell Solar) was printed, which provides an approximate thickness of 10−12 μm for the TiO2 electrode. The prepared electrodes were heated on a hotplate at 450 °C for 30 min to allow the crystallization of the TiO2 layer and to remove the binders available in the paste. After cooling down to 50 °C, the TiO2 electrodes were immersed in a 0.1 mM ruthenium 535 bis-TBA (N719) dye from GreatCell Solar in a 1:1 acetonitrile and ethanol solution at room temperature for 24 h to assure complete uptake of sensitizer. The Pt CE (Dyesol) was placed on a spin coating unit (SPIN 1200D, MIDAS System). A volume of 10 wt % (weight/volume) of a colloidal suspension of WO3 was prepared in a 2:1 isopropanol and propylene glycol medium. A volume of 1 mL of prepared WO3 colloidal solution was then spin-coated on the top of the FTO substrate at a speed of 3000 rpm for 30 s and dried at 200 °C for 30 min. Finally, the Platisol (T/SP, Solaronix ref. 41250) paste was drop-casted over the WO3 film and further calcined at 450 °C for 10 min to deposit a layer of Pt. The similar process has also been followed for nonporous WO3 layer deposition. The fabricated WO3−Pt CE was further placed over the dye-sensitized photoanode, and a plastic spacer was kept in between these two electrodes. The prepared redox electrolyte of I/I3 − was further infiltrated into the device and allowed for final testing. A scheme of the WO3 buffer layer-based DSSC device is given in Figure 2a. Fabrication of PSCs. The PSC devices were fabricated as per our previous report. However, for this case, 1 mL of the prepared WO3 colloidal solution, as used for DSSC fabrication, was spin-coated at a speed of 3000 rpm for 30 s on the top of Al2O3 layer deposition and further heated at 200 °C for 30 min during the PSC fabrication. The similar process has been also followed for nonporous WO3 layer deposition as well. The mesoscopic carbon layer was finally screen-printed with the asprepared carbon paste and sintered at 450 °C for 30 min. Finally, the perovskite precursor solution (CH3NH3PbI3) with an appropriate amount was infiltrated by drop-casting via the top of the carbon counter electrode and further spin-coating at a speed of 1000 rpm for kept at 50 °C for 1 h and finally used for further studies. The active area of the working electrode was selected as 0.25 cm. A schematic configuration of the WO3−carbon-based PSC is described in Figure 2b. Materials Characterization. The structural properties of the as-prepared and calcined powders are characterized by using a Bruker D8 Advance X-ray diffractometer (Billerica, MA, USA). The morphology of the synthesized WO3 was characterized using a high-resolution scanning electron microscope TESCAN VEGA3 SEM coupled with energy-dispersive spectroscopy (EDS; Oxford instrument elemental analysis) and a JEOL 2100 200 KV transmission electron microscope (TEM) providing a point resolution of 0.25 nm and a lattice resolution of 0.14 nm. The fluorescence spectrum was recorded at 22 ± 0.5 °C in 10 mm path length quartz cuvettes using a Quanta Master 400 unit (Horiba PTI, Canada) with a 150 W xenon lamp controlled by FelixGX spectroscopy software provided with the instrument. The X-ray photoelectron spectroscopy (XPS) measurements were carried out on a PHI 5000 Versa probe II scanning XPS microprobe (ULVAC-PHI, U.S.). The measurement was performed at room temperature at a base pressure higher than 6 × 10−10 mbar. All spectra were recorded with monochromatic Al Kα (hν = 1486.6 eV) radiation with a total resolution of about 0.7 eV and a beam size of 100 μm. Nitrogen physisorption measurements of all the samples were carried out by using a Quantachrome (iQ3) instrument after evacuation at 150 °C for 4 h. The specific surface area was calculated by the Brunauer−Emmett−Teller (BET) method, whereas desorption cumulative pore volume and pore size distribution were calculated by the Barrett−Joyner−Halenda (BJH) method. Device Testing. The photovoltaic measurements of both DSSC and PSC devices were performed under 1000 W/m of light from a Wacom AAA continuous solar simulator (WXS210S-20, AM1.5 G). The I−V characteristic of the devices was recorded using an EKO MP-160i I-V Tracer. Electrochemical impedance spectroscopy (EIS) measurements were carried out using an AUTOLAB frequency analyzer setup equipped with an AUTOLAB PGSTAT 10 and a Frequency Response Analyzer (FRA) Module. The measurements were performed under the same solar simulator conditions with a frequency Figure 1. Schematic representation of the formation of hierarchically porous WO3 assembly by the hydrothermal reaction. Figure 2. Schematic diagram of the WO3 buffer layer in (a) DSSCs and (b) PSCs, respectively. The Journal of Physical Chemistry A pubs.acs.org/JPCA Article https://dx.doi.org/10.1021/acs.jpca.0c02912 J. Phys. Chem. A 2020, 124, 5709−5719 5711 range from 0.1 Hz to 100 kHz. All the devices were measured at their corresponding open-circuit voltage. Incident photon to current efficiency (IPCE) was carried out using a Bentham PVE300 Photovoltaic EQE (IPCE) and IQE solution under 300−800 nm wavelength using a tungsten halogen lamp source. ■ RESULTS AND DISCUSSION Structural, Optical, and Surface Area Analysis of WO3. X-ray diffraction pattern, as shown in Figure 3, reveals the polycrystalline monoclinic WO3 structure as it matches well with the reference diffraction pattern of WO3 (JCPDS card number 83−0950). The main diffraction peaks at 23.1, 23.7, and 24.4 ascribed to Miller indices (002), (020), and (200), respectively, correspond to the monoclinic WO3 phase. This also rationalizes the synthesized WO3 free of orthorhombic WO3 and WO3·1/3H2O phases showing that the products are pure phase. High and sharp peak intensities indicate an excellent crystallinity of the sample. It is interesting to mention here that the orthorhombic phase of WO3 or WO3· 1/3H2O is quite often to be obtained via the hydrothermal treatment of tungstic acid H2WO4. H2O, although upon further heating >300 °C, all these metastable oxides lead to the stable monoclinic WO3 phase. 39,40 Here, through this dextrantemplating synthesis, the WO3 attained the monoclinic phase directly. In order to ascertain the effect of dextran addition on porosity improvement, the TEM investigation was performed with both the dextran-added and without dextran samples of WO3, respectively. Figure 4a,b indicate the TEM bright-field images of WO3 in the absence of dextran. It has been clearly observed that there are, as such, no porous characteristics formed on the WO3 surface. Moreover, without dextran employment, it leads to form a densely populated large hierarchical structure, whereas the addition of dextran assists to render the porous nature throughout the surface of WO3, as shown in the TEM bright-field images of Figure 4c,d at different magnifications. Besides, dextran addition influences to attain the morphology of WO3 toward a separated hierarchical platelet-like structure. During the hydrothermal treatment, the chainlike structure of the dextran preferentially combines with the hydroxide nanocrystals of WO3 anisotropically as building blocks thereby able to create patterned slits, which also facilitates the etching of the sites without the adsorption of dextran. This may cause the pore formation of WO3, allowing higher surface energy. In addition, Figure 4e shows the nitrogen adsorption−desorption isotherm of the WO3 sample in the absence of dextran and appeared as type III isotherm. It exhibits a BET surface area of 30.28 m g−1, involving no mesoporous characteristics, whereas the isotherm indicated mesoporous characteristic of the dextran-assisted WO3 samples and appeared as type IV isotherm as reported in Figure 4f. The appearance of type H-3 hysteresis indicating a stacked-plane (slit)-like pore shape. This exhibits comparatively a high surface area of 110.68 m g−1. To the best of our knowledge, Figure 3. X-ray diffraction pattern of the synthesized monoclinic WO3 sample. Figure 4. TEM bright-field images of (a,b) without dextran; (c,d) dextran-added WO3 samples at different magnifications, respectively; nitrogen adsorption−desorption isotherm plot (e) in the absence of dextran and (f) dextran-added WO3 samples (inset: corresponding pore size distribution plot), respectively. The Journal of Physical Chemistry A pubs.acs.org/JPCA Article https://dx.doi.org/10.1021/acs.jpca.0c02912 J. Phys. Chem. A 2020, 124, 5709−5719 5712 the obtained surface area of this WO3 is quite higher compared to the other similar reports of WO3 as shown in Table S1. 42−51 The pore diameter was found to be about 3.7 nm using the BJH method adsorption method (Figure S2). This value was artifice (pseudo peak) calculated from the BJH computation method. Further, the pore diameter was found to be about 20.2 nm using the BJH desorption method, which was further supported by TEM investigation, as shown in Figure 4f (inset). The corresponding pore volume was found to be 0.46 cc g−1. Again, stacked-plane pore geometry throughout the plate-like microstructural framework was fairly visible under a transmission electron microscope evidencing H3 hysteresis caused by the tensile strength effect during the gas condensation process and therefore resulted in a monomodal pore-size distribution as obtained from the BJH desorption method. Therefore, considering the high surface area and mesoporous characteristics of the WO3 sample, achieved by dextran addition, was selected for further studies in this manuscript. To further evaluate the morphology formation of the porous WO3 sample, the various microstructural analysis was investigated, as shown in Figure 5. During the gel formation, the dextran-added precursor solution creates amorphous networking of the sol as obtained from Figure 5a. The gel collapses into particles that precipitate from the solution during the hydrothermal treatment. This may look like the precursor solution is getting embedded inside the dextran polymeric network. On the other hand, Figure 5b,c exhibit the SEM images of the synthesized WO3 sample at different magnifications, respectively. The morphology exhibits as the hierarchical platelet-type shapes as obtained by the hydrothermal synthesis. Besides, the TEM bright-field images further reveal that the platelet-type morphology contains a wide range of size of 80−300 nm, having a predominate order of large distinct porosity (Figure 5d,e). The lattice fringes in highresolution TEM (HRTEM) image (Figure 5f) corresponds to the (020) reflection with an inter planner spacing of 0.379 nm, which confirms the single-crystalline nature of the monoclinic WO3 phase. The EDS analysis was also performed to confirm the elements of WO3 and the complete removal of Na, as shown in Figure S1. X-ray Photoemission Spectroscopic and Photoluminescence Analysis. The valence states of the fundamental elements W and O of the synthesized mesoporous WO3 have been identified through XPS measurements (Figure 6). The survey spectrum of the WO3 is shown in Figure 6a. First, two characteristic peaks at 36.0 ± 0.2 eV (Figure 6b) has been ascribed to the typical doublet of W with the binding energy of W (4f7/2 and 4f5/2). The peak ratio of W (4f7/2) to W (4f5/2) was found to be 4:3 with a spin−orbit splitting of a doublet value of 2.12 eV. A third broad peak of W (5p3/2) locates at ∼41.8 eV. This result reveals only W for its +6 oxidation state. However, the asymmetry of the W 4f7/2 indicates that WO3 readily becomes oxygen-deficient to form WO3−x, with the variable oxygen composition parameter x. On deconvolution, the asymmetric O 1s XPS spectrum evolves into two components centered at 529.21 (O1) and 530.65 eV (O2) as shown in Figure 6c. The dominant O1 component is a characteristic of O2− ions, whereas the O2 component at the higher binding side may be associated with the presence of chemisorbed oxygen in the grain boundaries or the existence of the hydroxyl group (−OH) because of the adsorption of water onto the surface of the WO3 during synthesis through the aqueous medium route. This oxygen deficiency greatly Figure 5. FESEM microstructural images of (a) the as-prepared sol, (b,c) mesoporous WO3 at different magnifications, respectively; (d,e) TEM bright-field images of mesoporous WO3 at different magnifications, respectively and (f) the corresponding HRTEM image. Figure 6. (a) XPS survey spectrum of the mesoporous WO3 sample, (b) XPS of spin−orbit split peaks of W 4f, and (c) spectral decomposition of the O 1s spectrum results in two different types of oxygen environment denoted as the O1and O2-fitted component in the WO3. The Journal of Physical Chemistry A pubs.acs.org/JPCA Article https://dx.doi.org/10.1021/acs.jpca.0c02912 J. Phys. Chem. A 2020, 124, 5709−5719 5713 influences the bulk of the electronic transport properties by introducing donor electronic states. Further, the photoluminescence (PL) spectrum of the aqueous dispersion of the WO3 sample was monitored at room temperature, as shown in Figure S3. A blue emission at 467 nm was observed against an excitation wavelength of 350 nm. It is anticipated that the reason for the visible emission exhibited by WO3 could be due to the presence of oxygenrelated defects as observed from the XPS analysis. The aqueous solution of synthesized WO3 resulted in a quantum yield (QY) of 3.25% at room temperature. Microstructural and Optical Measurement of WO3 Films. The SEM line-scale mapping was recorded on the WO3−Pt film, as shown in Figure S4a. This indicates a homogeneous distribution of W and O throughout the Pt layer. Further, the distribution of W and O over the Pt layer is very clear from the elemental mapping diagram, as shown in Figure S4b. Besides, in the case of WO3−carbon-based PSC film, a thin layer (<200 nm) of mesoporous WO3 has been formed as shown in Figure S4c, and the corresponding elemental mapping has been given in Figure S4d. The EDS result indicates that despite the different approach of the buffer layer formation for DSSCs and PSCs, both of them showed a proper distribution of W and O along with their respective back electrode material, indicating that the formulation of WO3 synthesis is precise and efficient in preparing the welldispersed sample. PV Performance of WO3 as a Cathode Buffer Layer in DSSCs. In order to evaluate the effect of mesoporous WO3 as a CBL in PV applications, the PV performances have been further investigated. An evaluation of the DSSC performances of the fabricated WO3-based devices were executed under 1SUN 1.5 AM (Figure 7). The performances of the J−V characteristic plot for the DSSC devices with Pt CE and WO3− Pt CE are shown in Figure 7a. The WO3−Pt device exhibited a maximum PCE of 7.43% with a short circuit current (JSC) of 16.25 mA cm−2, an open-circuit voltage (VOC) of 0.75 V, and a fill factor (FF) of 0.59. On the contrary, with only Pt CE-based devices exhibited a PCE of 5.01% with a JSC of 11.49 mA cm −2, a VOC of 0.72 V, and a FF of 0.64. The digital image of WO3− Pt-based DSSCs has been given in the inset of Figure 7a. The effect of mesoporous nature also influences the efficiency of the device. In the absence of the mesoporous WO3 layer, the maximum average PCE was achieved to be ∼6.12% (Figure S5a), which is ∼21% lesser than compared to mesoporous WO3 layer-added devices. These results strongly support the benefits of porous layer incorporation in such a DSSC device. Further, the recorded J−V characteristics parameters are compared in Table 1. In spite of this, using WO3 on Pt CE, the corresponding device exhibits ∼48% higher efficiency than only Pt CE. The enhancement of PCE was resulted by keeping insignificant alteration in VOC except for the JSC and FF, which signifies the presence of mesoporous WO3 gets an advantage as a catalyst for Pt CE and, therefore, accelerate the rate of high electron mobility and hole-extraction leading toward high Figure 7. (a) J−V characteristic and (b) corresponding Nyquist plot (inset: corresponding equivalent circuit) for Pt and WO3−Pt-based DSSC devices, respectively. Table 1. PV and Impedance Spectroscopic Parameters for Different DSSC Devices cell JSC (mA cm −2) VOC (mV) FF PCE ± 0.05 (%) RS (Ω cm−2) RCT1 (Ω cm−2) Cp (F) without WO3 11.49 722 0.59 5.01 18.52 32.68 1.52 × 10−6 Pt−WO3 (nonporous) 14.68 753 0.55 6.12 18.65 26.86 3.57 × 10−4 Pt−WO3 (mesoporous) 16.55 753 0.64 7.43 18.23 27.12 2.48 × 10−3 Figure 8. (a) J−V characteristic and (b) corresponding Nyquist plot (inset: corresponding equivalent circuit) for carbon and WO3−carbon-based PSC devices, respectively. The Journal of Physical Chemistry A pubs.acs.org/JPCA Article https://dx.doi.org/10.1021/acs.jpca.0c02912 J. Phys. Chem. A 2020, 124, 5709−5719 5714 efficiency. The EIS measurements were carried out to understand the transport properties at different interfaces in the DSSC assembly. The EIS spectra (Nyquist plot) and the corresponding equivalent circuit diagram (inset) of cells for the DSSC devices, as shown in Figure 7b. As shown in Table 1, the DSSCs show almost closure of sheet resistance (Rs), which is due to the use of the same Pt counter electrodes and I/I3 − electrolyte in all the devices. On the other hand, resistance at the Pt/electrolyte interface has been found significantly higher for the only Pt CE device compared to WO3−Pt-based devices. The CT resistance of 32.68 Ω cm−2 at the electrolyte/dye/ TiO2 interface for the Pt CE device is considerably higher than the WO3−Pt device (27.12 Ω cm−2). This result clearly shows that the catalytic activity of WO3 reduces the CT resistance of Pt producing higher efficiency. The EIS spectrum of the without mesoporous WO3-based DSSC device has been given in Figure S5a (inset), and the corresponding data are shown in Table 1. PV Performance of WO3 as a Cathode Buffer Layer in PSCs In addition, incorporation of the synthesized WO3 as a cathode buffer layer, it was further evaluated in the case of PSC devices. The corresponding J−V characteristics plot is shown in Figure 8a. It was noted that the addition of WO3−carbon back contact exhibited the highest PCE of 12.42%, which is ∼17 and ∼29% higher than the without mesoporous WO3 (PCE: ∼10.58%) and only carbon-based devices (∼9.65%), respectively. The J−V characteristics plot for the without mesoporous WO3-added device has been given in Figure S5b. Interestingly, the VOC does not affect the incorporation of the WO3 layer, and for both the cases, it shows quite high value as 0.9 V. The addition of WO3 significantly boosts up the JSC of the WO3−carbon-based device to 23.16 mA cm−2, which is quite high and impressive to generate higher efficiency. The digital image of the WO3−carbon-based PSC has been given in the inset of Figure 8a. The Nyquist plots derived from the impedance data, Figure 8b, revealed a clear difference between the two devices. The WO3−carbon-based PSC shows the smallest radius of the semicircle in the high-frequency region, which indicates the lowest CT resistance at the CH3NH3PbI3/carbon interface. This result implies that the film addition of WO3 plays an important role in DSSC electron recombination. All the measured parameters have been summarized in Table 2. RS of PSCs with WO3 is less than the device without WO3, which suggests that the hole-extraction process at the carbon electrode with WO3 is more efficient. Moreover, a lower charge exchange resistances (RCT) at a bias of −0.90 V was observed for WO3-additive PSCs compared to the devices without additive. However, the WO3 additive layer may play the role of a hole collector along with the appropriate energy level to the perovskite CH3NH3PbI3 sensitizer and lead to an acceleration of the hole extraction efficiency. The WO3 layer can also significantly reduce the series resistance and significantly improve the JSC of the PSC device. Besides, the EIS measurement was also performed for the absence of mesoporous WO3-based PSC devices, as shown in Figure S5b (inset), and the corresponding data are summarized in Table 2. Probable Mechanism of the WO3 Buffer Layer in Solar Cell Devices. The efficiency was measured for 10 numbers of cells for DSSCs and PSCs with an error bar plot, as shown in Figure 9a. This result implies that the film addition of WO3 plays an important role in DSSC electron recombination. Therefore, we suggest an apparent mechanism to understand the effect of the porous WO3 structure as an interfacial cathode buffer layer on the improved crystalline and enhanced PCEs of the solar cell devices. WO3 treatment may also play a similar role in affecting the electronic structure of carbon and can be explored as a hole-extracting layer for PSCs. It is assumed that a small polaron hopping conduction mechanism can be influenced to operate in WO3 as a cathode buffer layer for solar cell devices. This mechanism possibly suggests hopping can be a faster way compared to the traditional way of hole diffusion with an extremely fast rate constant that leads to exhibit enhanced device performance. Besides, the photoluminescence spectrum of WO3 predicts partial oxygen vacancies in substoichiometric WO3−x which creates localized defect states: W or W and W. These mixed valance states may further promote the CT and enhance electrical conductivity. The high surface area facilitates to collect and regenerate a higher number of holes, which is ultimately Table 2. PV and Impedance Spectroscopic Parameters for Different PSC Devices cell JSC (mA cm −2) VOC (mV) FF PCE ±0.05 (%) RS (Ω cm−2) RCT1 (Ω cm−2) Cp (F) without WO3 16.82 900 0.59 9.65 32.25 227.3 48.59 × 10−9 carbon−WO3 (nonporous) 20.15 891 0.59 10.58 36.92 225.7 32.67 × 10−7 carbon−WO3 (mesoporous) 24.21 900 0.60 12.42 20.92 102.2 16.32 × 10−6 Figure 9. (a) Efficiency measurement error plot for different devices and a schematic representation of (b) in the absence and (c) in the presence of highly porous WO3 cathode buffer layer for solar cell devices provides some added beneficial features such as hole hopping diffusion, restriction to water vapor/moisture, and less corrosion of the counter electrode. The Journal of Physical Chemistry A pubs.acs.org/JPCA Article https://dx.doi.org/10.1021/acs.jpca.0c02912 J. Phys. Chem. A 2020, 124, 5709−5719 5715 beneficial for rapid and smooth electron−hole mobility for the devices. At the same time, it is also assumed that because of distinct porous nature of WO3, that it may be suitable to entrap/adsorb any moisture or oxygen molecule, which may retard the overall performance of the device. Pt counter electrodes are usually affected by high corrosion because of the use of liquid I/I3 − electrolyte, which can retard the hole conduction process, although using the WO3 buffer layer may restrict the rate of the corrosion and facilitate the hole conduction process for the device. Therefore, the high porosity could be beneficial for the WO3 and favor as a buffer layer for the solar cell devices. The overall concept has been schematically illustrated in Figure 9b,c. The enhanced PCE has been mainly influenced by the enhancement of the JSC factor of the device. The increased JSC of the porous WO3 probably resulted in a thicker active layer, which is helpful to improve optical-trapping properties. Incident Photon to Current Conversion Efficiency Measurements. IPCE stands for the external quantum efficiency of the DSSC and PSC devices including the effects of optical losses caused by transmission and reflection. The IPCE measurements were also performed as a function of wavelength by comparing the WO3−Pt and Pt-based devices for DSSCs and WO3−carbon and carbon-based devices for PSCs as shown in Figure 10. The IPCE curve for all the DSSCs shows a broad peak over the range of 300−800 nm with a maximum value of ∼83% at 536 nm for the WO3−Pt-based device, whereas the Pt-based DSSC exhibits an IPCE value of 61% at the same wavelength as shown in Figure 10a. In the case of PSCs, incorporation of the WO3 layer exhibits predominant incident photon to current conversion efficiency (IPCE) value (∼90% at 400 nm and 76% at 560 nm wavelength) than the without WO3 layer device value (∼76% at 400 nm and 60% at 560 nm wavelength) as shown in Figure 10b. This also strongly recommends that the incorporation of a WO3 layer improves the IPCE in a great manner to minimize the electron recombination or electron entrapment within surface trap states or electron loss within the hole transporting section of the device. Therefore, the EIS and IPCE results predict that the CT process is taking place at the interface of Pt/carbon and WO3, leading to an acceleration of the hole mobility, and as a result, the short circuit current increases remarkably. This is quite competitive with organic interfacial materials where it cannot be readily achieved with the later. All the above-discussed results and the fundamental insights could be useful in the development of achieving solution-processed successful incorporation of the WO3 cathode buffer layer for highefficiency devices. Although the use of WO3 presented here are far from being fully optimized, but the profound advantages along with ease of preparation and applicability for potential performance make it a good candidate for the cathode buffer layer for solar cell devices. This work also provides insights into the design of the interfacial hole transport layer for the fabrication of solution-processed solar cell devices. The obtained results are quite encouraging and open up ample scope to develop new and economically viable synthetic approaches to enhance the performance of the solution-processed solar cell devices. In spite of this, there are very few reports available on the buffer layer-incorporated DSSC and PSC devices. In comparison to the reported results on the performance of various materials as incorporated buffer layer in solar cell devices (Table 3), the results obtained in this work are quite competitive and interesting to execute dextran template mesoporous WO3 as a promising buffer layer candidate for both DSSC and PSC devices, respectively. Figure 10. IPCE spectra for (a) WO3−Pt and Pt-based DSSC and (b) WO3−carbonand carbon-based PSC devices, respectively. Table 3. Reported Performances of Different Buffer Layer Used for Various Types of Solar Cells CBL type of solar cell PCE (%) reference PAMPS-Na nonconjugated polymer/Al OSC 9.16 22 PCBM/bis-C60/Ag PSC 15.44 23 C60/bis-C60/Ag PSC 13.37 23 TiO2/Pt DSSC 2.84 24 CdSe QDs/LiF/AuAg inverted-PSC 15.1 27 SnO2/Ag inverted-OSC 2.89 30 MoOx/Au all inorganic PSC 5.52 31 ZnO-DOL/Ag ISC 3.48 62 Zr-isopropoxide/Al BHJ-PSC 8.07 63 PCBB-2CN-2C8/Au PSC 17.35 64 a-ZrAcac/Al poly-SC 8.75 65 PCBM, C60, and LiF/Al planar-PSC 14.69 66 ZnO/PEI/Al OSC, PSC 11.76 67


■ INTRODUCTION
In the recent past, there has been a great impetus to develop and design functional materials for energy-harvesting applications in the field of dye-sensitized solar cells (DSSCs) and perovskite solar cells (PSCs). 1−4 While researching new ways to increase the efficiency of utilizing solar energy, the technology is constantly being developed into new and better-advanced products. DSSCs stand out as one of the front runners in view of the basic novelty of the concept derived from nature's principles, which allows facile and costeffective processing alternatives. 5,6 PSCs are thus considered as an advancement over DSSCs, where the photoanode thickness is reduced to a few microns and the dye is replaced by a superior light-absorbing organometal halide perovskite supported by a hole-transporting layer, which can be deposited directly over this cell architecture. 7 Among the research and development versatility of the third-generation solar cell photovoltaic (PV) area, PSCs are the best technology developed during the last few years that promise a cheaper and more efficient alternative to the existing technologies for converting light to electric power. 8,9 Implementation of various structures and fabrication strategies such as nanostructured materials (core−shell, mesoporous, one-dimensional, composite), 10 advanced light-harvesting materials (quantum dots, NIR sensitizers), 11,12 solid-state electrolytes, 13 elimination of hole-transport materials (HTM), 14 flexible substrates, 15 scattering layers, 16 up-conversion materials, 17 and encapsulation methods, 18 therefore, is envisaged as a potential effort to achieve prolonged stability and high efficiency in DSSC and PSC devices. The motive of all these strategies is efficient collection and extraction of charges. Additionally, introducing a buffer layer can provide enormous support, depending on where it has been used in the device. In general, a cathode buffer layer (CBL) excels as a supportive or optional layer for the cathode or back electrode material, which protects the cell from air and moisture and hence induces enhanced perform-ance and stability of the device. 19 Besides, as a protective layer, CBL employment further catalyzes the electron−hole separation too. Similarly, the primary function of an anode buffer layer in a heterojunction is to form a junction with the absorber layer while admitting a maximum amount of light to the junction region and absorber layer. 20 Several commonly used CBLs were first established for organic light-emitting diodes and then transferred to organic solar cells (OSCs). 21 Traditionally, CBL imparts as an inert, solution-processable substitute of calcium oxide; it is a hot spot in the field of OSCs and has achieved a significant efficiency of 9.2% in this field. 22 In this regard, CBL can no more be considered as an optional layer but is an essential layer for achieving maximum performance in OSCs. In OSCs, there are also several studies on the application of fullerene derivatives as buffer layers.  showed an enhanced efficiency of 15.44% using the fullerene buffer layer for CH 3 NH 3 PbI 3 -based PSCs, 23 whereas using TiO 2 buffer layer, there is a significant and exceeds the enhancements of ∼42% efficiency has been achieved by Lungu et al. (2019), for the DSSC device. 24 The effective hole-blocking role of the TiO x layer in the poly(3-hexylthiophene) (P3HT) and phenyl-C 61butyric acid methyl ester (PC 61 BM) system was investigated, and significant improvements in the efficiency were obtained: from 2.3 to 5.0% and from 1.4 to 4.0%, respectively. 25,26 The implementation of CBL for DSSC-or PSC-based devices has been overlooked because of other predominatant alternative strategies, as a result, realization of CBL's performance further opens up ample scope for further research. Lithium fluoride (LiF)-based CBL in PSCs has been demonstrated to enhance the device performance through the formation of an interfacial dipole at the cathode interface. 27 However, ZnO is the most extensively investigated one for CBLs among the metal oxidebased CBL for the inverted PSCs, mainly because of its suitable energy levels, high electron mobility, good transparency, environmental stability, and low cost. 28 Recently, Rajbhandari et al. (2020) reported the effect of three different oxide CBLs, such as Al 2 O 3 , ZnO, and TiO 2 , deposited by atomic layer deposition method in PSCs. 29 Tran et al. (2016) demonstrated about the implementation of SnO 2 nanoparticles as a CBL for an inverted OSC, which revealed excellent longterm device stability, and the photoconversion efficiency (PCE) was retained as ∼95% (2.89%) of its initial value after 10 weeks in ambient air. 30 Using MoO x as a CBL could also be able to improve the efficiency from 1.3 to 5.52% for an all-inorganic CsPbIBr 2 -based PSC. 31 Aside from rapid progress in the field of PSCs, originated from a variety of device configurations, metal-oxide buffer layer implementation is still a less-explored research interest.
Along with this, mesoporous characteristics in the devices stand out as a skillful technique, leading to heightened efficiency because of greater availability of active sites in the substrate. The effect of mesoporosity has been demonstrated not only for the metal oxide scaffold but also for the counter electrode. 32 The redox functionality of the devices is intimately related to the dynamics of charge transfer (CT)/electron transport across the mesoporous layer, which, in turn, is strongly influenced by the nanostructured morphology of the film. 33 In this work, a biocompatible, water-soluble dextran has been employed as a soft templating agent for porosity improvement in the WO 3 structure. Dextran has been found to be a promising soft templating agent for metallic sponge and high-surface-area developing and structure-directing synthesis. WO 3 is an extensively studied potential candidate for PV devices. 34,35 With a view to enhance the PCE, research focused on the exploration of WO 3 as a photoanode material retarding the efficiency because of high band gap and less electron mobility than that of DSSCs based on the other semiconducting metal oxides, such as TiO 2 and ZnO, without employing any surface modification. 36 On the other hand, efforts to replace organic polymers with WO 3 , as a hole injection layer (HTL) in PV cells leads to exhibit higher device performance, which are currently receiving increased attention. However, the methods such as sputtering, thermal evaporation, and pulsed laser deposition, involve depositing the WO 3 layer, which is incompatible with a low-cost and solution-processed layer for future scalable manufacturing.
In this study, our target is to integrate the mesoporous WO 3 as an n-type metal oxide-based CBL for DSSC and PSC devices. To achieve a high surface area and considerable mesoporosity features of WO 3 , a dextran-templating hydrothermal synthesis process has been attempted. The synthesized WO 3 has been further characterized by various physiochemical techniques and finally introduced as a CBL for DSSC and PSC devices. Further, our subsequent agenda is to imply this WO 3 -based PV devices for electrochromic (EC) switchable glazing applications. The glazing has the ability to tune its color in the presence of applied direct current electric power. By changing the transparency, an EC switchable window attenuates glare and undesired solar gains, which provides improved indoor comfort.

■ MATERIALS AND METHODS
Materials. Sodium tungstate (Na 2 WO 4 ·2H 2 O), HNO 3 , and HCl were purchased from Merck, and dextran (molecular weight ca. 75,000) was acquired from Alfa Aesar, Ward Hill, MA. Isopropanol and propylene glycol were purchased from Sigma-Aldrich, UK. All the chemicals were used without any further purification.
Synthesis of Porous WO 3 . Hierarchically porous WO 3 assemblies were synthesized by the hydrothermal method. In a typical experiment, 1.98 g of sodium tungstate (Na 2 WO 4 · 2H 2 O) was first dissolved in 40 mL of distilled water, and then, 1.5 mL of HNO 3 aqueous solution was added dropwise into the above solution under constant stirring to prepare tungstic acid H 2 WO 4 . H 2 O (8 mol L −1 , pH 2). To this solution, 10 g of dextran has been added to further prepare an aqueous viscous solution to control the growth, leading to porous textures at room temperature. To allow the dextran-formatting network throughout the aqueous solution of sodium tungstate for texture formation and further porosity development, the entire viscous solution was further kept completely under an undisturbed condition overnight. After this, the solution was transferred into a 50 mL Teflon-lined autoclave for the hydrothermal treatment at 160°C for 24 h. The yellowish product as obtained from the hydrothermal reaction was accordingly centrifuged at 10,000 rpm and successively washed with 0.1 M HCl to neutralize the solution pH to 7. Besides, HCl treatment further helps to remove any unreacted Na + ion also. The collected product was dried under an infrared lamp at ∼80°C for 12 h followed by calcination at 300°C for a period of 1 h to get the final product. The same synthesis process has been carried out without dextran to understand the role of dextran in controlling the morphology and porosity. The overall synthesis scheme has been represented in Figure 1. In this synthesis, dextran excels as a porosity developer and The Journal of Physical Chemistry A pubs.acs.org/JPCA Article morphology-texturing agent, which is required to obtain a high surface area of WO 3 . It is anticipated that the dextranformatting network behaves like a mould where the WO 3 precursor was attached. During the hydrothermal treatment, the removal of dextran further leads to the creation of pores on the particles. 37 Fabrication of DSSCs. Pieces with an area of 2 × 2 cm 2 of transparent conducting oxide [fluorine-doped tin oxide (FTO)] glass substrates were first ultrasonicated in deionized water and ethanol. Next, they were subjected to surface cleaning by using an UV-Ozone cleaner (Ossila UV Ozone Cleaner, UK) for 15 min to remove any dirt, patches on the surface of the FTO glasses. Next, a layer of transparent TiO 2 paste (18NR-T, GreatCell Solar) was allowed for screen printing (120 T mesh/inch) on the UV-ozone-treated FTO glasses and dried for 6 min on a hotplate at 125°C. After this, another layer of a similar transparent TiO 2 was deposited by the same method. The screen-printed active area of the working electrode was selected as 0.28 cm 2 . Then, a final layer of opaque TiO 2 paste (18NR-O, GreatCell Solar) was printed, which provides an approximate thickness of 10−12 μm for the TiO 2 electrode. The prepared electrodes were heated on a hotplate at 450°C for 30 min to allow the crystallization of the TiO 2 layer and to remove the binders available in the paste. After cooling down to 50°C, the TiO 2 electrodes were immersed in a 0.1 mM ruthenium 535 bis-TBA (N719) dye from GreatCell Solar in a 1:1 acetonitrile and ethanol solution at room temperature for 24 h to assure complete uptake of sensitizer. 38 The Pt CE (Dyesol) was placed on a spin coating unit (SPIN 1200D, MIDAS System). A volume of 10 wt % (weight/volume) of a colloidal suspension of WO 3 was prepared in a 2:1 isopropanol and propylene glycol medium. A volume of 1 mL of prepared WO 3 colloidal solution was then spin-coated on the top of the FTO substrate at a speed of 3000 rpm for 30 s and dried at 200°C for 30 min. Finally, the Platisol (T/SP, Solaronix ref. 41250) paste was drop-casted over the WO 3 film and further calcined at 450°C for 10 min to deposit a layer of Pt. The similar process has also been followed for nonporous WO 3 layer deposition. The fabricated WO 3 −Pt CE was further placed over the dye-sensitized photoanode, and a plastic spacer was kept in between these two electrodes. The prepared redox electrolyte of I − /I 3 − was further infiltrated into the device and allowed for final testing. 37 A scheme of the WO 3 buffer layer-based DSSC device is given in Figure 2a.
Fabrication of PSCs. The PSC devices were fabricated as per our previous report. 35 However, for this case, 1 mL of the prepared WO 3 colloidal solution, as used for DSSC fabrication, was spin-coated at a speed of 3000 rpm for 30 s on the top of Al 2 O 3 layer deposition and further heated at 200°C for 30 min during the PSC fabrication. The similar process has been also followed for nonporous WO 3 layer deposition as well. The mesoscopic carbon layer was finally screen-printed with the asprepared carbon paste and sintered at 450°C for 30 min. Finally, the perovskite precursor solution (CH 3 NH 3 PbI 3 ) with an appropriate amount was infiltrated by drop-casting via the top of the carbon counter electrode and further spin-coating at a speed of 1000 rpm for kept at 50°C for 1 h and finally used for further studies. The active area of the working electrode was selected as 0.25 cm 2 . A schematic configuration of the WO 3 −carbon-based PSC is described in Figure 2b.
Materials Characterization. The structural properties of the as-prepared and calcined powders are characterized by using a Bruker D8 Advance X-ray diffractometer (Billerica, MA, USA). The morphology of the synthesized WO 3 was characterized using a high-resolution scanning electron microscope TESCAN VEGA3 SEM coupled with energy-dispersive spectroscopy (EDS; Oxford instrument elemental analysis) and a JEOL 2100 200 KV transmission electron microscope (TEM) providing a point resolution of 0.25 nm and a lattice resolution of 0.14 nm. The fluorescence spectrum was recorded at 22 ± 0.5°C in 10 mm path length quartz cuvettes using a Quanta Master 400 unit (Horiba PTI, Canada) with a 150 W xenon lamp controlled by FelixGX spectroscopy software provided with the instrument. The X-ray photoelectron spectroscopy (XPS) measurements were carried out on a PHI 5000 Versa probe II scanning XPS microprobe (ULVAC-PHI, U.S.). The measurement was performed at room temperature at a base pressure higher than 6 × 10 −10 mbar. All spectra were recorded with monochromatic Al Kα (hν = 1486.6 eV) radiation with a total resolution of about 0.7 eV and a beam size of 100 μm. Nitrogen physisorption measurements of all the samples were carried out by using a Quantachrome (iQ3) instrument after evacuation at 150°C for 4 h. The specific surface area was calculated by the Brunauer−Emmett−Teller (BET) method, whereas desorption cumulative pore volume and pore size distribution were calculated by the Barrett−Joyner−Halenda (BJH) method.
Device Testing. The photovoltaic measurements of both DSSC and PSC devices were performed under 1000 W/m 2 of light from a Wacom AAA continuous solar simulator (WXS-210S-20, AM1.5 G). The I−V characteristic of the devices was recorded using an EKO MP-160i I-V Tracer. Electrochemical impedance spectroscopy (EIS) measurements were carried out using an AUTOLAB frequency analyzer setup equipped with an AUTOLAB PGSTAT 10 and a Frequency Response Analyzer (FRA) Module. The measurements were performed under the same solar simulator conditions with a frequency   ■ RESULTS AND DISCUSSION Structural, Optical, and Surface Area Analysis of WO 3 . X-ray diffraction pattern, as shown in Figure 3, reveals the polycrystalline monoclinic WO 3 structure as it matches well with the reference diffraction pattern of WO 3 39,40 Here, through this dextrantemplating synthesis, the WO 3 attained the monoclinic phase directly.
In order to ascertain the effect of dextran addition on porosity improvement, the TEM investigation was performed with both the dextran-added and without dextran samples of WO 3 , respectively. Figure 4a,b indicate the TEM bright-field images of WO 3 in the absence of dextran. It has been clearly observed that there are, as such, no porous characteristics formed on the WO 3 surface. Moreover, without dextran employment, it leads to form a densely populated large hierarchical structure, whereas the addition of dextran assists to render the porous nature throughout the surface of WO 3, as shown in the TEM bright-field images of Figure 4c,d at different magnifications. Besides, dextran addition influences to attain the morphology of WO 3 toward a separated hierarchical platelet-like structure. During the hydrothermal treatment, the chainlike structure of the dextran preferentially combines with the hydroxide nanocrystals of WO 3 anisotropically as building blocks thereby able to create patterned slits, which also facilitates the etching of the sites without the adsorption of dextran. 37 This may cause the pore formation of WO 3, allowing higher surface energy. In addition, Figure 4e shows the nitrogen adsorption−desorption isotherm of the WO 3 sample in the absence of dextran and appeared as type III isotherm. It exhibits a BET surface area of 30.28 m 2 g −1 , involving no mesoporous characteristics, whereas the isotherm indicated mesoporous characteristic of the dextran-assisted WO 3 samples and appeared as type IV isotherm as reported in Figure 4f. The appearance of type H-3 hysteresis indicating a stacked-plane (slit)-like pore shape. 41 This exhibits comparatively a high surface area of 110.68 m 2 g −1 . To the best of our knowledge,   Table S1. 42−51 The pore diameter was found to be about 3.7 nm using the BJH method adsorption method ( Figure S2). This value was artifice (pseudo peak) calculated from the BJH computation method. Further, the pore diameter was found to be about 20.2 nm using the BJH desorption method, which was further supported by TEM investigation, as shown in Figure 4f (inset). The corresponding pore volume was found to be 0.46 cc g −1 .
Again, stacked-plane pore geometry throughout the plate-like microstructural framework was fairly visible under a transmission electron microscope evidencing H3 hysteresis caused by the tensile strength effect during the gas condensation process and therefore resulted in a monomodal pore-size distribution as obtained from the BJH desorption method. 52,53 Therefore, considering the high surface area and mesoporous characteristics of the WO 3 sample, achieved by dextran addition, was selected for further studies in this manuscript.
To further evaluate the morphology formation of the porous WO 3 sample, the various microstructural analysis was investigated, as shown in Figure 5. During the gel formation, the dextran-added precursor solution creates amorphous networking of the sol as obtained from Figure 5a. The gel collapses into particles that precipitate from the solution during the hydrothermal treatment. This may look like the precursor solution is getting embedded inside the dextran polymeric network. On the other hand, Figure 5b,c exhibit the SEM images of the synthesized WO 3 sample at different magnifications, respectively. The morphology exhibits as the hierarchical platelet-type shapes as obtained by the hydrothermal synthesis. Besides, the TEM bright-field images further reveal that the platelet-type morphology contains a wide range of size of 80−300 nm, having a predominate order of large distinct porosity (Figure 5d,e). The lattice fringes in highresolution TEM (HRTEM) image (Figure 5f) corresponds to the (020) reflection with an inter planner spacing of 0.379 nm, which confirms the single-crystalline nature of the monoclinic WO 3 phase. The EDS analysis was also performed to confirm the elements of WO 3 and the complete removal of Na, as shown in Figure S1.
X-ray Photoemission Spectroscopic and Photoluminescence Analysis. The valence states of the fundamental elements W and O of the synthesized mesoporous WO 3 have been identified through XPS measurements ( Figure 6). The survey spectrum of the WO 3 is shown in Figure 6a. First, two characteristic peaks at 36.0 ± 0.2 eV (Figure 6b) has been ascribed to the typical doublet of W 6+ with the binding energy of W (4f 7/2 and 4f 5/2 ). The peak ratio of W (4f 7/2 ) to W (4f 5/2 ) was found to be 4:3 with a spin−orbit splitting of a doublet value of 2.12 eV. A third broad peak of W (5p 3/2 ) locates at ∼41.8 eV. This result reveals only W for its +6 oxidation state. 54 However, the asymmetry of the W 4f 7/2 indicates that WO 3 readily becomes oxygen-deficient to form WO 3−x , with the variable oxygen composition parameter x. On deconvolution, the asymmetric O 1s XPS spectrum evolves into two components centered at 529.21 (O1) and 530.65 eV (O2) as shown in Figure 6c. The dominant O1 component is a characteristic of O 2− ions, whereas the O2 component at the higher binding side may be associated with the presence of chemisorbed oxygen in the grain boundaries or the existence of the hydroxyl group (−OH) because of the adsorption of water onto the surface of the WO 3 during synthesis through the aqueous medium route. This oxygen deficiency greatly  The Journal of Physical Chemistry A pubs.acs.org/JPCA Article influences the bulk of the electronic transport properties by introducing donor electronic states. 55 Further, the photoluminescence (PL) spectrum of the aqueous dispersion of the WO 3 sample was monitored at room temperature, as shown in Figure S3. A blue emission at 467 nm was observed against an excitation wavelength of 350 nm. It is anticipated that the reason for the visible emission exhibited by WO 3 could be due to the presence of oxygenrelated defects as observed from the XPS analysis. The aqueous solution of synthesized WO 3 resulted in a quantum yield (QY) of 3.25% at room temperature.
Microstructural and Optical Measurement of WO 3 Films. The SEM line-scale mapping was recorded on the WO 3 −Pt film, as shown in Figure S4a. This indicates a homogeneous distribution of W and O throughout the Pt layer. Further, the distribution of W and O over the Pt layer is very clear from the elemental mapping diagram, as shown in Figure S4b. Besides, in the case of WO 3 −carbon-based PSC film, a thin layer (<200 nm) of mesoporous WO 3 has been formed as shown in Figure S4c, and the corresponding elemental mapping has been given in Figure S4d. The EDS result indicates that despite the different approach of the buffer layer formation for DSSCs and PSCs, both of them showed a proper distribution of W and O along with their respective back electrode material, indicating that the formulation of WO 3 synthesis is precise and efficient in preparing the welldispersed sample.  Figure 7a. The effect of mesoporous nature also influences the efficiency of the device. In the absence of the mesoporous WO 3 layer, the maximum average PCE was achieved to be ∼6.12% ( Figure  S5a), which is ∼21% lesser than compared to mesoporous WO 3 layer-added devices. These results strongly support the benefits of porous layer incorporation in such a DSSC device. Further, the recorded J−V characteristics parameters are compared in Table 1. In spite of this, using WO 3 on Pt CE, the corresponding device exhibits ∼48% higher efficiency than only Pt CE. The enhancement of PCE was resulted by keeping insignificant alteration in V OC except for the J SC and FF, which signifies the presence of mesoporous WO 3 gets an advantage as a catalyst for Pt CE and, therefore, accelerate the rate of high electron mobility and hole-extraction leading toward high    The Journal of Physical Chemistry A pubs.acs.org/JPCA Article efficiency. The EIS measurements were carried out to understand the transport properties at different interfaces in the DSSC assembly. The EIS spectra (Nyquist plot) and the corresponding equivalent circuit diagram (inset) of cells for the DSSC devices, as shown in Figure 7b. As shown in Table 1, the DSSCs show almost closure of sheet resistance (R s ), which is due to the use of the same Pt counter electrodes and I − /I 3 − electrolyte in all the devices. On the other hand, resistance at the Pt/electrolyte interface has been found significantly higher for the only Pt CE device compared to WO 3 −Pt-based devices. The CT resistance of 32.68 Ω cm −2 at the electrolyte/dye/ TiO 2 interface for the Pt CE device is considerably higher than the WO 3 −Pt device (27.12 Ω cm −2 ). This result clearly shows that the catalytic activity of WO 3 reduces the CT resistance of Pt producing higher efficiency. The EIS spectrum of the without mesoporous WO 3 -based DSSC device has been given in Figure S5a (inset), and the corresponding data are shown in Table 1.

PV Performance of WO 3 as a Cathode Buffer Layer in PSCs
In addition, incorporation of the synthesized WO 3 as a cathode buffer layer, it was further evaluated in the case of PSC devices. The corresponding J−V characteristics plot is shown in Figure 8a. It was noted that the addition of WO 3 −carbon back contact exhibited the highest PCE of 12.42%, which is ∼17 and ∼29% higher than the without mesoporous WO 3 (PCE: ∼10.58%) and only carbon-based devices (∼9.65%), respectively. The J−V characteristics plot for the without mesoporous WO 3 -added device has been given in Figure S5b. Interestingly, the V OC does not affect the incorporation of the WO 3 layer, and for both the cases, it shows quite high value as 0.9 V. The addition of WO 3 significantly boosts up the J SC of the WO 3 −carbon-based device to 23.16 mA cm −2 , which is quite high and impressive to generate higher efficiency. The digital image of the WO 3 −carbon-based PSC has been given in the inset of Figure 8a.
The Nyquist plots derived from the impedance data, Figure  8b, revealed a clear difference between the two devices. The WO 3 −carbon-based PSC shows the smallest radius of the semicircle in the high-frequency region, which indicates the lowest CT resistance at the CH 3 NH 3 PbI 3 /carbon interface. This result implies that the film addition of WO 3 plays an important role in DSSC electron recombination. All the measured parameters have been summarized in Table 2. R S of PSCs with WO 3 is less than the device without WO 3 , which suggests that the hole-extraction process at the carbon electrode with WO 3 is more efficient. Moreover, a lower charge exchange resistances (R CT ) at a bias of −0.90 V was observed for WO 3 -additive PSCs compared to the devices without additive. However, the WO 3 additive layer may play the role of a hole collector along with the appropriate energy level to the perovskite CH 3 NH 3 PbI 3 sensitizer and lead to an acceleration of the hole extraction efficiency. The WO 3 layer can also significantly reduce the series resistance and significantly improve the J SC of the PSC device. Besides, the EIS measurement was also performed for the absence of mesoporous WO 3 -based PSC devices, as shown in Figure S5b (inset), and the corresponding data are summarized in Table 2.
Probable Mechanism of the WO 3 Buffer Layer in Solar Cell Devices. The efficiency was measured for 10 numbers of cells for DSSCs and PSCs with an error bar plot, as shown in Figure 9a. This result implies that the film addition of WO 3 plays an important role in DSSC electron recombination. Therefore, we suggest an apparent mechanism to understand the effect of the porous WO 3 structure as an interfacial cathode buffer layer on the improved crystalline and enhanced PCEs of the solar cell devices. WO 3 treatment may also play a similar role in affecting the electronic structure of carbon and can be explored as a hole-extracting layer for PSCs. It is assumed that a small polaron hopping conduction mechanism can be influenced to operate in WO 3 as a cathode buffer layer for solar cell devices. 56 This mechanism possibly suggests hopping can be a faster way compared to the traditional way of hole diffusion with an extremely fast rate constant that leads to exhibit enhanced device performance. Besides, the photoluminescence spectrum of WO 3 predicts partial oxygen vacancies in substoichiometric WO 3−x which creates localized defect states: W 4+ or W 5+ and W 6+ . These mixed valance states may further promote the CT and enhance electrical conductivity. 57,58 The high surface area facilitates to collect and regenerate a higher number of holes, which is ultimately  The Journal of Physical Chemistry A pubs.acs.org/JPCA Article beneficial for rapid and smooth electron−hole mobility for the devices. At the same time, it is also assumed that because of distinct porous nature of WO 3 , that it may be suitable to entrap/adsorb any moisture or oxygen molecule, which may retard the overall performance of the device. Pt counter electrodes are usually affected by high corrosion because of the use of liquid I − /I 3 − electrolyte, which can retard the hole conduction process, although using the WO 3 buffer layer may restrict the rate of the corrosion and facilitate the hole conduction process for the device. Therefore, the high porosity could be beneficial for the WO 3 and favor as a buffer layer for the solar cell devices. The overall concept has been schematically illustrated in Figure 9b,c. The enhanced PCE has been mainly influenced by the enhancement of the J SC factor of the device. The increased J SC of the porous WO 3 probably resulted in a thicker active layer, which is helpful to improve optical-trapping properties.
Incident Photon to Current Conversion Efficiency Measurements. IPCE stands for the external quantum efficiency of the DSSC and PSC devices including the effects of optical losses caused by transmission and reflection. 37 The IPCE measurements were also performed as a function of wavelength by comparing the WO 3 −Pt and Pt-based devices for DSSCs and WO 3 −carbon and carbon-based devices for PSCs as shown in Figure 10. The IPCE curve for all the DSSCs shows a broad peak over the range of 300−800 nm with a maximum value of ∼83% at 536 nm for the WO 3 −Pt-based device, whereas the Pt-based DSSC exhibits an IPCE value of 61% at the same wavelength as shown in Figure 10a. In the case of PSCs, incorporation of the WO 3 layer exhibits predominant incident photon to current conversion efficiency (IPCE) value (∼90% at 400 nm and 76% at 560 nm wavelength) than the without WO 3 layer device value (∼76% at 400 nm and 60% at 560 nm wavelength) as shown in Figure  10b. This also strongly recommends that the incorporation of a WO 3 layer improves the IPCE in a great manner to minimize the electron recombination or electron entrapment within surface trap states or electron loss within the hole transporting section of the device. Therefore, the EIS and IPCE results predict that the CT process is taking place at the interface of Pt/carbon and WO 3 , leading to an acceleration of the hole mobility, and as a result, the short circuit current increases remarkably. This is quite competitive with organic interfacial materials where it cannot be readily achieved with the later. All the above-discussed results and the fundamental insights could be useful in the development of achieving solution-processed successful incorporation of the WO 3 cathode buffer layer for highefficiency devices. 20, 28,59 Although the use of WO 3 presented here are far from being fully optimized, but the profound advantages along with ease of preparation and applicability for potential performance make it a good candidate for the cathode buffer layer for solar cell devices. 60,61 This work also provides insights into the design of the interfacial hole transport layer for the fabrication of solution-processed solar cell devices. The obtained results are quite encouraging and open up ample scope to develop new and economically viable synthetic approaches to enhance the performance of the solution-processed solar cell devices.
In spite of this, there are very few reports available on the buffer layer-incorporated DSSC and PSC devices. In comparison to the reported results on the performance of various materials as incorporated buffer layer in solar cell devices (Table 3), the results obtained in this work are quite competitive and interesting to execute dextran template mesoporous WO 3 as a promising buffer layer candidate for both DSSC and PSC devices, respectively.  The Journal of Physical Chemistry A pubs.acs.org/JPCA Article In our previous report, different concentrations of commercial WO 3 nanoparticles as a HTM incorporated in carbon counter electrodes were tested for PSCs. 35 However, the current study mainly signifies about the mesoporous layer involvement of WO 3 as a CBL for both the DSSC and PSC devices. Of course, controlled synthesis, optimum pore size distribution, various surface treatments, and optimization of film fabrication and device fabrication techniques are required, which may further lead to the enhancement of efficiency of the device for futuristic applications. We believe this study seems to be applicable for the synthesis of other related materials with controllable morphology and tunable textural properties for various applications such as catalysis, adsorption, gas sensing, separation, and so forth. [42][43][44][45]48,51 ■ CONCLUSIONS Hydrothermal synthesis of mesoporous WO 3 using the dextran templating process has been performed in this study. Dextran can effectively help to generate pore on the surface of WO 3 . Hierarchical platelet-like porous morphology of WO 3 exhibited a high surface area of ∼110 m 2 g −1 and an average monomodal pore size diameter of 20 nm. Taking advantage of this mesoporous WO 3 , it has been further explored as a solutionprocessed cathode buffer layer on Pt in the case of DSSCs and on carbon for the PSC device performance evaluation. Incorporation of WO 3 interestingly enhanced the PCE of a Pt-based DSSC device by ∼48%, whereas in the case of PSCs, the PCE enhancement was recorded ∼29% compared in the absence of buffer layer-employed devices. It is anticipated that the WO 3 layer significantly boosts up the short circuit current of the PV devices, leading to enhance the PCE. Besides, an apparent hole hopping mechanism associated with the porous WO 3 layer structure credit some added benefits for the devices. On the basis of our results, we are able to establish porosity and buffer layer incorporation as an important factor for PV efficiency enhancement for DSSC and PSC devices. Moreover, the outcome from this work is highly promising and renders a potential for future self-powered electrochromic device integration, where WO 3 will enhance the power generation and switching color in a tandem structure concomitantly.
Reported surface area and porosity results of different morphology-based WO 3 for various applications; EDS spectrum of the mesoporous WO 3 sample; pore size distribution plot of WO 3 as obtained from the BJH adsorption model; PL spectrum of the aqueous solution of the mesoporous WO 3 sample at room temperature; SEM elemental and line scale mapping of WO 3 −Pt and WO 3 −carbon films; and J−V characteristics plot along with corresponding fitted EIS plot of the nonporous (without dextran added) WO 3 buffer layer for DSSC and PSC devices (PDF)