Integrated Electronic, Optical, and Structural Features in Pseudo-3D Mesoporous TiO2-X Delivering Enhanced Dye-Sensitized Solar Cell Performance.

The performance of a dye-sensitized solar cell (DSSC) is strongly affected by optical, structural, and electronic features of a photoanode. In this article, meso-TiO2-X was prepared by a solution combustion method and hydrogenation at high pressure. The properties of DSSCs with meso-TiO2-X photoanodes were investigated by photocurrent-voltage, incident photon-to-current conversion efficiency, and electrochemical impedance spectroscopy (EIS) measurements. The meso-TiO2-X materials exhibit new electronic states and aided to absorb in the visible region because of the narrow band gap. Facile charge transfer from the N719 dye to the TiO2 photoanode was assisted by low-lying mid-gap states. Electrically integrated nanoparticles, with a small-channel mesoporous framework, facilitates fast charge transport across the material. Furthermore, EIS has shown that chemical capacitance, recombination resistance, and electron lifetime were affected by hydrogenation, thus indicating an effect on the photoanode material charge dynamics of DSSCs. An η of 7.2% under AM 1.5G illumination is obtained and an improvement by 75.6% over Degussa P25 titania. This is attributed to improved light harvesting and charge collection by the meso-TiO2-X photoanode obtained via simple combustion synthesis.

Introduction

Energy needs for humanity in the coming times due to economic growth and rising population will increase.[1] Conventional energy sources are exhaustible and cause environmental pollution, which are not suitable for the climate. Sunlight is an abundant and sustainable power source[2] and can be converted to electric power via photovoltaics.[3,4] Ever since dye-sensitized solar cells (DSSCs) were reported by Grätzel and co-workers in 1991, they have gathered attention throughout the world because of their low cost and ease of preparation.[5,6] DSSCs are composed of a dye-sensitized nanocrystalline film as a photoanode, a redox couple in an electrolyte, and a platinized transparent conducting oxide glass as a counter electrode.[7] Here, the photoanode serves several essential functions such as it provides surface for supporting dye molecules, provides medium for transporting photoexcited electrons via a transparent conductive oxide substrate into an external circuit, absorbs light, and allows electrolyte diffusion and dye regeneration.[8,9] To reach the market and compete with conventional costly silicon solar cells, understanding and improvement are needed in DSSCs.[10]

TiO2 along with few other metal oxide semiconductors has been used for various practical applications such as energy conversion and storage, photocatalysis, sensing, electrochromic, and so forth.[11,12] It is required to fabricate novel nanostructures with desired morphology, crystallinity, composition, texture, and porosity for improved light harvesting and charge transport.[13−15] Of particular interest are pseudo-three-dimensional (p3D) mesoporous materials, which provide small diffusion length and have electrically connected particles with high crystallinity for fast electron transport, thus making them a good candidate for DSSCs, photocatalysis, and other light-harvesting processes.[16−19] In comparison, ordered mesoporous materials used for anode preparation in DSSCs have a 20 nm pore wall and a long pore of several micrometers in length, thus creating large diffusional constraint for charge transport.[20,21] Hard and soft templates such as P123 and CTAB surfactants are needed for synthesis and later have to be removed by calcination or other methods.[22] Further, photoelectrochemical properties in TiO2 are controlled via doping a foreign ion,[23] band gap engineering,[24] defect engineering,[25] and so forth. Hydrogenation of titania is attracting interest because of improved light absorption in TiO2, and much work has been done in reducing TiO2.[26] The density of oxygen vacancies results in higher charge carrier density, enhancing electrical conductivity, as well as charge separation and transportation, and also mid-gap electronic states improve the optical absorption.[27,28] Good repeatability is essential in investigating the performance of each component; thus, it requires great skill and know-how in DSSC fabrication, and nonetheless, it does not need expensive equipment.[29] These integrated meso-TiO2– anodes are expected to simultaneously affect electro-optical and electron transport features, thereby increasing photon-to-current conversion efficiency.

In this article, meso-TiO2– is prepared via solution combustion synthesis with urea and titanium nitrate in a 1:1 ratio at 400 °C and hydrogenation at different pressures.[30] During combustion, interconnected nanocrystallites are generated in a self-propagating manner at high temperature. Mesoporosity in the material is developed because of evolved gases during combustion. New mid-gap states are introduced and optimized in nanocrystallites via hydrogenation. By employing these meso-TiO2– anodes, 7.2% η was obtained under simulated sunlight. This work is a part of the research effort from the author to develop a photocatalyst material suitable under sunlight for photovoltaics.

Results and Discussion

UV–visible absorption spectra of the as-prepared meso-TiO2– have been measured, and the results compared with T0 are shown in Figure . Pure T0 exhibits a well-defined wavelength absorption cutoff edge at 400 nm. A continuous absorption spectrum beyond 400 nm with a poor absorption cutoff edge is observed for meso-TiO2–. These new lower energy features demonstrate the creation of mid-gap states via hydrogenation between a conduction band and a valence band (VB), thus narrowing the band gap with absorption extending into the visible region. Indeed, the above-mentioned observation supports the conclusion that the oxygen vacancy (Vo) and Ti3+ are far removed from the O 2p VB and suggests the formation of the Vo-derived mid gap in the forbidden gap of TiO2. The new band in the visible region displays an intensity comparable to the band in the UV regime, proposing the large density of states in the new band at higher hydrogenation pressure. The hydrogenation has formed a shallow acceptor and a donor level and decreases the band gap to 2.8 eV in meso-TiO2–, depending on the pressure of reduction. The band gap for meso-TiO2– materials is calculated using the Tauc plot from the diffusive reflectance UV–vis spectroscopy analysis and is shown in the Supplementary Information Figure S1 and Table . The optical band gap values for T0, 10B, 20B, and 30B are 3.1, 3.1, 2.82, and 2.89 eV, respectively. The band gap decreased to 2.82 eV for 20B, comparative to 3.1 eV for T0. Absorption in the visible, near-infrared region of the spectrum due to defect introduction can be seen as the white color for T0, light yellow for 10B, dark brown for 20B, and gray color for 30B; digital images are shown as the inset in Figure S1. Interestingly, the rise in hydrogenation pressure displays lowering of difference at absorption cutoff and extending into the visible and IR region. Thus, larger fraction of photons in the solar spectrum is absorbed and helps to generate an additional number of the charge carrier in the meso-TiO2– photoanode material. It is to be noted that improved light harvesting by the materials is necessary for increased solar cell efficiencies.

Figure 1

UV–vis absorption spectra of meso-TiO2– materials and TiO2. Absorption in the visible and near-IR region increases with increased pressure of hydrogenation.

Table 1

Electro-Optical and Structural Properties of Meso-TiO2– Materials

photoanode materials	crystallite size (nm)	rutile content (%)	band gap (eV)	surface area (m2/g)	pore diameter (nm)	pore volume (cm3/g)
T0	8.8	3.7	3.10	154	2.5	0.29
10B	7.4	3.9	3.10	104	3.3	0.19
20B	10.8	14.2	2.82	68	3.6	0.15
30B	33.2 (70.8)a	68.0	2.89	21	3.4	0.05
P25b	21.0	19	3.2	55	1.6	0.25

Crystallite size for the rutile phase.

Pore size and the surface area obtained from ref (39).

Figure shows transmission electron microscopy (TEM) images for 20B, T0, and 30B materials. Distorted spherical nanocrystallites of 10–15 nm, electrically well-interconnected with each other over long range, are observed.[30,37] All other meso-TiO2– materials exhibit similar morphologies. TEM shows pseudo-3D porosity in meso-TiO2–, and same morphologies were obtained from different batches. Disordered (wormholelike) mesopores can be seen in all the TEM images. It is likely that disordered mesoporosity arises because of intergrowth of fundamental particles, and the same leads to aggregates with significant extra framework void space. During solution combustion, ammonia released via urea decomposition passes through the developing material crystallite, which is attributed to the mesoporosity in the material. It is observed predominantly that (101) faceted particles are observed with lattice fringes corresponding to the (101) (d(101) = 0.352 nm) crystallographic plane of an anatase phase. The selected area electron diffraction (SAED) pattern was shown as an inset in the figure, diffraction spots demonstrate the polycrystalline nature of the material, and (101) facets are the predominant planes. Particles (60 nm) for 30B are observed, as well as an increase in the particle size with hydrogenation, supporting X-ray diffraction (XRD) results. Field emission scanning electron microscopy (FESEM) images were taken to understand the morphology of meso-TiO2– material films. FESEM images for T0 and 10B are represented in the Supplementary Information Figure S2a,b. High porosity is observed in scanning electron microscopy (SEM) images for materials because of the release of combustion gases during material synthesis. The 3D disordered mesoporous framework of these meso-TiO2– materials reduces the diffusional barriers and facilitates mass transport through them, which makes them attractive in the photoanode in DSSCs. Disordered mesoporous materials have small pore depths of few nanometers unlike ordered mesoporous materials such as SBA-15 and MCM-41 having long pore channels of few micrometers and create diffusional constraint for the electrolyte movement.[18,19]

Figure 2

TEM images of meso-TiO2– materials: (a) 20B, (b) interconnected T0 particles, (c) high-resolution TEM (HRTEM) showing predominantly (101) faceted crystallites, and (d) 30B. The SAED patterns are shown as the inset.

To understand the surface electronic nature of meso-TiO2– materials, X-ray photoelectron spectroscopy (XPS) studies are carried out. The XPS spectrum for the core-level Ti 2p is shown in Figure . The binding energy (BE) for the Ti 2p3/2 core level in 20B is observed at 458.1 eV, whereas BEs are observed at 458.4 eV and 458.8 eV for 30B and 10B or 459.2 eV for P25 (Figure a). This BE value demonstrates that 20B differs electronically from other meso-TiO2– materials. A significant decrease in the BE of Ti 2p3/2 in 20B shows that there is an increase in electron density around Ti. The peak in lower BE indicates the presence of lower valence states of Ti, and the core-level Ti3+ 2p3/2 appears at 457 eV as reported in the literature.[38] The O 1s core level also displays similar changes. 20B exhibits the O 1s peak at 529.4 eV, in contrast to 530.3 eV for T0 (Figure b). It can be correlated to two factors. Mainly due to the reduction of the lattice with Ti3+ formation in 20B than in 30B and 10B, due to electrically interconnected nanoparticles (EINPs) and textural properties leads to higher electron density on the surface. Under hydrogenation, systematic removal of oxygen from the lattice to maintain charge neutrality leads to increased electron density on the neighboring oxygen atoms in the lattice. Thus, the above factors improved effective electron density and increased electron transport between particles in the mesoporous framework of 20B.

Figure 3

XPS spectra obtained for (a) Ti 2p and (b) O 1s core levels of meso-TiO2– materials and the parent TiO2 material. A clear increase in the electron density on 20B is evident from the above results.

Figure shows powder XRD (PXRD) pattern of the meso-TiO2– materials. Structural properties calculated from the XRD pattern for meso-TiO2– are listed in Table and compared with Degussa P25. The diffraction patterns are indexed to JCPDS ICDD # 21-1272 for the anatase phase of TiO2 for 20B, 30B, 10B, and T0.[18,19] The peak intensity of 101 facets is the highest, and thus, they are the most abundant planes in the material. Additional features are observed as indexed to JCPDS ICDD # 21-1276, corresponding to the rutile phase in 30B. New crystalline rutile phase peaks are sharp and narrow. The 110 peak of the rutile phase is used for calculating the rutile content in 30B. The broad nature of peaks indicates the nanocrystalline nature of the particles. Particle sizes are calculated from XRD data using the Debye–Scherrer equation, and the same are listed in Table . It is to be noted that 20B exhibits a small particle size (10.8 nm) with high crystallinity and mesoporosity. 30B exhibits the crystalline anatase phase particle with 33.2 nm size, whereas the rutile phase has particles with a size of 70.8 nm.

Figure 4

Wide-angle PXRD patterns of meso-TiO2– materials at different hydrogenation pressures. 30B exhibits mainly the rutile content and η of 1.7%, whereas other three materials exhibit more than 5.0% η.

Textural properties of materials were examined via N2 physisorption studies. Measured N2 adsorption and desorption isotherms and pore size distribution for meso-TiO2– are shown in the Supplementary Information (Figure S3a,b), while calculated values are mentioned in Table . The meso-TiO2– materials exhibit the type IV isotherm with an H2 hysteresis loop which is a characteristic of the mesoporous materials.[37] T0 exhibits a surface area of 154 m2/g which is the highest among meso-TiO2– materials. With the increase in hydrogenation, the pressure surface area decreased. The pore size increased with hydrogenation. 20B exhibits the highest pore diameter of 3.6 nm. A trade-off has to be made between defect density and textural properties so that the overall efficiency by meso-TiO2– is high.

The current density–voltage (J–V) curves under AM 1.5 illumination for the DSSC with meso-TiO2– photoanode material and Degussa P25 are shown in Figure . Photovoltaic features such as short-circuit current (Jsc), open-circuit voltage (Voc), fill factor (FF), and power conversion efficiency (η) were calculated from J–V studies and are given in Table . The DSSC with the 20B photoanode exhibits 7.2% efficiency. Higher efficiency is attributed to improved charge carrier density or optical absorption because of band gap reduction due to the formation of Ti3+ species in the meso-TiO2– material. In addition, electrically interconnected particles lead to fast charge conduction in 20B, generating higher Jsc of 16.6 mA/cm2 and 0.69 V Voc with 67% FF. The small pore depth in a disordered mesoporous framework with preferentially (101) facets leads to fast charge transport across 20B, thus increasing η. The lower density of mid-gap states in 10B leads to η of 5.8%. Because of the absence of optical and electrical features, with mere textural features in the parent mesoporous (T0)-based device exhibited η of 5.0%. 20B exhibits 75.6% improvement in η over P25 titania. Improved light harvesting is attributed to higher charge carrier density and minimized recombination across the 20B structure. In fact, new species improved conductivity, affecting electron lifetimes. High surface area and small pore depth provide easy access to the electrolyte for regeneration at the interface. Improved optical, electrical, and structural features allow higher density and fast transport across the material, enhancing charge transfer and electrolyte regeneration, yielding better DSSC parameters. In contrast, the 30B material shows η of 1.7%, a 58.5% decrease in η compared to P25, underscoring the necessity of textural characteristics for various functions such as dye absorption, charge diffusion, and electrolyte regeneration for next cycle. With bigger 30B particles and less surface area for dye anchoring, poor electrical connectivity between particles results in solar cells with reduced efficiency. Primarily, the rutile phase (80%) in 30B shows high recombination; P25 having a 70% anatase content is responsible for η of 4.1%. The suitable phase alignment for the unidirectional charge transfer and the interface structure to minimize interface charge trap states. The presence of the anatase–rutile phase junction in 20B and P25 leads to efficient charge separation and transport across the interface.[39]

Figure 5

J–V characteristics for DSSCs prepared with meso-TiO2– photoanode materials and P25 for comparison under AM 1.5G irradiation.

Table 2

Photovoltaic Properties of DSSCs with Meso-TiO2– Photoanodes

photoanode materials	Rct (Ω)	Cμ (μF)	τn (ms)	Jsc (mA/cm2)	Voc (V)	FF (%)	η (%)	% change in efficiency
T0	5.0	42.8	9.4	11.4	0.66	66	5.0
10B	5.8	59.0	9.7	14.6	0.68	61	5.8	15.0
20B	1.9	90.4	14.7	16.6	0.69	67	7.2	45.6
30B	3.2	18.7	1.9	3.87	0.64	65	1.7	65.4
P25	6.2	20.4	2.3	8.3	0.65	63	4.1

Electrochemical impedance spectroscopy (EIS) is a useful technique for studying electron transport and charge recombination across the interfaces in a DSSC.[40] EIS results as Nyquist plots are shown in Figure for DSSCs. The physical interpretation of the device is done by fitting the EIS plot in terms of the resistor–capacitor circuit and measured values are marked in Table . A typical Nyquist plot for DSSCs consists of three arcs. The response at the high-frequency range (500 000–1000 Hz) is attributed to the charge-transfer process at the electrolyte/Pt interface. The mid-frequency range between 1000 and 1 Hz represents the charge-transfer process at the TiO2/dye/electrolyte interface. In the low-frequency region of 1–0.0002 Hz, the arc represents the mass transport resistance of ions in the electrolyte. Generally, the high- and low-frequency arcs are not easily distinguishable and appear as a deformation of the central arc. The chemical capacitance (Cμ) as obtained from fitting the central arc helps in describing the fundamental mechanism whereby photogenerated carriers store free energy and produce a voltage and current in the external circuit.[41]Cμ values of 20.4, 42.8, 59.0, 90.4, and 18.7 μF for P25, T0, 10B, 20B, and 30B are observed, respectively. 20B exhibits the highest Cμ value. Additional density of mid-gap states is available to accept more charges accessible with the high specific area. This indicates that 20B has produced high donor density in the framework, resulting in high Jsc. Oxygen vacancy (Vo) is generated in 20B under reducing conditions. The Vo plays an important role in determining the surface and electronic properties of meso-TiO2–.[27,28] The density of Vo results in enhanced electrical conductivity, separation, and transportation. 20B possesses long-range electrically interconnected nanocrystalline particles in the disordered mesoporous framework ensuing improved conductivity. Reduced Cμ values for 30B suggest poor dye loading. Charge separation relies on good connectivity into each of the macroscopic plates, thus depending on the specific surface area of the interface. Increase in the particle size limits charge transfer. Recombination resistance (Rct) is also calculated, which describes the charge-transfer resistance across the interface.[42]Rct values for P25, T0, 10B, 20B, and 30B are 6.2, 5.0, 5.8, 1.9, and 3.2 Ω, respectively. 20B exhibits an Rct of 1.9 Ω, a 69.3% lesser resistance in comparison to that of P25 titania. The photoexcited dye can undergo facile charge transfer to the new states of TiO2. Good transfer kinetics leads to improvement in device output. Strong dye anchoring on preferential (101) facets of meso-TiO2– controls the limiting charge recombination process in the dye. Bode plots for DSSCs were obtained from the EIS data, and the plots are presented in the Supplementary Information, Figure S4. Effective electron lifetimes (τn) were calculated using eq , and the results are summarized in Table , where fm is the frequency at the maximum in the mid-frequency arc of the Bode plot.[43]

Figure 6

Electrochemical impedance spectra (Nyquist plot) for DSSCs based on meso-TiO2– photoanode materials and P25 at Voc, 100 mW/cm2. The equivalent circuit model of the DSSCs is shown as the inset.

In the Bode plot, the peak in mid-frequency shifts to lower values. It is observed that lifetime is significantly higher for 20B (14.7 ms) than for 10B and 30B. Higher effective diffusion length due to suppressed recombination and fast charge transport affects charge collection efficiency. The higher recombination rate in P25 results in short lifetimes.

Incident photon-to-current conversion efficiency (IPCE) for DSSC devices prepared with meso-TiO2– photoanodes was recorded as a function of the excitation wavelength and is represented in Figure .[44] In the long-wavelength region (600–780 nm) in which the N719 dye has a low absorption for incident light, the higher IPCE values of the 20B meso-TiO2– electrodes could be ascribed to the enhanced light absorption capacities of the film. Fast electron injection from the dye to the anode and reduced charge recombination over T0 are observed. A similar minor improvement is observed for 10B and 30B. The IPCE values for meso-TiO2– are in agreement with the variation of Jsc observed in J–V studies. 20B achieved 58% conversion, a 67% improvement over T0 at 530 nm, and 35% for 10B while IPCE decreased for 30B by 47%. New electronic states and EINP with the mesoporous framework improved charge density and charge utilization in 20B. Although 30B exhibits a good absorption spectrum, the lowest IPCE is due to low dye adsorption and charge transport across the material because of the poor surface area.

Figure 7

IPCE data for the DSSC made with meso-TiO2– photoanode materials and TiO2.

Conclusions

In summary, we have outlined a successful strategy for integrating electronic, optical, and structural features in the photoanode for DSSCs. The maximum η is found in the 20B sample, a 7.2% power conversion efficiency under sunlight and corresponds to a 45.6% improvement over Degussa P25 titania. The enhancement in DSSC performance in the meso-TiO2– samples over TiO2 samples is attributed to the higher visible light absorption, resulting in an increase in the charge carrier density in DSSCs. The meso-TiO2– photoanodes with new states generated by hydrogenation exhibit improved electron mobility. Also, fast electron transport is exhibited in EINP with high crystallinity, with preferentially (101) TiO2 facets in the small pore depth mesoporous framework. Improved DSSC efficiency in meso-TiO2– is due to ease in electrolyte regeneration via the large interface. This report shows facile integration of various anode features for the enhancement of DSSC efficiency. These meso-TiO2– materials having good visible light absorption and low charge recombination are expected to be a worthy candidate for photocatalytic applications.[31,32] Improvement in the photoanode parameters, that is, fast charge transport and good optical absorption via band gap reduction by means of various synthetic strategies and light-harvesting protocols, are in progress for similar photocatalytic applications.

Experimental Methods

Synthesis of Meso-TiO2– Materials

All the chemicals employed were of analytical grade and used as such without any further purification. Titanium nitrate (Sigma-Aldrich) as the Ti precursor as well as oxidizer and urea (Merck) as fuels were used. Urea–metal ratios were kept at 1:1. Solution combustion synthesis requires no hard or soft template molecules for structure direction, as in ordered mesoporous materials.[18,19] The required amounts of aqueous titanium nitrate and urea were placed in a 250 mL beaker and introduced into a muffle furnace maintained at 400 °C. Water evaporates in the first few minutes followed by smoldering type combustion that occurs for less than a minute. Solid products were obtained within 10 min of total preparation time and are named as T0. The T0 catalyst was reduced to obtain 10B, 20B, and 30B materials at 10, 20, and 30 bar hydrogen pressures, respectively, at 200 °C for 10 h in a high-pressure reactor. To study the effect of different hydrogen pressures on the material, the temperature is kept at a minimum of 200 °C.

Characterization of Materials

Diffuse reflectance UV–visible measurements were performed on a Shimadzu 2600 spectrophotometer with an empty sphere setup. Spectral grade BaSO4 was used as a reference material.[31,32] XPS measurements were performed in an ambient pressure photoelectron spectrometer (APPES from Prevac, Poland) under UHV conditions. XPS measurements were made with an Mg K X-ray source for X-ray generation and an R3000HP (VG Scienta) analyzer for energy analysis.[33,34] PXRD data of meso-TiO2– materials were collected from a PANalytical X’pert Pro dual goniometer diffractometer. The data were collected with a step size of 0.02° and a scan rate of 0.5°/min. The sample was spun slowly throughout the scan for better counting statistics. The radiation used was Cu Kα (1.5418 Å) with an Ni filter, and the data collection was carried out using a sample holder in Bragg–Brentano geometry. A FEI Nova NanoSEM 430 field emission scanning electron microscope was used for SEM measurements of materials. The field emission gun was operated at 20 kV potential. Thin films were prepared on fluorine-doped tin oxide (FTO) and placed on the stage for studies. A FEI TECNAI 3010 electron microscope operating at 300 kV (Cs = 0.6 mm, resolution 1.7 Å) was used for recording HRTEM images of all materials. Samples were crushed and dispersed in isopropanol before depositing onto a holey carbon grid.[34] Nitrogen adsorption–desorption isotherms for the materials were collected from a quantachrome autosorb automated gas sorption system (NOVA 1200). The Brunauer–Emmett–Teller equation was used to calculate the surface area from the adsorption branch. The pore size distribution was calculated using the Barrett–Joyner–Hollenda method on the desorption curve of the isotherm.

Fabrication of Solar Cell Devices

TiO2 paste was prepared by the following method described by Ito et al. in the literature.[35] TiO2 powder was mixed with α-terpineol and ethylcellulose in the ratio of 16.8:4.5:78.7. TiO2 thin films were prepared by applying the paste on the FTO glass plate using the doctor blade method.[36] The TiO2 electrodes were gradually heated under an air flow at 500 °C for 15 min. The working electrode was composed of a 12–13 μm thick TiO2 film (active area 0.16 cm2). The dye was absorbed on a TiO2 film by dipping into 0.5 mM N719 dye solution in ethanol at room temperature for 20 h to complete the loading with a sensitizer. The dye-adsorbed TiO2 electrode and Pt-deposited glass counter electrode were assembled together into a sealed sandwich-type cell by clips. A drop of electrolyte solution containing 0.6 M 1,2-dimethyl-3-propylimidazolium iodide, 0.05 M I2, 0.1 M LiI, and 0.5 M tert-butylpyridine in acetonitrile was injected into the assembled cell.

Photovoltaic Measurements

J–V characteristics of assembled DSSCs were measured under AM 1.5G illumination (100 mW/cm2) with a 100 W xenon lamp from a class AAA solar simulator, SS30AAA-TP (Photo Emission Tech. Inc.) over an active photoanode area of 0.16 cm2. The photocurrent generated from the device was measured using a Keithley 2400 source meter. J–V measurement assembly was calibrated using a reference silicon diode cell (NREL certified). IPCE spectra were recorded on a QEPVSI-b (Newport) quantum efficiency measurement system by changing the excitation wavelength at an interval of 10 nm over the range of 350–800 nm. The incident monochromatic light was chopped at a frequency of 20 Hz, and a bias light of 0.1 sun was provided during measurements. The generated photocurrent is measured on a Keithley 2400 source meter. IPCE measurement systems were calibrated with respect to a reference silicon solar cell (without bias illumination). EIS measurements were performed on an SP-300 potentiostat (Bio-Logic Science Instruments) in the frequency range of 100 mHz to 1 MHz. An ac sinusoidal signal of 10 mV amplitude was employed. The complex ac impedance spectra were analyzed with the EC lab software from Bio-Logic Science Instruments.