Cadmium chloride (CdCl₂), Vanadium chloride (VCl₃) (Sigma-Aldrich), Sodium hydroxide (Fisher chemical), Sodium carbonate (Na₂CO₃), Triton-X100 (TX100), EY, LiI-I₂ electrolyte, and conducting glass slides from Indium doped tin oxide (ITO) (8–12 Ω/sq) (Sigma-Aldrich). All the used materials were general grade chemicals.

Cd-V-LDH with Cd:V molar ratio = 1:1, was prepared by the co-precipitation method at pH 10, as shown in Figure 1. Two solutions were prepared, one containing [NaOH (1.5 M) + Na₂CO₃ (0.5 M)] and the other containing CdCl₃ + VCl₃ (total metal concentration was 0.2 M), which was gradually added to the first solution with continuous stirring, and the pH was adjusted to 10 by the addition of NaOH. After complete precipitation, the temperature of the mixture was raised to 60°C with continuous stirring for 16 h. The mixture was cooled down to an ambient temperature. The precipitate was separated and washed several times with deionized water until pH 7. The product was separated by centrifugation (10 min, 500 min⁻¹). The LDH precipitate was dried at 100°C for 24 h and calcined at 300°C.

A Pan Analytical Model X′ Pert Pro system with Cu-K radiation (λ = 0.1542 nm) was used to record the XRD patterns. The nitrogen adsorption-desorption isotherms of the synthesized samples were recorded at −196°C on the Quantachrome NOVA automated gas sorption system. A JEOL JEM 3500 electron microscope was used for the (FE-SEM) examination. (FTIR) was recorded on the Nicolet is 50, Thermo Fisher Scientific FT-IR spectrophotometer. A JASCO-750 UV-Vis spectrometer has been used to measurement optical absorption. The VOLTALAB PGZ301 potentiostat was used for cyclic voltammetry and impedance studies. A photocell test and a visible light lamp with an intensity of 100 mWcm⁻² were used to measure the photocurrent density-voltage (J-V) output of the solar cell, which was calibrated using a radiometer (International Light Technologies 1700).

ITO coated glass slides were cleaned and washed with ethanol, isopropanol, and deionized water in an ultrasonic bath for 20 min. The prepared Cd-V-LDH was mixed with X-100 and acetic acid to form a blend paste, and to get a stable photoanode, doctor blade printing was the most widely used coating technique for the preparation of DSSC layers due to its low cost, simple operation, and fewer moving parts. A paste of Cd-V-LDH is dropped on the substrate and spread using a sharp blade. Cd-V-LDH layer (4 cm²) spread onto ITO was heated in a muffle furnace at 450°C for 10 min. After being cooled off to room temperature, the film was then dipped in a solution of EY for 24 h at ambient temperature. For the opposite electrode, GN, (GO), and (CC) were applied as PCs by depositing on the conducting side of ITO as in the same previous method. PA and PC electrodes were held together as a sandawitch, and the liquid electrolyte of (0.2 M LiI + 0.5I₂) in ethanol solution was completely absorbed into the electrodes by capillarity.

Different parameters affecting PCE were studied, e.g., EY concentration and pH, active area of PA, and types of PC. Different molar concentrations of EY were utilised for this purpose (3 × 10⁻³, 1 × 10⁻³, 5 × 10⁻⁴, 3 × 10⁻⁴ and 5 × 10⁻⁵ M) and the Cd-V-LDH layer was doped in the EY bath for 24 h. Also, different pH of EY at 1.5, 3.5, 5.5, 9.5, and 13.5 were considered. Different types of PC were studied and applied, including GN, GO, and CC.

In a solar box with photocell test equipment, the electrical characteristics of J-V curves were assessed. To determine PCE, the fill factor (FF) must first be determined using all cell outputs according to Eq. 1. It is simple to obtain the PCE by Eq after determining FF (Eq. 2). F F = V p p × J p p V o c × J s c (1) P C E   % = V o c × J s c × F F I s (2) where Vpp and Jpp denote maximum power output voltage and current density, respectively. Whereas is the intensity of the incident light (mW·cm⁻²).

The XRD patterns of the synthesized sample are displayed in Figure 1. It shows that two phases of cadmium vanadium oxide are formed (Cd V₂O₃ and Cd _0.95 VO₃). The highest diffraction peaks at 2θ around 30.316° and 33.058° can be attributed to Cd _0.95 VO₃ phase, according to JCPDS card # 00-019-0195. Another diffraction peaks appear at 2θ values of 19.104°, 23.317°, 34.26°, 50.91°, and 57.53°. Peaks at 2θ values of 24.57°, 27.09°, 36.35°, 38.36°, 40.13°, 44.13°, 45.08°, 47.32°, 49.75°, 56.28°, 55.29°, 61.81°, 65.88°, 69.36°, and 74.85° correspond to Cd V₂O₃ (JCPDS card # 00-059-0581), while the main peak appears at 2θ value of 31.311°. There are two peaks at 2θ equal to 23.517° and 43.776° indicate the presence of vanadium oxide (V₂O₃) (JCPDS card-no-34-0187). Additional peaks at 2θ of 52.29° and 63.14° indicate the presence of cadmium oxide (CdO) according to JCPDS card # 01-104. The crystallite size (D) of PA was calculated using the Debye Scherrer Eq. 3. D = k   ʎ β COS θ (3)

Where D is the crystallite size of nm, k = 0.9 is a correction factor that accounted for particle shapes, β is the full width at half maximum (FWHM) of the most intense diffraction peak plane, λ is the wavelength of the Cu target = 0.15406 nm, and θ is the Bragg angle, and the average crystallite size of the Cd-V- LDH was 16 nm.

Figure 2B shows diffuse reflectance (DRS) spectra and band gap determination. Figure 2B displays UV–Vis DRS graphs of the Cd-V-LDH and the data showed a strong peak in the visible range of 430 nm. Cd-V-LDH band gap energy can be calculated using Eq. 4 (Nayak and Parida, 2019). α h ν 1 / n = A h ν − E g (4) where α is the absorption coefficient, h is the Planck’s constant, ν is the energy of incident light, A is an arbitrary constant, and Eg is the band gap energy of a material LDH. The nature of band gap transition depends upon the n value of a LDH material; n = 1/2 for direct transition and n = 2 for indirect transition. In this case, LDH is found to have direct transitions (Nayak and Parida, 2019). Therefore, the plot of (αhν)2 vs. hν (Kubelka–Munk function as a function of photon energy) gives the band gap energy value by extrapolating the straight line to the hν axis intercept (X-axis) as shown in Figure 2C. The estimated band gap energy value was found to be 2.1 eV. The band gap energy value of LDH exhibits a single band gap value, which is due to the strong coupling effect of constituent semiconductor components and simultaneous overlapping of band gap energies due to self-assembly of layered-to-layered structure. As a result, the band gap energy tuning of the as-fabricated heterostructure nanocomposite is due to the accessibility of quantum confinement effects, defect sites, or oxygen vacancies, which may increase the intensity of visible light absorption for superior photocatalytic activities (Zheng et al., 2019).

The FT-IR spectrum of Cd-V-LDH was displayed in Figure 3. The spectrum (a) shows bands at 3,421 and 1,637 cm⁻¹ were associated with the interlayer water molecules in Cd-V-LDH (Wang et al., 2015). According to the literature (Kloprogge et al., 2002; Li et al., 2017), the strong band at 1,389 cm⁻¹ was described as the carbonate anions’ mode of v3 asymmetric stretching. α, β-unsaturated aliphatic esters appears from 1,730 to 1,715 cm⁻¹. All the vibration bands from 800 to 400 cm⁻¹ were unique to the M-O, M-O-M, and O-M-O groups (Bazzan et al., 2011).

After immersing the Cd-V-LDH in EY at concentrations of 1 × 10⁻³ M for 24 h at pH 5.5, FTIR was used to categorise the adsorbed groups on the surface. The spectrum (b) shows the decrease in the intensity of the bands due to the implimination of these groups in the adsorption process (WangWang, 2008). Figure 3 shows also the formation of new functional groups on the surface of Cd-V-LDH after impregnation. The main bands were located in 1,097, 1,236, 1,355, 1,385, and 1,556 cm⁻¹. The presence of ether linkage C–O–C and C–O stretching was observed at 1,097 and 1,236 cm⁻¹. The band at 1,355, 1,385, and 1,556 cm⁻¹ is assigned to acetate carboxyl groups (WangWang, 2008). From these results, we conclude that EY is attached to the Cd-V-LDH surface by carboxylate group.

Raman spectroscopic analysis was performed in this study to determine the nature of the bonding of the Cd-V-LDH Raman modes at 148, 207, 253, 324, 342, 374, 489, 520, 642, 895, 905, and 1,100 cm⁻¹ are observed, as shown in Figure 4. The peaks between 100 and 314 cm⁻¹, which are associated with oxides of V modes, are primarily caused by the bending and lattice modes of V-O bonds. The peak at 320 cm⁻¹ in the lower wavenumber area is connected with the motion of the cadmium sub-lattice and oxygen atoms in CdO, and it is associated with the second order scattering Raman mode (E² _high − E_low) where E² is 2 non-polar. Due to imperfections such oxygen vacancies, the band with a wavelength of 374 cm⁻¹ is first-order A1 (LO) mode (Samanta et al., 2006). Raman spectra’s A1 or E1 modes show that the crystal lattice’s or the sample’s lattice vibrations are parallel to or perpendicular to the c-axis, respectively. Additionally, it is corroborated by the phonons’ longitudinal and transverse optical emission. E1 (TO) may be to blame for the noticeable shift in peak position. These modes are indicative of the V position. Due to the stretching mode of (V2O2)n, which corresponds to chain translation, some Raman peaks are also seen at 148 cm¹ at lower frequencies (Boruah and Misra, 2016). The broad breadth of the peak at 255 cm¹ may be caused by the V-O-V bending mode (Julien et al., 1997). The stretching band and triply coordinated oxygen (V₃-O) bond may be responsible for the peak at 315 cm¹ (Wang et al., 2001).

The Raman mode at 253 cm⁻¹ corresponds to both the Cd-O and V-O-Vbending mode. The vibrational modes with frequencies at 489 cm⁻¹ indicate the presence of Cd-V-O symmetry and 526 cm⁻¹ represents V-O-V stretching mode due to medium range force. The Raman modes at 905 correspond to the V₂O₅ phase and the lower intense peak at 1,100 indicate Cd-O.

895, 792, 693, 509, 460, 370, 204, and 142 cm⁻¹ indicate the presence of the CdV₂O₆ phase, where the mode at 696 cm⁻¹ is the strongest mode of CdV₂O₆ (Mounasamy et al., 2020). The Raman mode at 488 cm⁻¹ is associated with Cd–V–O symmetry of the V₂O₅–CdO system.

FEG-SEM and textural analysis were carried out in Figure 5A, which shows hexagonal platelet-like morphology with a size of about 2 µm in width and about 78 nm in thickness.

The textural properties of the prepared Cd-V-LDH showed the adsorption–desorption isotherms of N₂ were carried out in Figures 5B, C. That will provide a deeper insight into the porosity and specific surface area of the samples. The isotherms were classified according to IUPAC as reversible type IV. This is a feature of materials that contain mesoporosity and have high adsorption energy. These often include hysteresis attributed to the mesoporosity, and the appearance of a hysteresis loop of H3-type indicates the presence of mesoporosity, and pores are represented by slit-shaped pore channels that are the result of the accumulation of plate-like layers. The pore diameters determined using the BJH method from the desorption curves of the isotherms was 3.66 nm, and the Brunauere Emmette Teller (BET) surface area was 22.65 m²/g (Table 1).

Using cyclic voltammetric measurements, the electrochemical behaviour of the produced material was examined. The electrochemical experiments were conducted with (ITO) electrodes coated with Cd-V-LDH in 0.2 LiI M+0.5 M I2 as liquid electrolyte between 0.6 and −0.6 at a scan rate of 20 mVs⁻¹. No pairs of peaks that represent iodide (I) and triiodide (I₃ ⁻) oxidation and reduction were observed in Figure 6, confirming optimal electric double-layer capacitance behaviours (Dang et al., 2016). The Cd-V-LDH electrode has a large integrated area under the curve due to better electron conduction and a higher degree of redox pseudocapacitance. The specific capacitance value calculated from the CV curve at 20 mV·s⁻¹ is found to be 333.62 F·g⁻¹.

After studying the constructed module system, J-V curves were obtained for (Cd-V-LDH/EY/LiI-I₂/GN system) under regular illumination of solar light, as shown in Figure 7. The cell’s operation parameters are determined under various experimental conditions from the J-V curve characteristics., e.g., V_OC, J_SC, FF, and the overall PCE, which are calculated by Eqs.1, 2, respectively. Table 5 displays the cell’s obtained parameters.

Photosensitization plays a significant role in the electrical output. Table 2 displays the electrical output at varied concentrations of EY. It was found that the cell sensitized with 1 × 10⁻³ M showed the highest electrical output, e.g., VOC (705 mV), JSC (12.40 mA/cm²), and PCE (5.4%). Meanwhile, the cell utilizing at 5 × 10⁻⁵ M showed the lowest power generation; V_OC (330 mV), J_SC (8.88 mA/cm²), and PCE (1.6%). This result agrees with the UV–Vis absorption spectrum.

The UV-vis of the samples with the highest and lowest concentrations of EY is shown in Figure 7A, EY absorbed the light primarily at 530 nm, and the absorption spectrum of the 1 × 10⁻³ M sample seems to be the highest one, so the most suitable concentration for EY to be adsorbed on the Cd-V-LDH surface was 1 × 10⁻³ M. EY was less effective in penetrating the LDH layer at higher concentrations. On the other side the amount of Photosensitizer atoms expected to absorb photons and provide an electron in the cell was limited at a low concentration, which results a drop in the output of the cell.

Furthermore, a shift in the pH of EY has the same effect as a shift in concentration. Table 3 indicates the influence of pH value on the cell’s performance. The results indicated that the V_OC J_SC, FF and the overall PCE increase as the pH increases from 1.5 to 13.5 and the highest electrical output, i.e., V_OC of 705 mV, J_SC of 12.40 mA/cm² and PCE of 5.4%, was shown in an acidic medium at pH 5.5.

The electrical output was fall in strong acidic media (pH 1.5) due to fast degradation of EY, with resultant inefficient light harvesting by the dye (Isah et al., 2015). Additionally, if the pH drops, it is expected that probable acid leaching will degrade cell quality. While in alkaline media at pH 13.5, the lowest electrical output was observed, i.e., V_OC of 558 mV, J_SC of 10.03 mA/cm² and PCE of 3.1%. This could be attributed to fall in optical absorbance of EY as shown in Figure 7B, as well as the negative charge can reduce possibility of the dye adsorption to the Cd-V-LDH surface.

The active area was defined as the effective area of PA that was exposed to visible light, with its value often ranging from 2 to 5 cm². The effect of the active area on the cell’s performance is seen in Table 4. Because PCE is directly impacted by FF, which is one of the cell’s most important outputs, a DSSCS cell with high FF will also have high PCE. The internal resistance of the DSSCs had an impact on the FF. Therefore, the FF and PCE will be higher in an area that is more active.

A 4.2 cm² area yielded the highest PCE measurement. The FF reduced as the active area reached 5.8 cm², and the cause of this is attributed to the cell’s rising internal resistance (Kang et al., 2003). The charge conveyance processes at GN as PC, the charge conveyance at the LDH/dye/electrolyte interface, permeation in the electrolyte, and the sheet resistance of ITO are all also related to the internal resistance (Han et al., 2005).

Cd-V-LDH has been applied as a photoanode, and the PCE increased to a maximum of 5.4% when the PA area was increased to 4.2 cm². The variation in PCE was shown in Table 4 as a function of PA size, and as the area rose to 4.8 cm², the PCE reduced by nearly 25%. Table 4 also showed that V_OC, J_SC, and FF varied depending on the active area. The equation below demonstrates how JSC followed a similar pattern to PCE and made a significant contribution to it. (Gregg and Hanna, 2003). J n x = n µ x ∇ F n (6) where n denotes the density of the electron and m(x) denotes its motion, and ∇Fn represents the gradient in the Fermi scale of the material’s interface, and the reduction in J_SC indicates the loss of electrons for the ∇Fn and µ constants, meaning that as the area increases, the electrons’ path length will be enlarged. Finally, recombination with the hole-conducting species in the electrolyte may result in an energy loss. As the area increases, VOC rates also decrease. At stable condition, the V_OC depends on the following factors: the electron density injected, the rate constant of the iodide reduction, and the conduction band electron density in the dark (Gregg and Hanna, 2003). V o c − K T / e ⁡ ln J i n / n c b K r e c I 3 (7)

The PC has an important impact on the photovoltaic parameters of DSSCs. The PC undertakes some functions, e.g., as a catalyst, promotes the completion of mechanism operation, works as a positive electrode of primary cells; collects electrons from the external circuit and transmits them into the cell. So, the ultimate mission of the PC is to return the electrons from the external load back into the “circulation” inside the cell, and as a mirror, it reflects the unabsorbed light from the cell back into the cell to promote utilization of sunlight (Wu et al., 2017). Figure 8 depicts the J–V curves of DSSCs prepared by Cd-V-LDH as PA with different types of PC, i.e., GN, GO, and CC.

Table 5 displays the photovoltaic parameters from J–V curves and a brief comparison with other PCs. According to the data, the GN-based cell had a PCE of 5.4%, which is higher than the GO and CC electrodes. The PC activity affects on the cell’s PCE, which synergizes the reduction reaction at the electrolyte/PC interface, which in turn support the restoration of the ground state of the organometallic EY at the electrolyte/photoanode interface. GN can be thought of as a layered material similar to graphene with nitrogen substituted. GN has superior qualities to graphene due to the presence of nitrogen atoms, a strong electronegativity, and a triazine‐rich structure (Fu et al., 2014). Additionally, GN exhibits high electrocatalytically active sites, which increases its capability for the I₃ ⁻ reduction reaction. (Afshari et al., 2018). Finally, GN can retard backward transportation of electrons between PA and photosensitizer (Fan et al., 2017).

As shown in Table 5, the highest parameters for the fabricated cell were V_OC (705 mV),J_SC (12.40 mA/cm²), FF (0.62), and PCE(5.4%) for Cd-V/EY-GN compared with results for [Cd-V/EY-GO] VOC (683 mV), J_SC (11.91 mA/cm²), FF (0.61), and PCE (4.9%) and [Cd-V/EY-CC] VOC (645 mV), J_SC (10.22 mA/cm²), FF (0.59), and PCE (3.8%), respectively.

With an active area of 4.2 cm², 0.2 LiI M + 0.5 M I₂ as liquid electrolyte, and GN as PC, a long-term stability test for Cd-V-LDH was conducted at light intensity of 100 mW/cm². The results revealed that the DSSCs were stable for 10 days under usual conditions.

Figure 9 shows the values for V_OC, J_SC, FF, and PCE as they changed over time. The starting parameters for (V_OC, J_SC, FF, and PCE) are 705 mV, 12.40 mAcm⁻², 0.62, and 5.4%, respectively. All parameters were steady for the first 3 days of the test period. VOC varied between a slight increase during the course of the following 3 days to reach 712 mV and a slight decline by the end of the test to reach 701 mV, which only drops by 4 mV from the initial value (705 mV), demonstrating high adsorption stability of the Cd-V-LDH/GN system. J_SC showed a slight decline to 12.25 mAcm^−-2 and eventually reached 12.16 mAcm⁻² at the conclusion of the test (10 days), this may be due to slow EY deterioration. After absorbing visible light for 10 days, the PCE gradually deteriorated to 5.15% and the system stayed at 95% of its initial value. The high cell stability demonstrated that the EY was very stable on the surface of Cd-V-LDH and that the degradation was limited.