Effect of chenodeoxycholic acid as dye co-adsorbent and ZnO blocking layer in improving the performance of Rose Bengal dye based dye sensitized solar cells

Effective suppression of dye aggregation on the photoanode surface of dye sensitized solar cells plays a key role in improving solar cell efficiency. Chenodeoxycholic acid (CDCA) is a very popular anti dye aggregation material used in dye sensitized solar cells (DSSC). However, the selection of an improper concentration of CDCA may lead to decreased solar cell efficiency by lowering the open circuit voltage and short circuit current as a consequence of reduced dye loading. The influence of CDCA as a dye co-adsorbent on the performance of DSSCs fabricated using Rose Bengal dye was studied in this paper. The concentration of the CDCA solution was varied to identify the optimum value for the best device performance. Aside from this, the effect of a very thin and compact ZnO blocking layer was also investigated to reduce the charge recombination. With photovoltaic parameters such as short circuit current density (Jsc) = 1.98 mA/cm2, open circuit voltage (Voc) = 0.58 V, and fill factor (FF) = 0.43, the traditional cell displayed an overall conversion efficiency of 0.49%, while the power conversion efficiency was found to be increased to 1.00% (Jsc = 2.80 mA/cm2, Voc = 0.64, FF = 0.58) when CDCA was added at optimised concentration of 8 mM. Reduced dye aggregation and increased electron injection in the presence of CDCA may be accounted for the DSSC's remarkable improvement in the efficiency. Moreover, the combined effect of 8 mM CDCA and the compact ZnO blocking layer dramatically enhanced the efficiency further to 1.22% (Jsc = 3.09 mA/cm2, Voc = 0.66, FF = 60). Electrochemical impedance spectroscopic (EIS) analysis revealed that the addition of CDCA as a co-adsorbent in the dye solution and addition of ZnO blocking layer resulted in significantly improved electron lifetime and reduced electron recombination yielding improved Jsc, Voc and η.


Introduction
Ever growing global energy requirement and depleting level of fossil fuels have accelerated the demand for efficient power generation from solar photovoltaic (PV) cells in recent years (Bach 1997;Meadows et al. 1972;Hosenuzzaman et al. 2015). The environmental impact of the use of fossil fuels is another major concern (Barbir et al. 1990). The current production of photovoltaic (PV) modules is dominated by crystalline silicon modules based on bulk wafers. However, the use of toxic materials and the high production cost of these solar cells have motivated the researchers to find new kinds of less expensive and non silicon-based solar cells to harvest solar energy efficiently (Goetzberger et al. 2003;Alharbi et al. 2011;Lee and Ebong 2017;Yamamoto et al. 2001).
Dye-sensitized solar cells (DSSCs) are a non-conventional photovoltaic technology that has attracted significant attention because of their high conversion efficiencies and low cost. O'Regan. B. and Grätzel reported high efficiency cells using nanoporous titanium dioxide (TiO 2 ) semiconductor electrodes, ruthenium (Ru) metal complex dyes, and iodine electrolyte solutions in the journal of Nature in 1991 (O'Regan and Grätzel 1991). Since then, many studies have been actively carried out on DSSCs and revealed their performance comparable to amorphous silicon thin films (Chiba et al. 2006;Grätzel 2005). These DSSCs have the advantages of low cost, lightweight and easy fabrication, but issues include durability and further improvement of their properties. To respond to these issues, many attempts have been made, such as solidifying electrolytes and improving materials and structures, but there have been no great breakthroughs yet (Chung et al. 2012;Cai et al. 2011).
A dye-sensitized solar cell consists of two conducting glass electrodes in a sandwich arrangement. Each layer has a specific role in the cell. The transparent glass electrodes allow the light to pass through the cell. The titanium dioxide serves as a holding place for the dye and participates in electron transfer. The dye molecules collect light and produce excited electrons which cause a current in the cell. The iodide electrolyte layer acts as a source for electron replacement. The bottom conductive layer is coated with platinum which plays the role of the counter electrode. A schematic structure of a liquid electrolyte DSSC and its working principle is shown in Fig. 1. When light passes through the conductive glass electrode, the dye molecules absorb the photons and the electrons in the dye go from the ground state (HOMO) to an empty excited state (LUMO). This is referred to as photoexcitation. The excited electrons jump to the conduction band of the semiconducting dioxide and diffuse across this layer reaching the conductive electrode. Then they travel through the outer circuit and reach the counter electrode. The dye molecules become oxidized after losing an electron to the semiconductor oxide material. The red-ox iodide electrolyte donates electrons to the oxidized dye molecules thereby regenerating them. When the originally lost electron reaches the counter electrode, it gives the electron back to the electrolyte (O'Regan and Grätzel 1991;Grätzel 2003).
The photovoltaic performance of a DSSC highly depends on all of its components and the fabrication methodology. Therefore, the optimization of every component is extremely crucial to achieve the best performance. Since its introduction into the science community in 1991, the nanocrystalline photoanode in dye-sensitized solar cells has predominantly been comprised of titanium (TiO 2 ) nanoparticles as the semiconducting material (O'Regan and Grätzel 1991;Grätzel 2003;Shao et al. 2011). Many researchers became very interested in studying the dye-sensitized solar cell performance fabricated using alternative semiconducting nanomaterials (Tiwana et al. 2011;Biswas and Chatterjee 2020).
Specifically, Zinc Oxide (ZnO) has been an ideal alternative to TiO 2 because of having a similar conduction band edge that is appropriate for proper electron injection from the excited dyes; moreover, ZnO provides better electron transport due to its higher electronic mobility. Along with that, ZnO is also highly transparent, which allows greater light penetration (Zhang et al. 2009;Guillén et al. 2011;Quintana et al. 2007;Vittal and Ho 2017;Biswas et al. 2019).
In this study, ZnO nanoparticles were implemented to fabricate the photoanode of the DSSCs and rose bengal dye was utilized as a sensitizer. To obtain better efficiency, the dye molecules must bind tightly to the mesoporous ZnO photoanode surface with the assistance of their anchoring group to ensure proficient electron injection from the LUMO of the dye molecule to the conduction band (CB) of ZnO.
Ruthenium dyes have long been used as quite efficient sensitizers for the photoanodes of the DSSCs (Ito et al. 2006, Aghazada et al. 2018, Nazeeruddin et al. 2011). However, these dyes are expensive, difficult to synthesis, requires high production cost, toxic, rare and easily pollute the environment . Owing to these facts, the organic photosensitizer Rose Bengal (RB), emerges as a promising and alternative candidate. It is a xanthenes class photosensitizer having high absorbtion coefficient and absorbs a wide spectrum of solar radiation. It energetically matches the conduction band edge of ZnO and iodine/iodide redox couple for DSSC application (Pradhan et al. 2007, Duffy et al. 2000. Accordingly, ZnO based DSSC performes specifically well when sensitized with Rose Bengal. Although the efficiency of these type of organic sensitizer based DSSCs is less, production cost per watt will be less compared to the ruthenium based DSSCs even if we achieve moderate efficiency. As the RB dye is an organic dye and does not contain any toxic noble metal such as ruthenium, there are no environmental pollution related issues with it. It is widely used because of its high absorption coefficient in the visible region of solar spectrum and its molecular structure (Fig. 2) comprises of anchoring groups that can be adsorbed onto the semiconductor oxide surface. For the particular case of ZnO-RB combination, the interaction between the unfilled valance shell of the ZnO and the carboxyl groups present in the dye molecules leads to easy adsorption of the dye molecules on the ZnO surface. Such kind of bonding between the dye molecule and ZnO not only increases adsorptivity of dye but also facilitates electron injection because of the substantial overlap between the electron molecular orbitals of the dye and those of the semiconductor's conduction band (Rani et al. 2010).
However, in case of ZnO photoanode based DSSCs, the dye aggregation on the ZnO surface affects the photoelectron injection by increasing charge recombination and hence limits the overall device performance (Kim et al. 2014;Patwari et al. 2017;Zhang and Cole 2017). The use of additives such as Chenodeoxycholic acid (CDCA) is a very useful and widely used strategy in lowering the self-aggregation of dye molecules by suppressing unfavourable dye-dye interactions as shown in Fig. 3 and thereby enhances the photoconversion efficiency (Buene et al. 2020;Kumar et al. 2020;Ismail et al. 2018). However, the strong binding of CDCA molecules to the ZnO surface partially displaces dye molecules and consequently reduces photon harvesting. Therefore, to maximize the positive effect of the co-adsorbent, it is very crucial to carefully optimize the amount of CDCA (Li et al. 2011). Few researchers have studied the role of CDCA as an anti-aggregation agent in ruthenium and organic dye based DSSCs and found it to be very effective in reducing the aggregation of dye molecules over the semiconductor surface (Inoue et al. 2010;Lee et al. 2007;Yum et al. 2008;Lu et al. 2009). But there is no report available related to the application of CDCA on Rose Bengal dye. Herein, we report the investigation on the effect of CDCA as co-adsorbent in the performance of Rose Bengal (RB) dye based DSSCs. Different concentrations of CDCA were studied to identify the optimum value for achieving the best device performance.
On the other hand, the mesoporous nature of the ZnO film is very essential to tender high surface area offering more dye loading and thereby generating more photoelectrons. However, small pores present in the nanocrystalline ZnO layer of the photoanode may provide a path for the direct contact between the liquid electrolyte and the FTO substrate. This may allow the electrons of FTO to recombine with the I − 3 ion present in the electrolyte resulting in high recombination current and hence decreased cell performance (Yang et al. 2014;Yeoh and Chan 2019). Therefore, to inhibit the electron back transfer, a promising approach is to modify the FTO/electrolyte interface by adding a compact metal oxide blocking layer. A thin blocking layer (BL) of ZnO was deposited by a facile and cost-effective sol-gel spin coating process before depositing the mesoporous active ZnO layer. In this work, we reported the fabrication and characterization of DSSCs based on ZnO nanoparticles and Rose Bengal dye. The effect of CDCA concentration and the compact ZnO blocking layer in boosting the photovoltaic performance of the device was investigated in terms of photocurrent-voltage (J-V) characteristics and dark current measurement. In addition to that, electrochemical impedance spectroscopy (EIS) analysis was employed to investigate the charge transfer kinetics and electron back reaction of the fabricated cells.

Materials
ZnO nanopowder, Zinc acetate dehydrate (CH 3 COO) 2 Zn, 2H 2 O) and Monoethanolamine (MEA) were bought from Sigma Aldrich, India. ethylcellulose and terpineol were bought from TCI Chemicals, India. Transparent FTO coated glass (10 Ω/square), the high-performance liquid electrolyte (Iodolyte AN50), chenodeoxycholic acid (CDCA) as a dye co-adsorbent and liquid platinum paint (Platisol T) to prepare the platinum-coated counter electrode were purchased from Solaronix, Switzerland. Meltonix 1170-25 (25 µm) (Solaronix) was used as a spacer between the working and counter electrode to avoid short-circuiting. All the reagents utilized in the fabrication process were of analytical grades. So no further purification was required.

Preparation of conventional ZnO photoanode
To prepare the thin films of the photoanode materials, the FTO coated glass substrates were first cleaned with dilute HCl in an ultrasonic bath for 15 min and then thoroughly rinsed with deionized water to remove the HCL residues. The substrates were then cleaned with acetone and ethanol using an ultrasonic cleaning bath (Biswas and Chatterjee 2020;Biswas et al. 2019). The mesoporous ZnO photoelectrode of the DSSC was prepared by following the standard doctor blade method. The paste for doctor blading was prepared by mixing 0.5 g of ZnO nanopowder with α-terpineol as a solvent and 0.45 g of ethyl cellulose as a binder (Wong et al. 2012). The mixture was stirred continuously to obtain a smooth lumpfree slurry. The ZnO paste was then coated on the conductive side of the cleaned FTO glass substrate and subsequently annealed at 400 °C on a hot plate for 30 min to burn out the ethyl cellulose and other organic contents of the working electrode and to strengthening the bonding between the substrate and the ZnO film. In addition to that, the annealing procedure also helps to improve the surface quality of the thin film along with increasing the crystallinity of the sample (Elilarassi and Chandrasekaran 2010;Shivaraj et al. 2015;Al-Kahlout 2015;Pandey et al. 2017).

Preparation of photoanode with compact ZnO layer
In order to improve the photovoltaic performance of the cells further by preventing the direct contact between FTO and liquid electrolyte, a thin and compact ZnO layer was deposited on FTO coated glass substrate by employing a simple sol-gel spin coating method prior to deposition of mesoporous active ZnO nanoparticle layer. The precursor solution was prepared by mixing Zinc acetate dehydrate (CH 3 COO) 2 Zn, 2H 2 O) in 50 ml isopropanol as solvent and monoethanolamine (MEA) was used as a stabilizer. The precursor solution concentration was maintained at 0.05 M. The mixture was vigorously stirred at 60° C by a magnetic stirrer for 1 h. MEA was added dropwise under stirring, yielding a clear homogenous solution. The solution was left for 24 h at room temperature for aging before it could be used for film deposition. The aged solution was then spin-coated on a cleaned FTO glass substrate with a programmable spin coater (Apex Instruments Co. Pvt. Ltd, Model SpinNXG-P1)at 3000 rpm for 30 s and annealed at 200 °C for 20 min to form the ZnO blocking layer. Over this compact blocking layer, the mesoporous active layer was coated using the same doctor blade method and then annealed at 400 °C as done earlier.

Assembling the devices
One set of ZnO photoanodes were sensitized by immersing them in a 0.5 mM ethanolic solution of pure Rose Bengal dye for 12 h. Another set of photoanodes (both with and without ZnO blocking layer) were sensitized with the RB dye solution containing various concentrations (0-10 mM) of CDCA at room temperature for 12 h. The working electrodes were then removed from the solution and rinsed thoroughly with deionized water and ethanol to get rid of any excess dye from the thin film surface and left for air drying at room temperature. The platinum catalyst precursor solution (Platisol-T) was spin-coated on the conducting side of the cleaned FTO glasses and heated at 450 °C for 15 min on a hot plate to prepare the counter electrodes for the cells. The dye adsorbed working electrodes and platinum(Pt)-coated counter electrodes were assembled against the coated sides of each other in a sandwich manner using two binder clips with a Surlyn film (Meltonix 1170-25 µm, Solaronix) gasket as a spacer in between them. The liquid electrolyte used in the fabrication process was poured inside the cell through fine holes pre-drilled on the counter electrodes. The red-ox concentration of the electrolyte was 50 mM. The active area of the cells for illumination was adjusted to 0.16 cm 2 .

Characterization and measurements
X-ray diffraction (XRD) analysis is a technique used for the determination of the crystal structure of materials in the nanomaterial, thin-film, or bulk material form. In the XRD experiment, a monochromatic X-ray beam is allowed to incident on the sample and the diffraction occurs. Constructive interference is obtained for the glancing angles (θ) corresponding to those (hkl) planes only for which the path difference is equal to integral multiple (n) of wavelength (λ) of the X-ray used. This condition is given by Bragg's equation where d is the interplanar spacing. The schematic of the experimental arrangement is shown in Fig. 4. The reflected X-rays make an angle 2θ with the material surface. A typical XRD pattern consists of these reflection peaks along the y-axis with the diffraction angles 2θ along the x-axis (Epp 2016, Zhang et al. 2004). In our study, the X-ray diffraction analysis was employed using PAN-analytical X'Pert PRO X-ray diffractometer (CuKα radiation, 30 mA, 40 kV, λ = 1.5406 Å) to determine the crystalline structure of ZnO nanoparticles used in making the photoanode of the DSSC.
Absorbance spectrum measurement of the dye was carried out using a Perkin-Elmer Lambda-35 UV-VIS spectrophotometer. In this measurement, the sample is exposed to light within a selected range of wavelengths. Absorption occurs when the incident photon energy surpasses the energy gap between the lower energy orbital (highest occupied molecular orbital-HOMO) and the higher energy unoccupied orbital (lowest unoccupied molecular orbital-LUMO) of the materials, and then the spectrometer records the signal. The block diagram of the UV-VIS spectrophotometer is shown in Fig. 5. (Rocha et al. 2018).
Scanning electron microscopy (JEOL) was used to examine the surface morphology of the prepared ZnO thin films. The current-voltage (J-V) characterization of the cells was measured under 100 mW/cm. 2 illumination using a Keithley 2400 digital source meter which was controlled by Keithley LabTracer computer software. The overall photoconversion efficiency of the solar cell was calculated using the formula where P in, V oc , I sc and FF denote the incident photon power, open-circuit voltage, the short circuit current density and fill factor respectively. The fill factor was estimated using the following formula: where I max and V max , respectively, represent values of current and voltage at the maximum output power point of the solar cell. The area of the fabricated cells that was exposed to light was 1 cm 2 . The electrochemical impedance spectroscopy (EIS) of the cells was done in the frequency range of 0.1-190 kHz under open circuit conditions.

Results and discussion
3.1 UV-VIS absorption spectral analysis of the dye 0.5 mM ethanolic solution of Rose Bengal dye was prepared and its absorption property was studied using Perkin Elmer Lambda-35 UV-VIS spectrophotometer. UV-VIS absorption spectrum of the RB dye is shown in Fig. 7a. The value of λ max obtained from the absorption spectrum is a very important parameter as it demonstrates the potential of the molecular systems for significant usage as a functional material in DSSC. It can observed that the Rose Bengal dye absorbs a larger fraction of the solar spectrum in the visible where α is the absorption coefficient, h is the photon energy, E g is the optical energy gap of the material and value of m depends on type of transition. The value of m is 2 or 1 2 for indirect and direct transition respectively. The direct band gap of the Rose Bengal dye was found to be 2.19eV by extrapolating the linear part of the (αhν) 2 versus h plot as shown in Fig. 7b. The indirect band gap of the dye was calculated through extrapolation of the linear portion of the (αhν) 1/2 versus h plot as shown in Fig. 7c (Patni et al. 2020) which indicates indirect transition at E g = 2.13eV.
The HOMO-LUMO direct band gap of the dye affects the electron transfer from the LUMO of the dye molecule to the conduction band of the ZnO particles. It facilitates the vertical electron transition through the dye excitation (Henson et al. 2013). On the other hand, indirect transition is a phonon assited transition where change in momentum must be taken into account. When the photons having energy fairly above the indirect band gap of dye molecule is absorbed by the dye electron, phonons get emitted (Seo and Hoffmann 1999). As a result the direct band gap is utilized to determine the vertical transition during the course of photosesitization , Prima et al. 2016. Hence, the lowest electronic transition, which corresponds to the onset of absorption in the UV-visible absorption spectrum, was used to calculate the optical bandgap. It is the energy difference between HOMO and LUMO which is caused by the excitation of electrons from HOMO to LUMO.

Structural and phase characterization ZnO compact layer
The X-ray diffraction pattern of the ZnO compact blocking layer, shown in Fig. 8a Fig. 6 Possible transition mechanism in the Rose Bengal molecular system (Costantino et al. 1998;Oh et al. 2002). The XRD pattern for the commercial ZnO nanopowder is shown in Fig. 8b. It can be clearly seen that both the commercial ZnO nanopowder and synthesized ZnO blocking layer showed similar XRD patterns. The XRD pattern of sample with both the blocking and active layer is shown in Fig. 8c. This is very similar to the XRD pattern of the blocking layer. This is because the blocking layer is more crystalline in nature, which is evident from its XRD pattern with its sharper peaks.

Surface morphology study and energy dispersive spectroscopy of the photoanodes
Scanning electron microscopic (SEM) analysis of the ZnO active layer and the compact blocking layer on the FTO substrate was carried out to study the surface morphology and the particle size of the sample. The SEM images of the ZnO active and blocking layers on the FTO substrate are depicted in Fig. 9. It can be seen from Fig. 9a that the ZnO nanoparticles have a hexagonal structure. Figure 9b and c represent the SEM images of compact blocking layer at low and high magnifications respectively. The diameter of the spin-coated nanoparticles ranges from 150 to 180 nm. Further, the chemical composition and elemental percentage of the compact ZnO film are revealed by the Energy Dispersive X-Ray Spectroscopy (EDS) analysis which is shown in Fig. 9d. Predominating peaks of Zn and O 2 unveil that the synthesized blocking layer contains pure ZnO.

Photovoltaic characterization of the cells
The Current-Voltage (J-V) characteristic is a crucial measurement that reveals the value of the overall photovoltaic performance of a solar cell along with the key performance parameters like open circuit voltage and short circuit current density. Figure 10a depicts the J-V characteristics of the DSSCs based on different types of photoanodes used in this study and the obtained photovoltaic parameters for each of the cells are summarized in Table 1.

Effect of CDCA
The conventionally prepared DSSC with ZnO nanoparticles and Rose Bengal dye displayed a short circuit current density (J sc ) of 1.98 mA/cm 2 , an open circuit voltage (V oc ) of 0.58 V, and a fill factor (FF) of 0.43, resulting in a photoconversion efficiency(η) of 0.49%. However, under the same working conditions, the device performance was found to be highly influenced when CDCA solution was incorporated into the dye solution at various concentrations. From Table 1, it can be noted that the value of V oc , as well as J sc , increases with an increase in the concentration of CDCA. Optimum concentration (8 mM) provides the finest dye attachment to the ZnO surface. The best device performance was achieved for the optimized CDCA concentration of 8 mM when added with 0.5 mM RB dye solution. This improvement in the performance may be attributed to reduced dye aggregation along with uniform dye adsorption yielding better electron injection into the conduction band of ZnO.

Effect of compact ZnO blocking layer
To avoid the direct contact between the FTO and the liquid electrolyte through the pores present in the nanocrystalline ZnO film in a conventionally prepared DSSC, a thin and compact layer of ZnO was employed as shown in Fig. 11. From Table 1 it can be observed that the addition of a compact ZnO blocking layer in DSSC7 with 8 mM CDCA additive shows a remarkable enhancement in J sc (3.09 mA/cm 2 ) and V oc (0.66 V) and consequently the highest value of photoconversion efficiency η (1.22%) was obtained among all the fabricated cells. Such type of performance enhancement may be accredited to the consolidated effect of improved dye loading due to the addition of CDCA with proper concentration and increased charge collection along with decreased electron recombination at the FTO/ZnO/ electrolyte interface hindering the direct contact between FTO and electrolyte by the blocking layer.

Dark current measurement
To explore the effect of CDCA concentration and ZnO blocking layer in the process of electron back transfer, the J-V characteristics were also measured in the dark which is shown in Fig. 10b. It is regarded as a qualitative method to assess the degree of electron back transfer in DSSCs. The dark current generation is known to be partly due to the presence of exposed FTO sites having direct contact with the liquid electrolyte and the pores left between the ZnO nanoparticles and the FTO surface (Yang et al. 2014; Yeoh and Chan 2019). These exposed FTO sites and the pores of ZnO nanoparticle film would allow the liquid electrolyte to penetrate through ZnO film and directly come in contact with bare FTO sites resulting in recombination losses as shown in the energy band diagram for conventional ZnO NP based DSSC in Fig. 12a. It can be observed from Fig. 10b that for a particular value of the voltage on the X-axis (i.e. voltage axis) of the dark current characteristics, the corresponding Y-axis value (i.e. value of dark current) is lowest for DSSC7 for that particular voltage. It can also be seen that the dark current has the highest value for conventionally prepared cell (DSSC1) with bare FTO and decreases with an increase in CDCA concentration up to 8 mM. An enlarged plot of the dark current characteristics is provided as an inset in Fig. 10b so that a clear scenario is observed. For the cell (DSSC7) with compact ZnO BL and 8 mM CDCA solution as dye co-adsorbent, the dark current is reduced significantly for the same bias potential in comparison to all other cells. This demonstrates that the compact ZnO BL reduces the bare FTO site and thereby successfully suppresses the dark current by lowering the electron back transfer. It was also observed that the DSSC7 has the slowest rate of increase of dark current with an increase in bias voltage confirming excellent suppression of electron recombination and consequently reduced current loss.
The variations of different cell parameters with the concentration of CDCA solution for the fabricated DSSCs are depicted in Fig. 13. It can be observed that CDCA concentration highly influences the value of J sc . ZnO BL improves the current further. A small increase in the values of V oc and FF can also be noted from Fig. 10 due to these processes. The highest values of cell parameters were obtained for the cell DSSC7.

Electrochemical impedance spectroscopy study
To further gain an insight into the influence of CDCA concentration and the coating of compact ZnO blocking layer on the charge transfer and recombination kinetics of the prepared devices, the DSSCs were further investigated by electrochemical impedance spectroscopic (EIS) measurement in dark under V oc bias voltage with 10 mV AC perturbation amplitude. This gives a more precise understanding of the limiting factors for the cell performance parameters. In the EIS measurement done under the dark condition and with an applied bias voltage, electrons from FTO are injected into the conduction band of ZnO and then transported through the ZnO network. Some of the injected electrons recombine with the I 3 − ion present in the electrolyte giving rise to the recombination phenomenon (Liu et al. 2018). Figure 14a shows the Nyquist plot of all the prepared cells exhibiting two obvious semicircles. The curves are fitted using the equivalent circuit shown in the inset of Fig. 14a and the EIS measurement results obtained in terms of resistances and capacitances are summarized in Table 2. The charge transfer resistance (R pt ) and double layer capacitance (C pt ) at the Pt counter electrode/electrolyte interface is responsible for the first semicircle in the high-frequency range, while the second semicircle in the midfrequency range may be assigned to the charge transfer and recombination resistance (R rec ) Fig. 13 Effect of CDCA concentration and ZnO blocking layer (BL) on different cell parameters and chemical capacitance (C) at the ZnO/dye/electrolyte interface (Li et al. 2011, Mazloum et al. 2019, Zhao et al. 2017. The intercept at the real axis of the Nyquist plot represents the series resistance (R s ) of FTO and other ohmic contacts like connecting cables, clamps and clips used to connect the cells for measurement. (Ondersma and Hamann 2010;Chou et al. 2019).
The small semicircles in the high-frequency range are almost identical indicating nearly similar values of R pt for all four kinds of cells as all of them have similar Pt counter electrodes and the same electrolyte. On the contrary, a substantial dissimilarity can be observed in large semicircles in the mid-frequency range. This indicates that the charge transport and recombination behaviour at the ZnO/dye/electrolyte interface was extensively affected due to working electrode modification by the addition of CDCA and incorporation of the blocking layer. The middle arc of the Nyquist plot for the conventionally prepared DSSC has the lowest diameter indicating the lowest recombination resistance (R rec ) and thus representing the highest recombination process among all the cells. The diameter is evidently larger for the CDCA treated cells indicating its positive role in increasing the recombination resistance and hence lowering the recombination phenomena. It can be observed from Table 2 that the recombination resistance increases in the order of DSSC1 < DSSC2 < D SSC3 < DSSC4 < DSSC6 < DSSC5 < DSSC7 indicating that the recombination resistance increases with an increase in CDCA concentration from 0 to 8 mM and decreases at 10 mM  concentration. The larger R rec value indicates it is more difficult to transfer the injected electrons from the ZnO back to the electrolyte, and thus the back recombination can be suppressed in the cell, thus giving a higher J sc and V oc . The highest value of recombination resistance is obtained when the optimum CDCA concentration (8 mM) is combined with the ZnO blocking layer in DSSC7 leading to the highest J sc and V oc of 3.09 mA/cm 2 and 0.66 V respectively and consequently best device performance. The highest recombination resistance is obtained for the cell with the blocking layer as the blocking layer prevents the injected electrons to come in direct contact with the electrolyte and consequently reduces the direct capture of electrons by the I 3 − ions of the electrolyte. Furthermore, the CDCA addition with optimum concentration and blocking layer increases the number of electrons accumulated in the conduction band of ZnO which led to increased electron density. This creates a small shift of Fermi level for the electrons present in the ZnO. This rise in Fermi level slightly improves V oc which can also be observed from Table 1 (Li et al. 2011;Wei et al. 2015).
Apart from recombination resistance, the second semicircle also provides information about the electron lifetime in the conduction band of ZnO which gives the measure of the rate at which the recombination reaction occurs. This lifetime is inversely proportional to the oscillation frequency at which the peak on the second arc is obtained. But, since, the frequency information is missing in the Nyquist plot; the electron lifetime can be calculated from the phase bode plot using the formula where f peak represents the peak frequency of phase Bode plot in the mid-frequency range as shown in Fig. 14b. A shift in the peaks may be observed in the Bode plots of the DSSCs prepared following different procedures. Shifting of peak frequency towards lower frequency represents longer electron lifetime (τ e ) and slower recombination process. The calculated electron lifetimes for all the cells are summarised in Table 2. The highest electron lifetime of 24.35 ms is obtained for the cell fabricated with 8 mM CDCA concentration along with ZnO compact blocking layer (DSSC 7). The increased electron lifetime due to CDCA and blocking layer effectively enhances the photoconversion efficiency (PCE) which is in good agreement with the results obtained from J-V measurement. The reduced dye aggregation in presence of CDCA and inhibition of electron recombination by the blocking layer may be accounted for this.

Conclusion
Effects of co-adsorption of CDCA and ZnO blocking layer were investigated in Rose Bengal dye based DSSCs. The surface, photovoltaic and electrochemical properties of all the cells were extensively studied. The strong binding of CDCA molecules to the ZnO surface partially displaces dye molecules and consequently reduces photon harvesting. Excessive CDCA concentration implies significantly reduced dye attachment to the ZnO surface leading to decreased amount of light energy absorption. Therefore, to maximize the positive effect of the co-adsorbent, it is very crucial to carefully optimize the amount of CDCA. The amount of CDCA has been optimized by adjusting its concentration in the dye solution and found that the best device performance was obtained for 8 mM concentration. At optimized co-adsorbent concentration, the reduced dye loading due to the presence of CDCA (3) e = 1 2 f peak and consequently decreased light-harvesting was compensated by the increased electron injection efficiency leading to maximum device efficiency of 0.97%. The performance was further increased from 1.00 to 1.22% when a compact ZnO blocking layer was added to the FTO before depositing the mesoporous ZnO active layer. This was due to the suppression of electron back transfer from the FTO to the liquid electrolyte. These results indicate that the addition of CDCA as a dye co-adsorbent and the introduction of ZnO blocking layer is an effective way to boost the performance of Rose Bengal dye based DSSCs. The efficiency of the fabricated cells is low as the dye used in this study is rose bengal. Though the efficiency is low here compared to the ruthenium based cells, it lies in the range of efficiency of rose bengal dye based DSSCs obtained by other researchers. Higher efficiencies can be obtained by using high performance ruthenium dye.
Funding The project was funded by the Department of Physics, University of North Bengal.

Data availability
Datasets generated during the current study are available and shall be provided on reasonable request.

Conflict of interest
There is no financial and non-financial conflict of interest in this research work.

Consent for publication
All the authors of this manuscript agree with the final version of the manuscript and give their consent for it to be published, accepting all the ethical standards of the Journal of Optical and Quantum Electronics. The authors also certify that this manuscript has not been published elsewhere and that it is not under consideration for publication in any other journal.