Design, synthesis, and performance evaluation of TiO2-dye sensitized solar cells using 2,2′-bithiophene-based co-sensitizers

We report on the synthesis and characterization of six novel 2,2′-bithiophene-based organic compounds (3a–c and 5a–c) that are designed to serve as co-sensitizers for dye-sensitized solar cells (DSSCs) based on TiO2. The compounds are linked to various donor and acceptor groups, and we confirm their chemical structures through spectral analyses. Our focus is on enhancing the performance of metal based N3, and the compounds were designed to operate at the nanoscale. We performed absorption and fluorescence emission measurements in dimethylformamide (DMF), where one of our compounds 5a exhibited the longest maximum absorption and maximum emission wavelengths, indicating the significant impact of the para methoxy group as a strong electron-donating group. Our dyes 5a + N3 (η = 7.42%) and 5c + N3 (η = 6.57%) outperformed N3 (η = 6.16%) alone, where the values of short current density (JSC) and open circuit voltage (VOC) for these two systems also improved. We also investigated the charge transfer resistance at the TiO2/dye/electrolyte interface using electrochemical impedance spectroscopy (EIS), which is important in the context of nanotechnology. According to the Nyquist plot, the 5a + N3 cocktail exhibited the lowest recombination rate, resulting in the highest VOC. Our theoretical calculations based on density functional theory (DFT) are also in agreement with the experimental process. These findings suggest that our compounds have great potential as efficient DSSC co-sensitizers. This study provides valuable insights into the design and synthesis of new organic compounds for use as co-sensitizers in DSSCs based on TiO2 and highlights the potential of these compounds for use in efficient solar energy conversion.

The elemental and spectral analyses were used to confirm the structures of bithienylbenzonitrile derivatives 3a-c.The corresponding figures (Figures 1-9) are provided in the supplementary files. 1 H-NMR spectrum of compound 3b displayed two singlet signals of dimethoxy groups at δ 3.78 (para -OCH 3 ), 3.83 (meta OCH 3 ), two doublet signals, at δ 6.99 (J = 8.5 Hz), δ 7.22 (J = 2.5 Hz), along with doublet of doublet signal at δ 7.20 (J = 8.5, 2.5 Hz) corresponding for 1,3,4-trisubstituted benzene ring, and four signals as doublet referring to bithiophene-H's at δ 7.40 (1H), 7.41 (1H), 7.46 (1H), and 7.74 (1H).In addition to, singlet signal at δ 7.86 (4H) of benzonitrile-H's.Mass spectrum of compound 3b gave a molecular ion peak m/z at 403 (M + , 100), along with a fragment at 388 (M + -CH 3 ).Figure 3 outlines the preparation of the bithienylfluorobenzonitrile derivatives 5a-c starting with a Suzuki coupling reaction via cross-coupling reaction of bromo bithienyl derivative 4 31 with the proper phenylboronic acids 2a-c.Structures of bithienylfluorobenzonitriles 5a-c were confirmed based on their elemental and spectral analyses.The corresponding figures (Figures 10-18) are provided in the supplementary files.IR spectra of the bithienyl fluorobenzonitriles 5a-c indicated the presence of cyano group with stretching vibrations in the range of 2228-2229 cm −1 . 1 H-NMR spectrum of bithienyl flurobenzonitrile derivative 5a displayed singlet signal referring to para -OCH 3 at δ 3.78, AA'BB' splitting pattern of two doublet signals at δ 6.98, 7.61 (J = 8.5 Hz, two protons each) of Ar-H of 4-methoxyphenyl ring and two doublet signals one proton each (J = 3.5 Hz) of one thiophene moiety and multiplet signal referring to 2H of the other thiophene moiety, along with doublet of doublet at δ 7.67 (J = 8.0, 1.5 Hz) and multiplet signal corresponding for 1,3,4-trisubstituted benzene ring.Moreover, mass spectrum of compound 5a gave a molecular ion peak m/z at 391 (M + , 100) as a base peak, along with a fragment at 376 (M + -CH 3 ) as outlined in Fig. 4.
Optical measurements.The absorption and fluorescence emission measurements were performed in DMF.The bathochromic shift observed in the fluorescence emission spectra is not as pronounced as that observed in the corresponding absorption spectra.The wavelength of maximum absorption and E 0-0 transition of the fluorescence emission spectra are collected in Table 1.
As shown in Figure 19  www.nature.com/scientificreports/absorbance abs max and maximum a emission em max indicating the effect of methoxy group as strong electron donating group and enhanced strength of the electron withdrawing cyano group by the presence of fluorine atom in its ortho position.Bithiophene derivatives 3c and 5c are similar in the electron donating meta-dimethoxy groups which approximately shows same fluorescence emission maximum and few nanometres bathochromic shift with the fluorinated analogue.The fluorinated analogue of the ortho-dimethoxy derivatives were found to enhance the bathochromic shift as has been observed for the para-mono methoxy analogue.This can be attributed to the presence of ICT process from the meta methoxy in the electron donating moiety to the meta fluorine atom in the acceptor group which interfere with the ICT from ortho methoxy to the cyano pathway.The absorption spectrum of N3 is displayed at Figure 20 in the supplementary file.
We also studied the UV-Vis spectrum of the dyes on the surface of the TiO 2 (Fig. 5).Through the results, it became clear that the spectrum is broader than that of the solution.This confirms the absorption of the dyes on the surface of the TiO 2 .It was also observed that the dyes 5a-c had an absorption at a longer wavelength than that in the solution, which resulted from the J-aggregation 33 .The presence of two linkage groups (CN and F) results in the highest absorption of dyes 5a-c, whereas the cyano group is present alone in dyes 3a-c.The aforementioned results indicate that dyes 5a-c could give interesting results when used in DSSCs.The absorption spectrum of N3 anchored to TiO 2 is displayed at Figure 21 in the supplementary file.
Figure 6 provides additional insight into the mechanism by which our nitrile dyes are anchored to the TiO 2 surface.This anchoring process is supported by the fact that various compounds, including transition metal cyanides, have been observed to form visible charge-transfer (CT) complexes with Ti(IV) ions on surfaces [34][35][36] .This results in strong dye-to-TiO 2 charge-transfer (DTCT) bands, indicating that the nitrile group-containing compounds, (3a-c and 5a-c), serve as effective anchoring groups for TiO 2 .
The nitrile dyes can undergo one-step electron injection to the conduction band (CB) of TiO 2 in a DSSC, similar to transition metal cyanides.Therefore, developing dye systems for electron injection to TiO 2 through this pathway is of great interest academically and practically [36][37][38][39][40] .In this study, we focused on sensitizers with nitrile units that bind to the surface of TiO 2 through chelation of surface Ti(IV) ions 41 .Among the nitrile sensitizers, sensitizer 5a, which contains an electron-withdrawing group (F) and an electron-donating methoxy group, is the best model.The dipolar structure of 5a provides directionality to electronic orbitals in the excited state, which improves direct, "one-step" electron injection from the ground state of the dye to the CB of TiO 2 via photoinduced charge-transfer excitation of the dye-to-TiO 2 bands 42 .The acidic hydrogen relocates to the titanium complex during the shift to the titanium complex from the dyes Figure 22 in the supplementary file.The orbital  www.nature.com/scientificreports/distribution and energy levels of each dye following the interaction with the TiO 2 surface were investigated to study the electronic connection between the LUMO of the dye and the conduction band CB of the semiconductor.

Electrochemical measurements.
There are some requirements that must be met in order to use organic dyes in the field of DSSCs, which depend on the values of the ground state oxidation potential (GSOP) and excited state oxidation potential (ESOP) 34 .With the help of both the cyclic voltammetry and the UV-Vis spectrometer measurements, these values were calculated by the following equation 43 : From Fig. 7, It was found that the ESOP values of the bithiophene dyes 3a-c and 5a-c are categorical (− 3.34 eV, − 3.41 eV, − 3.36 eV, − 2.72 eV, − 2.75 eV, and − 2.74 eV) in order.Since it is higher than the value of the conduction band of the TiO 2 (− 4.2 eV), the electrons will automatically flow easily to the TiO 2 35 .On the other hand, the GSOP values for dyes were found as follows: − 6.10 eV (3a), − 6.13 eV (3b), − 6.20 eV (3c), − 5.37 eV (5a), − 5.63 eV (5b), and − 5.54 eV (5c).It was found at a lower level than the electrolyte (− 5.2 eV), which helps return electrons to the dye to start over [44][45][46] .All this means that the dyes met the conditions for their use in DSSCs.The 5a-c dyes achieved LUMO values at higher levels than the 3a-c dyes, which means that the flow rate of electrons in these dyes to the surface of the semiconductor is higher, and this also indicates that the dyes will give high efficiency when used in DSSCs.

Theoretical investigation. Theoretical calculations using the density functional theory (DFT) of the
Gaussian program were applied to delve deeper into the structural composition of the compounds under study.B3LYP/6-311G (d, p) was used as the functional model and basis set for the first optimization of the structures [47][48][49] .The values of GSOP, ESOP, E 0-0 , and simulated frontier molecular orbitals (FMOs) are displayed and calculated for the different affirmations, and they proved to be in agreement with the practical values that we obtained (Table 2).
Figure 8 shows the optimized structures, FMOS and MEP, of co-sensitizers 3a-c.From their FMOS, the distribution of the electrons in the HOMO is concentrated on the whole molecule, whereas in the LUMO, there was a shift to the acceptor part toward the cyano-moiety.The computed contributions of various electronic transitions, presented in Table 1 (supplementary file), suggest that the ICT in co-sensitizers 3a-c and 5a-c is characterized by a peak absorption in the 470-478 nm range, related to the large contribution of HOMO to LUMO transitions Meanwhile, the π-π* electronic transitions of the compounds, due to significant contributions of HOMO-n to LUMO + n transitions 50 .Using the same basis set and functional energy, we calculated the electrostatic potentials    www.nature.com/scientificreports/(MEPs) of the dyes under investigation.The MEP concentrates on anticipating physicochemical features like the reactive groups associated with the molecular structure 51 .Also, MEP is a graph of the electrostatic potential, plotted onto the constant electron density surface 52 .Different colors are used to show ESP levels.An electron-rich field is shown by the red area, while electron-poor sections are indicated by the blue region 53 .Also, Fig. 9 shows the optimized structures, FMOS and MEP, of co-sensitizers 5a-c.For co-sensitizers 5a-c, the distribution of the electrons in the HOMO is concentrated on the donor part, whereas in the LUMO, there was a shift to the acceptor part toward the acceptor-anchoring moiety, which indicated better charge transfer as shown in their FMO levels.MEP also demonstrated the high electron density parts and low charge parts according to their different colors, from red to blue.

Photovoltaic measurements:
. With an iodolyte electrolyte and simulated AM 1.5 G illumination, photocurrent density-voltage (I-V) curves of DSCs based on dyes 3a-b and 5a-b together with N3 were done.The supplementary file contains information on the fabrication process and instruments used for the photovoltaic measurements [54][55][56] .Because of what the previous measurements showed and the small size of the dyes 3a-c and 5a-c that were made, they were used as co-sensitizers in DSSCs to make the standard dye (N3) work better.Figure 10 displays the I-V curves of the co-sensitizers 3a-c, 5a-c and N3.The corresponding data is recorded in Table 3.
As shown in Table 3, dyes 5a + N3 (η = 7.42%) and 5c + N3 (η = 6.57%) achieved higher efficiencies than N3 (η = 6.16%) alone.The value of J SC and V OC for these two systems also improved.The dye absorbs the main dye, increasing the number of electrons flowing into the conductor and resulting in a high J SC value.As for the increase in V OC , it is due to the filling of the existing interstitial spaces due to the adsorption of the dyes, which led to a larger number of dyes that bore on the surface of the conductor, and this leads to a decrease in aggregation and an increase in the voltage.Also, the co-sensitization impact of 3a-c, 5a-c, and N3 was further investigated by measuring the loading onto the photoanode, which directly proportional to the photocurrent value.Table 3 summarizes the findings of measuring the concentration of co-sensitizers 3a-c, 5a-c, and N3 adsorbed on TiO 2 using 0.1 M NaOH in a dimethylformamide and water (1:1) combination.As a result of its improved J SC and suitably expanded light collecting capacities, 5a provided the optimum co-sensitization performance.The increase in V OC and J SC consequently led to an increase in cell efficiency.
It is also noted that the dyes 3a-c did not achieve a positive effect on the dye N3, and this was originally due to the nature of the co-sensitizer and its inability to adhere well to the surface of the TiO 2 , i.e., an aggregation of the dye occurred, and the recombination increased, so the voltage decreased.Also, the N3 didn't get well absorbed because the N3 and co-sensitizer were aggregated together.This made it harder for electrons to get into the TiO 2 , so the J SC went down.Table 2 (Supplementary file) compares the performance parameters of 3a-c, 5a-c and N3 with recently reported high-performance sensitizer and co-sensitizer.Furthermore, the electronic distribution for co-sensitizers 3a-c, from Fig. 8, the HOMO for co-sensitizers 3a-c is randomly distributed throughout the framework, whereas the LUMOs were located in the anchoring parts (cyano units) through the biphenyl-spacer, this means that the LUMO electron destination does not really tightly coincide with TiO 2 's conduction band (CB), which could lead to poor electron injection efficiency from the dyes' LUMO and so the J SC values went down comparing to the others co-sensitizers and increasing the charge recombination rate at the interface of the TiO 2 / dye/electrolyte, which can be related to the energy level alignment at the interface.The bithienyl co-sensitizers 5a-c exhibit efficient charge transfer between their highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) levels, as seen in Fig. 9.The electron density is primarily localized on the substituent anchoring groups, allowing for electronic interactions between the TiO 2 d-orbitals and the LUMO electron density.Additionally, the coexistence of the HOMO and LUMO levels facilitates rapid photo-induced electron transfer from the donating portion to the acceptor group.Because of this increased charge separation, 5a has a much higher photovoltaic efficiency than other sensitizers.Further dye 5b did not improve the performance of the N3 although the increased in the J SC , which may be attributed to the recombination of the dye on the surface that result in the decreased V OC value and hence the decrease of the efficiency.To learn more about the mechanism of the absorption of dyes 3a-c and 5a-c on the surface of TiO 2 , the IR-spectrum of the dyes on the surface of TiO 2 was measured, as shown in Figures 23-28 in the supplementary files.It was found that the value of the cyano group (CN) disappeared, or its absorption decreased in the samples, which confirms that the binding was through the cyano group.Figure 11 shows the incident photon-to-current conversion efficiency (IPCE) spectra of devices sensitized with N3 and co-sensitized with 2,2′-bithiophene-based 3a-c and 5a-c.The IPCE spectra for all these cells exhibited a broad response within a large wavelength range (300-650 nm), indicating that all the dyes can efficiently convert the visible light into photocurrent.The higher IPCE values of the devices co-sensitized with the bithiophene co-sensitizers 3a-c and 5a-c can be attributed to their improved photovoltaic characteristics.Additionally, N3-based cells fabricated with co-sensitizer 5a-c demonstrated higher quantum efficiency (IPCE value) than N719 alone, possibly due to the appropriate mixing of organic dye and Ru-II complex and presence of auxiliary electron withdrawing group.Furthermore, the increased IPCE values can be attributed to increased J SC values, the IPCE integral areas of DSSCs exhibit an order for dyes of 5a + N3 > 5c + N3 > 5b + N3 > N3 > 3a + N3 > 3b + N3 > 3c + N3, trend of is consistent with the trend of J SC .The cell co-sensitized by 5a + N3 had the highest IPCE response corresponding to its highest Jsc value of 18.14 mA cm −2 and gave over 65% IPCE values from 300 to 650 nm.This observation strongly advocates that during the dye loading process of co-sensitizers   www.nature.com/scientificreports/along with N3 should have interacted with each other.Such interaction generally tends to induce electron and energy transfer between the two kinds and hence causes deterioration in cell performance.The co-sensitized cell exhibits the highest Jsc value in the J-V measurements, which can be attributed to its broadest and highest IPCE response 57 .Additionally, the co-sensitized cell's improved Voc value, coupled with its highest Jsc value, contributes to its further increased PCE value when compared to that of 5a + N3, the improved IPCE response of (3a-c, and 5a-c is interpreted in terms of higher J SC value which showed the same order of 5a + N3 = 18.).These results implies that the structural optimization with 3a-c and 5a-c architecture is a key in getting greater efficiency, compared to the Jsc values obtained from the J-V data, the J sc IPCE values integrated from the IPCE spectra are quite consistent.As a result, the co-sensitizer 5a dye produces the most abundant IPCE spectrum, proving that it also has the greatest Jsc.The enhanced IPCE replies match the enhanced Jsc results.
the other hand, the value of cyano and F disappeared for 5a-c, which confirms the association of these dyes with cyano and fluorine together.Electrochemical impedance spectroscopy (EIS) was used to study the charge transfer resistance at the TiO 2 /dye/electrolyte interface 58,59 .This was done to learn more about the relationship between the structure of the molecules and their photovoltaic performance.The middle area shown in the Nyquist plot, as shown in Fig. 12, represents the relationship between the TiO 2 , the dye, and the electrolyte.The diameter of this part of the curve has a clear relationship with dye recombination; the larger the diameter, the less likely it is to be recombined, resulting in an increase in open circuit voltage 60 .On the Nyquist plots, the radius of the large semicircle in the middle frequency range was found to be 5a + N3 > 5c + N3 > N3 > 5b + N3 > 3b + N3 > 3a + N3 > 3c + N3, which shows the order of charge recombination resistance R rec at the TiO 2 /dye/ electrolyte interface.DSSCs with an increasing R rec value have slower charge recombination between the electron  (14.29 Ω), in good agreement with the order photovoltage data.As a result, the V oc increases.Accordingly, this sequence of V oc values appears to be consistent.As a result, the 5a + N3 system with the largest diameter has the lowest recombination rate and the highest V OC , which is consistent with the photovoltaic values.
In Bode frequency plots Fig. 13, by applying the following equation the electron lifetime for injected electrons into TiO 2 conduction band can be calculated by (τ eff = 1/2πf) 58 , where τ represents the electron lifetime injected into TiO 2 and f is the mid-frequency peak in bode plots, which is directly related to the electron lifetimes.The electron lifetimes of DSSCs that were sensitized using co-sensitizers were determined through Bode frequency plots.By analyzing the Bode phase plots, the frequency of the peaks observed in the middle-frequency domain can be used to assess the electron lifetime within the semiconductor (τ r (ms).This provides additional insight into the charge recombination rate occurring at the interface of TiO  54 , as well as the dye adsorption behavior.These factors have a strong influence on the photovoltage of solar cells, as has been previously reported in the literature.In this study, the values of R rec and se for the new co-sensitizer dyes were found to be consistent with the corresponding V OC values obtained for the solar cells.The use of the 5a co-sensitizer in DSSCs resulted in higher V OC values compared to other sensitizers, including N3, due to the lower charge recombination rate at the TiO 2 /dye/ electrolyte interface, which is attributed to the strength of the donating moiety.
The stability of dye-sensitized solar cells (DSSCs) is a key issue that must be addressed to ensure their longterm viability as a renewable energy technology, the stability of co-sensitizers has been measured and added into the manuscript, As depicted in Fig. 14, the high photo-stability of the 3a-c and 5a-c/N3 dyes for DSSC applications is confirmed by the fact that the cell exhibits excellent stability with no discernible degradation of the initial performances even after 100 h of illumination.The superior long-term stability of co-sensitizer 5a may be attributed to the hydrophobic properties of the substituted methoxy group attached to bithienyl spacer addition to acceptors moieties (F and CN).

Conclusion
Our team accomplished significant success by designing and synthesizing six organic dyes (3a-c and 5a-c) using a simple yet effective D-π-A structure.By displaying the structural effects of electron donor and acceptor groups, these dyes have evaluated a new trend for improving the photovoltaic capabilities of dye-sensitized solar cells (DSSCs).To evaluate the efficiency of our dyes, we co-sensitized them with the standard dye N3.The absorption spectra of co-sensitized TiO 2 films become more intense and broader than the individual dyes.The EIS data indicates a reduced recombination of injected electrons with the triiodide ions and a longer electron lifetime.Therefore, an improvement of open circuit photovoltage (Voc) is achieved for 5a and 5c.Under optimal conditions, the power conversion efficiency ranges (4.13-7.42%)and short circuit current (J SC ) range (14.13-18.14mA cm −2 ).Our mixed dyes 5a and N3 exhibited a remarkable efficiency of 7.42%, and with further optimization, we obtained a short-circuit current density of 18.14 mA cm −2 and a high open-circuit voltage (V OC ) of 0.676 V.The exceptional J SC value can be attributed to the presence of two linkage groups (CN and F) that increase the absorption of dye 5a, as confirmed by the UV-Vis absorption over TiO 2 surfaces.The maximum V OC value, on the other hand, is due to the extensive surface area covered by both dyes 5a and N3.The increase in Jsc and Voc are strongly correlated to enhance light absorption and reduced charge recombination, respectively.Our findings suggest that tuning the electron-donating groups is critical to improving the light-harvesting efficiency of co-sensitizers.

Table 2 .
Electrochemical data for compounds 3a-c and 5a-c.Compound Practical results (