Unless otherwise specified, all the reactions were performed under nitrogen atmosphere using standard Schlenk techniques. All solvents used were purified by standard procedures, or purged with nitrogen before use. ¹H NMR spectra were recorded on a Bruker 300-MHz or 400-MHz spectrometer. Absorption spectra were recorded on a Cary 50 probe UV-Vis spectrophotometer. All chromatographic separations were carried out on silica gel (60 M, 230–400 mesh). Mass spectra (FAB) were recorded on a VG70-250S mass spectrometer. The photoelectrochemical characterizations on the solar cells were carried out using an Oriel Class A solar simulator (Oriel 91195A, Newport Corp.). Photocurrent-voltage characteristics of the DSSCs were recorded with a potentiostat/galvanostat (CHI650B, CH Instruments, Inc.) at a light intensity of 100 mW cm⁻²calibrated by an Oriel reference solar cell (Oriel 91150, Newport Corp.). The monochromatic quantum efficiency was recorded through a monochromator (Oriel 74100, Newport Corp.) at short circuit condition. The intensity of each wavelength was in the range of 1–3 mW cm⁻². Electrochemical impedance spectra (EIS) were recorded for DSSC under illumination at open-circuit voltage (V_OC) or dark at −0.55 V potential at room temperature. The frequencies explored ranged from 10 mHz to 100 kHz. The TiO₂ nanoparticles and the reference compound, N719, were purchased from Solaronix, S.A., Switzerland.

All the new dyes were prepared via Knoevenagel condensation reaction by reacted corresponding aldehyde derivatives and cyanoacetic acid in the presence catalytic amount of ammonium acetate.

2-Cyano-3-(5-(6-(diphenylamino)-9,10-bis(hexyloxy)anthracen-2-yl)thiophen-2-yl)acrylic acid (An-1). Spectroscopic data for An-1: ¹H NMR (400 MHz, acetone-d₆): δ 8.63 (d, J = 1.6 Hz, 1H), 8.46 (s, 1H), 8.24 (d, J = 8.8 Hz, 1H), 8.22 (d, J = 8.8 Hz, 1H), 8.04 (d, J = 4.0 Hz, 1H), 7.68-7.84 (m, 4H), 7.32 (dd, J = 8.8, 2.4 Hz, 1H), 7.79 (d, J = 8.0 Hz, 2H), 7.53 (d, J = 4.0 Hz, 1H), 7.41–7.27 (m, 10H), 7.24–7.16 (m, 6H), 4.25 (t, J = 6.4 Hz, 2H), 3.95 (t, J = 6.4 Hz, 2H), 1.85–1.64 (m, 4H), 1.46–1.40 (m, 6H), 0.95–0.90 (m, 6H). ¹³C NMR (CDCl₃): δ 168.1, 156.5, 148.8, 147.5, 147.2, 145.8, 139.9, 134.5, 129.4, 128.3, 125.2, 124.6, 123.4, 123.8, 123.5, 123.1, 122.8, 121.0, 115.5, 96.2, 75.6, 31.8, 30.6, 30.3, 26.0, 25.6, 22.7, 22.5, 14.1. MS (FAB): m/z 722.1 (calcd [M]⁺). HRMS m/z [M + H]⁺ calculated for C₄₆H₄₆N₂O₄S, 723.3256 found: 723.3282.

3-(5-(9,10-Bis(hexyloxy)-6-(naphthalen-1-yl(phenyl)amino)anthracen-2-yl)thiophen-2-yl)-2-cyanoacrylic acid (An-2). Spectroscopic data for An-2: ¹H NMR (400 MHz, acetone-d₆): δ 8.61 (d, J = 1.2 Hz, 1H), 8.45 (s, 1H), 8.21 (d, J = 8.0 Hz, 1H), 8.18 (d, J = 8.0 Hz, 1H), 8.05–8.00 (m, 3H), 7.97 (d, J = 8.0 Hz, 1H), 7.84–7.82 (m, 2H), 7.63 (t, J = 7.2 Hz, 1H), 7.53 (t, J = 7.2 Hz, 1H), 7.46 (d, J = 8.0 Hz, 1H), 7.43–7.34 (m, 5H), 7.25 (d, J = 8.0 Hz, 2H), 7.13 (t, J = 7.2 Hz, 1H), 4.24 (t, J = 6.8 Hz, 2H), 3.90 (t, J = 6.8 Hz, 2H), 1.81–1.71 (m, 4H), 1.49–1.21 (m, 6H), 0.92 (t, J = 7.8 Hz, 6H). ¹³C NMR (CDCl₃): δ 168.2, 156.5, 148.8, 147.5, 147.4, 146.2, 145.7, 143.1, 139.9, 135.2, 134.4, 130.9, 129.2, 128.4, 127.8, 127.3, 127.1, 126.8, 126.4, 126.3, 126.1, 125.1, 124.5, 124.0, 123.9, 123.6, 123.4, 123.2, 123.1, 122.7, 122.5, 121.0, 115.5, 109.5, 96.1, 75.7, 31.7, 31.6, 30.6, 30.1, 26.0, 25.4, 22.6, 22.5, 14.1, 14.0. MS (FAB): m/z 772.1 (calcd [M]⁺). HRMS m/z [M]⁺ calculated for C₅₀H₄₈N₂O₄S, 772.3177 found: 772.3193.

3-(5-(6-(Bis(9,9-diethyl-9H-fluoren-2-yl)amino)-9,10-bis(hexyloxy)anthracen-2-yl)thiophen-2-yl)-2-cyanoacrylic acid (An-3). Spectroscopic data for An-3: ¹H NMR (400 MHz, acetone-d₆): δ δ 8.54 (d, J = 1.2 Hz, 1H), 8.45 (s, 1H), 8.25 (d, J = 8.4 Hz, 1H), 8.02 (d, J = 4.0 Hz, 1H), 7.98–7.78 (m, 7H), 7.73 (d, J = 2.0 Hz, 1H), 7.46 (dd, J = 8.4, 2.4 Hz, 1H), 7.41 (d, J = 7.2 Hz, 2H), 7.37–7.30 (m, 9H), 4.28 (t, J = 6.4 Hz, 2H), 3.95 (t, J = 6.8 Hz, 2H), 1.90–1.63 (m, 8H), 1.59–1.11 (m, 10H), 0.94 (t, J = 6.4 Hz, 3H), 0.74 (t, J = 6.4 Hz, 3H), 0.37 (t, J = 7.2 Hz, 12 H). ¹³C NMR (CDCl₃): δ 167.7, 156.6, 149.8, 148.8, 147.6, 146.7, 146.1, 141.2, 139.8, 137.5, 134.5, 128.1, 126.9, 126.5, 124.7, 124.4, 123.9, 123.0, 122.8, 121.2, 120.8, 120.3, 119.7, 119.2, 115.6, 96.4, 56.2, 32.7, 31.8, 31.6, 30.6, 30.3, 26.0, 25.7, 22.6, 22.4, 14.1, 13.9, 8.6. MS (FAB): m/z 1011.3 (calcd [M]⁺). HRMS m/z [M + H]⁺ calculated for C₆₈H₇₀N₂O₄S, 1011.5134 found: 1011.5176.

3-(5-(6-(Anthracen-2-yl(phenyl)amino)-9,10-bis(hexyloxy)anthracen-2-yl)thiophen-2-yl)-2-cyanoacrylic acid (An-4). Spectroscopic data for An-4: ¹H NMR (400 MHz, acetone-d₆): δ δ 8.65 (d, J = 1.6 Hz, 1H), 8.52 (s, 1H), 8.45 (s, 1H), 8.34 (s, 1H), 8.27 (d, J = 7.2 Hz, 1H), 8.25 (d, J = 7.2 Hz, 1H), 8.09 (d, J = 8.0 Hz, 1H), 8.06 (dd, J = 8.0, 4.0 Hz, 1H), 8.02 (d, J = 4.0 Hz, 1H), 7.97 (dd, J = 8.0, 4.0 Hz, 1H), 7.88 (dd, 8.0, 1.6 Hz, 1H), 7.85 (d, J = 4.0 Hz, 1H), 7.75 (d, J = 2.4 Hz, 1H), 7.72 (d, J = 1.2 Hz, 1H), 7.48–7.41 (m, 6H), 7.34 (d, J = 8.0 Hz, 2H), 7.25 (t, J = 7.2 Hz, 1H), 4.28 (t, J = 6.8 Hz, 2H), 3.96 (t, J = 6.8 Hz, 2H), 1.90–1.63 (m, 4H), 1.55–1.48 (m, 6H), 0.97–0.92 m, 6H). ¹³C NMR (CDCl₃): δ 167.2, 156.3, 148.9, 147.4, 147.0, 146.1, 145.5, 144.2, 139.7, 134.6, 132.5, 132.2, 131.2, 129.5, 129.3, 128.1, 127.8, 127.3, 126.0, 125.6, 125.2, 125.1, 124.9, 124.8, 124.6, 124.5, 124.2, 124.1, 123.8, 123.3, 123.1, 121.1, 120.1, 115.7, 112.9, 96.5, 76.2, 67.9, 31.8, 31.6, 30.6, 30.3, 26.0, 25.6, 22.6, 22.3, 14.1, 13.9. MS (FAB): m/z 823.3 (calcd [M]⁺). HRMS m/z [M]⁺ calculated for C₅₄H₅₀N₂O₄S, 822.3491 found: 822.3511.

The photoanode used was the TiO₂ thin film (12 μm of 20 nm particles as the absorbing layer and 6 μm of 400 nm particles as the scattering layer) coated on FTO glass substrate with a dimension of 0.5 × 0.5 cm² [35]. The film thickness measured by a profilometer (Dektak3, Veeco/Sloan Instruments Inc., USA). A platinized FTO produced by thermopyrolysis of H₂PtCl₆ was used as a counter electrode. The TiO₂ thin film was dipped into the THF solution containing 3 × 10⁻⁴ M dye sensitizers for at least 12 h. After rinsing with THF, the photoanode adhered with a polyester tape of 60 μm in thickness and with a square aperture of 0.36 cm² was placed on top of the counter electrode and tightly clipping them together to form a cell. Electrolyte was then injected into the space and then sealing the cell with the Torr Seal cement (Varian, MA, USA). The electrolyte was composed of 0.5 M lithium iodide (LiI), 0.05 M iodine (I₂), and 0.5 M 4-tert-butylpyridine that was dissolved in acetonitrile.

The computations were performed with Q-Chem 4.0 software [36]. Geometry optimization of the molecules were performed using hybrid B3LYP functional and 6-31G* basis set. For each molecule, a number of possible conformations were examined and the one with the lowest energy was used. The same functional was also applied for the calculation of excited states using time-dependent density functional theory (TD–DFT). There exist a number of previous works that employed TD–DFT to characterize excited states with charge-transfer character [37,38]. In some cases underestimation of the excitation energies was seen [36,39]. Therefore, in the present work, we use TD–DFT to visualize the extent of transition moments as well as their charge-transfer characters, and avoid drawing conclusions from the excitation energy.

Scheme 1 illustrates the synthetic route towards the target molecules (An-1 to An-4). Kumada cross-coupling [40] of thienyl magnesium bromide with 2,6-dibromo-9,10-dihexyloxyanthracene (1) generated 2 with a monosubstituent, which then underwent C–N coupling reaction [41,42] with diarylamine to afford 3a–3d. These intermediates were conveniently converted to aldehydes (4a–4d) by lithiation with n-butylithium followed by quenching with dimethylformamide. The aldehydes were converted to the dyes via Knoevenagel condensation with cyanoacetic acid in acetic acid in the presence of ammonium acetate as the catalyst. The final target dyes are obtained as dark red powders, and are soluble in common organic solvents such as THF, CH₂Cl₂, and CHCl₃.

The UV-Vis absorption spectra of the sensitizers An-1 to An-4 are shown in Figure 1, and their photophysical data are listed in Table 1. All dyes exhibit absorption bands in two distinct regions. The band(s) at 300–400 nm may be attributed to the π–π* electron transition of the conjugation backbone, and the band at around 450–550 nm has prominent intramolecular charge transfer (ICT) from the diarylamine and anthracene to 2-cyanoacrylic acid. The absorption band below 400 nm exhibited vibronic characteristics of anthracenyl entity for all dyes except An-3. The vibronic pattern may be buried under the intense absorption band due to bis(fluorenyl)amino moiety [18]. The λ_max value of the ICT band decreases in the order of An-3 > An-4 > An-1 ≈ An-2. The influence of the donor on the ICT band is evident: An-3 has the longest absorption wavelength with the strongest absorption because of strong electron-donating ability of bis-diethylfluorenyl amino group [43]. The significantly bathochromic and hyperchromic shifts of the ICT band in An-4 compared to An-1 and An-2 may also be attributed to the incorporation of electron excessive anthracenyl entity at the arylamino donor. Compound An-1 exhibits absorption peak at 479 nm (ε = 3.7 × 10⁴ M⁻¹cm⁻¹), which is almost the same as that of An-2 (477 nm, 3.48 × 10⁴ M⁻¹cm⁻¹), indicating that the phenyl and naphthyl rings at the N atom have similar influence on the arylamine donor. It is interesting to compare the UV absorption for the 2,6-difuntionalized anthracene entity as the spacer between the donor and the acceptor is advantageous for intramolecular charge transfer compared to the 9,10-difuntionalized anthracene spacer, as witnessed from comparison of the spectra between An-2 and a 9,10-difuntionalized anthracene-based dye (5 in Figure 2) with a similar structure [34]. Compound An-2 has red-shifted absorption with a higher molar extinction coefficients compared to 5 (434 nm, ε = 1.31 × 10⁴ M⁻¹cm⁻¹). Non-coplanar conformation of 9,10-difuntionalized anthracene undoubtedly jeopardizes the charge transfer.

Significant blue shift of the spectra of the dyes on TiO₂ surface (Figure 3) may be ascribed to the deprotonation of the carboxylic acid [16], which was also confirmed by measuring the absorption of the dyes in the presence of Et₃N. Certain degree of J-aggregation [44,45,46] may also occur, as evidenced from the prominent red tailing of the absorption spectra.

The energetic alignment of highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) energy levels is crucial for an efficient operation of the sensitizer in DSSCs. To judge the feasibility of electron transfer from the excited dye molecule to the conduction band of TiO₂ electrode, redox potentials of these dyes were investigated by cyclic voltammetry and the data are listed in Table 1. Representative cyclic voltammograms are shown in Figure 4. The redox potentials of these new dyes were measured in THF with 0.1 M tetra-butylammonium hexafluorophosphate. All the dyes exhibited one quasi-reversible redox wave due to the oxidation of arylamine. The oxidation potentials of An-1, An-2, An-3 and An-4 were measured to be 1.05, 1.06, 1.01 and 1.06 V versus NHE, respectively. The lowest oxidation potential of An-3 is consistent with the presence of strong electron-donating bis-diethylfluorenyl amino group. All the oxidation potential are more positive than that of I⁻/I₃⁻ (0.4 V vs. NHE) [47], ensuring favorable regeneration of dyes. The oxidation potentials of these materials were calculated with the energy gap (E_0-0, 2.12–2.29 eV) obtained from the cutoff wavelength of the absorption spectra was utilized to derive the excited state potentials. The excited state potential (vs. NHE) for sensitizers An-1, An-2, An-3, and An-4 were calculated to be −1.24, −1.23, −1.11 and −1.08 V, respectively, indicating enough driving force for electron injection from the photo-excited sensitizers to the conduction band of the TiO₂ electrode (−0.5 V vs. NHE) [44]. These results clearly demonstrate that the novel dyes are potentially efficient dyes for DSSCs.

The dye-sensitized solar cells were constructed by using these dyes as a sensitizer for nanocrystalline anatase TiO₂ particles, and the electrolyte composed of 0.05 M I₂/0.5 M LiI/0.5 M tert-butylpyridine in acetonitrile solution. The photovoltaic performance statistics under a solar condition (AM 1.5) illumination are collected in Table 2.

The photocurrent-voltage (J-V) curves, dark current and the incident monochromatic photo-to-current conversion efficiency (IPCE) plots of the cells were shown in Figure 5 and Figure 6, respectively. The short-circuit photocurrent density (J_SC), open-circuit voltage (V_OC) and fill factor (FF) of device are in the range of 4.52–7.30 mA/cm², 0.55–0.64 V and 0.65–0.67, respectively, corresponding to an overall conversion efficiency of 1.62%–2.88%. The reference standard of ruthenium dye N719-based cell (conversion efficiency = 7.11%) fabricated and measured under similar conditions. The cells exhibit a maximum IPCE between 450 and 500 nm and the IPCE spectra is broadened in the long wavelength side extending to 650 nm (Figure 6). The values of the dye density on TiO₂ film are also listed in Table 2 and found to decrease in the order ofAn-1 (3.97 × 10⁻⁷ mol/cm²) > An-2 (3.17 × 10⁻⁷ mol/cm²) > An-4 (3.14 × 10⁻⁷ mol/cm²)> An-3 (2.88 × 10⁻⁷ mol/cm²). Likely the bulkier bis-diethylfluorenyl amino group of the An-3 results in the lowest dye density. Despite of relatively low dye density and unfavourable charge recombination with the electrolyte (vide infra), the An-3 cell still exhibited the best performance. This outcome can be attributed to the good light harvesting efficiency (long absorption wavelength and high molar extinction coefficient) and better electron collection because of lower electron transport resistance (vide infra). In contrary, the An-4 cell exhibited the lowest performance. This might be due to the higher dark current (Figure 5) and the inferior electron injection (vide infra). The cell efficiency of An-1 is higher than that of 5 [34]. Apparently light harvesting plays a very important role in cell performance. The dark currents of DSSCs fabricated were also checked (Figure 5). The V_OC data of DSSCs decreases in the order of N719 > An-2 > An-1 > An-3 > An-4. Large dark currents of the cells from An-3 and An-4 are consistent with their lower V_OC values, though the influence of Fermi-level variation cannot be ruled out.

The electrochemical impedance spectroscopic (EIS) measurement under illumination was shown in Figure 7. Upon illumination of 100 mW cm⁻² under open circuit conditions, the radius of the intermediate frequency semicircle in the Nyquist plot (Figure 7) represents the electron transport resistance. The electron transport resistance (R_ct) (Table 2) decreased in the order of An-4 (107.6 Ω) > An-2 (58.5 Ω) > An-1 (44.0 Ω) > An-3 (40.0 Ω) > N719 (20.3 Ω). This is consistent with the cell performance. The aforementioned trend of dark current was also supported by the EIS studies carried out in the dark (Figure 8). The intermediate frequency semicircle represents the charge recombination resistance on the TiO₂ surface (R_rec), where the larger R_rec value implies the smaller dark current. The R_rec values decreases in the order of N719 > An-2 > An-1 > An-3 > An-4, respectively. This is also consistent with the V_OC values.

The series of dyes An-1−An-4 have been modeled by density functional calculations at B3LYP/6-31G* level of theory. In the optimized structure, the dihedral angle between the anthracene and the thiophene units is smaller than 23° (Figure 9), which is significantly smaller than that (37°–52°) observed in the organic sensitizers based on 1,4-naphthyl unit developed earlier by us [35]. The calculated An-3 molecule has the highest HOMO energy level (−4.77 eV) among all, which is consistent with the electrochemical data (Figure 10).

The time-dependent DFT calculations also confirm the absorption behavior of the dyes in the UV-Vis spectra (see Table 3). The S₀ → S₁ excitations of the dyes having the oscillator strength between 0.55–0.60 are mainly contributed from HOMO → LUMO transitions (99%). The HOMOs of the dye largely populate on the arylamine and the anthracenyl entity, and the LUMOs are mainly populated on the thienyl entity and 2-cyanoacrylic acid (see also Figure 11). Consequently, charge transfer from both arylamine and anthracent to 2-cyanoacrylic acid is prominent. This is also supported by the analysis of Mulliken charges in different fragments (PhN/NapN/FluN/AntN: amines; Ant: 2,6-dihexyloxyanthracene; T: thiophene; Ac: 2-cyanoacrylic acid) of the dyes for the S₁ state (Figure 12).