Titania Nanorods Embedded with 2-Bromo-3-(methylamino)naphthalene-1,4-dione for Dye-Sensitized Solar Cells.

In a recent study, TiO2 nanorod electrodes were prepared by the hydrothermal approach followed by calcination at various temperatures from 300 to 600 °C. The effects of calcination temperature on the morphological and structural properties were investigated. The novel analogue of aminonaphthoquinone(2R-(n-alkylamino)-1,4-naphthoquinone) photosensitizer, viz. BrA1, 2-bromo-3-(methylamino)naphthalene-1,4-dione was synthesized from 2,3-dibromonaphthalene-1,4-dione. X-ray crystallographic data collection and refinement confirm that BrA1 crystallizes in the triclinic space group P 1̅. After loading BrA1, the photosensitizer on the annealed TiO2 nanorod (TiO2NR) electrodes, the optical properties of the photoanodes showed broadbands in each of the UV and visible regions, which are attributed to the π →π* and n → π* charge-transfer transitions, respectively. The dye-sensitized solar cell (DSSC) system was formed by loading the BrA1 photosensitizer on TiO2NR. The electrochemical impedance spectroscopy (EIS) analyses confirm that calcination temperature improves the charge transportation by lowering the resistance path during the photovoltaic process in TiO2NR (400 °C) photoanode-based DSSCs due to the sufficient photosensitizer adsorption and fast electron injection. Due to the effective light harvesting by the BrA1 photosensitizer and charge transport through the TiO2 nanorod, the power conversion efficiencies (PCE) of the TiO2NR (400 °C/BrA1-based) DSSCs were improved for 2-bromo-3-(methylamino)naphthalene-1,4-dione.

Introduction

To meet the rising energy need, the search for new energy sources and easily fabricated photovoltaic devices has recently been of paramount importance. Solar light energy is the clean, most abundant, sustainable, and inexpensive source among the available renewable sources.[1−3] Photovoltaic devices have attracted great interest in achieving efficient solar energy utilization due to their effective solar energy to electrical energy conversion efficiency.[4−7] However, the photovoltaic performance of photovoltaic devices (DSSCs) significantly depends on the properties of the photoanode materials and their structure.[3,5,8,9] Till today, numerous semiconductors (ZnO, TiO2, Zn2SnO4, and SnO2) have been used as photoanode materials to drive reactions in solar cells.[10−16] Due to its specific properties such as nontoxicity and strong chemical stability, TiO2 is considered one of the capable photocatalyst candidates.[17,18] However, due to the wide band gap (3.0–3.2 eV), TiO2 can absorb only ultraviolet light from the solar spectrum. Thus, out of the total light available from the solar spectrum, TiO2 absorbs only 4–5% incident light, while nearly 43% of the visible light in the solar energy is wasted.[19] Thus, under visible-light irradiation, the overall efficiency of light absorption over TiO2 is limited.[20,21] However, in the case of the TiO2 nanorod photoanodes-based DSSCs, TiO2 acts as a framework to transport the electron photogenerated by light absorption and helps in adsorption of more dye molecules and regeneration.[22] Till today, various attempts have been made to advance TiO2 transport ability to increase DSSC solar energy conversion efficiencies.[23] TiO2 semiconductors with different morphologies of nanorods, nanotubes, spherical, and squares have been synthesized by various sol–gel and hydrothermal approaches.[24−26] This hydrothermal approach led to simple operation conditions to produce TiO2 with different shapes and controlled surface area with porosity. This assists in modification of the surface area, crystal size, and band-gap energy of the TiO2 semiconductor to improve DSSC efficiency.[27] Mathew et al. used the TiO2 electrode, porphyrin photosensitizer (SM 351), with the cobalt (II/III) redox shuttle to form DSSCs that exhibited 13% PCE at 1 sun illumination.[28] Further, Kakiage et al. also reported using the silyl-anchor photosensitizer of ADEKA-1 and the carboxy-anchor group organic photosensitizer LEG4 in DSSCs, and obtained 14.3% power conversion efficiency using TiO2 photoelectrodes and graphene nanoplatelets (GNP) as a counter electrode with the cobalt(II/III) redox shuttle.[29] However, the effects of various parameters such as the photosensitizer’s structure, electrolyte, photoanode morphology, and the counter electrode effect on the efficiency of DSSCs have been studied.[30−33] Furthermore, in addition to the reported well-known photosensitizers, organic photosensitizers are expected to have various advantages compared to inorganic photosensitizers.

Generally, these organic photosensitizers are made up of a donor-π-acceptor (D-π-A) system. When used to photosensitize the photoanode materials, the acceptor parts are attached to the surface of semiconductors.[34] Therefore, the Nb2O5 photoanode sensitized with N3 dye showed 2% efficiency,[35] whereas ZnO sensitized with N719 dye exhibited 5% PCE.[36] In this investigation, BrA1 was used as a donor-π-acceptor (D-π-A) photosensitizer in the DSSC structures. DSSCs fabricated using aminonaphthoquinone-based photosensitizers, viz. 2-bromo-3-propylamino-1,4-naphthoquinone and 2-bromo-3-butylamino-1,4-naphthoquinone, and mesoporous ZnO photoelectrodes were shown to have an impressive startup PCE of 0.13 and 0.20%, respectively.[37] As reported in the literature, the modifications of photoanode in terms of morphology, doping, and film thickness have significant influences on the PV performance of DSSCs.[38] On the one hand, the photosensitive dyes provide photoelectrons; on the other hand, the bulk of semiconductor materials is only used as a charge transporter.[39] However, due to their nanorod structures, even though the TiO2 nanorod-based electrodes exhibit effective charge transport pathways for the photogenerated electrons in DSSCs, the dye loading in the ZnO-based powder was higher in our previous work due to the nanograin structures,[37] which gave sufficient surface area for light harvesting. However, to improve the dye loading and light absorption, in the near future, we will modify the surface of TiO2NR-based photoanode materials, particularly by loading the TiO2 grains on TiO2 nanorods to improve the dye loading and light absorption simultaneously. Previous literature shows that excessive research has been done on DSSCs; however, in the 1970s, the ZnO-single crystals based on dye-sensitized solar cells[40] showed very poor efficiency due to the monolayer of dye molecules being able to absorb only 1% incident light. Furthermore, progress has been made in improving the efficiency by optimizing various parameters such as the porosity of the electrode, surface area, and optical transparency of the photoelectrode, by designing the sensitizers with functional groups such as −COOH, −PO3H2, and −B(OH)2.[41−43] In DSSCs, the semiconductor photoanodes are only used as charge transporter and the photoelectrons are provided by photosensitive dyes. Therefore, over the years, many experiments have been carried out on DSSCs to improve their efficiency by developing and studying various organic dyes such as silyl-anchor and carboxy-anchor dyes ruthenium dye, panchromatic black dye, etc.[44,45] Although organic sensitizers have become a good competitor due to their low-cost purification process and ability to tailor the absorption band in the solar spectrum, it is important to recognize the kinetics of photo-excited electrons during DSSC operation.[46] Mainly, the D−π–A configuration showed better light-harvesting capacity between the visible and far-red regions of the solar spectrum.[47] An organic dye’s electrochemical and optical properties are also influenced by intramolecular charge transfer.[48] Therefore, many efforts have been made to engineer the sensitizer (dye) to prohibit charge recombination, reinforce dye regeneration, and cover a broader absorption band in the solar spectrum.[49−51] As the performance of DSSCs has been strongly relying on the molecular structure of the photosensitizer, it is essential to engineer the structures of organic dyes to harvest solar light more efficiently. Consequently, inspired by previous literature, we have synthesized a novel analogue of the aminonaphthoquinone-based photosensitizer, viz. BrA1, 2-bromo-3-(methylamino)naphthalene-1,4-dione (molecular structure of BrA1 and used as a photosensitizer for stable TiO2 photoanodes). BrA1 was used as a donor-π-acceptor (D-π-A) photosensitizer in the DSSC structures. Thus, it is expected that further improvement of the power conversion efficiencies of TiO2-based DSSCs can be achieved by forming state-of-the-art devices by combining the hydrothermal syntheses of TiO2 nanorod-based electrode materials with a novel analogue of 2-bromo-3-(methylamino)naphthalene-1,4-dione as the photosensitizer.

Therefore, inspired by our previous work, herein, we have synthesized a novel analogue of aminonaphthoquinone-based photosensitizer, viz. BrA1, 2-bromo-3-(methylamino)naphthalene-1,4-dione (molecular structure of BrA1 shown in Figure a) for use as a photosensitizer for stable TiO2 photoanodes. After the photosensitizer is adsorbed on the TiO2NR photoanodes, the ligand to metal charge transfer occurs due to titanium (IV) chelation with the amine group. This further leads to the formation of the bidentate complex by adsorption of the amino group with titanium (IV). Additionally, the oxygen and nitrogen atoms of BrA1 are involved in the physical contact with the titanium (Ti) atom of TiO2. Thus, the interface between BrA1 and TiO2 is formed. The possible schematic interfacial contact structures of BrA1 and TiO2 are shown in Figure b.

Figure 1

(a) Molecular structure of the photosensitizer (2-bromo-3-(methylamino)naphthalene-1,4-dione) and (b) schematic of the interfacial contact structure of BrA1 and TiO2.

Furthermore, TiO2 nanorod (TiO2NR) photoanodes were synthesized by the hydrothermal technique and annealed at different temperatures of 300, 400, 500, and 600 °C for 2 h. Furthermore, the effect of calcination temperatures on the optical, structural, and morphological properties of TiO2NR has been investigated. Then, a DSSC device was developed by loading the BrA1 photosensitizer solution as BrA1 on the TiO2NR-based photoelectrodes in a separate set of experiments. Further, device characteristics in terms of the photocurrent, carrier recombination, and transport times of the fabricated photoanodes (TiO2NR (300 °C)/BrA1, TiO2NR (400 °C)/BrA1, TiO2NR (500 °C)/BrA1, and TiO2NR (600 °C)/BrA1) were investigated in detail. This study would first use BrA1 as a photosensitizer for the TiO2NR photoanode. Thus, our work will soon open a new approach to using the novel BrA1 photosensitizer for various metal oxide photoanodes’ light-harvesting and DSSC photoelectric conversion efficiency.

Experimental Section

Materials and Methods

The materials used, viz. 2-bromo-3-(methylamino)naphthalene-1,4-dione photosensitizers, were synthesized and characterized using the literature-reported procedure and purified by column chromatography with toluene:methanol (9:1) as eluent.[52] 2,3-Dibromonaphthalene-1,4-dione (DBrNQ) and fluorine-doped tin oxide (FTO) having a resistance of 16 Ω/□ (ohms per square) were purchased from Sigma-Aldrich. Methylamine (40%) was purchased from LOBA Chemicals India. Methanol (AR), ethanol (AR), dichloromethane (AR), and toluene (LR) were purchased from Merck Chemicals. Dimethyl sulfoxide (HPLC; Sigma-Aldrich), titanium (IV) butoxide (97%; Sigma-Aldrich), and hydrochloric acid (35–37%; Junsei, Japan) were used as received. Toluene was purified by standard reported methods,[53] and dry methanol was prepared by the literature-reported procedure.[54]

Preparation of Photoelectrodes (TiO2 Nanorod Arrays)

Preparation of TiO2 Nanorod Photoelectrodes and BrA1 Photosensitizer

Photoelectrodes of TiO2NR semiconductors were prepared on the surface of FTO. F-doped SnO2 has a resistance of 16 Ω/□. Initially, the conducting substrate of the FTO was cleaned with a soap solution for 10-15 min, followed by ultrasonication for 15 min, and then cleaned using double-distilled water (DDW).[55,56] The modified hydrothermal process was used to synthesize the TiO2 nanorods (TiO2 NR) on FTO.[57−59] In the typical synthesis process, 60 mL of 1:1 ratio of DDW and concentrated HCl were mixed under constant stirring for 5 min, adding 1 mL of titanium (IV) butoxide and stirring for 30 min. Finally, the resultant solution was transferred into a Teflon-lined stainless steel autoclave with two FTOs kept in the autoclave. The autoclave was tightly closed and placed in an electric oven at 150 °C for 4 h. After deposition, the TiO2NR photoanodes were washed with DDW and dried at room temperature (27 °C).[60,61] Further, these as-prepared TiO2NR films were annealed at different temperatures such as 300, 400, 500, and 600 °C, respectively, in a box furnace with a 5 °C/min ramp step. Moreover, the BrA1-photosensitized TiO2NR photoelectrodes were prepared via the chemical bath deposition method. Initially, a 0.01 M solution of photosensitizer (BrA1) in methanol was prepared in five separate beakers. Then, one hydrothermally prepared TiO2NR photoanode (300, 400, 500, and 600 °C) was dipped into the sensitizer solution for 72 h at 27 °C under the dark condition. After 72 h of adsorption of photosensitizers, the photoelectrodes were washed with ethanol for 10 s to remove the unloaded photosensitizer molecules. Finally, the BrA1-photosensitized TiO2NR photoelectrodes were denoted as TiO2NR (300 °C)/BrA1, TiO2NR (400 °C)/BrA1, TiO2NR (500 °C)/BrA1, and TiO2NR (600 °C)/BrA1.

Characterization of Photosensitizers and TiO2 Nanorod Array Photoelectrodes

The morphology of the photoanodes was further analyzed using the JEM-2010 field emission-scanning electron microscopy (FE-SEM) instrument (SUPRA40VP, Germany). The structure and crystallite size of the prepared TiO2NR-based photoelectrodes were calculated using a powder X-ray diffractometer, Bruker D8, with Cu Kα radiation source of wavelength (λ) = 0.154 nm in the angular range of 20–80°. The X-ray tube had a fixed current of 30 mA and a voltage of 40 kV. For X-ray photoelectron spectroscopy (XPS), a Thermo Scientific XPS spectrometer equipped with a monochromatic Al Kα X-ray source (hυ = 1486.6 eV) was used to study the elemental composition and valence state of the photoanodes. The UV–visible absorption spectra of the photosensitizer and photoanodes were collected using the Jasco Ultraviolet-Visible spectrophotometer model (V-670). The photosensitizer’s cyclic voltammetry (CV) studies were performed using a CH instrument with an electrochemical analyzer (CHI 6054E). Fourier-transform infrared spectroscopy (FT-IR) spectra of the photosensitizer (BrA1) and photosensitizers-loaded photoanodes were recorded in the range 4000–400 cm–1 using a BRUKER FT-IR spectrophotometer.

Single-crystal X-ray diffraction data for BrA1 were collected using a D8 Venture PHOTON 100 CMOS diffractometer using graphite-monochromatized Mo Kα radiation (λ = 0.7107 Å) with exposure/frame = 10 s. The X-ray generator was operated at 50 kV and 30 mA for Mo Kα radiation, and at 50 kV and 1 mA for Cu Kα. An initial set of cell constants and an orientation matrix were calculated from 24 and 60 frames for the Mo and Cu sources. The optimized strategy for the data collection consisted of different φ and ω scans with 0.5° steps of φ/ω. The crystal to detector distance was 5.00 cm with 512 × 512 pixels/frame, oscillation/frame = −0.5°, maximum detector swing angle = −30.0°, beam center = (260.2, 252.5), and in-plane spot width = 1.24. Data integration was carried out with a Bruker SAINT program, and empirical absorption correction for intensity data was carried out subsequently by Bruker SADABS. The programs were integrated into the APEX II package.[62] The data was corrected for Lorentz and polarization effect. The structures were solved by the Direct Method using SHELX-97.[53] The final refinement of the structure was performed by full-matrix least-squares techniques with anisotropic thermal data for the non-hydrogen atoms on F2. The non-hydrogen atoms were refined anisotropically, while hydrogen atoms were refined at the calculated positions as riding atoms with isotropic displacement parameters. Molecular diagrams were generated using the Mercury program.[63] Geometrical calculations were performed using SHELXTL[64] and PLATON.[65]

Results and Discussion

Single-Crystal X-ray Diffraction Studies of the BrA1 Photosensitizer

BrA1 crystallizes in the centrosymmetric triclinic space group 1̅. The ORTEP diagram is shown in Figure a. The crystal structure data are shown in Table S1 in ESI† and the hydrogen bonding parameters in Table S2 in ESI†. The asymmetric unit’s two BrA1 molecules are present in the asymmetric unit; these molecules differ by bond distances (Figure S1 in ESI†) and noncovalent interactions (Figure S2 in ESI†). Molecule A (Figure a, atom numbered as A) is in the vicinity of four similar molecules, while molecule B (Figure a, atom numbered as B) is connected to six nearest similar molecules. Only asymmetric molecule B shows C–H···π interaction (Figure S2 in ESI†). The quinonoid carbonyl distances of both the asymmetric molecules are similar to the oxidized form of naphthoquinone.[52,66−69] Both the asymmetric molecules form a polymeric chain-like structure via “head to tail” N–H···O and C–H···O (benzenoid proton) interactions, and methyl of n-alkylamino also supports C–H···O interaction. The planes of the polymeric chain of asymmetric molecules make an angle of 88.71° (Figure b). Figure c shows the intermolecular interaction between the polymeric chains of A and B molecules via Br···O and C–H···O interactions; the C–H···π interaction of molecule B is visible. The crystallographic data for BrA1 are shown in Tables S1–S7 in ESI†.

Figure 2

(a) ORTEP of BrA1 and (b, c) polymeric chains of asymmetric molecules of BrA1 down the a-axis.

Surface Morphology of TiO2NR Photoanodes

Figure shows the surface and cross-sectional FE-SEM of TiO2NRs grown on the FTO surface by the hydrothermal process and annealed at various temperatures ranging from 300 to 600 °C for 1 h. As shown in Figure a–d, the whole surface of FTO is covered with vertically aligned TiO2NRs. The upper faces of the nanorods are round-shaped, and some space is present between two hydrothermally deposited TiO2NR. This porous structure could be helpful during photosensitizer loading and photovoltaic measurements.[70] Additionally, during the DSSC measurements, more electrolyte was inserted into the porous structure of TiO2NR, and thus, the maximum active surface area of TiO2 can be utilized during the DSSC application.[59]Figure e–h shows that the thicknesses of the TiO2NR photoelectrodes (TiO2NR (300 °C), TiO2NR (400 °C), TiO2NR (500 °C), and TiO2NR (600 °C)) are approximately 1.48, 1.53, 1.22, and 1.86 μm, respectively. Generally, a simple single-step chemistry approach involves hydrothermally synthesized TiO2 nanorods, forming thin TiO2 grains on FTO as seed particles for TiO2 nanorod growth. This thin TiO2NR nanograins layer helps to keep contact between the TiO2NR and the FTO substrate. Also, the interface contact between the TiO2NR and the substrate improves as the calcination temperature increases. After increasing the calcination temperature above 400 °C, a thicker compact optimum TiO2 layer forms between the TiO2NRs and substrate interface. This is due to the nanocrystals gradually crystallizing with the annealing temperature. Further, due to high-temperature annealing, some nanocrystals split and are formed with more compact contact between the TiO2 and FTO interface. The SEM cross-sectional views of the interface between the TiO2 nanorods array and the FTO substrate are shown in Figure e–h. Generally, it is expected that the hydrothermally synthesized TiO2 nanorods would involve a simple single-step chemistry approach, which uses small TiO2 grains as seed particles for nanorod growth, as shown in the FTO/TiO2 interface. In this process, the hierarchical nonporous TiO2 spheres serve as scaffolds that provide plenty of conductive tunnels for efficient charge transfer and the subsequent growth of TiO2 nanorods. Further, as the annealing temperature increases to 400 °C, the amorphous nanocrystals gradually crystallize and act as cores, leading to the continuous growth of the TiO2 nanorods. This facilitates the formation of a tamping contact on the surface of FTO, which helps reduce the resistance between FTO and the TiO2 electrode and results in an average thickness of the interconnected nanograins layer of 50–75 nm. However, when the annealing temperature was extended to 500 °C when the nanocrystals gradually crystallized, some of the nanocrystals split and formed with more compact contact between the FTO interface, leading to the continuous transfer of photogenerated charge carriers across the interface. Interestingly, additional growth of the TiO2 nanocrystals was observed at the 600 °C annealing temperature, which enhanced the grain boundary resistance between the grains and decreased the DSSC performance. This continuous growth of nanocrystals occurred at the expense of a thick portion of the nanocrystal with a higher grain boundary resistance, thus resulting in a lower photocurrent. Ghicov et al. indicated that TiO2 nanotubes might collapse at specific heat-treatment conditions, such as high temperatures and long annealing times. Based on the observations from this study, the photocurrent decreased significantly for the TiO2 nanotubes annealed at 600 °C.[71]

Figure 3

Surface FE-SEM images of (a–d) hydrothermally deposited TiO2NR photoanodes; (e–h) cross-section views of the hydrothermally deposited TiO2NR photoanodes.

X-ray Diffraction Analysis of TiO2NR Photoanodes

Figure shows the XRD peaks of the FTO substrate; hydrothermally deposited TiO2NR’s photoanodes annealed at 300, 400, 500, and 600 °C exhibited the diffraction peaks at 2θ = 36.13, 41.2, and 54.31°, which correspond to the tetragonal phase of (rutile) TiO2NRs (reference code 98-002-4277). In contrast, the remaining peaks are assigned to the FTO. The results well agreed with the results obtained by Wu et al.[72] Further, to elucidate the annealing effect on the favored growth orientation of rutile nanorods, the discussion is included based on the available literature.[73,74] To study the influence of annealing temperature on the grain size, the crystallite size of the photoanodes annealed at different temperatures was measured. In the present case, it can be seen that among the diffractive reflection peaks (101), (112), and (200) of TiO2NR, (101) is the most prevalent and is used for the crystal size measurements. In the present case, the average crystallite size (D) of TiO2NR was calculated using the Debye–Scherrer formula[75]where λ is the X-ray wavelength in nanometers (nm), λ = 0.154 nm is the X-ray wavelength of Cu Kα, K is the Scherrer constant (0.9), β is the full width at half-maximum (FWHM) of the peak, and θ is the Bragg angle.[76] The (101) peak of the annealed TiO2NRs becomes stronger and sharper with increasing annealing temperature from 300 to 400 °C, indicating better crystallinity. The crystallite sizes of TiO2NR (300 °C), TiO2NR (400 °C), TiO2NR (500 °C), and TiO2NR (600 °C) are 37.69, 34.93, 39.69, and 41.57 nm, respectively. Furthermore, the full width at half-maximum (FWHM) of the (101) peaks of the TiO2NR’s photoanodes (TiO2NR (300 °C), TiO2NR (400 °C), TiO2NR (500 °C), and TiO2NR (600 °C)) are shown in Figure S3 in ESI†. However, the crystallinity of TiO2NRs increases at 500 to 600 °C annealing temperature, but the internal surface area of the TiO2NRs might decrease on enhancing the annealing temperature. Similar observations have been reported by Pengn et al. and Zhao et al.[77,78] Thus, in addition to the crystallinity at high-temperature annealing, the photoanodes are significantly affected by other superseding factors such as the internal surface area. Therefore, it can be summarized that the annealing process may not be heating all grains in the sample equally. This led to unequal growth of the grains (few grew in size, whereas others underwent the shrinking process). Thus, during the annealing process, the interface between the growing and shrinking grains undergoes ruptures and detaches the atoms from one another in the same materials. Therefore, the (112) peaks in the TiO2NR’s photoanodes annealed at 300 and 600 °C have different intensities than the photoanodes annealed at 400 and 500 °C.[79]

Figure 4

The XRD pattern of hydrothermally deposited TiO2NRs grown on the FTO substrate and annealed at 300, 400, 500, and 600 °C.

XPS of TiO2NR Photoanodes

To analyze the Ti’s elemental composition and valence state in the TiO2NR, TiO2NR (400 °C) photoanodes’ and XPS (X-ray photoelectron spectroscopy) analysis was carried out. The survey scan XPS spectra (Figure a) of TiO2-based photoanodes indicate the binding energy peaks at 458.4, 565.4, 529.4, and 284.6 eV, respectively, showing that the Ti, O, and C elements are obtained in the film predominantly. The carbon element may be identified by the “C 1s” peak, the carbon element from the organic residue precursors, or the measured environment that was not removed entirely during the hydrothermal process.[80]Figure b,c indicates the high-resolution XPS spectra of Ti 2p and O 1s, respectively, for TiO2NR photoanodes. The two strong peaks at around 465.2 and 459.5 eV can be attributed to Ti 2p1/2 and Ti 2p3/2, respectively.[60] The calculated binding energy difference between the Ti 2p doublet (Ti 2p3/2 and Ti 2p1/2) peaks was 5.7 eV, compatible with the binding energy separation observed for stoichiometric TiO2.[81] The broad O 1s peak, located at about 530.8 eV, is associated with the lattice oxygen of the TiO2NR.[82] This suggests the prepared photoanode is of pure TiO2.

Figure 5

(a) X-ray photoelectron spectroscopy (XPS) spectra of hydrothermally deposited TiO2NR (400 °C) photoanode; (b, c) high-resolution XPS spectra of Ti 2p and O 1s, respectively.

UV-Visible Absorption Spectra Photosensitizers (BrA1) and Photoanodes (TiO2NR (300 °C)/BrA1, TiO2NR (400 °C)/BrA1, TiO2NR (500 °C)/BrA1, and TiO2NR (600 °C)/BrA1)

Figure a depicts the UV–visible spectra of a 1 μM concentrated solution of the photosensitizer (BrA1). As shown in Figure a, the naphthoquinone exhibits the bands in the UV region at 282 and 330 nm, assigned to the (π → π*) electronic transition. Additionally, the broadband in the visible region was observed at 471 nm due to the weak n → π* transition or the charge transfer transition. The electron-donating effect of the substituted amino group shows the charge transfer band in BrA1. The delocalization of the electron lone pair of nitrogen requires its orthogonality in the plane of naphthoquinone. The nature of amines affects their basicity, which affects the extent of the bathochromic shift.[52] This absorption is a typical characteristic of amino-substituted quinone.[83] Generally, it is seen that the charge transfer transition in the visible region is due to the electron-donating effect of the substituted amine in BrA1.[52] The absorption spectra of BrA1 do not depend on the number of polymer-like chains. The absorbance spectra of one- and two-molecule chains were observed at nearly the same wavelength.[52,84] The polymer-like chains are observed via hydrogen bonding of the BrA1 photosensitizer molecules in solid states. The absorption spectra will not affect even though this bonding exists in the solution.[85] Further, Figure b indicates the UV–visible spectra of the unloaded TiO2NR and photosensitizer-loaded TiO2NR photoanodes (TiO2NR (300 °C)/BrA1, TiO2NR (400 °C)/BrA1, TiO2NR (500 °C)/BrA1, and TiO2NR (600 °C)/BrA1). The bathochromic shifts in the photosensitizers loaded on photoanodes (TiO2NR (300 °C)/BrA1, TiO2NR (400 °C)/BrA1, TiO2NR (500 °C)/BrA1, and TiO2NR (600 °C)/BrA1) compared to the band of photosensitizer BrA1 are presented in Figure , and the corresponding values are shown in Table . There is negligible change in diameters are observed to TiO2 nanorods with increase in temperature. Therefore, the absorbance is increased in the TiO2NR 300 °C/BrA1, TiO2NR 400 °C/BrA1, TiO2NR 500 °C/BrA1, and TiO2NR 600 °C/BrA1 photoanodes than the TiO2 bare samples in 400–550 nm region. However, the decrease in the absorption of TiO2 in annealed photoanodes might be due to disordered surfaces and the bulk of nanorods (defects in as-synthesized TiO2NR). The same effect has been reported in the literature.[86,87] It seems that the absorption depends on the dye absorption and the packing density of the nanorods. Therefore, to quantify the TiO2NR’s packing density, we calculated the area in each SEM image. We counted several nanorods at two different places, considering the 1 mM area of the SEM image of each sample, and averaged the values. Figure e–h shows the cross-sectional views of the nanorod films; the packing density of the array is different for TiO2NR 300 °C/BrA1, TiO2NR 400 °C/BrA1, TiO2NR 500 °C/BrA1, and TiO2NR 600 °C/BrA1 photoanodes. This causes the different absorption in the UV–visible absorption measurements; the TiO2NR 400 °C/BrA1 and TiO2NR 600 °C/BrA1 photoanodes exhibit a lower absorption, with the lowest packing density, whereas for TiO2NR 300 °C/BrA1 and TiO2NR 500 °C/BrA1, with an increase in the amount of packing density, the absorption becomes elevated. The increased packing density helps improve the interfacial contact with BrA1 and further increases the absorption. Further, as reported previously, in 2-(ethylamino)naphthalene-1,4-dione (LH-2) to 2-(hexylamino)naphthalene-1,4-dione (LH6) and 2-(octylamino)naphthalene-1,4-dione (LH8) series, the highest occupied molecular orbital (HOMO) electron density is distributed over the quinone ring and the adjacent nitrogen atom connected with the aliphatic chain and quinone ring. The lowest unoccupied molecular orbital (LUMO) electron density is distributed over the benzenoid ring (naphthalene-1,4-dione ring).[88] Similarly, the electron density distribution occurs in the HOMO and LUMO of the BrA1 photosensitizer. Therefore, to study the HOMO and LUMO energy gap, the intersection of the excitation and emission spectra of the BrA1 photosensitizer are shown in Figure c. The HOMO-LUMO energy gap was calculated by using the following formula[89]where E is the optical band gap in eV, h represents the Planck’s constant, λ represents the intersection of the excitation and emission spectra in nm, and c is the speed of the light. The HOMO-LUMO energy difference is between the vibrationally relaxed ground state and excited state and is denoted by E0-0.[90,91] The E0-0 of the BrA1 photosensitizer is calculated from the intersection of the absorbance and emission spectra of the BrA1 photosensitizer and is equal to 2.33 eV.

Figure 6

UV–visible spectra of (a) BrA1 photosensitizer, (b) unloaded TiO2 and TiO2NR (300 °C)/BrA1, TiO2NR (400 °C)/BrA1, TiO2NR (500 °C)/BrA1, and TiO2NR (600 °C)/BrA1, and (c) excitation and emission spectra of the BrA1 photosensitizer.

Table 1

UV-Visible Data of the Photosensitizer and Photoanodes

photosensitizers/photoanode	UV region (nm) (π → π*) transition	visible region (nm) (n → π*) transition
BrA1	330	471
TiO2NR	345
TiO2NR (300 °C)/BrA1	361	507
TiO2NR (400 °C)/BrA1	348	507
TiO2NR (500 °C)/BrA1	353	507
TiO2NR (600 °C)/BrA1	366	507

Electrochemical Investigation Using Cyclic Voltammetry

Cyclic voltammetry (CV) is one of the outstanding techniques for studying compounds. The cyclic voltammogram of the BrA1 photosensitizer is shown in Figure a. The HOMO-LUMO energy difference (E0-0), and HOMO and LUMO energy levels have been calculated using UV–visible, fluorescence spectroscopy, and CV, respectively. The HOMO-LUMO energy difference (E0-0) of the BrA1 photosensitizer is 2.33 eV (Figure c). The LUMO can be calculated from the first reduction potential (onset of the first reaction peak) obtained from Figure a. Further, the BrA1 photosensitizer reduction peak was specified using CV analysis, and the LUMO was calculated by the formula[92,93]where IP(FC) is the standard internal redox of the system and FC/FC+ is the Ferrocene/Ferrocenium couple. The ELUMO of the BrA1 photosensitizer is −3.64 eV, and EHOMO is −5.97 eV. The EHOMO was calculated from ELUMO and E0-0 values. Figure b shows the energy-level diagram of the conduction band edge of TiO2 with HOMO-LUMO values of the BrA1 photosensitizer. The conduction band of TiO2NR is below the LUMO level of the BrA1 photosensitizer. TiO2NR (400 °C)/BrA1 shows a higher Voc, as shown in Table . The TiO2NR (400 °C)/BrA1 shows a slower recombination of higher Voc. However, as the temperature increases above 400 °C, the Voc value of the TiO2NR-based photoanodes is decreased again. Figure b shows that the maximum voltage is the difference between the redox potential of the electrolyte and the Fermi level of TiO2NR. The latter is changed with change in temperature and forms a new Fermi level, i.e., quasi-Fermi level.[94] Therefore, due to the change in the Fermi level of TiO2NR at 400 °C, there is a change in Voc.

Figure 7

Cyclic voltammetry spectra of (a) BrA1 photosensitizer, and (b) energy diagram of the conduction band edge of TiO2 with the HOMO-LUMO values of the BrA1 photosensitizer.

Table 3

Photovoltaic Performance of BrA1-Sensitized TiO2NR Photoelectrodes at Different Annealed Temperatures

photosensitizers/samples	Voc (V)	Jsc (mA/cm2)	fill factor	efficiency (η) %
TiO2NR (300 °C)/BrA1	0.50	0.48	44	0.10
TiO2NR (400 °C)/BrA1	0.55	0.60	48	0.16
TiO2NR (500 °C)/BrA1	0.52	0.63	46	0.15
TiO2NR (600 °C)/BrA1	0.51	0.55	40	0.11

FT-IR Analysis of Photosensitizers and Photoanodes

To evaluate the photosensitizer (BrA1)/TiO2NR interaction, Fourier-transform infrared (FT-IR) measurements were performed on BrA1 and TiO2NR-based photoanodes (TiO2NR (300 °C)/BrA1, TiO2NR (400 °C)/BrA1, TiO2NR (500 °C)/BrA1, and TiO2NR (600 °C)/BrA1), respectively. Figure a,b shows the FT-IR spectra of BrA1, and Figure c,d presents the FT-IR of TiO2NR photoanodes (TiO2NR (300 °C)/BrA1, TiO2NR (400 °C)/BrA1, TiO2NR (500 °C)/BrA1, and TiO2NR (600 °C)/BrA1) in the range 3500–2900 cm–1 and 1750–1500 cm–1, respectively. Furthermore, for comparison, the FT-IR band frequencies of photosensitizers and photoelectrodes are well summarized in Table . It is observed that the νN-H band of photosensitizer BrA1 (3305 cm–1) was increased by 8, 11, 6, and 9 cm–1 in their photoanodes (TiO2NR (300 °C)/BrA1, TiO2NR (400 °C)/BrA1, TiO2NR (500 °C)/BrA1, and TiO2NR (600 °C)/BrA1), respectively. On the other hand, the νC=O frequencies of photosensitizers observed at 1680 cm–1 (BrA1) are shifted to higher wavenumbers (bathochromic shift) by 2, 5, 3, and 4 cm–1, respectively, in TiO2NR (300 °C)/BrA1, TiO2NR (400 °C)/BrA1, TiO2NR (500 °C)/BrA1, and TiO2NR (600 °C)/BrA1 with the annealing temperature. The BrA1 photosensitizer anchoring site is attached to TiO2NR through νN-H and νC=O bonding. The BrA1 photosensitizer shows νN-H and νC=O frequencies at 3305 and 1680 cm–1, respectively, whereas the smaller crystallite size and defect-free TiO2NR (400 °C) photoanode surface help form the attachment of photosensitizer through νN-H and νC=O. The cause of this blue shift might be the presence and interaction of different surface-active sites of TiO2 with a photosensitizer.[91] Thus, as explained earlier, the crystallite size of the TiO2NR (400 °C) photoanode is smaller than that of other photoanodes. The absorption spectra shifted from lower to higher frequencies as the particle size decreased.[95] However, the νC-N frequency observed at 1596 cm–1 (BrA1) is increased by 8, 12, 7, and 8 cm–1, respectively, in TiO2NR (300 °C)/BrA1, TiO2NR (400 °C)/BrA1, TiO2NR (500 °C)/BrA1, and TiO2NR (600 °C)/BrA1. Thus, the shift in the observed frequency in BrA1-loaded TiO2NR suggests the strong interaction between the photosensitizer and base photoanodes, which helps provide a channel for the transfer of electrons in the DSSC system.

Figure 8

(a, b) FT-IR of the photosensitizer (BrA1) and (c, d) FT-IR of photosensitizer-loaded TiO2NR photoelectrodes.

Table 2

FT-IR Band Frequencies of the Photosensitizer and Photoelectrodes

photosensitizers/photoanode	νN-H band frequencies (cm–1)	νC=O band frequencies (cm)	νC-N band frequencies (cm–1)
BrA1	3305	1680	1596
TiO2NR (300 °C)/BrA1	3313	1682	1604
TiO2NR (400 °C)/BrA1	3316	1685	1608
TiO2NR (500 °C)/BrA1	3311	1683	1603
TiO2NR (600 °C)/BrA1	3314	1684	1604

Photovoltaic Performance

TiO2NR/BrA1 for the Dye-Sensitized Solar Cell (DSSC)

The DSSC was fabricated with crucial parameters such as working electrodes (BrA1-sensitized TiO2NR photoanode). A polyiodide solution (mixture of 0.5 M tetra-n-propylammonium iodide ((CH3CH2CH2)4NI) and 0.05 M iodine (I2) in an ethylene carbonate:acetonitrile mixture of solvents in 80:20 proportion) was used as the electrolyte[96] and a Pt electrode was used as the counter electrode.[96] A spacer made up of four layers of cello tape (thickness ∼ 40 μm) was used between the working electrodes (0.25 cm2 active cell area). The aminonaphthoquinone BrA1 photosensitizer-loaded photoanodes were sandwiched with the counter electrode (Pt electrode). The polyiodide electrolyte was inserted between the working and counter electrodes; simultaneously, care was taken to avoid air bubbles inside the solar cell assembly. This prepared device was further used for photovoltaic measurement. The DSSC components and the experimental scheme of the BrA1-photosensitized TiO2NR-based DSSC are shown in Figure a. The photocurrent density voltage (J–V) characteristic curves of the DSSCs were measured using a Keithley 2400 source meter and solar simulator (ENLITECH model SS-F5-3A) under the incident light intensity of 100 mW cm–2. Generally, in DSSC, EIS measurements are performed under illumination at the voltage corresponding to open-circuit conditions.[97,98] Electrochemical impedance spectroscopy (EIS) was measured using a potentiostat (Vertex IVIUM Technologies Netherlands). The measurements were carried out at an applied potential of −0.5 V with 0.01 V amplitude over a frequency range of 1 Hz–1000 kHz under illumination with an intensity of light of 100 mW cm–2. All of the EIS spectra were fitted using ZView software.

Figure 9

(a) Experimental scheme of BrA1-photosensitized TiO2NR-based DSSC and (b) schematic representation for photovoltaic conversion mechanism of TiO2NR (400 °C)/BrA1-based DSSC.

The schematic picture of the charge transfer mechanism in a BrA1-sensitized TiO2NR dye-sensitized solar cell is shown in Figure b. BrA1 shows the lower energy absorption assigned in the visible region at 471 nm, the charge transfer transition, and the n → π* transition. The charge transfer band indicates the electron-donating effect of the amine group in BrA1.[52] Therefore, in the DSSC fabrication using the BrA1-loaded TiO2NR photoanodes, in addition to TiO2NR itself, the BrA1 absorbs the light and readily donates the electrons to the conduction band of the TiO2NR photoanode. Under light irradiation, the BrA1 dye in the TiO2NR (400 °C)/BrA1-based DSSC absorbs photons equal to or more than the HOMO-LUMO band-gap energy, and electrons are excited from HOMO to LUMO state (Dye*).[99] Further, the photogenerated electrons in the LUMO state of the photosensitizer are excited up into the conduction band of TiO2NR. Some photogenerated electrons revert to the ground state of the BrA1 photosensitizer, and some recombine with the I3– in the electrolyte.[100] However, most of the photoelectrons reach the interface of TiO2NR and FTO via TiO2NR channels.

Further, due to the optimum TiO2 compact layer, the photogenerated electrons are easily collected at the current collector and transferred to the platinum counter electrode and the electrolyte. A TiO2 compact layer between TiO2NR and FTO reduces the photoelectrons recombination process in the bulk of TiO2 and the FTO interface and prolongs the electron’s lifetime.[79] At the same time, the photogenerated electrons collected at the counter electrode help to prevent the oxidized dye from decomposing and speed up the circulation of the electrons in the DSSC. In addition, the absorption of dye molecules on the TiO2NR surface leads to improved light-harvesting capability. Hence, the significant improvement in the Jsc of the BrA1-loaded TiO2NR DSSC (TiO2NR 400 °C)/BrA1 is the synergistic effect of light absorption by BrA1 and effective charge transfer by TiO2NR.

The J–V curves of the TiO2NR (300 °C)/BrA1, TiO2NR (400 °C)/BrA1, TiO2NR (500 °C)/BrA1, and TiO2NR (600 °C)/BrA1 photoanodes-based DSSC are shown in Figure . The J–V curves of the DSSC device prepared using TiO2NR (400 °C) without the photosensitizer are shown in Figure S4 in ESI†. The Jsc and Voc of the TiO2NR (400 °C) without the photosensitizer are nearly equal to zero. During the DSSC photovoltaic analyses, the device was illuminated with 100 mW cm–2 (AM 1.5) light intensity. A polyiodide solution was used as the electrolyte. The DSSC parameters of TiO2NR/BrA1 photoanodes such as open-circuit voltage (Voc) V, short-circuit photocurrent density (Jsc) mA/cm2, fill factor (FF), and efficiency (η) % were calculated and are summarized in Table . It is observed that the Jsc value for TiO2NR (500 °C)/BrA1 (Jsc = 0.63 mA/cm2) is higher than those of other TiO2NR/BrA1-based photoanodes. The Jsc values for TiO2NR (300 °C)/BrA1, TiO2NR (400 °C)/BrA1, TiO2NR (500 °C)/BrA1, and TiO2NR (600 °C)/BrA1 are 0.48, 0.60, 0.63, and 0.55 mA/cm2, respectively. On the other hand, the corresponding Voc values of the photoelectrodes are 0.50, 0.55, 0.52, and 0.51 V, respectively. The FF values of the photoelectrodes are 44, 48, 46, and 40, respectively, and the corresponding η values of the photoelectrodes are 0.10, 0.16, 0.15, and 0.11%, respectively. Among the studied DSSCs, the TiO2NR (400 °C)/BrA1 photoelectrode-based DSSC exhibited a higher photovoltaic performance than other photoelectrodes. This is due to effective light absorption and efficient charge generation and separation. Also, in this investigation, the calculated Voc and FF values for the TiO2NR (400 °C)/BrA1 photoelectrode are higher than those of other photoelectrodes (TiO2NR (300 °C)/BrA1, TiO2NR (500 °C)/BrA1, TiO2NR (600 °C)/BrA1); for this reason, the photovoltaic performance of TiO2NR (400 °C) is higher than that of other photoelectrodes. Finally, the TiO2NR (400 °C)/BrA1 photoelectrodes exhibited a power conversion efficiency (PEC) of 0.16%.

Figure 10

J–V characteristic curve of DSSCs based on TiO2NR photoelectrodes annealed at 300, 400, 500, and 600 °C.

Electrochemical Impedance Spectroscopy (EIS) Analysis of TiO2NR (300 °C)/BrA1, TiO2NR (400 °C)/BrA1, TiO2NR (500 °C)/BrA1, and TiO2NR (600 °C)/BrA1 Photoanodes

Electrochemical impedance spectroscopy (EIS) was used to investigate the charge transport properties and recombination of electrons in TiO2NR/BrA1-based DSSCs. The EIS plots of the TiO2NR (300 °C)/BrA1, TiO2NR (400 °C)/BrA1, TiO2NR (500 °C)/BrA1, and TiO2NR (600 °C)/BrA1-based DSSCs under dark conditions and fitted by ZView software are shown in Figure . The inset shows the equivalent circuit used to fit the Nyquist plot of the TiO2NR/BrA1-based DSSCs. EIS analysis was carried out to determine the effect of annealing temperature on the charge transfer resistance of TiO2. The equivalent circuit consists of three resistance values and two capacitances. Rs denotes the resistance between the FTO and TiO2NR interface. The Rs values are summarized in Table ; the TiO2NR (400 °C) photoanode shows 51.5 Ω resistance. In contrast, R1 and R2 are the bulk charge transfer and charge transport resistance (R2) between the TiO2NR-based photoelectrodes and the electrolyte interface. CPE1 is the double-layer capacitance corresponding to R1, and CPE2 is the double-layer capacitance between the TiO2NR photoelectrodes and the electrolyte interface.[70,101,102] The charge transfer resistance R2 at the FTO-TiO2 and electrolyte interfaces is very high in the as-grown photoanode, suggesting that it contains many defects and impurities. Further, as shown in Table , at 400 °C annealing temperature, the R2 value drastically decreases from 11 979 to 4445 Ω, indicating the reduced grain boundary defects at the interface and enhancement of the charge transport efficiency in TiO2 nanorods. However, FTO annealed at 500 and 600 °C showed high resistance; as Sn diffusion starts at that annealing temperature, the increased resistance of the FTO, as well as the random nature of the TiO2NRs in TiO2NR (500 °C)/BrA1 and TiO2NR (600 °C)/BrA1 photoanodes introduce several boundary resistances at both the interfaces.[79,103] This unexpected behavior might be explained using the packing density of the photoanodes, as shown in the FE-SEM cross-sectional view in Figure . The electron transport-recombination properties of the BrA1-sensitized TiO2NR-based DSSCs are calculated from the equivalent circuit, Nyquist plot, and Bode plot and are represented in Table . It was reported that a decrease in the charge recombination rate resulted in an enhancement in device efficiency.[104]Figure S5 in ESI † shows the Bode phase plot, which is used to calculate the electron lifetime (τeff) in DSSC[101] using the formulawhere f is the maximum peak frequency of the semicircle; the calculated electron lifetime values of the TiO2NR-based DSSCs are summarized in Table .

Figure 11

Nyquist plots of different annealed photoanodes of the DSSCs; the inset shows the equivalent circuit diagram used for EIS fitting.

Table 4

Electron Transport Properties of the BrA1-sensitized TiO2NR-based DSSCs

samples/DSSC parameters	Rs (Ω)	R1 (Ω)	R2 (Ω)	R2/R1	frequency (Hz)	τeff (m s)	Ln	τD (m s)	Deff (m s–1)	Leff (mm)
TiO2NR (300 °C)/BrA1	51.5	47	11 979	249	2454	0.06	15	0.0002	8.49	0.7371
TiO2NR (400 °C)/BrA1	51.7	161	4445	27	354	0.44	5	0.0159	0.1465	0.2538
TiO2NR (500 °C)/BrA1	51.5	17	10 290	577	602	0.26	24	0.0004	3.30	0.9262
TiO2NR (600 °C)/BrA1	51.9	52	10 803	207	794	0.20	14	0.0096	3.53	0.8402

Conclusions

2-Bromo-3-(methylamino)naphthalene-1,4-dione (BrA1) has been successfully synthesized using 2,3-dichloronaphthalene-1,4-dione. The as-synthesized BrA1 presents the centrosymmetric triclinic space group 1̅, and the polymeric chains of asymmetric molecules of BrA1 have been systematically presented using the ORTEP diagram. Further, the synthesized dye was loaded on hydrothermally prepared TiO2NR photoanodes annealed at different temperatures. The powder XRD confirms the rutile TiO2NR’s tetragonal phase, and with an increase in calcination temperature of the photoanodes, the average crystallite size increases from 37.69 to 41.57 nm. The SEM micrographs confirm the improved interface contact between the TiO2NR and substrate, increasing the calcination temperature. The presence of Ti 2p, Ti 2s, and O 1s XPS peaks at binding energies of 458.4, 565.4, and 529.4 confirms the formation of TiO2NR. The red shift in the absorption edge is observed in the BrA1-loaded TiO2NR as the calcination temperature advances. Optimum TiO2 nanorods annealed at 400 °C temperature loaded with the optimum photosensitizer were used as photoanodes to fabricate the DSSCs. It was further observed that on increasing the calcination, TiO2NR-based DSSC cells ((400 °C)/BrA1, TiO2NR (500 °C)/BrA1, and TiO2NR (600 °C)/BrA1) exhibited significantly increased electron lifetimes than TiO2NR (300 °C)/BrA1.