Teng et al. have used Triphenylamine (TPA) group as a donor moiety previously in 2010, however, power conversion efficiency of that dye was not appreciable [35] that may because of limited conjugation and hence little electron transfer activity. However, in a recent report the idea of more conjugation via double and triple bond was introduced by Wang et. al. in 2017 which could take-up the power conversion efficiency to 6.04%. [31] Therefore, in this report, we employed DTF and TD-DFT computational approach aimed at finding the best metal-free sensitizer by decreasing energy gap Egap that could boost up power conversion efficiency by the introduction of different substitutions at the donor end. We have designed eight new structures of metal-free sensitizers by substituting different hydrogens with different groups as shown in Fig. 1. The designed sensitizers (D1, D2, D3, D4, D5, D6, D7, and D8) vary in their structures at positions (R1, R2, R3, and R4) from the parent structure. [31] Different substitutions with donating inductive effect including; methyl, -C(CH3)3, amine, and width drawing effect such as NO2, Fluorine and -C6H5 have been investigated for their HOMO and LUMO levels, optoelectric properties and absorption spectra.
TiO2 is gaining much attention in DSSCs as an inert and non-toxic semiconducting metal oxide. It is widely used in paint industries, ceramics, medicine, and in food dyes. Moreover, application in batteries and solar cells make it a more fascinating compound. [39] Keeping in mind the computational cost, we have considered a simplest TiO2 cluster due to rationally simplifying the model being studied. This simplest model comprises three Ti atoms connecting six oxygens so that all Ti atoms are sharing an equivalent environment. This type of smaller cluster approach has been using previously by analyzing the photo-catalytic reactions [40] as well as in small molecular adsorption analysis. [41–42]
Frontier Molecular Orbital Patterns and Optoelectric Properties:
Among the most requisite features deciding the performance of the DSSCs, the most promising are HOMO and LUMO levels as well as HOMO-LUMO gap (Egap). For more clarity of geometry and energy levels and distributions of molecular orbitals, sensitizer were analysed as isolated and also adsorbed on the surface of Titanium cluster (TiO2)3. The simulated energy levels of the LUMO and HOMO and their Egap (EHOMO-ELUMO) are shown in Table 1. The TiO2 conduction band was adopted from an earlier report as -4.00 eV. [43] All the energies of the molecular orbitals LUMO and HOMO were calculated by using CAM-B3LYP functional with 6-31G basis sets for sensitizer’s isolated-phase and, LANL2DZ/6-31G basis set and CAM-B3LYP functional for sensitizer and TiO2 chelating mode. After a close observation of the molecular HOMO and LUMO levels profile, it was observed that all the HOMOs level of all the designed sensitizers (D1-D8) were uniformly populated over the donating group’s structure, Fig. 2. However, the LUMOs remained mostly delocalized on anchoring group (acceptor group) which is a clear indication of the electron transfer from donor to acceptor and eventually from acceptor to connected TiO2.
Table 1
Molecular orbitals energies (eV) and energy gap Egap (eV) for studied isolated sensitizers (D1-D8) using CAM-B3LYP/6–31 G and CAM-B3LYP functional and LANL2DZ /6–31 G) basis set for sensitizers and semiconductor (TiO2).
Sensiti-zers | LUMO | HOMO | Egap | LUMO Dye+(TiO2)3 | HOMO Dye+(TiO2)3 | Egap |
1 | -1.189 | -7.353 | 6.164 | -3.325 | -7.672 | 4.347 |
2 | -2.071 | -7.599 | 5.528 | -5.562 | -8.00 | 2.438 |
3 | -1.900 | -7.451 | 5.551 | -4.157 | -7.798 | 3.641 |
4 | -1.802 | -7.323 | 5.521 | -3.769 | -7.636 | 3.867 |
5 | -1.813 | -7.278 | 5.465 | -4.248 | -7.594 | 3.346 |
6 | -2.337 | -7.569 | 5.232 | -4.00 | -7.953 | 3.953 |
7 | -1.983 | -7.224 | 5.241 | -5.224 | -7.514 | 2.290 |
8 | -1.988 | -7.495 | 5.507 | -5.558 | -7.88 | 2.322 |
Since the orbital spatial distribution is a key factor responsible for intramolecular charge transfer and absorption spectra. Egap of the donor itself and between the donor and acceptor plays an important role in improving photocurrent efficiency and, generation and recombination of the charges, respectively. HOMO and LUMO energy levels of the isolated molecule of sensitizer D1 as we observed showed remarkably increase (-7.353 eV and − 1.189 eV), with a large Egap of 6.164 eV as shown in Table 1 and in Fig. 3. However, connecting with the semiconductor TiO2 a clear decrease in HOMO and LUMO levels (-7.672 eV and − 3.325 eV) as well as in Egap 4.347 eV could be seen (Table 1). This means the energy level can be changed by the addition of methyl group at donor moiety aiming to reduce the energy gap due to its donating property results in an enhancing the efficient transfer of an electron to the semiconductor. On the other hands, the introduction of fluorine to the donor group (triphenylamine) at position R1 in D2 decreases dramatically the energy gap between the HOMO and LUMO levels after connecting to the semiconductor. Nevertheless, the LUMO level (-5.562 eV) further pushed beyond the conduction band edge of semiconductor (4.0 eV), lower by 1.562 eV, where the transfer of electron is not feasible.
Methyl group provided the healthier affect as a substitution at the para position (R1) of donor group in D1, therefore, another resonating position R3 (Ortho) for donating effect was further explored by the same substitution. However, methyl group at ortho position could not provide a prominent decrease in Egap (5.551 eV) alone but in chelation with semiconductor cluster a prominent decrease in Egap (3.641 eV) was observed, although, the LUMO levels as estimated by TD-DFT lied down (-4.157 eV) to conduction band edge level (4.0 eV), Table 1. It can be concluded that the para position (R1) is more prone to inductive effect than ortho (R3) of donor group in charge distribution and efficient electron transfer to acceptor unit. In sensitizer D4 at position R1 another donating group (C(CH3)3 was substituted in order to see a more deeper insight to geometry effect on Egap dependence of frontier molecular orbitals.
The sensitizer D4 alone provide the HOMO (-7.323 eV) and LUMO (-1.802 eV) levels with the Egap of (5.521 eV) as in Fig. 3 and, after chelating to cluster the HOMO level remains the almost same with − 7.636 eV see the Fig. 3. However, a dramatic change in LUMO energy level (-3.769 eV) and Egap level (3.867 eV) was observed as expected due to donating inductive effect of (C(CH3)3, shown in Table 1. The LUMO level of D4 is in an excellent position to transfer of charge and electron to conduction edge band of semiconductor through acceptor group. After the absorption of the sensitizer on the semiconductor (TiO2)3, an interaction exists between acceptor and the semiconductor which, ultimately, lowers the LUMO level of the sensitizer [44] since LUMO of the sensitizer remain close to acceptor group. The donating effect of the methyl was further explored in D5 by substituting hydrogen by –CH3 at R1, R2, and R4, but surprisingly, the decrease in Egap (5.465 eV) as we had expected remains too high. The LUMO levels in chelating mode lower even down to conduction band edge of TiO2. However, the insertion of –NO2 group at position R1 to donor moiety in D6 provides lowering in LUMO levels (-2.337 eV) to (4.00 eV) in chelating mode. This decrease is in an excellent position for the efficient transfer of an electron from donor to semiconductor via acceptor. [45] Similarly, the introduction of –C6H5 and –NH2 in D7 and D8 respectively could make the intramolecular charge transfer due too far from conduction band edge in chelating mode. Therefore, the FMO and the optoelectronic properties of these sensitizers (D1-D8) show that just three molecules (D1, D4 and D6) are in a position to deliver intramolecular charge transfer and electron regeneration.
Electronic Absorption Properties:
The absorption properties of metal-free sensitizers can be tuned by careful consideration of the HOMO and LUMO energy levels and their respective Egap. For the development metal-free sensitizers for DSSCs, it is desirable for a new sensitizer that it could show absorbance in near visible region with relatively large coefficient. Therefore, to take deeper insight into optimized ground geometries, electronic transitions and optical properties of the studied sensitizers, TD-DFT calculations were tested in gas-phase by employing CAM-B3LYP functional and 6-31G basis set. The simulated electronic spectra (UV-Visible) are depicted in the Fig. 4.
Since all sensitizers (D1-D8) possess the same basic skeleton, therefore, it is desirable to get the almost same absorption behavior, although some shoulders may present due to introductions of substitutions. It is noteworthy that almost all studied sensitizers presented an identical absorbance behavior as in Fig. 4. The newly designed sensitizers showed little absorption bands in near UV region (around 300 nm) and almost all absorption peaks in the visible region, which is deemed to be an important region for photo-to-current conversions. [45] For understanding electronic transitions, present results are further explored for main configurations, excitation energies, absorption behavior and oscillator strength are depicted in Table 1. The strength of the oscillator is quite well in case of the dye D8, on the other hand, the LUMO level of this dye localized, although near to HOMO level but below the conduction band edge (4.0 eV) of semiconductor i-e TiO2 as shown in Table 1 and Fig. 5. Therefore, the possibilities of the efficient electron injection to anode can be overruled for D8. Since the л-conjugation system has the enormous effect on the absorption band of the dye therefore, we suggest D8 sensitizer with extra conjugation system of phenyl ring showed higher absorption bands due to extended л-conjugation system and extended planarity. From the Table 1, it can be seen that all transitions are л-л* related showing the absorption bands in visible region. The –C6H5 (D8) group substituted sensitizers have provided the maximum absorption bands at 450nm, an example of clear red-shifted bands. On contrary, the –CH3 substituted (D1), -C (CH3)3 substituted (D4) and –F substituted (D2), each showed two absorption bands one longer and other shorter. The longer bands laid in the range of 360–600 nm while the shorter bands range of 260–360 nm as shown in Fig. 5.
The larger band of these sensitizers, we assigned the intramolecular charge transfer between the donor to acceptor through л-bridge, while the shorter bands are the result of л-л* transitions in the whole molecule of sensitizers. Another important factor responsible for larger absorption is the position of the donor with respect to л-bridge. The proximity of the donor to the л-conjugation system increases the chance to molecular orbital to contribute more. This results in increasing the contribution of electron transfer from HOMO to LUMO. The substituted donor at para position (R1) has greater influence as compared to the other positions (R2, R3 and R4), for instance, a larger absorption bands and little Egap was found in case of D1, D4, because of direct involvement of electron resonance from this position and hence in ease in electron transfer.
On the other hand, in case of substituted sensitizer D6, when –CH3 was at position R2, and R3 no as such decrease in Egap level was observed as shown in Table 1 and Fig. 5. As both positions of the substitution, do not allow the direct electron delocalization to the donor moiety. The extended planarity is the main reason of larger absorption band in D6, where a –NO2 group at position R1, could help the sensitizer to show maximum absorption at 438 nm as depicted in Fig. 5 and Table 1. In conclusion, when the donor is close to the accepter the intramolecular charge transfer occurs more efficiently thereby, lead to high absorption band caused by the narrow Egap level and maximum contribution from the HOMO to LUMO level as in the case of D1, D2, D4 and D6. All most all studied sensitizers showed absorption bands in the visible region that is a necessary condition for the development of the DSSCs from organic dye, however, due to the influence of the planarity, proximity and the dihedral angle involvement the Egap level does not decease significantly and limits the electron transfer. Therefore, the substituted donor D2, D3, D7, D5, and D8 sensitizers may cause impediments in electron transfer to acceptor and possibilities of regeneration.
Adsorption of metal-free Sensitizer on the (TiO2)3 surface:
Apart from the geometries, energy levels of HOMO and LUMO in an isolated state, we also consider the binding of the studied sensitizers to the small oxide cluster of Titanium (TiO2)3. Where, all the titanium atoms share equivalent environment. By the introduction of a donating group to the donor, the absorption band can be further extended to longer wavelengths. A enhance spectral response can be achieved from the sensitizer by maintaining the electron injection and regeneration. Thus, a sufficient control of energy gap (Egap) by structural modification is extremely important in the designing a high-performance DSSCs. The optimized structure of (TiO2)3 has been adopted from the an earlier report, [42] and optical and electronic properties of adsorbed sensitizers were calculated by using TD-DFT theory by using CAM-B3LYP functional and hybrid basis set of LANL2DZ /6-31G on Gaussian 9.0. The adsorption was performed by two most commonly accepted model—a monodentate binding mode and second a bidentate mode. [46] However, we have decided to choose bidentate mode because in our test result it provide lower total energy than monedentate binding mode. Therefore, our designed sensitizers were optimized using bidentate-chelating mode. The binding configurations (HOMO-LUMO of sensitizer-(TiO2)3), adsorption energy and bidentate-chelation distances were shown in Fig. 6 while Egap (HOMO-LUMO level) are presented in Table 1 and Fig. 5. The calculated bond length between the Titanium atom and the Oxygen atoms of these optimized structures were found in range of 1.983–2.187 Å. Different types of electronic transitions have contributed to major adsorption wavelength. For example, a HOMO + 2 to LUMO prevails for D1 and HOMO + 4 for D4, in absorption maxima. However, it is noteworthy that all HOMOs level were localized around the donor moiety and all LUMOs level covered the acceptor as well as the TiO2 cluster. Thus, from these calculated data we derive the conclusion that these sensitizers would help the efficient interfacial electron injection from sensitizer to the semiconductor surface and eventually, would provide best performance in power conversion efficiencies. Furthermore, the LUMO level of the studied sensitizers decreased dramatically in chelation with semiconductor as compared to the isolated ones, as shown in Table 1. This drastic decrease is a result of the strong coupling between the LUMO of the sensitizers and surface of the TiO2. If we observe the LUMO level difference when isolated to chelating mode, D1 lowered by 2.136 eV, D4 by 1.967 eV and D6 by 1.663 eV as in Fig. 5. This massive decrease in LUMO values in chelation is another proof of the excellent manifestation to show good results in power conversion efficiencies of DSSC applications.