Bioorganic dye-sensitized solar cell of carotenoid–pheophytin a–TiO₂

Abstract

The isolated carotenoid and pheophytin a, the complex of carotenoid–pheophytin a and a carotenoid–pheophytin a–TiO₂ film have been investigated using the time-dependent density functional theory. Charge transfer mechanisms were revealed by the three-dimensional visualization technology. The photoinduced electron transfer, energy transfer and photoelectric conversion efficiency were systematically investigated in the above systems. Furthermore, the charge transfer mechanism and the energy transfer mechanism were studied when the investigated systems were optically excited in electronic transitions.

---

Introduction

Organic solar cells benefit from the advantages of a low cost fabrication,¹ physical flexibility, suitability for mass production and causing little damage to the environment,^2–5 are supposed to be a potential technology for next generation clear energy. Compared to inorganic silicon solar cells, the great superiorities make organic solar cells a huge development prospect. Organic solar cells are mainly used to realize photovoltaic conversions based on semiconductors as the active material. In the past few years, theoretical and experimental investigations have been mainly focused on finding novel and low-cost methods to convert solar radiation to electric energy. The photoelectric conversions of various semiconductor electrodes and electrolyte systems have been widely studied.^6–8 Since the 1990s, nanoscale semiconductor organic composite materials have turned out to be the main research object of photoelectric conversion, due to the appearance and rapid development of novel active materials. Unfortunately, the organic solar cell has not been widely adopted in people's lives because of a low photoelectric conversion efficiency, narrow spectral response range and instability of the cell performance.^9–11 In recent years, despite organic solar cells have experienced rapid developments, they still remain in the developing stages, and thus lots of questions on their science and technology still needs to be answered and their performance needs to be improved.¹²

Dye-sensitized solar cells (DSSC) have a number of attractive features: they can be easily prepared using conventional roll-printing techniques, they are semi-flexible and semi-transparent, which offers a variety of uses that are not applicable to glass-based systems, and most of the materials used in dye-sensitized solar cells are low-cost.¹³ For bioorganic systems, no pollution occurred after the cell was destructed, and the low-cost make them better active materials for the organic solar cell. In photosynthesis, the role of pheophytin a is to absorb and transfer solar radiation. Because of the structural characteristics of porphyrins, pheophytin a has a long life in excited states that can ensure effective charge separation. Due to the special π system in the porphyrin molecule, pheophytin a can introduce some functional groups into the molecule system, which can broaden the spectrum response range, and then improve the photoelectric conversion efficiency. The spectral response of carotenoids is particularly important to external solar radiation because carotenoids serve as the electron donor to absorb photons in the carotenoid–pheophytin a molecule system. However, the spectral response of carotenoids is around 400 nm, which can directly affect the photoelectric conversion efficiency. Wu and co-workers found that a red-shift phenomenon of the absorption of the S band can be induced while introducing an electronic group into the middle position of the porphyrin, which lowers the HOMO–LUMO energy levels of the molecule.¹⁴

TiO₂ is a low-cost, non-toxic semiconductor material with high stability and excellent corrosion resistance. Because of its high roughness, solar radiation can be reflected several times by its surface and repeatedly absorbed by a donor, which largely improves the utilization rate of solar radiation. However, the spectral response range of TiO₂ is mainly in the ultraviolet region and the forbidden band is very broad, which significantly limits the absorption of visible light. However, TiO₂ can be linked to an acceptor by chemical bonds. Ma and co-workers studied the influence of the connection of TiO₂ to an acceptor on the properties of solar cells.¹⁵ Their results show that the photoelectric conversion efficiency of solar cells with chemical bond connections is 9 and 60 times than that for weak interaction connections. Under specific conditions, a carotenoid and pheophytin a can link to the surface of TiO₂ to form a supermolecule, which not only improves the light absorption, but also broadens the spectral response region. Based on this supermolecule, electrons can directly transfer to the surface or interior of TiO₂. It can efficiently increase the charge injection efficiency and even realize an instantaneous charge transfer. The photoinduced electron transfer between a carotenoid and TiO₂ nanoparticles has been experimentally reported.¹⁶ Furthermore, the self-assembly and photoinduced electron transfer for the system of a carotenoid and pheophytin on a TiO₂ semiconductor surface have also been reported.¹⁷

In this paper, appropriate bioorganic molecules of carotenoid, pheophytin a and TiO₂ were selected to study the photoinduced charge transfer and energy transfer mechanism, while they are optically excited in electronic transitions. The three-dimensional (3D) visualization technology^18–20 was adopted to reveal the process of photoinduced charge transfer. Visualization of the charge transfer process directly demonstrates that the electrons transfer from the carotenoid to TiO₂ for the carotenoid–pheophytin a–TiO₂ film in a wider spectroscopy response, which is the key factor to a high efficiency bioorganic (DSSC) solar cell. Furthermore, our results show that the method of the chemical bond connection is more suitable for the organic solar cell.

Calculation methods

The isolated carotenoid,¹⁶ pheophytin a,¹⁷ and the complex of carotenoid–pheophytin a were selected as active materials for the bioorganic solar cells. All the quantum chemical calculations were performed by the Gaussian 09 software.²¹ The ground-state geometries were optimized using the density functional theory (DFT),²² B3LYP functional,²³ and the 6-31G (d) basis set. The electronic transitions in the optical absorption were calculated using the time-dependent DFT (TD-DFT),²⁴ CAM-B3LYP functional²⁵ and the 6-31G(d) basis set. The vertical excitation energies and oscillation strengths were also calculated at the geometries of the ground-state. In addition, a rutile nanocrystal TiO₂ cluster²⁶ was introduced to build the complex of carotenoid–pheophytin a–TiO₂, which was optimized using the DFT, B3LYP functional and 6-31G (d) basis set for C, O, N, H, and the LANL2DZ basis set²⁷ for the Ti atom. Charge difference density^18–20 was calculated by the three-dimensional (3D) visualization technology to study the relationship between the structure and performance of the carotenoid–pheophytin a–TiO₂ film. Furthermore, the visualization of the molecular system at excited state was realized, which revealed the mechanisms of the photoinduced charge transfer and energy transfer.

Results and discussion

The carotenoid (reans-8′-apo-β-caroten-8′-oic acid) and pheophytin a are two of the most common materials in biological nature. Despite the strong absorption, which exists for carotenoid and pheophytin a, their spectral responses are limited in a narrow region (420–550 nm). Surprisingly, once they combine to form a complex molecule system, the spectral absorption range is largely expanded.

The carotenoid is a common nature pigment and the molecule structure is shown in Fig. 1(a). The carotenoids mainly absorb blue-violet light and transfer the energy to pheophytin a in photosynthesis. Because of the long conjugate C=C bond, the carotenoid has strong absorption peaks in the visible spectral region. Fig. 1(b) shows the structure of pheophytin a, which consists of a tetrapyrrole ring and four methylene groups. In order to broaden the pheophytin a spectrum response to the visible light region, the carotenoid molecule was linked to the pheophytin a based on coulombic interactions.

Fig. 1 The optimized chemical structure of (a) carotenoid, (b) pheophytin a, and (c) carotenoid–pheophytin a.

The excited states of the isolated carotenoid, pheophytin a and the complex system of carotenoid–pheophytin a were also calculated (see Table 1). Fig. 2(a) shows that the strong absorption peak of the carotenoid, which is located at 456.85 nm, in which the oscillator strength reaches the maximum intensity. In general, the oscillator strength can be used to reflect the ability to absorb light, as well as the transition probability, which reveals that the electric transition of molecules occur in the strong absorption region, and thus improve the photoelectric conversion efficiency. Fig. 2(b) shows that the absorption spectra of pheophytin a are mainly located in range from 250 nm to 450 nm. Fifteen excited states were calculated for pheophytin a, and the oscillator strength of the third and fourth excited state reached to the maximum intensities of f = 0.79 and 1.00, respectively (see Table 1). This reveals that the third and fourth excited state have a larger electronic transition probability compared to other excited states. Furthermore, pheophytin a has a strong absorption, at which there is maximum oscillator strength, and thus the electronic transition can occur. Fig. 2(c) shows that there are two absorption peaks for the complex of carotenoid–pheophytin a. The absorption peak located at 385.07 nm arises from the spectral response of pheophytin a, whereas the peak located at 473.58 nm arises from the spectral response of carotenoid. Interestingly, the spectral response region is largely broadened while the carotenoid and pheophytin a formed a complex molecule. The largest oscillator strength in Fig. 2(c) reveals that the S₄ excited state has strong electron transition ability and light absorption. On the basis of the above results, pheophytin a and carotenoid were selected as the electron donors in this work.

Table 1 Calculated transition energies (eV, nm) and oscillator strengths (f) for the isolated carotenoid and pheophytin a

	States	E^a (eV)	f^b	CI^c
a The numbers in parentheses are the transition energy in wavelength.b Oscillator strength.c H stands for HOMO and L stands for LUMO, CI coefficients are in absolute values.
Carotenoid	S1	2.7139	3.5762	0.68439(H → L)
	S2	3.8593	0.0793	0.67021(H-1 → L)
	S3	4.0016	0.0541	0.67042(H → L + 1)
Pheophytin a	S1	2.0943	0.0231	0.39267(H → L)
	S2	2.3112	0.0027	0.37604(H → L)
	S3	3.2415	0.7904	0.24173(H → L)
	S4	3.3997	1.0063	0.36101(H → L)
	S5	3.6079	0.1600	0.61475(H-2 → L + 1)
	S6	3.7711	0.0117	0.51434(H-6 → L)
	S7	3.8336	0.3109	0.55229(H-2 → L)
	S8	3.9926	0.0170	0.59530(H-3 → L)
	S9	4.0317	0.4418	0.44364(H-3 → L + 1)
	S10	4.1161	0.0723	0.21271(H-4 → L)
	S11	4.2727	0.1671	0.60836(H → L + 2)
	S12	4.3025	0.0688	0.67389(H-1 → L + 2)
	S13	4.3699	0.0011	0.33489(H-8 → L + 1)
	S14	4.4958	0.0009	0.38301(H-10 → L)
	S15	4.5826	0.1316	0.64125(H-7 → L)

---

Fig. 2 The optical electronic state absorption spectra of (a) carotenoid, (b) pheophytin a, and (c) carotenoid–pheophytin a.

To better understand the excited-state properties of the isolated carotenoid and pheophytin a, a theory analysis was performed using the three-dimensional (3D) real space analysis method and the two-dimensional (2D) sit representation. Fig. 3 shows the 3D charge difference densities of carotenoid. Charge difference densities, defined as the difference between the charge density after bonding and the charge density of the corresponding point, is the charge redistribution after the atoms form a system.^18–20 The direction of the charge transfer in the bonding electron coupling process can be seen clearly. As is shown in Fig. 3, electrons and holes are localized at the spine of carotenoid, which reveals that S₁, S₂ and S₃ are localized excited states, and the electrons transfer to the sides of the spine whereas the holes gradually transfer to the central chains.

Fig. 3 Charge difference densities for different electronic excited states of carotenoid, where green and red stand for the holes and electrons, respectively.

The charge difference density of pheophytin a is shown in Fig. 4. S₃ and S₄ are strong absorptions located at 382.49 nm and 364.69 nm, respectively. The electrons and holes localized on the entire porphyrin reveal a strong charge localization in pheophytin a. S₁₃ and S₁₄ are weak absorptions for which electrons and holes are also localized in the entire porphyrin, which further proved the charge localization in pheophytin a. Once the carotenoid is placed parallel close to the pheophytin a, electrons will rapidly transfer between the carotenoid and the pheophytin a.

Fig. 4 Charge difference densities of pheophytin a, in which green and red stand for the holes and electrons, respectively.

In the transition process of electrons, the binding state is created by the coulombic force between the electrons in the conduction band and the holes in the valence band. The separation of excitons is to generate electrons and holes by overcoming the coulombic force. The exciton binding energy is one of the key factors directly determining the separation of electrons and holes in the organic solar cell. The exciton binding energy is mainly from the coulombic force between donor cations and acceptor anions.

where q_d and q_a are the partial charges in atoms d and a from donor cations and acceptor anions, respectively. ε is the dielectric constant. Δr index is a quantitative indicator of the electron excitation mode, which is a measure of the CT length (see Table 2). The value of the exciton binding energy determines whether the separation of electrons and holes is successfully realized. As shown in Table 2, the binding energy of carotenoid–pheophytin a reveals that we can get the minimum Δr of 0.279 Å. In this case, the electrons and holes are completely located in pheophytin a, in which they hardly separate and transfer. For the eleventh excited state, Δr reaches the maximum of 4.924 Å. In this case, the electrons are located in pheophytin a, and the holes are not completely located in carotenoid. Generally, the separation of electrons and holes is easily realized for a low coulombic force. In other words, the binding energy decreases as Δr increases, and thus increasing the separation probability of electrons and holes.

Table 2 Selected electronic transition energies (eV) and the corresponding oscillator strengths (f), main compositions and CI coefficients of carotenoid–pheophytin a

States	Transition energy^a (eV)	f^b	Δr^c (Å)	Excited-state property^d
a The numbers in parentheses are the transition energy in wavelength.b Oscillator strength.c Δr index is a quantitative indicator of the electron excitation mode, which is a measure of the CT length.d PC61BM and BT in parentheses present that the density are localized on the fullerene and polymer, respectively.
S1	2.0850 (594.65 nm)	0.0127	0.279202	LE
S2	2.3051 (537.88 nm)	0.0036	0.426212	LE
S3	2.4014 (516.31 nm)	0.0012	4.449655	ICT
S4	2.6180 (473.58 nm)	1.9146	2.981704	ICT
S5	2.7042 (458.49 nm)	0.3422	4.000988	ICT
S6	3.2198 (385.07 nm)	1.0314	0.316830	ICT
S7	3.3606 (368.94 nm)	1.2517	0.380151	ICT
S8	3.4436 (360.04 nm)	0.0791	4.042327	ICT
S9	3.5201 (352.22 nm)	0.0236	4.071160	ICT
S10	3.5928 (345.09 nm)	0.1652	0.525675	LE
S11	3.6468 (339.98 nm)	0.0012	4.923839	ICT
S12	3.7996 (326.31 nm)	0.0595	4.314745	ICT
S13	3.8029 (326.03 nm)	0.1976	1.611009	ICT
S14	3.8732 (320.11 nm)	0.0988	3.367877	ICT
S15	3.9024 (317.72 nm)	0.0223	0.232065	LE
S16	3.9406 (314.63 nm)	0.0851	4.283794	ICT
S17	3.9752 (311.89 nm)	0.0458	0.707571	ICT
S18	4.0259 (307.97 nm)	0.3523	1.388938	ICT
S19	4.1096 (301.69 nm)	0.0719	3.029693	ICT
S20	4.2026 (295.02 nm)	0.1199	3.456301	ICT

---

Fig. 5 shows the charge difference density of the complex of carotenoid–pheophytin a. Two types of excited states exist in the absorption band of the complex of carotenoid–pheophytin a: one is the strong resonance charge transfer (CT) excited state, in which electrons transfer takes place from the main chain of the carotenoid to the pheophytin a. For the S₇, Δr is 0.380 Å, which indirectly reveals a large binding energy, and the electrons are located in both the carotenoid and pheophytin a. For the S₈, Δr is 4.042 Å, which indirectly reveals a small binding energy in which the electrons and holes are easy to separate, and then the electrons completely transfer to the carotenoid. The other is the weak resonance charge transfer process, which shows that electrons transfer from the pheophytin a to the main chain of the carotenoid. S₁₀ is not a pure intermolecular charge transfer excited state because electrons and holes are localized on pheophytin a. S₁₁ is a pure intermolecular charge transfer excited state because electrons and holes are localized on pheophytin a and carotenoid, respectively.

Fig. 5 Charge difference densities of carotenoid–pheophytin a, where green and red stand for the holes and electrons, respectively.

The 3D visualization technology can clearly reveal the transfer direction of the electric charge in the coupling process of bonding and bonding electrons, but cannot provide the information of electronic coherence and the source of electrons and holes. As a result, the 2D site representation is introduced to analyze the electronic coherence. Note that, the larger value along the off-diagonal element in the 2D site representation demonstrates a stronger electronic coherence where it reveals the higher transfer probability of electric transfer. The diagonal elements reveal a small transfer region of electronic transfer, whereas the points far away from the diagonal reveal a wider transfer region of electronic transfer. These points with large values, which are close to the diagonal, demonstrate a strong electronic coherence between an atom with the adjacent atom.

As shown in Fig. 6, the observed S₁, S₁₅ are the localized excited states. For S₁, both electrons and holes are localized on the pheophytin a molecule and the electron–hole coherence is strong at the left bottom of pheophytin a. For S₁₅, both electrons and holes are localized on the carotenoid molecule. The electron–hole coherence is strong at the top right corner of the carotenoid. S₈ is an intermolecular electric charge transfer state, in which electrons and holes cohere between carotenoid and pheophytin a.

Fig. 6 The 2D contour plots of the selected transition density matrix of the pheophytin a linked carotenoid with different excite-state properties. The color bars are shown at the right of the figure.

TiO₂ was linked to the complex of carotenoid–pheophytin a based on chemical bonds and coulombic interactions, which were used to study their excited state properties. Simply, the carotenoid–pheophytin a–TiO₂ was renamed C–P–T. Fig. 7(a) shows the connection by coulombic interactions (C–P–T-1), which is a π–π connection. The strong absorption peaks (S₃, S₁₇) are located in the UV-vis region, and S₁, S₂, are located in the visible region. The strongest oscillator strength is f = 2.7349, which is located at 505.58 nm whereas the weakest oscillator strength is f = 0.0001, which is located at 337.63 nm. Fig. 7(b) shows the connection by chemical bonds (C–P–T-2). The strong absorption peaks (S₁₁, S₃₈, S₄₁) are located in the UV-vis region, and S₁, S₂, S₃, S₄ are located in the visible region. The strongest oscillator strength is f = 2.7349, which is located at 506 nm, whereas the weakest oscillator strength is f = 0.0001, located at 338 nm. Compared to Fig. 6(a) and (b), the spectral response range for C–P–T-2 is broader than that for C–P–T-1, in which the former is in the range from 300 nm to 750 nm while the latter is in the region between 300 nm and 650 nm.

Fig. 7 The optimized chemical structure and the optical electronic state absorption spectra of carotenoid–pheophytin a–TiO₂.

The study of the excited states of carotenoid–pheophytin a–TiO₂ shows that there are three types of charge transfer: one is the electrons transfer from carotenoid to TiO₂; the second is the electrons transfer from carotenoid and pheophytin a to TiO₂; the last is the electrons transfer form pheophytin a to TiO₂. For S₂₀, electrons and holes are localized in the TiO₂ and carotenoid, in which the electron–hole coherence is strong at the crossing of TiO₂ and carotenoid, as shown in Fig. 8(a). For S₂₇, electrons and holes are localized in TiO₂ and carotenoid–pheophytin a, respectively, in which the electron–hole coherence is strong at the crossing of TiO₂ and carotenoid–pheophytin a. For S₃₁, electrons and holes are localized on TiO₂ and pheophytin a, respectively, in which the electron–hole coherence is strong at the crossing of TiO₂ and pheophytin a. As shown in Fig. 8(b), for S₃₆, electrons and holes are localized in TiO₂ and carotenoid, respectively, in which the electron–hole coherence is strong at the center of TiO₂. For S₄₃, electrons and holes are localized in TiO₂ and carotenoid–pheophytin a, respectively, in which the electron–hole coherence is strong at the center of TiO₂. For S₅₂, electrons and holes are localized in TiO₂ and pheophytin a, respectively, in which the electron–hole coherence is strong at the left side of TiO₂.

Fig. 8 The 2D contour plots of the selected transition density matrix for C–P–T-1 and C–P–T-2. The color bars are shown at the right of the figure.

Fig. 9 shows the charge difference densities for P–C–T-1 and P–C–T-2. The electric charge transfer for C–P–T-1 in the strong resonance CT excited states occurs between the main chain of carotenoid and TiO₂. S₁₃ is the pure intermolecule electron transition excited state, in which the electrons are localized on the entire TiO₂ and the holes are localized on the main chain of the carotenoid and pheophytin a. However, S₁₀ and S₁₂ are not pure intermolecular charge transfer excited states, in which the electrons transfer from pheophytin a to carotenoid, and then to TiO₂. Fig. 9 (b) shows that the oscillator strengths of S₁₁ (f = 2.6159) and S₄₁ (f = 1.0594) in C–P–T-2, which are considered as strong absorbing states. In addition, S₂₂ and S₄₃ are pure intermolecule electronic transfer excited states, because the electronic for the C–P–T-2 is localized. For S₈ and S₄₁, electrons transfer first from pheophytin a to carotenoid, and then to TiO₂. S₁₁, is mainly formed by the electronic transfer from the HOMO to LUMO, which is the strongest absorption excited state, which was revealed by the maximum probability of electron transition. By comparing C–P–T-1 with C–P–T-2, the electronic absorption and injection can be directly affected by the manner in which carotenoid is linked to TiO₂. For the chemical bond connection, electrons immediately inject into TiO₂, which can effectively decrease the loss of electrons during the electric charge transfer process. The chemical bond connection can also help us to improve the light absorption and the photoelectric conversion efficiency.

Fig. 9 Charge difference densities of carotenoid–pheophytin a–TiO₂, where green and red stand for the holes and electrons, respectively.

Conclusion

In this work, the excited state properties of the isolated carotenoid, pheophytin a, the complex of carotenoid–pheophytin a and carotenoid–pheophytin a–TiO₂ film were studied, using quantum chemical methods. The vertical excited energies and absorption spectra were also investigated. Depending on the three-dimensional visualization technology, the visualization of charge transfer can be clearly realized. The complex of carotenoid–pheophytin a not only shows a broader spectrum response than the isolated carotenoid and pheophytin a, but also demonstrates that the strong resonance and weak resonance electric charge separation excited state exists in the absorption band. The visualization of the charge transfer process directly demonstrates that the electrons transfer from the carotenoid to TiO₂ for the carotenoid–pheophytin a–TiO₂ film in a wider spectroscopy response, which is the key factor to a high efficiency bioorganic solar cell. Chemical bond and coulombic interaction based connections were used to link the carotenoid to TiO₂. Further, our results show that the chemical bond connection is more suitable for the organic solar cell.

Acknowledgements

This work was supported by National Key Basic Research Program of China (2014CB921001), National Natural Science Foundation (11474141, 91436102, 11374353, 11304135, 11474239 and 21173171), the Special fund based research new technology of methanol conversion and coal instead of oil and the china postdoctoral science Foundation (2014M550158), the Program of Liaoning Key Laboratory of Semiconductor Light Emitting and Photocatalytic Materials, the Scientific Research Foundation for the Returned Overseas Chinese Scholars, State Education Ministry and the Scientific Research Foundation for the Doctor of Liaoning University.

Notes and references

1. R. H. Friend, R. W. Gymer, A. B. Holmes, J. H. Burroughes, R. N. Marks, C. Taliani, D. D. C. Bradley, D. A. D. Santos, J. L. Bredas and M. Logdlund, Nature, 1999, 397, 121.
2. H. E. A. Huitema, G. H. Gelinck, J. B. P. H. Vanderputten, K. E. Kuijk, C. M. Hart, E. Cantatore and D. M. deLeeuw, Adv. Mater., 2002, 14, 1201.
3. P. Song, Y. Z. Li, F. C. Ma, T. Pullerits and M. T. Sun, J. Phys. Chem. C, 2013, 117, 15879.
4. C. J. Brabec, N. S. Sariciftci and J. C. Hummelen, Adv. Funct. Mater., 2001, 11, 15.
5. Q. Zhou and T. M. Swager, J. Am. Chem. Soc., 1995, 117, 12593.
6. G. Yu, J. Gao, J. C. Hummelen, F. Wudl and A. J. Heeger, Science, 1995, 270, 1789.
7. S. Mukamel, S. Tretiak, T. Wagersreiter and V. Chernyak, Science, 1997, 277, 781; S. Tretiak and S. Mukamel, Chem. Rev., 2002, 102, 3171.
8. S. Tretiak, A. Saxena, R. L. Martin and A. R. Bishop, Phys. Rev. Lett., 2002, 89, 097402; S. Tretiak, A. Saxena, R. L. Martin and A. R. Bishop, Proc. Natl. Acad. Sci., 2003, 99, 2185.
9. M. Chegaar, N. Nehaoua and A. Bouhemadou, Energy Convers. Manage., 2008, 49, 1376.
10. J. Chandrasekaran, D. Nithyaprakash, K. B. Ajjan, S. Maruthamuthu, D. Manoharan and S. Kumar, Renewable Sustainable Energy Rev., 2011, 15, 1228.
11. K. Bouzidi, M. Chegaar and M. Aillerie, Energy Procedia, 2012, 18, 1601.
12. S. E. Shaheen, C. J. Brabec, N. S. Sariciftci, F. Padinger, T. Fromherz and J. C. Hummelen, Appl. Phys. Lett., 2001, 78, 841.
13. B. Regan and M. Grätzel, Nature, 1991, 353, 737.
14. S. L. Wu, H. P. Lu, H. T. Yu, S. H. Chuang, C. L. Chiu, C. W. Lee, E. W. G. Diau and C. Y. Yeh, Energy. Environ. Sci., 2010, 3, 949.
15. T. Ma, K. Inoue, H. Noma, K. Yao and E. Abe, J. Photochem. Photobiol., A, 2002, 152, 207.
16. J. Pan, G. Benko, Y. H. Xu, T. Pascher, L. C. Sun, V. Sundstrom and T. Polívka, J. Am. Chem. Soc., 2002, 124, 13949.
17. J. X. Pan, Y. H. Xu, L. C. Sun, V. Sundstrom and T. Polivka, J. Am. Chem. Soc., 2004, 126, 3066.
18. M. T. Sun, J. Chem. Phys., 2007, 127, 084706.
19. Y. Li, T. Pullerits, M. Zhao and M. Sun, J. Phys. Chem. C, 2011, 115, 21865.
20. L. Xia, M. Chen, X. Zhao, Z. Zhang, J. Xia, H. Xu and M. Sun, J. Raman Spectrosc., 2014, 45, 533.
21. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, et al.Gaussian 09, Revision A. 02, Gaussian, Inc., Wallingford, CT, 2009.
22. P. Hohenberg and W. Kohn, Phys. Rev., 1964, 136, 864.
23. C. Lee, W. Yang and R. G. Parr, Phys. Rev. B: Condens. Matter Mater. Phys., 1988, 37, 785.
24. E. K. U. Gross and W. Kohn, Phys. Rev. Lett., 1985, 55, 2850.
25. T. Yanai, D. Tew and N. Handy, Chem. Phys. Lett., 2004, 393, 51.
26. Z. W. Qu and G. J. Kroes, J. Phys. Chem. C, 2007, 111, 16808.
27. P. J. Hay and W. R. Wadt, J. Chem. Phys., 1985, 82, 270.

---

Footnote

† Contributed equally.

---