Chlorophyll-a Derivatives with Various Hydrocarbon Ester Groups for

Apr 9, 2010 - Ritsumeikan University, Kusatsu, Shiga 525-8577, Japan,. ) Faculty of Science and Technology, Kwansei Gakuin. University, Gakuen 2-1, ...
0 downloads 0 Views 1017KB Size
Download PDF
pubs.acs.org/Langmuir © 2010 American Chemical Society

)

)

Chlorophyll-a Derivatives with Various Hydrocarbon Ester Groups for Efficient Dye-Sensitized Solar Cells: Static and Ultrafast Evaluations on Electron Injection and Charge Collection Processes Xiao-Feng Wang,*,†,‡ Hitoshi Tamiaki,*,§ Li Wang, Naoto Tamai, Osamu Kitao,‡ Haoshen Zhou,‡ and Shin-ichi Sasaki§,^

)

† Environmental and Renewable Energy Systems Division, Graduate School of Engineering, Gifu University Yanagido 1-1, Gifu 501-1193, Japan, ‡Energy Technology Research Institute, National Institute of Advanced Industrial Science and Technology (AIST), Umezono 1-1-1, Tsukuba, Ibaraki 305-8568, Japan, §Department of Bioscience and Biotechnology, Faculty of Science and Engineering, Ritsumeikan University, Kusatsu, Shiga 525-8577, Japan, Faculty of Science and Technology, Kwansei Gakuin University, Gakuen 2-1, Sanda 669-1337, Hyogo, Japan, and ^Nagahama Institute of Bio-Science and Technology, Nagahama, Shiga 526-0829, Japan

Received July 27, 2009. Revised Manuscript Received March 25, 2010 Five chlorophyll-a derivatives, chlorins-1-5 possessing C32-carboxy and O174-esterified hydrocarbon groups including methyl, hexyl, dodecyl, 2-butyloctyl, and cholesteryl were synthesized. Their performance as sensitizers in dye-sensitized solar cells (DSSCs) was compared. These sensitizers have similar surface coverage on the unit surface of TiO2 film and their absorption spectra on transparent TiO2 films were identical. On the basis of DFT and TD-DFT calculations of these sensitizers in ethanol, a major difference between them was the geometry of the hydrocarbon ester group, to affect their electron injection and charge recombination with the TiO2 electrode rather than the energy level of their molecular orbitals. DSSC based on chlorin-3 with a dodecyl ester group gave a solar energy-to-electricity conversion efficiency of 8%, which was the highest among all the chlorophyllous sensitizers. The large photocurrent in the chlorin-3 sensitized solar cell can be explained by the least impedance in the electrolyte-dye-TiO2 interface in electrical impedance spectroscopy measurements. Subpicosecond time-resolved absorption spectroscopic studies have also been carried out to evaluate the electron injection and charge recombination dynamics in the dye-TiO2 interface. For the electron injection and charge recombination processes, a charge separated state of the dye-TiO2 complex has been found to be free from the type and concentration of dye sensitizer, reflecting the same type of electron transfer process for all the five chlorin sensitizers. A new quenching pathway of the dye excitation, which is probably from the exciton annihilation, in addition of the charge recombination has been observed for chlorin-1 and chlorin-5, but not for chlorin-3. The higher open-circuit photocurrent observed in the present dyes with larger ester groups can be attributed to the reduced leaking of charges in the TiO2-electrolyte interface, which was supported by the longer electron lifetimes.

Introduction Chlorophylls (Chls) are naturally occurring photosynthetic pigments playing important roles in light-harvesting pigmentprotein complexes and electron-transferring reaction centers.1,2 Recently, Chl derivatives as well as synthetic porphyrins (Pors) *To whom correspondence should be addressed. E-mail: charles1976110@ hotmail.com (X.-F.W.); [email protected] (H.T.). (1) Scheer, H. In Advances in Photosynthesis and Respiration; Grimm, B., Porra, R. J., R€udiger, W., Scheer, H., Eds.; Springer: Dordrecht, The Netherlands, 2006; Vol. 25, Chapter 1. (2) Tamiaki, H.; Shibata, R.; Mizoguchi, T. Photochem. Photobiol. 2007, 83, 152. (3) (a) Wang, X. F.; Xiang, J.; Wang, P.; Koyama, Y.; Yanagida, S.; Wada, Y.; Hamada, K.; Sasaki, S.; Tamiaki, H. Chem. Phys. Lett. 2005, 408, 409. (b) Wang, X.-F.; Kakitani, Y.; Xiang, J.; Koyama, Y.; Rondonuwu, F. S.; Nagae, H.; Sasaki, S.; Tamiaki, H. Chem. Phys. Lett. 2005, 416, 229. (c) Wang, X.-F.; Arihiro, M.; Koyama, Y.; Nagae, H.; Sasaki, S.; Tamiaki, H.; Wada, Y. Chem. Phys. Lett. 2006, 423, 470. (d) Wang, X.-F.; Koyama, Y.; Wada, Y.; Sasaki, S.; Tamiaki, H. Chem. Phys. Lett. 2007, 439, 115. (e) Wang, X.-F.; Zhan, C. H.; Maoka, T.; Wada, Y.; Koyama, Y. Chem. Phys. Lett. 2007, 447, 79. (f) Wang, X.-F.; Koyama, Y.; Nagae, H.; Wada, Y.; Sasaki, S.; Tamiaki, H. J. Phys. Chem. C 2008, 112, 4418. (g) Wang, X.-F.; Kitao, O.; Zhou, H.; Tamiaki, H.; Sasaki, S. Chem. Commun. 2009, 1523. (h) Wang, X.-F.; Kitao, O.; Zhou, H.; Tamiaki, H.; Sasaki, S. J. Phys. Chem. C 2009, 113, 7954. (i) Wang, X.-F.; Tamiaki, H. Energy Environ. Sci. 2010, 3, 94. (j) Wang, X.-F.; Kitao, O.; Hosono, E.; Zhou, H.; Sasaki, S.; Tamiaki, H. J. Photochem. Photobiol. A: Chem. 2010, 210, 145. (k) Wang, X.-F.; Koyama, Y.; Kitao, O.; Wada, Y.; Sasaki, S.; Tamiaki, H.; Zhou, H. Biosens. Bioelectron. 2010, 25, 1970.

6320 DOI: 10.1021/la1005715

are being developed as sensitizers for efficient dye-sensitized solar cells (DSSCs), and the highest solar energy-to-electricity conversion efficiency (η) using Chl and Por sensitizers to date were reported to be 6.6% and 7.1%,3,4 respectively. Compared to other inorganic and organic sensitizers, such cyclic tetrapyrrole type Chl sensitizers possess at least the following six advantages: (1) The positions and the intensities of Soret, Qx, and Qy transition bands (major electronic absorption bands) are readily controllable by chemical modification of the macrocycle. (2) Basic studies (4) (a) Campbell, W. M.; Burrell, A. K.; Officer, D. L.; Jolley, K. W. Coord. Chem. Rev. 2004, 248, 1363. (b) Wang, Q.; Campbell, W. M.; Bonfantani, E. E.; Jolley, K. W.; Officer, D. L.; Walsh, P. J.; Gordon, K.; Humphry-Baker, R.; Nazeeruddin, M. K.; Gr€atzel, M. J. Phys. Chem. B 2005, 109, 15397. (c) Campbell, W. M.; Jolley, K. W.; Wagner, P.; Wagner, K.; Walsh, P. J.; Gordon, K. C.; Schmidt-Mende, L.; Nazeeruddin, M. K.; Wang, Q.; Gr€atzel, M.; Officer, D. L. J. Phys. Chem. C 2007, 111, 11760. (d) Rochford, J.; Chu, D.; Hagfeldt, A.; Galoppini, E. J. Am. Chem. Soc. 2007, 129, 4655. (e) Stromberg, J. R.; Marton, A.; Kee, H. L.; Kirmaier, C.; Diers, J. R.; Muthiah, C.; Taniguchi, M.; Lindsey, J. S.; Bocian, D. F.; Meyer, G. J.; Holten, D. J. Phys. Chem. C 2007, 111, 15464. (f) Eu, S.; Hayashi, S.; Umeyama, T.; Matano, Y.; Araki, Y.; Imahori, H. J. Phys. Chem. C 2008, 112, 4396. (g) Hayashi, S.; Tanaka, M.; Hayashi, H.; Eu, S.; Umeyama, T.; Matano, Y.; Araki, Y.; Imahori, H. J. Phys. Chem. C 2008, 112, 15576. (h) Park, J.; Lee, H.; Chen, J.; Shinokubo, H.; Osuka, A.; Kim, D. J. Phys. Chem. C 2008, 112, 16691. (i) Lee, C.; Lu, H.; Lan, C.; Huang, Y.; Liang, Y.; Yen, W.; Liu, Y.; Lin, Y.; Diau, E. W.; Yeh, C. Chem.;Eur. J. 2009, 15, 1403. (j) Imahori, H.; Umeyama, T.; Ito, S. Acc. Chem. Res. 2009, 42, 1809. (k) Lu, H.; Mai, C.; Tsia, C.; Hsu, S.; Hsieh, C.; Chiu, C.; Yeh, C.; Diau, E. W. Phys. Chem. Chem. Phys. 2009, 11, 10270.

Published on Web 04/09/2010

Langmuir 2010, 26(9), 6320–6327

Wang et al.

Article

on Chl-based DSSCs have brought us vast insights and principles to deepen our knowledge of the interfacial electron- and energytransfer mechanism. (3) The natural resource of Chls is so large that future large-scale production of Chl dyes will be easy. (4) Chl sensitizers have low toxicity, so that such DSSCs will cause less serious environmental contamination when they are thrown away. (5) By modification of the tetrapyrrole macrocycle with different central metals (Zn, Mg, Ni, Cu, Pd, Fe, etc.), the excitedstates lifetime of Chls can largely be affected, and the LUMO orbital of the sensitizers can also be adjusted to match the energy level of the conduction band edges of different types of semiconductors. (6) Chl sensitizers strongly absorb solar energy in both shorter and longer wavelength regions, making them the most suitable partners in tandem use with an Ru complex. It has been found that, by introducing longer alkyl chains into a cyanine molecule, such dyes gave higher η values.5 The same strategy could also be used to improve the performance of DSSCs based on Chl sensitizers. Actually, a natural Chl-a molecule (Figure S6) contains a long phytyl ester group at the C17propionate residue of the chlorin macrocycle, which interacts with neighboring Chl molecules and surrounding oligopeptides and directs the arrangement of each Chl molecule to facilitate efficient energy transfer followed by light-driven charge separation. In the present study, we synthesized a set of Chl sensitizers, chlorins-1-5 (Figure 1), with different aliphatic hydrocarbon ester groups at the propionate residue to suppress the undesired back electron transfer in DSSC. We measured the performance of DSSCs based on these synthetic sensitizers and elucidated the mechanism that was responsible for the changes in the photocurrent and photovoltage. We obtained a η value of 8% for DSSC based on chlorin-3 under the irradiation of 100 mW cm-2 air mass 1.5 global (AM1.5G) sunlight, which is, to our best knowledge, the highest performance in Chls and Pors sensitized solar cells reported so far. The interfacial electron injection process between Chl and TiO2 is similar for all the five sensitizers, while the charge collection processes are the most efficient in chlorin-3, both based on the static DFT and TD-DFT calculations and the electric impedance measurements, and the transient subpicosecond timeresolved absorption spectroscopy

Experimental Section Synthesis of Sensitizers. Chlorin-1,3g methyl pyropheophorbide-d,6 methyl pyropheophorbide-a (chlorin-R1),6 and chlorinR27 were prepared according to reported procedures. Chlorins2-5 were synthesized by transesterification of methyl ester of methyl pyropheophorbide-d with ROH,8 followed by Wittig reaction of the 3-formyl group with Ph3PdCHCOOC(CH3)3, and acidic cleavage of the resulting tert-butyl ester. Their synthtetic route is shown in the Supporting Information, Figure S6. Synthetic procedures of dodecyl trans-32-carboxypyropheophorbidea (chlorin-3) as well as analytical data of other chlorins-2, -4, -5, and -R3 are described below. Synthesis of Dodecyl trans-32-Carboxypyropheophorbide-a (Chlorin-3). A mixture of methyl pyropheophorbide-d (110 mg, 0.20 mmol), 1-dodecanol (186 mg, 1.0 mmol), and bis(dibutylchlorotin) oxide (27 mg, 0.04 mmol) in toluene (20 mL) was refluxed for 3 h. The mixture was cooled to room temperature, and subjected to silica gel chromatography (Et2O-CH2Cl2, 1:19) (5) Sayama, K.; Tsukagoshi, S.; Hara, K.; Ohga, Y.; Shinpo, A.; Abe, Y.; Suga, S.; Arakawa, H. J. Phys. Chem. B 2002, 106, 1363. (6) Tamiaki, H.; Amakawa, M.; Shimono, Y.; Tanikaga, R.; Holzwarth, A. R.; Schaffner, K. Photochem. Photobiol. 1996, 63, 92. (7) Wang, Q.; Sasaki, S.; Tamiaki, H. Chem. Lett. 2009, 38, 648. (8) (a) Otera, J.; Dan-oh, N.; Nozaki, H. J. Org. Chem. 1991, 56, 5307. (b) Sasaki, S.; Tamiaki, H. Tetrahedron Lett. 2006, 47, 4965. (c) Tamiaki, H.; Michitsuji, T.; Shibata, R. Photochem. Photobiol. Sci. 2008, 7, 1225.

Langmuir 2010, 26(9), 6320–6327

Figure 1. Chemical structures of chlorin sensitizers. to give dodecyl pyropheophorbide-d (138 mg, 98%) as a reddish brown solid: mp 83-85 C; vis (THF) λmax 691 (relative intensity, 82%), 631 (8), 551 (15), 520 (15), 426 (100), 397 nm (80); 1H NMR (CDCl3) δ 11.48 (1H, s, CHO), 10.17 (1H, s, 5-H), 9.49 (1H, s, 10-H), 8.81 (1H, s, 20-H), 5.34, 5.19 (each 1H, d, J = 19 Hz, 131-CH2), 4.58 (1H, dq, J=2, 8 Hz, 18-H), 4.39 (1H, dt, J=9, 2 Hz, 17-H), 3.74 (3H, s, 2-CH3), 3.67 (3H, s, 12-CH3), 3.64 (2H, m, 8-CH2), 3.23 (3H, s, 7-CH3), 2.73, 2.33 (each 1H, m, 17-CH2), 2.58, 2.29 (each 1H, m, 171-CH2), 4.00, 3.94 (each 1H, m, 172COOCH2), 1.46 (2H, m, 172-COOCCH2), 1.25 (2H, m, 172COOC2CH2), 1.18 (16H, m, 172-COOC3(CH2)8), 1.86 (3H, d, J=8 Hz, 18-CH3), 1.67 (3H, t, J=8 Hz, 81-CH3), 0.85 (3H, t, J= 7 Hz, 172-COOC11CH3), -0.27, -2.19 (each 1H, s, NH  2); MS (APCI) m/z 705 (MHþ). The above mixture of 3-formyl-chlorin (128 mg, 0.34 mmol) and (tert-butoxycarbonylmethylene)triphenylphosphorane (339 mg, 0.90 mmol) in toluene (20 mL) was refluxed for 2 h. The mixture was cooled to room temperature, and subjected to silica gel chromatography (Et2O-CH2Cl2, 3:97) to give dodecyl trans32-(tert-butoxycarbonyl)-pyropheophorbide-a (130 mg, 95%) as a black solid: mp 120-122 C; VIS (CH2Cl2) λmax 682 (relative intensity, 56%), 622 (8), 546 (11), 515 (13), 421 nm (100); 1H NMR (CDCl3) δ 9.54 (1H, s, 10-H), 9.48 (1H, s, 5-H), 9.08 (1H, d, J=16 Hz, 3-CH), 8.68 (1H, s, 20-H), 7.00 (1H, d, J=16 Hz, 31CH), 5.30, 5.15 (each 1H, d, J=19 Hz, 131-CH2), 4.54 (1H, dq, J= 2, 7 Hz, 18-H), 4.35 (1H, dt, J=9, 2 Hz, 17-H), 3.70 (2H, q, J= 8 Hz, 8-CH2), 3.69 (3H, s, 12-CH3), 3.53 (3H, s, 2-CH3), 3.27 (3H, s, 7-CH3), 2.70, 2.54 (each 1H, m, 17-CH2), 2.33, 2.26 (each 1H, m, 171-CH2), 3.98, 3.94 (each 1H, m, 172-COOCH2), 1.45 (2H, m, 172-COOCCH2), 1.25 (2H, m, 172-COOC2CH2), 1.17 (16H, m, 172-COOC3(CH2)8), 1.84 (3H, d, J=7 Hz, 18-CH3), 1.74 (9H, s, COOC(CH3)3), 1.70 (3H, t, J=8 Hz, 81-CH3), 0.85 (3H, t, J= 7 Hz, 172-COOC11CH3), 0.15, -1.87 (each 1H, s, NH  2); MS (APCI) m/z 804 (MHþ). The above tert-butyl ester (100 mg, 0.12 mmol) was dissolved in trifluoroacetic acid (5 mL), and the solution was stirred for 30 min at room temperature. The solution was then poured into water and extracted with CH2Cl2. The extract was washed with brine, dried over anhydrous Na2SO4, filtered, and concentrated. The residue was washed with hexane to give the titled carboxylic acid chlorin-3 (86 mg, 92%) as a black solid: mp 240-242 C; vis (THF) λmax 682 (relative intensity, 56%), 621 (8), 543 (10), 513 (13), 418 nm (100); 1H NMR (1% pyridine-d5/CDCl3) δ 9.47 (1H, s, 10-H), 9.40 (1H, s, 5-H), 9.10 (1H, d, J=16 Hz, 3-CH), 8.66 (1H, s, 20-H), 7.09 (1H, d, J=16 Hz, 31-CH), 5.30, 5.14 (each 1H, d, J= 19 Hz, 131-CH2), 4.54 (1H, q, J=7 Hz, 18-H), 4.34 (1H, br-d, J= 9 Hz, 17-H), 3.66 (3H, s, 12-CH3), 3.64 (2H, q, J=8 Hz, 8-CH2), 3.50 (3H, s, 2-CH3), 3.20 (3H, s, 7-CH3), 2.72, 2.34 (each 1H, m, DOI: 10.1021/la1005715

6321

Article 17-CH2), 2.56, 2.28 (each 1H, m, 171-CH2), 3.99, 3.94 (each 1H, m, 172-COOCH2), 1.46 (2H, m, 172-COOCCH2), 1.23 (2H, m, 172-COOC2CH2), 1.17 (16H, m, 172-COOC3(CH2)8), 1.84 (3H, d, J=7 Hz, 18-CH3), 1.67 (3H, t, J=8 Hz, 81-CH3), 0.84 (3H, t, J= 7 Hz, 172-COOC11CH3), -1.90 (1H, s, NH) [another NH was not observed]; IR (KBr) ν 1732, 1697, 1618 cm-1; MS (TOF) m/z 747 (MHþ); HRMS (FAB) found, m/z 747.4476, calcd for C46H59N4O5, MHþ, 747.4485.

Hexyl trans-32-Carboxypyropheophorbide-a (Chlorin-2).

Black solid: mp 230-233 C; vis (THF) λmax 682 (relative intensity, 58%), 621 (8), 544 (11), 514 (13), 419 nm (100); 1H NMR (3% pyridine-d5/CDCl3) δ 9.46 (1H, s, 10-H), 9.42 (1H, s, 5-H), 9.08 (1H, d, J=16 Hz, 3-CH), 8.63 (1H, s, 20-H), 7.09 (1H, d, J=16 Hz, 31-CH), 5.26, 5.10 (each 1H, d, J=19 Hz, 131-CH2), 4.49 (1H, q, J = 7 Hz, 18-H), 4.30 (1H, br-d, J = 9 Hz, 17-H), 3.63 (3H, s, 12-CH3), 3.61 (2H, q, J=8 Hz, 8-CH2), 3.46 (3H, s, 2-CH3), 3.19 (3H, s, 7-CH3), 2.67, 2.30 (each 1H, m, 17-CH2), 2.52, 2.23 (each 1H, m, 171-CH2), 3.95, 3.89 (each 1H, m, 172-COOCH2), 1.41 (2H, m, 172-COOCCH2), 1.18-1.23 (6H, m, 172-COOC2(CH2)3), 1.80 (3H, d, J=7 Hz, 18-CH3), 1.64 (3H, t, J=8 Hz, 81-CH3), 0.77 (3H, t, J=7 Hz, 172-COOC5CH3), 0.13 (1H, br, NH), -1.91 (1H, s, NH); IR (KBr) ν 1732, 1697, 1616 cm-1; MS (TOF) m/z 662 (Mþ); HRMS (FAB) found, m/z 663.3555, calcd for C40H47N4O5, MHþ, 663.3546.

2-Butyloctyl trans-32-Carboxypyropheophorbide-a (Chlorin-4).

Black solid: mp 235-238 C; VIS (THF) λmax 682 (relative intensity, 56%), 621 (8), 544 (11), 513 (13), 418 nm (100); 1H NMR (3% pyridine-d5/CDCl3) δ 9.45 (1H, s, 10-H), 9.40 (1H, s, 5-H), 9.08 (1H, d, J=16 Hz, 3-CH), 8.62 (1H, s, 20-H), 7.08 (1H, d, J=16 Hz, 31-CH), 5.25, 5.10 (each 1H, d, J=19 Hz, 131-CH2), 4.49 (1H, br-q, J=7 Hz, 18-H), 4.30 (1H, br-d, J=8 Hz, 17-H), 3.88, 3.82 (each 1H, m, 172-COOCH2), 3.62 (5H, br, 12-CH3, 8-CH2), 3.45 (3H, s, 2-CH3), 3.17 (3H, s, 7-CH3), 2.67, 2.30 (each 1H, m, 17-CH2), 2.53, 2.25 (each 1H, m, 171-CH2), 1.79 (3H, d, J= 7 Hz, 18-CH3), 1.63 (3H, t, J=7 Hz, 81-CH3), 1.14 (1H, br-m, 172COOCCH), 1.10-1.16 (16H, m, 172-COOC2[(CH2)5][(CH2)3]), 0.76, 0.75 (each 3H, t, J=7 Hz, 172-COOC2(C3CH3)(C5CH3)), 0.11 (1H, br, NH), -1.93 (1H, s, NH); IR (KBr) ν 1732, 1697, 1618 cm-1; MS (TOF) m/z 746 (Mþ); HRMS (FAB) found, m/z 747.4438, calcd for C46H59N4O5, MHþ, 747.4485.

Cholesteryl trans-32-Carboxypyropheophorbide-a (Chlorin-5).

Black solid: mp 205-208 C; VIS (THF) λmax 682 (relative intensity, 57%), 621 (8), 544 (11), 514 (13), 419 nm (100); 1H NMR (2% pyridine-d5/CDCl3) δ 9.49 (1H, s, 10-H), 9.43 (1H, s, 5-H), 9.11 (1H, d, J=16 Hz, 3-CH), 8.66 (1H, s, 20-H), 7.10 (1H, d, J=16 Hz, 31-CH), 5.30, 5.14 (each 1H, d, J=19 Hz, 131-CH2), 5.27 (1H, m, CdCH in cholesteryl unit), 4.54 (1H, dq, J=2, 7 Hz, 18-H), 4.49 (1H, m, 172-COOCH), 4.34 (1H, dt, J=8, 2 Hz, 17-H), 3.66 (3H, s, 12-CH3), 3.65 (2H, q, J=8 Hz, 8-CH2), 3.50 (3H, s, 2-CH3), 3.22 (3H, s, 7-CH3), 2.70, 2.32 (each 1H, m, 17-CH2), 2.51, 2.25 (each 1H, m, 171-CH2), 2.18-0.14 (46H, m, cholesteryl unit), 1.84 (3H, d, J=7 Hz, 18-CH3), 1.67 (3H, t, J=8 Hz, 81CH3), -1.89 (1H, s, NH) [another NH was not observed]; IR (KBr) ν 1732, 1697, 1616 cm-1; MS (TOF) m/z 948 (Mþ); HRMS (FAB) found, m/z 947.6056, calcd for C61H79N4O5, MHþ, 947.6050. trans-32-Carboxypyropheophorbide-a (chlorin-R3) was prepared by hydrolysis of 172-methyl ester of chlorin-1 in concentrated HCl.9 Black solid: mp >300 C; VIS (THF) λmax 682 (relative intensity, 58%), 622 (8), 544 (11), 514 (13), 419 nm (100); 1 H NMR (10% pyridine-d5/CDCl3) δ 9.52 (1H, s, 10-H), 9.43 (1H, s, 5-H), 9.16 (1H, d, J=16 Hz, 3-CH), 8.73 (1H, s, 20-H), 7.16 (1H, d, J = 16 Hz, 31-CH), 5.44, 5.22 (each 1H, d, J = 19 Hz, 131-CH2), 4.64 (1H, q, J=7 Hz, 18-H), 4.46 (1H, br-d, J=10 Hz, 17-H), 3.68 (3H, s, 12-CH3), 3.66 (2H, q, J=8 Hz, 8-CH2), 3.47 (3H, s, 2-CH3), 3.20 (3H, s, 7-CH3), 2.89, 2.53 (each 1H, m, 17-CH2), 2.78, 2.39 (each 1H, m, 171-CH2), 1.92 (3H, d, J=7 Hz, (9) Tamiaki, H.; Miyata, S.; Kureishi, Y.; Tanikaga, R. Tetrahedron 1996, 52, 12421.

6322 DOI: 10.1021/la1005715

Wang et al. 18-CH3), 1.71 (3H, t, J=8 Hz, 81-CH3), 0.17 (1H, br, NH), -1.89 (1H, s, NH); IR (KBr) ν 1695, 1618 cm-1; MS (APCI) m/z 579 (MHþ); HRMS (FAB) found, m/z 579.2580, calcd for C34H34N4O5, MHþ, 579.2607. DFT and TD-DFT Calculations. All systems were optimized by the use of DFT calculation with the CAM-B3LYP10-12 exchange-correlations functional and the 6-31G(d,p)13 basis set, and with solvation effect for the free base sensitizer described by CPCM (ethanol).14 The TD-DFT15 with the CAM-B3LYP exchange-correlations functional and the 6-31G(d,p) basis set calculations are based on the optimized structure with solvation effect described by CPCM (ethanol) for the set of sensitizers. All calculations were done by Gaussian 0916 using the Research Center for Computational Science, Okazaki, Japan. Cyclic Voltammetry. Cyclic voltammetry was carried out by using a potentiostat (HA1010 mM1A, Hokuto Denko, Japan). The working electrode and counter-electrode consisted of platinum wire (0.5 mm in diameter), and the reference electrode was a Ag/AgCl electrode. The scan rate was 100 mV/s. The supporting electrolyte, tetrabutylammonium hexafluorophosphate (TBAHFP, polarographic grade, Sigma), was used without further purification. The concentration of TBAHFP was ∼0.1 M, and those of dyes were in the region 10-3-10-4 M in CH2Cl2. Electrical Impedance Spectroscopy (EIS). EIS data were obtained under a constant light illumination of 100 mW cm-2 provided by a halogen lamp under the open-circuit condition in an experimental setup consisting of an electrochemical interface (1287N, Solartron, U.K.) and impedance/gain-phase analyzer (1260N, Solartron, U.K.). The frequency range was 10-2-107 Hz.

Measurement of Surface Coverage (Γ) and Absorption Spectra. TiO2 films 0.8  0.8 cm in size, ∼10 μm in thickness, and

1.7 g 3 cm-3 in density were dipped into an ethanolic solution containing each dye sensitizer for 3 h and then washed with ethanol to remove free dye sensitizers on the surface. The adsorbed dye sensitizers were estimated when dissolved in 5 mL of 5 mM KOH aqueous solution. The absorption spectrum of the KOH solution of each sensitizer was measured. Thus, the Γ value could be obtained by a standard method as described in our previous paper.3f Absorption spectra of the sensitizers deposited on TiO2 films were measured with a Shimadzu UV-3150 spectrometer. The TiO2 films with a typical thickness of 4 μm were dipped into 0.3 mM ethanol solution of Chl dye for 3 h. The dye-adsorbed films were rinsed with ethanol three times and dried in air before measurement of the absorption spectra.

Fabrication of DSSC and Photovoltaic Measurements.

The optically transparent electrode with a 0.25 cm2 working area contains 18 and 400 nm TiO2 nanoparticles with a thickness of 10 and 4 μm obtained from Catalysts & Chemicals Ind. Co., for light-harvesting and light scattering, respectively. Details of the (10) Becke, A. D. J. Chem. Phys. 1993, 98, 5648. (11) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785. (12) Yanai, T.; Tew, D. P.; Handy, N. C. Chem. Phys. Lett. 2004, 393, 51. (13) Ditchfield, R.; Hehre, W. J.; Pople, J. A. J. Chem. Phys. 1971, 54, 724. (14) Miertus, S.; Scrocco, E.; Tomasi, J. Chem. Phys. 1981, 55, 117. (15) Eric Stratmann, R.; Scuseria, G. E.; Frisch, M. J. J. Chem. Phys. 1998, 109, 8218. (16) Gaussian 09, Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A.; Vreven, T., Jr.; Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian, Inc.: Wallingford CT, 2009.

Langmuir 2010, 26(9), 6320–6327

Wang et al. fabrication of DSSCs were described earlier.3a For addition of a coadsorbent, 1 mM chenodeoxycholic acid (CDCA) was mixed with a solution of 0.3 mM Chl dye in ethanol. The TiO2 films were dipped in this solution for 5 h for dye adsorption. Each DSSC used the counter electrode of Pt-sputtered FTO glass (Nippon Sheet Glass, 10 Ω cm-2) and the electrolyte containing 0.05 M LiI, 0.03 M I2, 1.0 M 1-propyl-3-methylimidazolium iodide, 0.5 M tert-butylpyridine, and 0.1 M guanidiniumthiocyanate (GNCS) in a mixture of acetonitrile-valeronitrile (85:15, v/v, this electrolyte was also used for time-resolved absorption spectroscopic measurements); in contrast, the reference N719 dye used an electrolyte containing 0.6 M 1-butyl-3-methylimidazolium iodide, 0.03 M I2, 0.1 M GNCS and 0.5 M 4-tert-butylpyridine in a mixture of acetonitrile-valeronitrile (85:15). The electron lifetimes (τ) were measured by the use of commercially available setup SLIM-PCV (PSL-100, EKO, Japan) with stepped laser beam (YL-331M, Yamamoto Photonics, Japan).

Subpicosecond Time-Resolved Absorption Spectroscopy. Transient absorption spectra were measured with a conventional pump-probe method. Briefly, the light source was a femtosecond regenerative amplifier system (Spectra-Physics, Spitfire, 1 kHz, 80 fs, 800 nm) seeded by a mode-locked Ti:sapphire oscillator (Spectra-Physics, Tsunami). The pump beam was the second harmonic generation 400 nm with energies of 0.2 μJ/pulse. The probe beam was the supercontinuum white light generated from a 5 mm thick sapphire plate in visible/near IR region. The signal and reference beams were acquired by using a chopper in the pump light path with 500 Hz repetition rate. The transmitted probe beam was collected by an optical fiber and dispersed by a grating monochromator (Acton Research, Spectra Pro 275/2150i). Finally, the signal was acquired by a liquid nitrogen cooled CCD detector (Princeton Instruments, Spec-10-100LN) or an InGaAs detector (OMA V), and the data were recorded by a personal computer. The samples were contained in a 2 mm cuvette and the experiments were performed without flowing. The dye-sensitized TiO2 slides were prepared by immersion in dye solutions with the absorbance of 1.8 at the pump laser wavelength.

Results and Discussion Design of Hydrocarbon Functionalized Sensitizers and Their Orientation on TiO2 Films. Figure 1 shows the chemical structures of chlorophyllous sensitizers chlorins-1-5, together with three reference sensitizers chlorins-R1-R3. Figure S6 in the Supporting Information shows the synthetic routes for each sensitizer starting from a chlorophyll a-enriched photosynthetic organism, Spirulina geitleri. Here, four major alterations have been made on the common precursor Chl-a molecule to produce the set of major sensitizers, chlorins-1-5. (1) A carboxy group was attached to the C32 position for binding to a semiconductor surface. (2) The unstable Mg in the central metal was removed from the chlorin macrocycles. (3) The C132-COOCH3 group was replaced by hydrogen due to the avoidance of undesired reaction at the C132 position in the β-keto-ester. (4) The reactive phytyl ester was replaced by various stable hydrocarbon ester groups. More specifically, chlorins-2 and -3 have straight alkyl chains with different lengths (C6 and C12), while chlorin-4 has a branched alkyl chain. On the other hand, chlorin-5 possesses a bulky cholesteryl unit. It has been found by our previous studies that addition of cholic acid derivatives to Chl sensitizers can enhance the performance of DSSCs by reducing the aggregate formation of the dye sensitizers.3a,3g The molecular design of the present chlorin-5 covalently bonded with a cholesteryl unit was based on these findings. As shown in Figure S6 (Supporting Information), the synthetic route of chlorins-1-5 mainly involved in transesterification of methyl pyropheophorbide-d6 with the corresponding alcohol using the tin catalyst,7 followed by the Wittig reaction of Langmuir 2010, 26(9), 6320–6327

Article

the 3-formyl group with Ph3PdCHCOOC(CH3)3, and acidic cleavage of the resulting tert-butyl ester.3g The molecular structures of chlorins-1-5 commonly consist of two active sites that are potentially able to bind with TiO2, i.e., the free COOH group at the C32 position and the esterified COOR group at the C172 position. In the case where these dyes only bind to TiO2 at the C32 position, the different hydrocarbon esters must likely face to electrolyte and this makes our study on the mechanism relatively simpler. In contrast, if these dyes can bind to TiO2 at both the C32 and C172 positions, the mechanism of different esterification effect at the C172 position can be extremely complicated. In order to clarify this issue, we tried to compare chlorin-1 with three reference sensitizers, chlorins-R1-R3 (see Figure 1 for their chemical structures). Chlorin-R1 is similar to chlorin-1, but without a carboxy group at the C32 position. Chlorin-R2 is the free acid form of methyl ester chlorin-R1. Chlorin-R3 has two carboxy groups at both the C32 and C172 positions. In Figure 2a, we measured the absorption spectra of the three reference sensitizers and chlorin-1 deposited on transparent TiO2 films. Figure 2b shows the incident photon-to-current conversion efficiency (IPCE) profiles of solar cells based on these dye-sensitized TiO2 films. Chlorin-R1 has a negligible absorption on TiO2, reflecting the extremely weak binding ability of the C172 ester group, although this kind of binding cannot be completely ruled out. Indeed, the soaking time of TiO2 films in dye solution is merely 3-5 h, and this may be too short for active oxygens of the C172 ester and C13 ketone to react with TiO2, even though there are still possibilities. The absorbances of chlorins-R2 and -R3 were stronger than that of chlorin-R1, yet still weaker than that of chlorin-1. The difference in the absorption spectra affected the light-harvesting capability of each DSSC, and finally caused the difference in the IPCE profiles. The lower electronic combination of either chlorin-R2 or -R3 with TiO2 through the C172 carboxylate group is another reason for the low photocurrent generation. In this article, we will not give any more details of studying chlorins-R1-R3 in DSSC, due to their lower efficiencies. Static Evaluations on the Light-Harvesting, ElectronInjection, and Charge Collections. It has earlier been reported that different C17-propionate residues do not alter the absorption spectra of chlorin pigments in their monomeric states.7c In Figure S7 in the Supporting Information, we measured the absorption spectra of this set of sensitizers deposited on transparent TiO2 films with a thickness of 4 μm. A typical dipping time of 3 h was applied for all these sensitizers. The terminal C172 ester group contains no auxochromes, so that almost the same absorption spectrum was observed for each sensitizer. These absorption spectra indicate that chlorins-1-5 possess a similar light-harvesting efficiency (LHE) when used in DSSCs. Figure S8 (Supporting Information) depicts the molecular orbitals of chlorins-1-5 including HOMO-1, HOMO, LUMO, and LUMOþ1 that were obtained from DFT calculations. The electron densities at the position of carboxy groups of these sensitizers are practically the same. Therefore, a similar combination between the dye sensitizer and TiO2 electrode can be expected. Also, the energy levels of these four major molecular orbitals and the molecular Fermi level (MFL)17 of chlorins-1-5 were unchanged for each sensitizer (see Table 1). The high similarity of these calculated energy levels of HOMO and LUMO orbitals was also supported by the experimental results of the one electron-oxidation potentials (Eox) and the absorption spectra of these sensitizers, which were almost equal. (17) Kitao, O. J. Phys. Chem. C 2007, 111, 15889.

DOI: 10.1021/la1005715

6323

Article

Wang et al.

Figure 3. (a) IPCE profiles and (b) I-V curves of DSSCs based on chlorins-1-5 and N719. Table 2. Monolayer Surface Coverage (Γ) of Chlorins-1-5 on the TiO2 Films and Photovoltaic Performance of DSSCs Using Chlorins-1-5 and Ru Complex N719

Figure 2. (a) Absorption spectra of chlorins-1 and -R1-R3 on TiO2 films and (b) their IPCE profiles of DSSCs.

Eox/V sensitizer HOMO-1 HOMO MFL LUMO LUMOþ1 vs NHE -6.45 -6.45 -6.45 -6.45 -6.44

-6.14 -6.14 -6.14 -6.14 -6.13

-3.98 -3.98 -3.98 -3.98 -3.97

-1.92 -1.92 -1.92 -1.93 -1.91

-1.10 -1.10 -1.10 -1.10 -1.09

1.13 1.13 1.13 1.13 1.12

The photocurrent of a solar cell can be calculated by integrating the IPCE in the full absorption region, against the standard solar irradiation spectrum. IPCE was determined by the following equation, IPCE ¼ APCE  LHE ¼ Φinj  ηcol  LHE,

ð1Þ

where APCE is absorbed photon-to-current conversion efficiency that should be divided into two terms, i.e., the overall electron injection efficiency (Φinj) between dye sensitizer and semiconductor electrode and the overall charge collective efficiency over solar cell (ηcol). The LHEs for all the sensitizers were similar, as supported by their absorption spectra in Figure S7 (Supporting Information). The Φinj values for this set of sensitizers should also be similar, because either the energy levels of MFL or the electron densities of the carboxylate group of LUMO and LUMOþ1 orbitals for the these sensitizers are the same. Thus, any difference in the photocurrent among DSSCs based on these sensitizers could be attributed to the difference in the forward and backward charge transfer to cause different ηcol. Here, difference in the forward charge transfer can be measured by the use of EIS, while the backward charge transfer, especially 6324 DOI: 10.1021/la1005715

Jsc/ mA cm-2

Voc/V

FF

η/%

-8

Table 1. Major Energy Levels (in eV) of Chlorins-1-5 Calculated Based on TD-DFT with CAM-B3LYP/6-31G(d,p) with CPCM (Ethanol), and Their One Electron-Oxidation Potentials (Eox) Measured Electrochemically

chlorin-1 chlorin-2 chlorin-3 chlorin-4 chlorin-5

sensitizer

Γa / mol cm-2

15.0 ( 0.3 0.60 ( 0.03 0.72 ( 0.05 6.5 ( 0.1 chlorin-1 3.6  10 chlorin-2 4.0  10-8 15.5 ( 0.6 0.62 ( 0.05 0.73 ( 0.02 7.0 ( 0.1 -8 17.4 ( 1.1 0.64 ( 0.02 0.72 ( 0.03 8.0 ( 0.2 chlorin-3 3.7  10 chlorin-4 4.0  10-8 13.9 ( 1.5 0.65 ( 0.05 0.74 ( 0.02 6.6 ( 0.3 chlorin-5 4.2  10-8 13.6 ( 0.2 0.64 ( 0.03 0.70 ( 0.05 6.1 ( 0.1 N719 17.9 ( 1.2 0.77 ( 0.04 0.68 ( 0.04 9.3 ( 0.3 a This data was obtained by desorption of dye sensitizers from TiO2 surface by KOH solution. See Experimental Section for the details.

between TiO2 and electrolyte, should be evidenced by measurements of τ values in the solar cells. From these viewpoints, the bulky cholesteryl ester group in chlorin-5 may cause a serious shielding effect to prevent the useful contact of chromophore with TiO2 and electrolyte to reduce both forward and backward charge transfer. Figure 3a shows the IPCE profiles and Figure 3b shows the photocurrent-photovoltage (I-V) curves of DSSCs based on each Chl sensitizer. Table 2 lists the surface coverage (Γ) of TiO2 electrodes by the set of sensitizers, and relevant parameters of DSSCs that were obtained from the I-V curves. To evaluate our results, a solar cell fabricated using a well-known ruthenium complex, cis-bis(isothiocyanato)bis(2,20 -bipyridyl-4,40 -dicarboxylato)ruthenium(II) bis(tetrabutylammonium) (N719) was used as the standard cell. The Γ values for chlorins-1-5 were almost identical, suggesting that almost the same amount of dye molecules was adsorbed on the TiO2 films in each Chl sensitizer. In the IPCE profiles, chlorin-3 with the longest alkyl ester group gives rise to the highest values, but with slightly smaller values in the shoulder at the longer wavelength region of the redmost band. The other sensitizers give rise to similar IPCE values in the major absorption region. Compared to the IPCE profile of DSSC based on N719 dye, the IPCE values of chlorin-3 sensitized solar cell were larger in both shorter and longer wavelength regions, but still smaller at the absorption maxima of N719 dye. In Figure 3b and Table 2, DSSC based on chlorin-3 gives the highest short-circuit current (Jsc) and η values, while the open-circuit photovoltage (Voc) and Langmuir 2010, 26(9), 6320–6327

Wang et al.

Article

Figure 4. EIS Nyquist plots for DSSCs based on chlorins-1-5 under AM1.5 illumination.

fill factors (FF) were similar to other Chl sensitizers, except for the one based on chlorin-1 that gave lower Voc value. Chlorin-3 sensitized solar cell gives a η value of 8%, which is the highest value not only among the present Chl sensitizers, but also among all the Chl and Por sensitizers published to date. The high Jsc value was at the top level of organic sensitizer based DSSCs. The η values for DSSCs based on the other Chl sensitizers also exceed 6%, and in a ranking order: longer linear alkyl ester (chlorin-3)>shorter linear alkyl ester (chlorin-2)>branched alkyl ester (chlorin-4)> small methyl ester (chlorin-1)>bulky cholesteryl ester (chlorin-5). The best DSSC based on chlorin-3 still exhibits lower photovoltaic performance compared to that based on the standard N719 dye, due to the relatively lower Voc value, when the optimized electrolyte for both sensitizers were used. Future studies on the semiconductor structure and electrolyte may lead to higher performance of DSSCs with the present chlorin sensitizers. To understand the difference in the Jsc values, we applied the EIS to study the electron flow in the solar cells. Figure 4 shows EIS Nyquist plots for DSSCs based on the set of sensitizers under standard 100 mW cm-2 illumination. The data have been recorded under an open-circuit condition. According to previous studies,18,19 the larger semicircle at the left-hand side and the smaller semicircle at the right-hand side correspond to the electron transfer processes in TiO2-dye-electrolyte interface (Ret) and in electrolyte, respectively. Another semicircle, which corresponds to the electron transfer at the Pt electrode and should appear at the high-frequency region, was overlapped by a larger semicircle in the midfrequency region. A smaller Ret should reflect a very smooth electron transfer in the TiO2-dye-electrolyte interface, and should lead to a larger ηcol, and vice versa. The increase in the alkyl group length at the C17-propionate residue leads to the smaller Ret value, except for the case in chlorin-4. Compared to Ret values of isomeric chlorins-3 and chlorin-4 possessing C12H25 ester group, the straight dodecyl chain in chlorin-3 induced a smaller value than the branched chain in chlorin-4. This steric factor would affect the interaction of the dye molecules with TiO2 surface and electrolyte. Moreover, the bulky cholesteryl ester group in chlorin-5 giving the largest Ret value caused a serious steric shielding effect to prevent the useful contact of chromophore with TiO2 and electrolyte to reduce interfacial charge collection toward TiO2. Thus, both reversed similar ranking in the Jsc values, i.e., chlorin-3 > chlorin-2 > (18) Longo, C.; Nogueira, A. F.; De Paoli, M.-A.; Cachet, H. J. Phys. Chem. B 2002, 106, 5925. (19) Kern, R.; Sastrawan, R.; Ferber, J.; Stangl, R.; Luther, J. Electrochim. Acta 2002, 47, 4213.

Langmuir 2010, 26(9), 6320–6327

Figure 5. Subpicosecond time-resolved absorption spectra of chlorin-1.

chlorin-1 and chlorin-3>chlorin-4>chlorin-5 can be seen in the Ret values. Evaluation of Electron Injection and Charge Collection with Subpicosecond Transient Absorption Spectroscopy. In previous section, we have concluded that the Φinj values are the same for the set of sensitizers, based on the electron densities and energy levels of the LUMO and HOMO orbitals of the sensitizers, and the ηcol values are different for these sensitizers based on the measurements of the EIS. In order to further clarify these issues, we applied subpicosecond time-resolved absorption spectroscopy (TAS) to study charge transfer in the TiO2-dye interface. Figure 5 shows the TAS spectra of chlorin-1 in ethanol solution, and after its adsorption on a TiO2 nanocrystalline film with and without an electrolyte. After excitation of the dye at 400 nm in ethanol, a broad absorption signal in 450-650 nm region and a bleaching signal at 650-720 nm were appeared, and both of them increase in intensity until 100 ps. The broad absorption could be asighed to internal conversion from Soret f Qx f Qy, while the bleaching signal is due to the vacancy of electrons in the ground state. These two TAS paterns can persist for a long period of up to several nanoseconds (data was not shown), indicating that the excited state electrons of the free sensitizer has a long lifetime in solution. When chlorin-1 molecules were deposited on a TiO2 thin film, two distinguishable spectral changes in TAS can be described as follows: (1) an absorption peak newly generated at 610 nm immediately after excitation; (2) the decay time of the broad absorption and the recovery time of the bleaching singnal become much shorter. During the experiment, we also found that after the TAS measurement, the dye sensitizers have some degree of degradation. To protect the dye-sensitizers during the measurment, we introduced an electrolyte consisting I-/I3- redox couple as detailed in Experimental Section. A comparison of the spectral changes after TAS measurments with and without the electrolyte has been presented in Figure S11 (Supporting Information). Apart from the above effect of protecting dye sensitizer by electrolyte, it is important to know whether this will also affect the excited state dynamics in TAS. Figure S12 (Supporting Information) DOI: 10.1021/la1005715

6325

Article

Wang et al.

Figure 6. Subpicosecond time-resolved absorption spectra of TiO2 films sensitized with chlorins-1, -3, and -5 with different amount of dye molecules. An electrolyte was used for the dye protection.

shows the kinetic decay trace at 610 nm and bleach recovery kinetics at 700 nm of the chlorin-1 sensitized TiO2 thin film with and without the electrolyte. Here, each decay or recovery dynamics could be exponentially fitted with two time constances. When the electrolyte was introduced, the two decay time constants at 610 nm become larger, while the two recovery time constants at 700 nm keep almost unchanged. In previous studies on porphyrins,20,21 the 610 nm singal has been assigned to the charge separated state of the dye-TiO2 complex, and the decay of this singal was due to the charge-recombination process. Our results have good agreement with this assignment, since the change of the time constants can be from the shift of the conduction band edge (CBE) upon addition of electrolyte, and this can cause a reduced charge-recombination efficiency between TiO2•- and dye•þ. On the other hand, the bleaching signal at 700 nm can be more complicated. The recovery of this signal is due to the de-excitation of dye molecules that should mainly be attributed to charge-recombination,22 exciton annihilation,23 and exciton migration.24 The fact that the presence of the electrolyte has negligible effect on the dynamic signal at 700 nm may suggest that this signal was less affected by the charge recombination process. To understand deeper about the meaning of these TAS signals at 610 and 700 nm, we measured the TAS spectra in Figure 6 for chlorins-1, -3, and -5 bound on TiO2 both with high and low surface coverage (see Figure S13 (Supporting Information) for the details). We also ploted the excitation dynamics of the three (20) Ramakrishna, G.; Verma, S.; Amilan Jose, D.; Krishna Kumar, D.; Das, A.; Palit, D. K.; Ghosh, H. N. J. Phys. Chem. B 2006, 110, 9012. (21) Chang, C.-W.; Luo, L.; Chou, C.-K.; Lo, C.-F.; Lin, C.-Y.; Hung, C.-S.; Lee, Y.-P.; Diau, E. W.-G. J. Phys. Chem. C 2009, 113, 11524. (22) Tachibana, Y.; Haque, S. A.; Mercer, I. P.; Durrant, J. R.; Klug, D. R. J. Phys. Chem. B 2000, 104, 1198. (23) Wang, X.-F.; Koyama, Y.; Nagae, H.; Yamano, Y.; Ito, M.; Wada, Y. Chem. Phys. Lett. 2006, 420, 309. (24) Wang, X.-F.; Fujii, R.; Ito, S.; Koyama, Y.; Yamano, Y.; Ito, M.; Kitamura, T.; Yanagida, S. Chem. Phys. Lett. 2005, 416, 1.

6326 DOI: 10.1021/la1005715

dyes with high and low surface coverage at both 610 and 700 nm in Figure S14 (Supporting Information). At 610 nm, all three dyes gave two similar decay time constants at ∼4 ps and 100-200 ps, both at high and low surface coverage. Since the 610 nm signal comes from the binding state of dye with TiO2, it is therefore free from the concentration effect of dye sensitizers. The high similarity of the two time constants strongly support this assignment. In contrast, the recovery dynamics at 700 nm shows clear difference in the three dye sensitizers. For chlorin-1 having the small C172 ester group and chlorin-5 having the most bulky C172 ester group, increasing the surface coverage causes shorter recovery time constants, i.e., 8.0 f 2.6 ps and 350 f 127 ps for chlorin-1, and 6.2 f 2.3 ps and 318 f 200 ps for chlorin-5. In the case of chlorin-3 having a linear alkyl C174 substituent, the recovery dynamics become more similar at different surface coverage, i.e., 2.2 f 2.4 ps and 280 f 147 ps. Especially, the concentration of dye has larger influence on the short time constants. As we discussed above, the recovery dynamics at 700 nm is affected by the rate of charge-recombination, the exciton annihilation, and the exciton migration. The higher dye concentration would mainly affect the later two processes. The larger dependence of the time constants on the dye concentration can be attributed to a more serious quenching of the excitation. Since the solar cell performance of chlorins-1 and -5 is lower than that of chloirin-3, the above serious quenching effect of excitation in the former two sensitizers is expected to be a major reason to limit the performance. Moreover, each solar cell has been fabricated at a high surface coverage with dye sensitizer to reach a high LHE, the excitation dynamics at that condition is more important to understand the mechanism. At high surface coverage, the second of each two time constands is the smallest in chlorin-1 with the smallest C172 ester group, indicating that the charge-recombinating should also be most serious in this dye. Since both chargerecombination and exciton annihilation can be regarded as a part of factors to determine the charge collective efficiency of a solar cell, the high charge collective efficiency in chlorin-3 sensitized Langmuir 2010, 26(9), 6320–6327

Wang et al.

Article

Figure 7. Dependency of τ on Jsc in DSSCs based on chlorins-1, -3, and -5 and Ru complex N719.

solar cell therefore corresponds to the high overall conversion efficiency in solar cell. The above observation in the TAS spectra has good agreement with our observation in the EIS spectra, in which chlorin-3 sensitized solar cell gave smallest impedance value in the TiO2-dye interface. It has been found that by introduction of hexyl chains to a trithiophene-based dye sensitizer, the τ of the DSSC was elongated to give a larger Voc value.25 The larger Voc values of chlorins-2-5 compared to chlorin-1 could be understood by the same principle. In Figure 7, we measured the τ values for DSSCs based on chlorins-1, -3, and -5 and N719 using SLIM-PCV with stepped laser beam. Compared to N719 sensitized solar cell, DSSC based on chlorin-1, which has the smallest methyl ester group in dimensions gives clearly lower τ value throughout the entire photocurrent region. In contrast to chlorin-1, chlorins-3 and -5 give longer τ values than N719 in the small photocurrent region, or in other words, under lower light intensity illumination. However, when the light intensities were increased, the τ values for chlorins-3 and -5 became smaller than that for N719. The decrease of τ value in chlorin-5 was especially rapid with increasing light intensities, among the three Chl sensitizers. The higher τ values in the cases of chlorins-3 and -5 compared to chlorin-1 can be explained by the bulky hydrocarbon ester group in the former that could partially avoid the direct contact of electrolyte with TiO2 electrode to reduce the generation of dark current, and this can lead to higher Voc values. When the light intensity or the Jsc value was increased, the τ value decreased more slowly in chlorin-3 than in chlorin-5, suggesting a more stable charge separation ability in the former under different magnitite of light illumination.

Conclusion A set of Chl-a derivatives chlorins-1-5 with different ester groups in the C17-propionate residue was synthesized and (25) Koumura, N.; Wang, Z.-S.; Mori, S.; Miyashita, M.; Suzuki, E.; Hara, K. J. Am. Chem. Soc. 2006, 128, 14256.

Langmuir 2010, 26(9), 6320–6327

compared as sensitizers of DSSCs. The Chl sensitizers have similar absorption spectra and almost equal surface coverage on TiO2 electrodes. DFT and TD-DFT calculations suggest these sensitizers possess the same electron injection efficiency when they are bound to the TiO2 surface. The different Jsc values were attributed to the difference in forward charge transfer efficiency supported by different impedance values at the TiO2-dyeelectrolyte interface. Subpicosecond time-resolved absorption spectroscopy on the dye-TiO2 systems suggests that the set of dye sensitizers may processes similar electron injection efficiency, but different charge collective efficiency due to the changes in the charge recombination and exciton annihilation. The larger Voc values of DSSCs based on Chl sensitizers with hydrocarbon ester groups were attributed to the lengthened τ values by suppressed backward charge recombination between TiO2 and electrolyte. The chlorin-3 sensitized solar cell gives the highest performance, i.e., Jsc = 17.4 mA cm-2, Voc=0.64 V, FF=0.72, and η=8.0%, under standard AM 1.5 (100 mW cm-2) sunlight illumination, which is the highest performance among the present and reported DSSCs based on cyclic tetrapyrrole type sensitizers. Finally, the durability of DSSCs based on the set of chlorin sensitizers has not been studied in the present manuscript. However, we did some very preliminary test of the durability against the high temperature (60 C) in our previous study on similar compounds, and we found that durability of chlorin dyes was better than that of organic dye D149, but similar to that of N719.3h The introduction of solid-state or quasi-solid-state electrolyte into the present system would be able to substantially improve the durability of DSSCs. These questions should be answered in future investigations. Acknowledgment. The calculations were done at the Research Center for Computational Science, Okazaki. We thank Dr. Tomohiro Miyatake of Ryukoku University for measurements of HRMS. This work was partially supported by Grantsin-Aid for Young Scientists (B) (No. 20760606 and 21750154) (to X.-F.W and S.S) from the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of the Japanese Government and for Scientific Research (B) (No. 19350088) (to H.T) from the Japan Society for the Promotion of Science (JSPS). Supporting Information Available: Text giving experimental details and figures showing 1H NMR spectra of chlorins2-5 and -R3, synthetic route of the set of sensitizers, EIS spectrum of N719-DSSC, electronic absorption spectra of chlorin-1-TiO2, DFT, cyclic voltammetry of chlorin-1, and fitting of the TAS results, and tables of TD-DFT calculation data of chlorins-1-5. This material is available free of charge via the Internet at http://pubs.acs.org.

DOI: 10.1021/la1005715

6327