Synthesis and electron-transfer properties of benzimidazole-functionalized ruthenium complexes for highly efficient dye-sensitized solar cells

(1)

8992 Chem. Commun., 2010, 46, 8992–8994 This journal is c The Royal Society of Chemistry 2010

Synthesis and electron-transfer properties of benzimidazole-functionalized

ruthenium complexes for highly eﬃcient dye-sensitized solar cellsw

Wei-Kai Huang,

a

Chi-Wen Cheng,

a

Shu-Mei Chang,

b

Yuan-Pern Lee

a

and Eric

Wei-Guang Diau*

ac

Received 27th August 2010, Accepted 4th October 2010 DOI: 10.1039/c0cc03517c

Novel heteroleptic ruthenium complexes—RD1, RD5, RD10 and RD11—with ligands based on benzimidazole were synthesized and characterized for application to dye-sensitized solar cells (DSSC); the remarkable performance of RD5-based DSSC is understood for its superior light-harvesting ability and slower charge-recombination kinetics.

Dye-sensitized solar cells (DSSC) have attracted attention because of their efficient performance, ease of fabrication and economy of production.1_{Much effort has been devoted} to the synthesis and characterization of various sensitizers for DSSC, such as ruthenium complexes,2 zinc porphyrins3and metal-free organic dyes.4 The devices made of ruthenium polypyridyl complexes, such as N3 and N719 dyes, attained an efficiency B11% of power conversion under one-sun illumination.5 To improve the efficiency of light harvesting, heteroleptic ligands based on thiophene were designed to make ruthenium complexes with large absorption coefficients, thus enhancing the efficiency of the device to 11.5%.6The synthesis of those thiophene-based ligands was, however, more elaborate than for the commercial N719 dye. Because their electron mobilities are great, benzimidazole derivatives have been developed as layer materials for electron transport and hole blocking in organic light-emitting diode devices.7 Bearing a similar idea in mind, we designed heteroleptic ruthenium complexes containing benzimidazole substituents in a series for which the corresponding ligands can be synthesized in a simple two-step procedure.

Here we report four novel heteroleptic ruthenium complexes, RD1 [Ru(dcbpy)(1-methyl-2-(pyridine-2-yl)benzoimidazole)-(NCS)2], RD5 [Ru(dcbpy)(1-benzyl-2-(pyridine-2-yl)benzo-imidazole)(NCS)2], RD10 [Ru(dcbpy)(1-decyl-2-(pyridine-2-yl)-benzoimidazole)(NCS)2] and RD11 [Ru(dcbpy)(1,10 -dimethyl-2,20_{-bibenzoimidazole)(NCS)}

2], for which one 4,40 -dicarboxylic-2,20_{-bipyridine (dcbpy) ligand in N3 dye was replaced by a}

heterocylic ligand (Chart 1). The DSSC device made of RD5 sensitized on the TiO2ﬁlm exhibits photovoltaic performance comparable with the speciﬁcations of a device made of N719

dye. The results obtained from photocurrent and photovoltage decays indicate that the electron lifetimes of the devices display a systematic trend RD5 > RD10 > RD1 > RD11, which is consistent with the cell performance in the same order. Femtosecond measurements of infrared transient absorption (TA) of the samples as thin ﬁlms indicate that, for RD-series sensitizers there exists a decay component on a ns time scale representing the back-electron transfer (BET) occurring at the interface between TiO2and dye, whereas for dye N719 such a process was not observable. Combining results from both electron-transfer kinetics reasonably explains the order of VOC to be N719 > RD5 > RD10 > RD1 > RD11. In contrast, the IPCE action spectra of the devices account for the order of JSC to be RD5 > N719 > RD10 > RD1 > RD11. The superior performance of RD5-based DSSC is thus understood for its superior light-harvesting ability and charge-recombination kinetics.

Fig. 1 shows the absorption spectra of RD1, RD5, RD10, RD11 and N719 in DMF; the corresponding spectral (Fig. S1w) and electrochemical (Fig. S2w) properties of these dyes are listed in Table S1.w The UV-visible absorption spectra of these complexes exhibit a band to the red of the corresponding band for dye N719. Even though the molar absorption coefficients (e/M1 cm1) of RD1 (7560 at 539 nm), RD5 (8005 at 537 nm), RD10 (7796 at 537 nm) and RD11 (6483 at 538 nm) are significantly smaller than that of N719 (13610 at 524 nm), the absorbances of the corresponding thin-film spectra on TiO2 films are slightly greater than that of N719 (inset of Fig. 1). This property reflects that the amounts of the RD series dyes adsorbed on TiO2 surface are signifi-cantly greater than that of N719 (Table 1).

Chart 1 Molecular structures of RD1, RD5, RD10 and RD11. a

Department of Applied Chemistry and Institute of Molecular Science, National Chiao Tung University, Hsinchu 300, Taiwan. E-mail: [email protected]; Fax: +886 3-572-3764; Tel: +886 3-513-1524

b_{Institute of Organic and Polymeric Materials,}

National Taipei University of Technology, Taipei 106, Taiwan c_{On sabbatical leave in Department of Chemistry,}

University of Copenhagen, DK-2100 Copenhagen, Denmark w Electronic supplementary information (ESI) available: Supplementary ﬁgures (Fig. S1–S3), tables (Tables S1–S2) and experimental details. See DOI: 10.1039/c0cc03517c

COMMUNICATION www.rsc.org/chemcomm | ChemComm

Published on 20 October 2010. Downloaded by National Chiao Tung University on 25/04/2014 06:46:50.

(2)

This journal is c The Royal Society of Chemistry 2010 Chem. Commun., 2010,46, 8992–8994 8993 Fig. 2a compares the current–voltage characteristics of these

four dyes on TiO2films of thickness 12 + 3 mm with that of N719 under the same conditions of device fabrication; the corresponding photovoltaic parameters are listed in Table 1. The values of VOC/V are 0.694 for RD1, 0.737 for RD5, 0.710 for RD10 and 0.685 for RD11, showing the effect of various substituents on the imidazole ring and the effect of the bibenzimidazolyl ligand. The values of JSC/mA cm2 are 12.744 for RD1, 15.084 for RD5, 13.612 for RD10 and 11.159 for RD11. This variation of JSCis inexplicable solely as the effect of dye loading. The trend of JSC is, however, consistent with the variation of IPCE shown in Fig. 2b. As mentioned for the absorption spectra, the IPCE action spectra of dyes in the RD series exhibit also a red-shifted spectral feature beyond 800 nm, in particular for RD5 showing greater efficiency and breadth than those properties of N719. The total efficiency of power conversion of RD5 attains 7.7%, which is comparable with that of N719 (Z = 7.8%). The RD5 device performed slightly better than the N719 device with the thinner TiO2films (Fig. S2w and Table S2w) because of the superior dye-loading effect of RD5.

Fig. 3a and b show the electron-transport kinetics of the corresponding devices obtaining from an analysis of the photovoltage (Fig. 3a) and photocurrent (Fig. 3b) decay data.8 The electron-diffusion coefficients (D) are similar for each dye, but the time coefficients for electron recombination (tR) display a systematic trend RD5 > RD10 > RD1 > RD11. This trend indicates that charge recombination between the dye (or I3) and the TiO2 surface might occur near the imidazolyl ligands. The existence of a hydrophobic chain in the imidazolyl ligand (RD10) longer than for dye RD1,

impedes the charge recombination, but the existence of the benzyl substituent in the imidazolyl ligand (RD5) might involve a hindrance to retard the charge recombination. In contrast, the system with a dimeric imidazolyl ligand (RD11) might increase the chance for charge recombination, so as to diminish the electron density in the conduction band of TiO2. As a result, both VOCand JSCshow the same order as we observed for tR.

Our results indicate that charge recombination plays a key role in cell performance. There are two charge recombinations: one is the reaction between the electrons in the conduction band of TiO2and I3in the electrolyte (electron interception), and the other is the reaction between the conduction-band electrons and the dye cations (BET).9 To investigate the kinetics of back transfer of electrons without involving an electrolyte, we measured the infrared TA kinetics10for each dye adsorbed on a thin-ﬁlm sample. The dye molecules were excited with a fs pulse at 625 nm, and the conduction-band electrons were probed with another, delayed fs pulse at 4.9 mm. The resulting TA signals were obtained on varying the delay between the visible pump pulse and the IR probe pulse. Fig. 4 displays the normalized TA traces for RD1, RD5, RD11 and N719. The transient of N719 involves a rapid rise and then a slow rise approaching asymptotically an oﬀset level within a ns range. The transients of the RD-series dyes exhibit, however, a rapid rise and a slow decay. The existence of a slow rise for the transient of N719 indicates that there might exist an energy barrier between the excited state of the dye and the conduction band of TiO2, and the excitation occurred near the band edge, whereas the absence of such a feature for the transients of all Fig. 1 Absorption spectra of RD1, RD5, RD10, RD11 and N719 in

DMF. Inset shows absorption spectra of TiO2ﬁlms (active layer of thickness 2 mm) sensitized with the indicated dyes.

Table 1 Photovoltaic parameters and amounts of dye loaded on DSSC with TiO2ﬁlms sensitized with RD1, RD5, RD10, RD11 and N719 under simulated AM-1.5G illumination (100 mW cm2) and active area 0.16 cm2

Dye Dye loading/nmol cm2 JSC/mA cm2 VOC/V FF Z (%) RD1 295 12.744 0.694 0.68 6.0 RD5 216 15.084 0.737 0.69 7.7 RD10 253 13.612 0.710 0.70 6.8 RD11 291 11.159 0.685 0.67 5.1 N719 149 14.157 0.783 0.70 7.8

Fig. 2 (a) Current–voltage characteristics of RD1, RD5, RD10, RD11 and N719 measured under thick TiO2conditions (12 mm active layer +3 mm scattering layer); (b) corresponding IPCE action spectra.

Fig. 3 Electron-transport properties: (a) recombination coefficient vs. VOC and (b) diffusion coefficient vs. JSCfor DSSC devices made of RD1, RD5, RD10 and RD11. The wavelengths are 632.8 nm for the bias light and 430 nm for the probe (ns pulse).

(3)

8994 Chem. Commun., 2010, 46, 8992–8994 This journal is c The Royal Society of Chemistry 2010 other RD dyes indicates that the excitation energies were much

above the energy barrier. The cyclic voltammetry measure-ments indicate that the LUMO level of N719 is signiﬁcantly below those of the RD dyes (Fig. S3w), consistent with our assumption.

The transient of N719 does not decay in the observed ns region, whereas the transients of RD1, RD5 and RD11 exhibit a slow-decay feature with an offset for which the decay characteristics cannot be resolved on this time scale. The slow-decay components of the transients were fitted with time coefficients 1.0 ns for RD1, 3.6 ns for RD5 and 0.42 ns for RD11. These results indicate that there exists a ns-BET process in the RD dyes that was non-observable in N719. The observed BET kinetics are consistent with HOMO levels showing a sequence of RD11 > RD1 > RD5 > N719 (Fig. S3w). Because the dye regeneration of a Ru system occurs typically on a ms scale,11the observed BET of the RD dyes on the ns scale are expected to play a role in cell performance. The BET kinetics were observed for thin-film samples that involved no electrolyte, but the trend of the BET decays is consistent with the variation of VOC showing the order N719 > RD5 > RD1 > RD11 for the corresponding devices. Because RD5 has a greater light-harvesting feature to enhance its JSC, it com-pensates its VOCloss to yield a cell performance comparable with that of N719.

In conclusion, we designed new heteroleptic ruthenium complexes containing benzimidazole substituents for applica-tion to dye-sensitized solar cells. These Ru complexes were synthesized according to a standard one-pot procedure with the corresponding heteroleptic ligands produced in only two simple steps. The corresponding devices show performances with the order RD5 > RD10 > RD1 > RD11; the eﬃciency of power conversion of RD5 is comparable with that of N719. The results obtained from photocurrent and photovoltage decays indicate that the electron lifetimes of the devices display a systematic trend RD5 > RD10 > RD1 > RD11, which is consistent with the cell performance showing the same order. The observed charge-recombination kinetics reasonably explain

the order of VOC to be N719 > RD5 > RD10 > RD1 > RD11. The IPCE action spectra of the devices account for the order of JSCto be RD5 > N719 > RD10 > RD1 > RD11. Note that the molecular structure of RD10 is similar to a highly efficient Ru sensitizer (CBTR) recently reported,12but the latter benzimidazyl ligand was coordinated with the N–C atoms involving more synthetic steps. We emphasize here that the ease of synthesis is an important factor to be considered in making a highly efficient sensitizer for future commerciali-zation. Work is in progress along this line to design and to synthesize more efficient heteroleptic ruthenium complex sensitizers with superior light-harvesting ability and slower charge recombination.

We thank Prof. Michael Gra¨tzel for many helpful discus-sions and Ms. Yu-Sin Liu for her preliminary work on synthesis. National Science Council of Taiwan and Ministry of Education of Taiwan, under the ATU program, provided support for this project.

Notes and references

1 (a) M. Gra¨tzel, Acc. Chem. Res., 2009, 42, 1788; (b) H. J. Snaith, Adv. Funct. Mater., 2010, 20, 13.

2 (a) F. Gao, Y. Wang, D. Shi, J. Zhang, M. Wang, X. Jing, R. Humphry-Baker, P. Wang, S. M. Zakeeruddin and M. Gra¨tzel, J. Am. Chem. Soc., 2008, 130, 10720; (b) C.-Y. Chen, J.-G. Chen, S.-J. Wu, J.-Y. Li, C.-G. Wu and K.-C. Ho, Angew. Chem., Int. Ed., 2008, 47, 7342; (c) J.-J. Kim, H. Choi, C. Kim, M.-S. Kang, H. S. Kang and J. Ko, Chem. Mater., 2009, 21, 5719.

3 (a) C.-W. Lee, H.-P. Lu, C.-M. Lan, Y.-L. Huang, Y.-R. Liang, W.-N. Yen, Y.-C. Liu, Y.-S. Lin, E. W.-G. Diau and C.-Y. Yeh, Chem.–Eur. J., 2009, 15, 1403; (b) H.-P. Lu, C.-L. Mai, C.-Y. Tsia, S.-J. Hsu, C.-P. Hsieh, C.-L. Chiu, C.-Y. Yeh and E. W.-G. Diau, Phys. Chem. Chem. Phys., 2009, 11, 10270; (c) H.-P. Lu, C.-Y. Tsai, W.-N. Yen, C.-P. Hsieh, C.-W. Lee, C.-Y. Yeh and E. W.-G. Diau, J. Phys. Chem. C, 2009, 113, 20990.

4 (a) S. Ito, H. Miura, S. Uchida, M. Takata, K. Sumioka, P. Liska, P. Comte, P. Pechy and M. Gra¨tzel, Chem. Commun., 2008, 5194; (b) G. Zhang, H. Bala, Y. Cheng, D. Shi, X. Lv, Q. Yu and P. Wang, Chem. Commun., 2009, 2198; (c) Z. Ning and H. Tian, Chem. Commun., 2009, 5483.

5 (a) M. K. Nazeeruddin, F. De Angelis, S. Fantacci, A. Selloni, G. Viscardi, P. Liska, S. Ito, B. Takeru and M. Gra¨tzel, J. Am. Chem. Soc., 2005, 127, 16835; (b) Q. Wang, S. Ito, M. Gra¨tzel, F. Fabregat-Santiago, I. Mora-Sero´, J. Bisquert, T. Bessho and H. Imai, J. Phys. Chem. B, 2006, 110, 25210.

6 (a) C.-Y. Chen, M. Wang, J.-Y. Li, N. Pootrakulchote, L. Alibabaei, C. Ngoc-Ie, J.-D. Decoppet, J.-H. Tsai, C. Gra¨tzel, C.-G. Wu, S. M. Zakeeruddin and M. Gra¨tzel, ACS Nano, 2009, 3, 3103; (b) Y. Cao, Y. Bai, Q. Yu, Y. Cheng, S. Liu, D. Shi, F. Gao and P. Wang, J. Phys. Chem. C, 2009, 113, 6290.

7 Y. Li, M. K. Fung, Z. Xie, S.-T. Lee, L.-S. Hung and J. Shi, Adv. Mater., 2002, 14, 1317.

8 (a) L. Luo, C.-J. Lin, C.-Y. Tsai, H.-P. Wu, L.-L. Li, C.-F. Lo, C.-Y. Lin and E. W.-G. Diau, Phys. Chem. Chem. Phys., 2010, 12, 1064; (b) L. Luo, C.-J. Lin, C. S. Hung, C.-F. Lo, C.-Y. Lin and E. W.-G. Diau, Phys. Chem. Chem. Phys., 2010, 12, 12973. 9 C.-W. Chang, C. K. Chou, I.-J. Chang, Y.-P. Lee and E. W.-G. Diau,

J. Phys. Chem. C, 2007, 111, 13288.

10 C.-W. Chang, L. Luo, C.-K. Chou, C.-F. Lo, C.-Y. Lin, C.-S. Hung, Y.-P. Lee and E. W.-G. Diau, J. Phys. Chem. C, 2009, 113, 11524.

11 A. B. F. Martinson, T. W. Hamann, M. J. Pellin and J. T. Hupp, Chem.–Eur. J., 2008, 14, 4458.

12 W.-C. Chang, H.-S. Chen, T.-Y. Li, N.-M. Hsu, Y. S. Tingare, C.-Y. Li, Y.-C. Liu, C. Su and W.-R. Li, Angew. Chem., Int. Ed., 2010, DOI: 10.1002/anie.201001628, in press.

Fig. 4 Time-resolved proﬁles of infrared transient absorption of TiO2 ﬁlms (thickness 2 mm) sensitized with RD1, RD5, RD11 and N719. The excitation and probe wavelengths are 625 nm and 4.9 mm, respectively.