Mutsumi
Kimura
*,
Yuki
Tohata
,
Takuro
Ikeuchi
and
Shogo
Mori
*
Division of Chemistry and Materials, Faculty of Textile Science and Technology, Shinshu University, Ueda 386-8567, Japan. E-mail: mkimura@shinshu-u.ac.jp; shogmori@shinshu-u.ac.jp; Fax: +81-268-21-5499; Tel: +81-268-21-5499
First published on 23rd September 2015
Red-absorbing zinc phthalocyanine (ZnPc) PcS25 with double carboxylic acid anchors has been synthesized and used as the sensitizer for dye-sensitized solar cells with cobalt(II/III)-based redox electrolyte. Double anchored ZnPc sensitizer PcS25 exhibited higher conversion efficiency than single-anchored sensitizers.
Metallophthalocyanines (MPcs) have been extensively investigated as a class of molecular materials with applications from dyes/pigments to semiconductors for organic electronic devices.5 Red/near-IR light absorbing MPcs have been used as sensitizers for DSSCs, and the PCEs of MPc-sensitized DSSCs have improved significantly by systematic molecular engineering such as steric suppressions of aggregation, electronic push–pull structure through unsymmetrical substitutions, and optimization of adsorption sites.6,7 Recently, we reported that unsymmetrical ZnPc PcS20 bearing propoxy groups in the 2 and 6 positions of peripheral phenoxy units showed a PCE of 6.4% when used with iodide/triiodide redox electrolyte.8 However, the DSSCs with PcS20 using Co(bpy)32+/3+-based redox electrolyte displayed an inferior PCE value compared to iodide/triiodide redox electrolyte system.8 The introduction of alkyl chains around a ZnPc core was not effective in DSSCs with Co(bpy)32+/3+-based redox electrolyte. Therefore, the other molecular design for ZnPc sensitizers is required to enhance the PCE of DSSCs with Co(bpy)32+/3+-based redox electrolyte. In this study, we describe the structural effect of adsorption sites in ZnPc sensitizers on the photovoltaic properties of DSSCs with Co(bpy)32+/3+-based redox electrolyte. Double-anchored sensitizers have been synthesized to enhance optical density and binding strength of dye on TiO2.9 We found that the adsorption of ZnPc sensitizers possessing two carboxylic acids on the TiO2 surface enhanced the PCE in DSSC with Co(bpy)32+/3+-based redox electrolyte.
Two phthalocyanine precursors having ester units were synthesized from methyl 4-iodosalicylate through three steps (see ESI†).10 The unsymmetrical ZnPcs PcS25 and PcS26 (Fig. 1) were prepared by mixed cyclotetramerization of phthalocyanine precursors with 1,2-dicyano-3,4-bis(2,6-diphenylphenoxy)benzene7a in 1
:
3 ratio by refluxing in 2-(dimethylamino)ethanol in the presence of Zn(CH3COO)2 and subsequent hydrolysis with an aqueous alkaline solution. All final dyes and their corresponding intermediates were fully characterized by using standard spectroscopic techniques. As previously reported by us,7a,b the attachment of six bulky 2,6-diphenylphenoxy groups can prevent the intermolecular aggregation among ZnPcs. Absorption spectra of PcS25 and PcS26 adsorbed onto TiO2 electrodes revealed sharp and intense Q bands, indicating the prevention of molecular aggregation within the adsorbed monolayer of ZnPcs on TiO2 surface (Fig. 1).11 While the phenolic hydroxyl group in the adsorption site of PcS26 is protected with benzyl ester, PcS25 has two types of carboxylic acids on a peripheral substituent. The carboxylic acids can form an ester linkage with the surface of TiO2 to provide a strong anchoring dye as well as a pathway of electron transfer from the lowest occupied molecular orbitals (LUMO) of the dye to the TiO2 conduction band. Cid et al. reported the effect of the spacer between zinc phthalocyanine ring and carboxylic acid anchoring group on the photovoltaic properties.12 They found that ZnPc sensitizer having an aliphatic acid revealed a poor performance due to the intrinsic property of the spacer. In the case of PcS25, the photoinduced electron transfer from excited ZnPc to TiO2 may occur through the benzoic acid instead of the aliphatic acid (Fig. 2).
ZnPcs PcS25 and PcS26 exhibited one oxidation peak for oxidation of phthalocyanine ring (ZnPc(−1)/ZnPc(−2)) at +0.85 and +0.86 V vs. normal hydrogen electrode (NHE).13 The highest occupied molecular orbital (HOMO) levels of PcS25 and PcS26 are more positive than the redox potential of the Co(bpy)32+/3+ redox couple (0.54 V vs. NHE).4 The LUMO levels of PcS25 and PcS26 were estimated from the optical band gaps and the HOMO levels. The LUMO levels (PcS25: −0.93 V vs. NHE; PcS26: −0.94 V vs. NHE) are more negative than the conduction band (CB) of TiO2 (−0.5 V vs. NHE), revealing that the electron injection from the excited ZnPc to the CB of TiO2 is thermodynamically feasible.
FT-IR spectrum of PcS25-stained TiO2 film showed carboxylate units with asymmetric and symmetric stretching bands at 1600 and 1385 cm−1, and no absorption peaks corresponding to carboxylic acid were observed. This implies that two anchoring groups in PcS25 adsorbed onto the TiO2 surface via the bidentate binding mode.14 The adsorption densities of PcS25 and PcS26 on the TiO2 films were determined to be 2.4 × 10−5 and 5.7 × 10−5 mol cm−3 by measuring the absorbance of ZnPcs dissociated from the ZnPc-stained TiO2 films. These values for PcS25 and PcS26 were lower than the previously reported value of PcS6 lacking a side chain at the adsorption site.7a The introduction of side chain decreased the packing density of ZnPcs on the TiO2 surface. Since two carboxylate units in PcS25 occupy two adsorption sites on the TiO2 surface, the adsorption density of PcS25 was almost a half value of PcS26.
The DSSC performance of PcS25 and PcS26 was examined using double-layered TiO2 electrodes with iodide/triiodide redox electrolyte containing 0.6 M 1,2-dimethyl-3-propylimidazolium iodide (DMPImI), 0.1 M LiI, 0.05 M I2, 0.5 M tert-butylpyridine (TBP) in acetonitrile. Fig. 3a shows the photocurrent density–voltage (J–V) curve of the DSSCs using PcS25 and PcS26 under a standard AM 1.5 solar condition (100 mW cm−2), and the short-circuit photocurrent density (Jsc), open-circuit voltage (Voc), fill factor (FF), and PCE of PcS25 and PcS26 cells are listed in Table 1. The PCEs of the PcS25 and PcS26 cells were lower than that of the reported dye Pc6 (Jsc = 11.0 mA cm−2, Voc = 610 mV, FF = 0.70, PCE = 4.7%),7b and the PCE value was in the order of PcS25 < PcS26 < PcS6. This is mainly due to the low Jsc because of the lower adsorption densities of PcS25 and PcS26 on the TiO2 surface and lower incident photon-to-current conversion efficiency (IPCE) values, as shown in Fig. 3b, relative to PcS6. The difference in Jsc between PcS25 and PcS26 cells is also due to different amount of dye loading and IPCE values.
| Sensitizers | Adsorption densitya × 10−5/mol cm−3 | ELb | V oc/mV | J sc/mA cm−2 | FF | PCE/% |
|---|---|---|---|---|---|---|
| a Adsorption densities were determined by measuring the absorbance of dyes released from the TiO2 films by immersing into THF containing tetrabutylammonium hydroxide methanoic solution. b Electrolyte (EL): type A: 0.6 M DMPImI, 0.1 M LiI, 0.05 M I2, 0.5 M TBP in acetonitrile; type B: 0.2 M [CoII(bpy)3](B(CN)4)2, 0.02 M [CoIII(bpy)3](B(CN)4)3, 0.5 M TBP, and 0.1 M LiClO4 in acetonitrile. | ||||||
| PcS25 | 2.4 | A | 610 | 7.3 | 0.74 | 3.3 |
| B | 686 | 6.7 | 0.65 | 3.0 | ||
| PcS26 | 5.7 | A | 600 | 9.9 | 0.73 | 4.3 |
| B | 629 | 3.6 | 0.60 | 1.4 | ||
| PcS6 | 7.9 | B | 566 | 2.9 | 0.5 | 0.8 |
We also fabricated DSSCs using PcS25 and PcS26 with Co(bpy)32+/3+-based redox electrolyte containing 0.2 M [CoII(bpy)3](B(CN)4)2, 0.02 M [CoIII(bpy)3](B(CN)4)3, 0.5 M TBP, and 0.1 M LiClO4 in acetonitrile.4 The J–V curves of both cells are shown in Fig. 4a. Double anchored PcS25 gave a higher PCE of 3.0% compared with 1.4 and 0.8% for single anchored PcS26 and PcS6. The PCE of PcS25 cell with Co(bpy)32+/3+-based redox electrolyte was slightly lower than that with iodide/triiodide redox electrolyte (Table 1). The Voc values for the PcS25 cells employing Co(bpy)32+/3+-based redox electrolyte increased by 76 mV as compared with iodide/triiodide redox electrolyte-based cells, because of the more positive redox potential of cobalt complex. Furthermore, the introduction of aliphatic carboxylic acid as the adsorption site induced a large IPCE improvement. The PcS25 cell showed higher IPCE values over the whole visible light range than PcS26 cell (Fig. 4b).
Initially, we expected that PcS25 could increase the PCE by retarding the charge recombination by covering TiO2 surface more effectively by two carboxylic acid anchoring groups of PcS25. Actually, it seems PcS25 somehow reduced the recombination based on the J–V curves in the dark, although the amount of adsorption was decreased in comparison to single anchored PcS26. To check if the larger IPCE values from PcS25 were due to increased electron diffusion length, we also fabricated DSSCs with 4.9 μm thick TiO2 without scattering layer (see Fig. S1 in ESI†). We expected higher IPCE values with thinner film if the diffusion length was not sufficient. The absorbance of the thin films around 680 nm was more than 4 for the both dyes, but the IPCE values at the wavelength was about 30 and 15% for PcS25 and PcS26, respectively. Fig. 4b shows higher IPCE values. This means absorbed photon-to-current conversion efficiency was increased with the increase of the film thickness. At this stage, we have no explanation for this result. However, at least, the result suggests that the diffusion length was not the main reason for the higher IPCE values of PcS25. Secondly, we thought the lower IPCE values of PcS26 was due to insufficient reduction rate of dye cation. Since PcS26 was adsorbed more densely on the TiO2 surface, Co(bpy)3 complex may have less chance to hit dye's HOMO due to smaller scattering cross section. Then, we prepared the cell with less amount of adsorbed PcS26. The IPCE was not decreased by decreasing the amount of the PcS26 to roughly half. The result partially supports the hypothesis. However, we could not increase the IPCE of PcS26 cells. Thus, the difference in the reduction rate can not explain fully the difference in the IPCE values in Fig. 4b. The last possibility is that some of excited electrons in the dyes transfer directly to Co(bpy)33+ complexes and the degree of the transfer is different between two dyes, but this hypothesis should be substantiated further.
Footnote |
| † Electronic supplementary information (ESI) available: Synthetic procedures and DSSC fabrication. See DOI: 10.1039/c5ra16610a |
| This journal is © The Royal Society of Chemistry 2015 |