Benzo [ 1 , 2-b : 4 , 5b 0 ] difuran-based sensitizers for dye-sensitized solar cells †

Dye-sensitized solar cells (DSSCs) as low cost alternatives to traditional photovoltaics have attracted the interest of scientists from different disciplines, and their overall performances developed very rapidly day by day. Despite the great achievement of Ru(II)-based dyes showing a power conversion efficiency (PCE) of up to 12%, much effort has been devoted to the attainment of metal-free organic dyes due to their low cost, ease of synthesis, structural diversity, and remarkably high optical extinction coefficients. Consequently, a large number of organic dyes, typically with a donor–p–acceptor (D–p–A) conguration, have extensively been investigated for DSSCs. In practice, a judicious variation of the molecular architectures of the donor fragments as well as of the p-linkers between the donor and acceptor fragments of the dyes, has been the most popular strategy to tailor the frontier orbital energy levels, which nally leads to the formation of dyes with broad and intense optical absorption patterns. Various molecular scaffolds such as triarylamine, carbazole, porphyrin, and indoline, have been used as attractive components in organic sensitizers. To the best of our knowledge, within this context there has been no report on systems featuring a benzo[1,2-b:4,5-b0]difuran (BDF) chromophore. BDFs have proven to be excellent components in organic eld-effect transistors (OFETs) and organic light-emitting diodes (OLEDs) by virtue of their p-type semiconductor characteristics, intrinsic optical properties and high hole mobility. It is noteworthy that their cationic radicals formed by a one-electron oxidation, show an intriguing stability. Thus, the BDF-based dyes are expected to have an enhanced stability in photo-induced electron transfer processes from the dyes to TiO2. For an efficient design of conjugated D–p–A dyes, one of the key issues relates to the synthesis of properly functionalized BDFs, which has rarely been exploited due to the associated synthetic challenges. In previous studies we have developed efficient synthetic routes to fully functionalized BDF derivatives. On this basis, we have been able to obtain two p-conjugated dyes (Fig. 1) that feature the BDF donor linked to one (Dye-1) and two (Dye-2) standard cyanoacrylic acid acceptors. To date, studies on the effect of the number and the positioning of a cyanoacrylic acid anchoring group on the cell performance are quite scarce.We report herein the synthesis and DSSC performance of Dye-1 and Dye-2. This work elaborates on the impact of the number of cyanoacrylic acid anchoring groups on the overall efficiencies of solar cells based on BDF donors. As illustrated in Scheme 1, two new dyes were synthesized in good yields as red solids via a Knoevenagel reaction with cyanoacetic acid. The corresponding aldehyde precursor II was readily obtained from I in 74% yield by reaction with one equivalent of 4-bromobenzaldehyde under standard Sonogashira conditions. Similarly, a Sonogashira reaction of III with 4-ethynylbenzaldehyde was accomplished to afford the dialdehyde IV. The target dyes and intermediate compounds have been fully characterized. Their NMR spectroscopic and Departement für Chemie und Biochemie, Universität Bern, Freiestrasse 3, CH-3012 Bern, Switzerland. E-mail: liu@iac.unibe.ch; Fax: +41 31 631 43 99; Tel: +41 31 631 42 96 Laboratory of Photonics and Interfaces, Institute of Chemical Science and Engineering, École Polytechnique Fédérale de Lausanne (EPFL), Station 6, CH-1050 Lausanne, Switzerland. E-mail: michael.graetzel@ep.ch; Fax: +41 21 693 61 00; Tel: +41 21 693 31 12 Computational Chemistry Lab, Department of Chemistry, University of Fribourg, Ch. du Musée 9, CH-1700 Fribourg, Switzerland † Electronic supplementary information (ESI) available: Experimental procedure and characterization data for Dye-1 and Dye-2, copies of their H and C NMR spectra, their detailed DFT calculations as well as their optical spectra on TiO2.

2][3][4][5] Despite the great achievement of Ru(II)-based dyes showing a power conversion efficiency (PCE) of up to 12%, 5 much effort has been devoted to the attainment of metal-free organic dyes due to their low cost, ease of synthesis, structural diversity, and remarkably high optical extinction coefficients.2][3][4][5][6][7] In practice, a judicious variation of the molecular architectures of the donor fragments as well as of the p-linkers between the donor and acceptor fragments of the dyes, has been the most popular strategy to tailor the frontier orbital energy levels, which nally leads to the formation of dyes with broad and intense optical absorption patterns. 8,9Various molecular scaffolds such as triarylamine, carbazole, porphyrin, and indoline, have been used as attractive components in organic sensitizers.To the best of our knowledge, within this context there has been no report on systems featuring a benzo[1,2-b:4,5-b 0 ]difuran (BDF) chromophore.
BDFs have proven to be excellent components in organic eld-effect transistors (OFETs) and organic light-emitting diodes (OLEDs) by virtue of their p-type semiconductor characteristics, intrinsic optical properties and high hole mobility. 10,11It is noteworthy that their cationic radicals formed by a one-electron oxidation, show an intriguing stability. 12,13hus, the BDF-based dyes are expected to have an enhanced stability in photo-induced electron transfer processes from the dyes to TiO 2 .For an efficient design of conjugated D-p-A dyes, one of the key issues relates to the synthesis of properly functionalized BDFs, which has rarely been exploited due to the associated synthetic challenges.3][14][15][16] On this basis, we have been able to obtain two p-conjugated dyes (Fig. 1) that feature the BDF donor linked to one (Dye-1) and two (Dye-2) standard cyanoacrylic acid acceptors.To date, studies on the effect of the number and the positioning of a cyanoacrylic acid anchoring group on the cell performance are quite scarce. 17,18We report herein the synthesis and DSSC performance of Dye-1 and Dye-2.This work elaborates on the impact of the number of cyanoacrylic acid anchoring groups on the overall efficiencies of solar cells based on BDF donors.
As illustrated in Scheme 1, two new dyes were synthesized in good yields as red solids via a Knoevenagel reaction with cyanoacetic acid.The corresponding aldehyde precursor II was readily obtained from I 14 in 74% yield by reaction with one equivalent of 4-bromobenzaldehyde under standard Sonogashira conditions.Similarly, a Sonogashira reaction of III 12,13 with 4-ethynylbenzaldehyde was accomplished to afford the dialdehyde IV. 19 The target dyes and intermediate compounds have been fully characterized.Their NMR spectroscopic and The optical and redox properties of the new dyes are listed in Table 1.Their electrochemical properties in CHCl 3 were investigated by cyclic voltammetry (CV).3][14][15][16] In contrast, in the case of Dye-2 these processes are not electrochemically fully reversible (see ESI †).This compound shows two distorted anodic peaks with E 1 pa ¼ 0.51 V and E 2 pa ¼ 0.77 V, and two successive cathodic peaks at E 1 pc ¼ 0.33 V and E 2 pc ¼ 0.43 V.This observation is indicative of an electrontransfer reaction, most probably coupled with a complex sequence of chemical follow-up processes.Upon the addition of one more cyanoacrylic acid anchoring group, the rst oxidation potential is substantially positive-shied, which reects the electronic interaction between D and A within these dyes.The rst oxidation potential in both cases is more positive than the Co(II)/Co(III) redox couple in use, ensuring efficient regeneration of the oxidized dye.Moreover, the HOMO and LUMO energy levels of both dyes were estimated according to the spectral analyses and the CV data (Table 1).Their relatively low-lying HOMO energy levels are expected to reveal good air stability and a high open circuit voltage (V oc ) in the DSSC device.Compared to Dye-1, the HOMO and LUMO levels of Dye-2 are energetically lowered due to the strong electron-withdrawing effect of two cyanoacrylic acid anchoring groups.
Density-functional theory (DFT) calculations for both dyes using the Coulomb Attenuated Methods at the B3LYP 6-31G(d,p) level of the theory, support the directional movement of charge upon photoexcitation (for details, see ESI †).As expected, the HOMO and LUMO of Dye-1 and Dye-2 are mainly located on the BDF and 2-cyano-3-(4-ethynylphenyl)acrylic acid moieties, respectively.An energetic stabilization of the HOMO when going from Dye-1 (À5.09 eV) to Dye-2 (À5.11 eV) is predicted, which is in line with the increase of the oxidation potential from Dye-1 to Dye-2; see Table 1.
The electronic spectra of the red colored Dye-1 and Dye-2, recorded in THF solution (see ESI †), show intense optical absorptions over the UV-vis spectral part with absorption onset energies at about 17 900 cm À1 (559 nm) and 17 100 cm À1 (585 nm), respectively.As expected, both of them exhibit a quite similar absorption pattern.Based on the detailed spectral analysis of the related BDF derivative with pending 4-ethynylpyridine groups 13 and DFT calculations, the main characteristics of the electronic transitions can be explained.Firstly, at low wavelength a strong and broad absorption band appears around 20 960 cm À1 (477 nm) for Dye-1 and 19 380 cm À1 (516 nm) for Dye-2, which can only be observed in such p-extended chromophores and is attributed to intramolecular p-p* charge-transfer (ICT) transitions from the BDF unit (HOMO, HOMO À 1 and HOMO À 2) to the 2-cyano-3-(4-ethynylphenyl)acrylic acid unit(s) (LUMO, LUMO + 1 and LUMO + 2) (for the relevant orbitals, see ESI †).The intense absorptions between 23 000 cm À1 (435 nm) and 31 000 cm À1 (323 nm) originate to a larger extent from BDFbased p-p* transitions.Compared to Dye-1, the red-shi of 1580 cm À1 and an increase in the molar extinction coefficient (3) for the ICT band in Dye-2 are attributed to the extension of the pconjugation of the BDF unit.
The absorption spectra of Dye-1 and Dye-2 adsorbed on a TiO 2 lm (see ESI †) show that the ICT absorption bands, related to those in THF, are red-shied by 11 nm and blue-shied by 12 nm, respectively.
Despite similar optical and redox properties for the two dyes, in the presence of a prototype coabsorbent, chenodeoxycholic acid (CDCA), which prevents dye aggregation on the TiO 2 surface, DSSC experiments revealed an enhanced performance of the Dye-1-based cells relative to Dye-2-based cells.The discrepancy in their cell performances can be rationalized by the incident photon-to-current conversion efficiency (IPCE) spectra depicted in Fig. 2. The IPCE depends on the light-harvesting efficiency, the net charge injection efficiency, and the electron collection efficiency. 20The integrated IPCE values between 400 nm and 540 nm exceed 60% for Dye-1, and they are higher than those for Dye-2 by a factor of 1.2.Both the shortcircuit photocurrent density (J sc ) and the power conversion efficiency (PCE) values for Dye-1 (Table 1, Fig. 2) are consistently higher than those for Dye-2 by a factor of 1.4.It can be therefore deduced that the presence of four hexyl groups prevents the leakage of electrons from TiO 2 to the electrolyte/the oxidized dye, and thus effectively suppresses the recombination processes, leading to the large FF (0.81 and 0.77) and high V oc values.As a result, the PCE values decrease progressively from 5.5% (Dye-1) to 3.8% (Dye-2).Based on all these observations and previously reported results in other DSSCs, 17,18,20,21 the differences between Dye-1 and Dye-2 in IPCE behavior are most likely ascribed to their net injection efficiencies.
In conclusion, we have presented a new type of dye sensitizers endowed with a BDF p-conducting group for DSSCs, namely, uorescent Dye-1 and non-emissive Dye-2 with one and two cyanoacrylic acid anchoring groups, respectively.For the rst time, BDF-sensitized DSSCs are described, showing PCEs of up to 5.5% in the presence of the coabsorbent CDCA.Detailed kinetic studies of DSSCs based on these two dyes will be carried out to determine their quantum yields for electron injection.Also a study on the structural modication of the BDF dyes capable of extending the spectral response to the long wavelength region and further enhancing the DSSC performance, is currently underway. .e HOMO level is calculated from the onset of the rst oxidation potential in cyclic voltammetry, according to the equation E HOMO ¼ [Àe(E onset + 4.8)] eV, where 4.8 eV is the energy level of ferrocene below the vacuum level.f LUMO level is estimated according to the equation E LUMO ¼ [E g,opt + E HOMO ] eV.g The cells were tested with a solution of dye (0.1 mM) in the presence of CDCA using a [Co(II/III)(bpy) 3 ] electrolyte under standard air mass 1.5 and simulated sunlight at 1000 W m À2 intensity.

Fig. 1
Fig. 1 Chemical structures of the synthesized dyes.

Fig. 2
Fig. 2 Photovoltaic performance of two BDF-based dyes in the presence of CDCA.(a) Photocurrent action spectra for Dye-1 (red) and Dye-2 (black), showing the IPCE as a function of excitation wavelength.(b) Photocurrent density (J) as a function of voltage (V) for Dye-1 (red) and Dye-2 (black) measured under standard air mass 1.5 and simulated sunlight at 1000 W m À2 intensity.

Table 1
Optical and electrochemical data, HOMO and LUMO energy levels, and photovoltaic parameters of Dye-1 and Dye-2 after optimization Optical band gap is estimated from UV-vis absorption onset.b The redox potentials vs. Fc + /Fc were recorded in CHCl 3 -Bu 4 NPF 6 (0.1 M) solution.c It corresponds to half-wave oxidation potential E 1/2 1 .d It corresponds to an anodic peak E pa 1 a