Visible and Near-infrared Organic photosensitizers comprising isoindigo derivatives as chromophores : synthesis , optoelectronic properties and factors limiting their efficiency in dye solar cells

The development of ruthenium-free organic photosensitizers showing panchromatic absorption up to the near-infrared (NIR) region for application in dye-sensitized solar cells (DSSCs) is still scarce. Among the sensitizers with absorption beyond 700 nm and developed for DSSCs, only zinc-phthalocyanine and boron-dibenzopyrromethene-based dyes have been able to reach efficiencies as high as 6%. Here we report metal-free organic dyes based on isoindigo, thieno-isoindigo or benzo-thieno-isoindigo chromophores that absorb in the UV-visible and NIR spectral range up to 900 nm. These molecules, that exhibit purple, blue, or green hues, were used to sensitize TiO2 mesoporous electrodes in order to fabricate DSSCs with an iodide/triiodide-based electrolyte. Advanced photophysical characterizations, including charge extraction, transient photovoltage, and laser transient absorption spectroscopy experiments, combined with density functional theory modeling and computational investigations allow us to fully unravel the interfacial processes at the origin of the solar cell performances and to identify the limiting factors. A power conversion efficiency as high as 7% associated with a Jsc close to 19 mA cm−2 was obtained with one of the dyes, which is comparable to those of the best panchromatic organic dyes reported so far. We also demonstrate in this work that the Voc of the solar cells is linearly correlated to the dipolar moments of the oxidized dyes, the molecules possessing larger dipoles leading to the highest Voc values.


I. Instruments and methods
UV-vis absorption spectra were recorded in solution on a Cary 60 (wavelength range: 190 to 1100 nm; 1.5 nm fixed spectral bandwidth, full spectrum Xenon pulse lamp single source), Agilent.
Electrochemical studies of the synthesized molecules were carried out in a one compartment, three-electrode electrochemical cell equipped with a flat platinum working electrode (7 mm 2 ), a platinum wire counter electrode, and a silver wire pseudo-reference electrode, whose potential was checked using the Fc + /Fc couple as an internal standard.The electrolyte consisted of 0.2 M tetrabutylammonium hexafluorophosphate (Bu4NPF6) solution in dichloromethane containing 2 x 10 -3 M of the dye.Cyclic voltammetry measurements were conducted with a sweep rate of 100 mV.s -1 .

II. Dyes Sensitized Solar Cells fabrication and characterizations Fabrication
The devices were prepared as followed: the various layers of TiO2 films were screen printed.The electrode total active area was 0.36 cm 2 .A first layer (12 µm) of transparent titania was deposited with a TiO2 nanoparticles paste (Ti-Nanoxide HT/SP) purchased from Solaronix, Switzerland.To further increase the light-harvesting capacity of these devices, a reflective layer (Solaronix' Ti-Nanoxide R/SP) of 4 µm was added on top.The total thickness of the titania working electrode was around 17 µm.In order to optimize adhesion, titania layer porosity and specific area a pre and post TiCl4 treatment was performed.After sintering at 500°C and cooling down to 80°C, the sintered TiO2 electrodes were sensitized by immersion in a solution of the dye in indicated solvent with or without chenodeoxycholic acid (CDCA) for 18 h, and then assembled using a thermally platinized FTO/glass (TCO 22-7, Solaronix) counter electrode.The working and counter electrodes were separated by a 25 µm thick hot melt gasket (Meltonix 1170-25, Solaronix) and sealed by heating.The heating was minimized to avoid dye thermal degradation.The cell was then filled with a volatile electrolyte (Solaronix Iodolyte HI-30) through a pre-drilled hole using a vacuum pump.The electrolyte injection hole on the thermally platinized FTO glass counter electrode was finally sealed with a thin glass cover.Devices using a non-volatile ionic liquid based electrolyte (Solaronix Mosalyte TDE-250) were prepared following the previously described procedure.The devices were characterized using a Solaronix SolarSim 150 previously calibrated.The current-voltage characteristics of the cell measured under AM 1.5G, 100% sun, were obtained by applying external potential bias to the cell and by measuring the generated photocurrent with a Keithley model 2400 digital source meter (Keithley, USA).The devices were masked prior to measurement according to a procedure previously described to attain an illuminated active area of 0.16 cm 2 .

Charge Extraction
As the name indicates, with this technique, we extract and measure the charge accumulated in the system under operating conditions.The Charge Extraction (CE) measurements were carried out with a system similar to that employed by O'Regan et al. 1 The cells were simultaneously in open-circuit and illuminated with LEDs until they reach to the steady state.Right after, the LEDs were switched off and the cells were short-circuited.Under these conditions, all the accumulated charges under open-circuit conditions can be collected, allowing the measurement to calculate the electron density. 2 By changing the LEDs intensity, the electron density can be estimated as a function of the cell voltage.The obtained value determines the conduction band edge shift of the semiconductor.The CE system consists in six white LEDs that generate pulses controlled by the Trigger (TGP, from Thrurlby Thandar Instruments).The voltage decay is measured by an oscilloscope TDS 2022 from Tektronix©.

Transient Photovoltage
Transient photovoltage (TPV) measurements were also carriedout with a system similar to the one employed by O'Regan et al. 1 This technique provides information on the electron recombination rate between the photoinjected electrons in the semiconductor band and the redox couple presents in the electrolyte under operating conditions.In TPV measurements, a constant background voltage is applied to the cells with a set of white LEDs until they reach to a steady state condition.Afterwards, an ultra-short laser pulse (660 nm, 10 mW) is applied and induces extra-injected electrons in the conduction band of the TiO2.Those extra charges recombine with the oxidized electrolyte due to the open circuit conditions, producing a transient decay.By modifying the voltage (changing the illumination intensity of the white LEDs) it is possible to record different transient decays.The TPV equipment is the same used for CE measurements.

Laser-Transient Absorption Spectroscopy
The measurement of Laser-Transient Absorption Spectroscopy (L-TAS) measurements were similar to those carried out previously in our group. 3It provides us information about the electron injection constant by monitoring the formation of the dye cation species (D + ) which appears after the electron injection.L-TAS can determinate both the undesired charge recombination of these electrons in the semiconductor with the oxidized dye and the regeneration of the dye cation by the redox electrolyte. 4A sample is irradiated constantly at a determinate wavelength, which corresponds to the maximum absorption wavelength of the excited state of the sample.At the same time, it is excited with a short light pulse, producing a change in the sample optical density.The change in optical density is monitored during a short period to record the variations.

V. DFT calculations and Bond Length Alternation determination
Density Functional Theory (DFT), implemented in ADF 2017, has been used to examine the frontier orbitals of every molecules.The estimated lowest energy conformations using Chem3D software were firstly submitted to a more accurate geometry optimization, using a combination of the Local Density Approximation VWN (Vosko, Wilk, Nusair) and the Generalized Gradient Approximation PBE (Perdew-Burke-Ernzerhof) functionals corrected for dispersion using Grimme methodology (Grimme 3) with the TZ2P basis sets (triple zeta + 2 polarization functions) in a solvent phase modeled through the COSMO model set for dichloromethane.Then, single-point calculations using the hybrid functional B3LYP (B3: Becke's 3-parameters, Eexchange functional + LYP: Lee, Yand, Parr, Ecorrelation functional) corrected with dispersion Grimme 3 with TZ2P sets in dichloromethane solvent phase were performed on the optimized conformations to modeled the HOMO and LUMO energy levels and their spatial localizations. 1       [CDCA] (mmol.L -1 ) Voc (V) Jsc (mA.cm -2 ) FF ƞ (%)

Computational details
All the calculations have been carried out using Gaussian09 software. 1The HOMO/LUMO energies have been obtained after geometry optimizations of the neutral and cationic radical species of the studied dyes in gas phase using the B3LYP functional 2 in combination with the 6-31G* basis set 3 for all atoms.This choice has been reported to produce a good agreement between the experimental and computed energies of frontier orbitals. 4The dye/iodide interaction energies have been computed using the unrestricted formalism of the dispersion corrected M062X (D3) functional. 5The geometries have been reoptimized in acetonitrile using the PCM method 6 and employing the SMD solvation model. 7The cc-pVDZ basis set 8 was employed to describe all atoms except iodide, for which the aug-cc-pVDZ-PP basis set was used. 9These settings have been employed before for describing this kind of interactions in similar systems. 10To ensure that charges were distributed appropriately on the interacting pairs, a fragment guess was first generated specifying a negative charge on iodide and all subsequent calculations were performed reading their initial guess from a checkpoint file based off this fragment guess.The basis set superposition error (BSSE) for the dye-iodide interactions is considered to be negligible 10 and thus it was not computed for the species in this report.Table S3: Optical and electrochemical properties of TPAT8-Iso and 6OTPA-Iso

Figure S7a -
Figure S7a-Frontier molecular orbital spatial repartition of the dyes calculated using hybrid B3LYP functional.

Figure S12 -
Figure S12-Photovoltaic parameters and IV curves of the best 6OTPA-Iso based solar cells.

Table S1 -
Summary of the different dihedral angles and BLA (Bond Length Alternation) measurements of the 6 synthesized dyes calculated from DFT calculations.
Photovoltaic parameters and IV curves of 6OTPA-Iso based solar cells obtained with different CDCA ratio and various tBP concentrations in the electrolyte.