Contents lists available at ScienceDirect Solar Energy journal homepage: www.elsevier.com/locate/solener Neutral organic redox pairs based on sterically hindered hydroquinone/ benzoquinone derivatives for dye-sensitized solar cells Natalie Flores-Díaza,b, Andrea Soto-Navarroa, Marina Freitagd, Guy Lamoureuxb,c,⁎, Leslie W. Pinedaa,b,⁎ a Centro de Electroquímica y Energía Química (CELEQ), Universidad de Costa Rica, 2060 San José, Costa Rica b Escuela de Química, Universidad de Costa Rica, 2060 San José, Costa Rica c Centro de Investigación en Productos Naturales (CIPRONA), Universidad de Costa Rica, 2060 San José, Costa Rica d Department of Chemistry, Ångström Laboratory, Uppsala University, 75126 Uppsala, Sweden A R T I C L E I N F O Keywords: Dye-sensitized solar cells Liquid electrolyte Organic redox pair Hydroquinone Benzoquinone A B S T R A C T Substituted derivatives of hydroquinone/benzoquinone were studied as organic redox mediators in the elec- trolyte for dye-sensitized solar cells (DSSCs). Thus, di-tert-butylhydroquinone (DTHQ), thymohydroquinone (ThymHQ) and phenylhydroquinone (PhHQ), were combined with their oxidized counterparts to form the pairs DTHQ/DTBQ, ThymHQ/ThymBQ, and PhHQ/PhBQ. In general, the characteristic parameters of the DSSCs with the substituted derivatives surpassed those of the DSSC with the unsubstituted hydroquinone/benzoquinone electrolyte. The short-circuit current (JSC) of the devices using DTHQ/DTBQ and ThymHQ/ThymBQ (13.61mA cm−2 and 12.56mA cm−2, respectively) are comparable to the JSC obtained for cobalt(II/III) tris (bipyridine) as a reference electrolyte (14.54mA cm−2). However, parameters such as open-circuit voltage (VOC) and fill factor (FF) (547mV and 0.48, respectively) are far from competitive. The best photovoltaic performance was obtained for the pair ThymHQ/ThymBQ using a triphenylamine (TPA)-based organic dye (LEG4) as sen- sitizer and a hybrid counter electrode with poly(3,4-ethylenedioxythiophene) (PEDOT) and graphene. These experimental conditions give under 1 sun (98%) the highest efficiency (η=3.19%); low-light intensities of 12.3% and 51.8% suns lead to efficiencies of 3.34% and 3.29%, respectively. Electrochemical impedance spectroscopy (EIS) revealed that the main cause for loss in photocurrent is the low recombination resistance compared to Co(II/III) as reference electrolyte. Based on the EIS analysis, a down-shift of the conduction band of TiO2 was found for all assembled devices containing the organic redox mediators, which explains the low VOC values for these derivatives. 1. Introduction Dye-sensitized solar cells (DSSCs) are third generation solar cells that can be produced in a variety of colors and shapes and can work under diffuse light conditions (O’Reagan and Grätzel, 1991; Zhang et al., 2016). They are comprised of a photoanode – typically a trans- parent conductive glass substrate of fluorine-doped tin oxide (SnO2:F, FTO) – a thin layer of a nanostructured semiconductor (e.g., TiO2), which is then sensitized with dye molecules, and a counter electrode (CE) with a thin layer of a catalyst such as platinum (Pt) or poly(3,4- ethylenedioxythiophene) (PEDOT) (Kalyanasundaram et al., 2010). Importantly, as dye molecules excite upon photon absorption followed by electron injection, the oxidized state of the dye is reduced to its ground state by an electron transfer from a redox couple in a liquid electrolyte filled between the electrodes. Together, these components contribute to the overall performance of the cell; the redox couple in the liquid electrolyte plays an important role to attain high energy con- version efficiencies (Feldt et al., 2010; Wu et al., 2016). The redox couple − −I I/ 3 is widely used because it presents large re- combination resistance and provides high electron lifetimes (Ito et al., 2005; Teuscher et al., 2014). Nonetheless, this redox couple has several disadvantages such as corrosiveness, absorption of light in the visible range, sublimation, and a large potential drop due to the mismatch of the redox potential of − −I I/ 3 compared to the highest occupied mole- cular orbital (HOMO) level of the sensitizer dye (Sun et al., 2015; Yum et al., 2012). In the pursuit of alternative redox shuttles, cobalt com- plexes such as [Co(bpy)3]2+/3+ (where bpy=2,2′-bipyridine) have shown a rather good match of the redox potential with the HOMO of the dye. Cobalt complexes require a lower driving force for dye re- generation, leading to higher open circuit voltages (VOC), favorable https://doi.org/10.1016/j.solener.2018.03.084 Received 14 January 2018; Received in revised form 17 March 2018; Accepted 30 March 2018 ⁎ Corresponding authors at: Escuela de Química, Universidad de Costa Rica, 2060 San José, Costa Rica. E-mail address: leslie.pineda@ucr.ac.cr (L.W. Pineda). Solar Energy 167 (2018) 76–83 Available online 06 April 2018 0038-092X/ © 2018 Elsevier Ltd. All rights reserved. T http://www.sciencedirect.com/science/journal/0038092X https://www.elsevier.com/locate/solener https://doi.org/10.1016/j.solener.2018.03.084 https://doi.org/10.1016/j.solener.2018.03.084 mailto:leslie.pineda@ucr.ac.cr https://doi.org/10.1016/j.solener.2018.03.084 http://crossmark.crossref.org/dialog/?doi=10.1016/j.solener.2018.03.084&domain=pdf power conversion efficiencies (η), and other characteristic parameters (e.g., short circuit current density, JSC, and fill factor, FF) (Safdari et al., 2016; Bella et al., 2016). However, the large size of such coordination complexes decreases the diffusion coefficient of the active species in the electrolyte, the rate of dye regeneration, and the efficiency of the cell. Another unwanted issue with these colored complexes relates, in some cases, with the absorption of light in the visible range competing with the dye (Aghazada et al., 2016). This issue can be solved by optimi- zation of organic redox couples, as shown with the use of disulfide/ thiolate redox couples, which exhibit low absorption in the visible range (Wang et al., 2010). Alternatively, some interesting outcomes regarding iodine/iodide redox mediators in DSSCs made use of a hy- drogel electrolyte consisting of carboxymethylcellulose and − −I I/ 3 in an aqueous systems (Bella et al., 2017), or membrane-based separators (Nair et al., 2015; Shanti et al., 2016). On the other hand, the organic redox pair hydroquinone/benzo- quinone (HQ/BQ) has spurred a great body of research due to its im- portance in biological processes (Alligrant et al., 2010). Interestingly, the electron transfer of the redox couple is a thermodynamically re- versible process (Shaidarova et al., 2003). In non-aqueous media and in the absence of proton donor/acceptors, the hydroquinone dianion un- dergoes two successive one-electron steps (Fig. 1a). In the presence of proton donor/acceptors, the electron transfer process is coupled with the proton transfer processes (Fig. 1b) (Guin et al., 2011). Previous experiments have shown that quinones and their electron transfer processes are coupled with proton transfer processes of the corresponding hydroquinones, in the absence or presence of proton donors or acceptors (Bhat, 2012). When there are proton acceptors in the media, the mechanism starts with the deprotonation of the hydro- quinone (QH2) compound, then an oxidation step forms the radical QH %, which is further deprotonated leading to the radical anion Q%− that, upon oxidation, yields a quinone (Q) (Fig. S5 in Supporting Info). The overall reaction proceeds as depicted in Fig. 1b, with a two-electron/ two-proton process between the neutral hydroquinone and benzoqui- none, suggesting that HQ/BQ-like redox pairs in their neutral forms can undergo electron transfer steps and could regenerate the dye in DSSC. When hydroquinone is treated with tetramethylammonium hydro- xide (TMAOH), it gives the bis-tetramethylammonium hydroquinone dianion (TMAHQ) ionic species, shown in Fig. 1a. In previous reports, the anionic hydroquinone species (TMAHQ/BQ) was used as a redox mediator in DSSCs with dye N719 as sensitizer and Pt as CE; these systems showed promising photovoltaic characteristics: VOC=750mV, JSC= 17.2mA cm−2, FF=0.663, and 8.4% conversion efficiency. With the same redox mediator, but with PEDOT as CE and the organic dye CM-309, the following parameters were achieved: VOC=755mV, JSC= 12.10mA cm−2, FF 0.678, and η=6.2% (Cheng et al., 2012, 2013). One drawback of using TMAHQ in organic redox mediators is the limitation of the lifetime of the solar cell (Yu et al., 2013). Deproto- nated hydroquinones are prone to decomposition in open air condi- tions, which compromises the overall performance of the DSSC. Independently and innovatively, we examine the photovoltaic performance of a series of DSSCs using neutral HQ derivatives bearing various bulky substituents on the aromatic ring. Our molecular design provides neutral, stable organic redox shuttles. The HQ derivatives, and the corresponding BQ pair, are stable in air. Our approach also sim- plifies the chemical composition of the electrolyte by decreasing the number of charged species in the liquid phase. To this end, the deri- vatives 2,5-di-tert-butylhydroquinone, phenylhydroquinone and 2-iso- propyl-5-methylhydroquinone (thymohydroquinone) are paired with their benzoquinone counterparts to render the redox pairs DTHQ/ DTBQ, PhHQ/PhBQ and ThymHQ/ThymBQ, respectively (Fig. 2). In solution, these redox pairs were optically characterized by UV–Vis spectroscopy and by cyclic voltammetry, followed by determination of their photovoltaic parameters (for instance, JSC, VOC, FF, and η) from current density-voltage plots once the DSSC were assembled. Likewise, a series of internal processes of the assembled DSSCs were determined by electrochemical impedance spectroscopy (EIS). 2. Experimental Details of the synthesis and characterization of the derivatives are described in the Supporting Information. 2.1. Materials All chemicals were purchased from Sigma-Aldrich unless otherwise noted. The organic dyes 3-{6-{4-[bis(2′,4′-dibutyloxybiphenyl-4-yl) amino-]phenyl}-4,4-dihexyl-cyclopenta-[2,1-b:3,4-b']dithiophene-2- yl}-2-cyanoacrylic acid, LEG4, and 3-{6-{4-[bis(2′,4′-dihexylox- ybiphenyl-4-yl)amino-]phenyl}-4,4-dihexyl-cyclopenta-[2,1-b:3,4-b'] dithiphene-2-yl}-2-cyanoacrylic acid, Y123, were purchased from Dyenamo AB (Stockholm, Sweden). 2.2. Solar cell preparation Glass substrates with fluorine-doped tin oxide (FTO, Pilkington, TEC15) were cleaned in an ultrasonic bath for 1 h in the following order of solvents: water, ethanol, and acetone. The FTO substrates were pretreated by immersion for 30min in a 40mM aqueous TiCl4 solution at 70 °C, and then washed with water, to form a thin blocking layer. Mesoporous TiO2 films of 0.25 cm2 were prepared as follows: a trans- parent active layer was made with colloidal TiO2 paste (Dyesol DSL 30 NRD-T) by screen-printing technique and dried at 120 °C for 6min between each layer of TiO2 applied. Subsequently, a light-scattering layer (Dyesol WER2-0) was deposited on top by screen-printing. The substrates were then gradually heated in an air atmosphere oven (Nabertherm controller P320), applying a four-level program: 125 °C (10min), 250 °C (10min), 350 °C (10min), and 450 °C (30min). After sintering, the electrodes were treated in aqueous titanium tetrachloride (TiCl4) at 70 °C for 30min, then washed with water and ethanol. The thicknesses of the TiO2 films were measured with a profilometer (Veeco Dektak 3); the thicknesses are 6 μm and 2 μm for transparent and scattering layers, respectively. A final heating step at 500 °C (30min) Fig. 1. Organic redox pair hydroquinone/benzoquinone (HQ/BQ) electron transfer steps: (a) absence of a proton donor/acceptor and (b) in presence of a proton donor/acceptor. N. Flores-Díaz et al. Solar Energy 167 (2018) 76–83 77 was performed followed by overnight immersion of the electrodes in the dye bath solution. The dye bath composition was 0.2 mM LEG4 in tert-butyl alcohol and acetonitrile (1:1), and for Y123 the concentration was 0.2 mM in THF/EtOH (1:4). Then, after immersion, all films were rinsed in acetonitrile to remove the excess of dye. Solar cells were as- sembled using a 25 μm thick thermoplastic Surlyn frame as sealant and spacer between electrodes, with a (PEDOT)-coated counter electrode (TEC8). The PEDOT electrodes were prepared by electro-polymeriza- tion of 3,4-ethylenedioxythiophene (EDOT) from a micellar aqueous solution of 0.1M sodium dodecyl sulfate (SDS) and 0.01M EDOT. The electrolyte solution was introduced under vacuum through a hole pre- drilled in the counter electrode and sealed with thermoplastic Surlyn and a glass coverslip. The electrolyte composition was 0.2 M of the HQ derivative and 0.04M of the corresponding benzoquinone derivative, 0.1 M lithium bis(trifluoromethanesulfonyl)imide (LiTFSI), and 0.5 M 4-tert-butylpyridine (TBP) in acetonitrile. Reference electrolytes were prepared for comparison, by using 0.22M Co(bpy)3(PF6)2, and 0.05M Co(bpy)3(PF6)3, with the same additives and solvent used in the case of the organic pairs. 2.3. Electrochemical measurements Cyclic voltammetry (CV) measurements were performed with a Potentiostat/Galvanostat PGSTAT 128N Metrohm/Autolab with a three-electrode setup cell. A glassy carbon electrode as the working electrode, a graphite rod as the counter electrode, and a non-aqueous reference electrode of Ag/AgCl (2M in ethanol) were used. The elec- trolyte solutions contained 2mM of the hydroquinone derivative, 0.1 mM of the corresponding benzoquinone, and 0.1M of LiClO4 as supporting electrolyte in dry acetonitrile. The scan rate was 100mV s−1, formal potentials were determined versus ferrocenium/ ferrocene as a reference system, and then versus normal hydrogen electrode (NHE) with a value established for Fc+/Fc= 0.64 V versus NHE in acetonitrile and 25 °C (Pavlishchuk and Addison, 2000). 2.4. Solar cell characterization Current–voltage (I–V) characteristic curves and photocurrent-dy- namics were measured using a Xenon lamp of 450W (Oriel USA) as light source, with a filter Schott K113 Tempax and matched to AM 1.5G solar standard conditions using a reference Si photodiode. The current and voltage were measured and controlled by a Keithley 2400 digital source meter (Keithley, USA) and the current measurement was set up to be delayed 80ms from applying voltage. A set of metal mesh filters was used to adjust the light intensity to a desired level. A black metal mask with a 0.16 cm2 aperture was used to define the active area. 2.5. Electrochemical impedance spectroscopy (EIS) Analysis of DSSC with the respective organic pairs as mediators in the electrolyte was measured in dark conditions using a SP-300 bipo- tentiostat (Biologic Science Instrument). A range of potentials was se- lected, 18 steps linearly spaced between 0 and 1.0 V, but only from 0 V to approximately 0.1 V beyond the VOC of every respective DSSC were taken in account for the simulation and data analysis. The AC amplitude voltage perturbation was 10mV, and the range of frequencies was from 7MHz to 0.1 Hz at forward applied bias. Data analysis of the obtained impedance spectra was realized using Zview software from Scribner Associate Inc. The data was fitted into the corresponding equivalent circuit, using for the corresponding potentials the transmission line defined by DX type 11- Bisquert #2. 3. Results and discussion The hydroquinone and benzoquinone compounds under study did not decompose over several months in open air conditions, hence all the electrolytes were prepared and manipulated under normal conditions (without excluding atmospheric oxygen, water or light). UV–Vis spectroscopy was employed to analyze the optical behavior for the electrolytes containing the organic redox pairs: HQ/BQ, PhHQ/ PhBQ, DTHQ/DTBQ, and ThymHQ/ThymBQ (Fig. 3). The formulation of the electrolytes consists of the hydroquinone derivative (0.4M) and the corresponding oxidized species (benzoquinone, 0.02M), TBP (0.5M), and LiTFSI (0.1M) in dry acetonitrile; this composition was further used during the photovoltaic characterizations of the assembled devices. For the electrolytes with HQ, ThymHQ, and DTHQ derivatives the absorbance in the visible range is almost negligible, whereas PhHQ weakly absorbs between 400 and 500 nm. Since the maximum ab- sorption peak of the selected dyes LEG4 (Ellis et al., 2013) and Y123 (Tsao et al., 2011) are 541 and 530 nm, respectively, we found that competition for light absorption of our derivatives is minimum, thus high injection rates of photogenerated electrons can be expected. The electrochemical features for the HQ derivatives are measured using cyclic voltammetry (CV) experiments (Fig. 4). To provide the same proportion of HQ/BQ for the assembled solar devices (20:1, re- spectively), the chemical composition of all organic redox pairs is comprised of hydroquinone derivative (2mM), benzoquinone Fig. 2. Organic redox pairs based on derivatives of the pair hydroquinone/benzoquinone, (a) DTHQ/DTBQ, (b) PhHQ/PhBQ, (c) ThymHQ/ThymBQ. 400 500 600 700 800 0.0 0.5 1.0 1.5 2.0 2.5 A bs or ba nc e /a .u . Wavelength / nm PhHQ/PhBQ DTHQ/DTBQ ThymHQ/ThymBQ HQ/BQ Fig. 3. Absorption spectra of the electrolytes containing the organic redox pairs with a composition of 0.4M of the hydroquinone derivative and 0.02M of the benzoquinone counterpart, 0.5 M TBP and 0.1M LiTFSI. N. Flores-Díaz et al. Solar Energy 167 (2018) 76–83 78 counterpart (0.1 mM), and LiClO4 (0.01M) as supporting electrolyte in dry acetonitrile. Previous reports disclosed that HQ in aprotic solvents undergoes facile and reversible electron transfer steps; the nature of the sub- stituent groups affect the general trend of displacements in the reduc- tion and oxidation peaks (Gupta and Linschitz, 1997; Uchimiya and Stone, 2009). By contrast, other electrochemical measurements (also in aprotic media) revealed two-electron irreversible oxidation steps and two-electron irreversible reduction steps for hydroquinones (Gamboa- Valero et al., 2016). In fact, the specific mechanisms in non-aqueous media depend strongly on the presence or absence of proton donor/ acceptors. For reversible one-electron reactions, the thermodynamic separa- tion of the oxidation and reduction peaks (peak-to-peak separation) at 25 °C should be approximately 59mV and the height of both peaks should be same (the ratio of the anodic current, Ipa, and the cathodic current, Ipc, equals 1 for a reversible redox process) (Girault, 2005). For hydroquinone/benzoquinone couples, two single-electron transfer steps should be expected, following the pathway as shown in Fig. 1a, or one transfer step with 2-electrons with a peak-to-peak separation of 29.5 mV, as shown in the mechanism Fig. 1b (Girault, 2005). The cyclic voltammograms of the HQ derivatives are shown in Fig. 4; they exhibit an irreversible behavior, since the peak-to-peak separation is greater than 59mV. Such electron transfer processes are usually followed by proton transfers (Bhat, 2012), and it is possible that the first electron- transfer step and the first proton-transfer step are kinetically controlled (Astudillo et al., 2007). Furthermore, the height of the oxidation peak for all derivatives is greater than the reduction peak, which indicates a dimerization of the intermediates (Staley et al., 2014). Anionic inter- mediates of quinones can interact strongly by hydrogen bonding with other species (Zhu et al., 2010), forming a quinhydrone-like complex (dimer) via strong hydrogen bonds (Gamboa-Valero et al., 2016). Interestingly, irreversible features in the cyclic voltammograms cannot be explained using the aforementioned mechanisms (see Fig. S5); however, these features can be explained by an irreversible proton- coupled electron transfer in HQ/BQ-like systems, as follows (Astudillo et al., 2007; Wang et al., 2013; Gamboa-Valero et al., 2016): QH2 ⇌ QH2 %+ + e− (1) QH2 %+ ⇌ QH% + H+ (2) QH% ⇌ QH+ + e+ (3) QH% + QH% ⇌ Q + QH2 (4) Even though the oxidation and reduction processes shown by the organic pairs are seemingly not related through the same mechanism path, they both undergo well-defined reduction and oxidation process (as seen in the voltammograms). In terms of the DSSC, the redox mediator should undergo a reduction at the counter electrode in the presence of the dye in the oxidized form, and an oxidation process to regenerate the dye molecules in their ground state. Although the ex- perimental redox pairs do not show Nernstian reversible behavior, the well-defined and separated peaks observed in the cyclic voltammo- grams could represent charge transfer processes and charge regenera- tion in the DSSC. In Table 1, the electrochemical parameters of the HQ derivatives obtained by the technique of CV are reported. The redox potentials “E1/ 2” in this case represent only a mid-value between the oxidation and reduction peaks, since the processes are not reversible. Both DTHQ/ DTBQ and ThymHQ/ThymBQ pairs exhibited oxidation and reduction peaks shifted to less positive values, which are in accordance with the expected behavior of substituted quinones, where electron-donating groups tend to lower the reduction potential. For the PhHQ/PhBQ, the reduction peak value is greater than the value for HQ/BQ, suggesting that phenyl group acts as an electron-withdrawing group (Uchimiya and Stone, 2009). To further study the electrochemical characteristics of liquid elec- trolytes containing organic redox mediators based on HQ/BQ deriva- tives in solar-cell devices, symmetric cells were assembled with iden- tical electrodes of PEDOT and the organic electrolytes to obtain the limiting current plots (Fig. S6 in Supporting Info), and diffusion coef- ficient values were measured (Table 2). Highest values of Jlim were obtained for DTHQ and ThymHQ, re- sulting in higher diffusion coefficients for these redox pairs, with a maximum value of 1.62×10−6 cm2 s−1 for DTHQ/DTBQ. Indeed, one would expect to obtain a better performance in the DSSC with the former mediator from J-V curve measurements (vide infra). In the case of the obtained diffusion coefficients, they are comparable to previously reported values for HQ with PEDOT electrodes (Monge-Romero and Suárez-Herrera, 2013), iodide, and cobalt. Hence, diffusion limitations of the redox mediators in the DSSC are expected to be low (Park et al., 2014). The open-circuit voltage of DSSCs is determined by the difference between the Fermi level of TiO2 and the redox potential of the redox 0.0 0.5 1.0 1.5 -2.0x10-5 0.0 2.0x10-5 4.0x10-5 6.0x10-5 8.0x10-5 C ur re nt /A Potential / V vs. NHE DTHQ/DTBQ ThymHQ/ThymBQ PhHQ/PhBQ HQ/BQ Fig. 4. Cyclic voltammograms of the different organic redox pairs with a con- centration of the hydroquinone derivative of 2mM, benzoquinone counterparts 0.1 mM+LiClO4 0.1M in dry acetonitrile. A three-electrode cell with a glassy carbon electrode as the working electrode, a graphite rod as the counter elec- trode, and a non-aqueous reference electrode of Ag/AgCl (2M in ethanol), and scan rate of 100mV s−1. Table 1 Electrochemical data for the hydroquinone derivatives in dry acetonitrile. Redox pair Oxidation peak (V) NHE Reduction peak (V) NHE “E1/2” (V)a NHE HQ/BQ 1.256 0.715 0.985 DTHQ/DTBQ 1.238 0.292 0.765 PhHQ/PhHQ 1.218 0.754 0.986 ThymHQ/ThymHQ 1.136 0.561 0.848 a “E1/2” represents the average or mid-value between the reduction and oxidation peaks observed in the voltammograms, but it does not indicate the reversible redox potential. Table 2 Electrochemical parameters for the studied symmetrical cells with electrolytes containing organic mediator pairs based on hydroquinone derivatives. Redox pair Jlim (mA cm−2) D (cm2 s−1) HQ/BQ 7.63 1.235E−06 DTHQ/DTBQ 10.03 1.624E−06 PhHQ/PhBQ 8.95 1.450E−06 ThymHQ/ThymBQ 9.75 1.579E−06 N. Flores-Díaz et al. Solar Energy 167 (2018) 76–83 79 mediator in the electrolyte. The data of mid-potentials obtained from the oxidation and reduction peaks, “E1/2”, are an estimation of the redox potentials of these separated processes (Table 1). These E1/2 va- lues are greater than the − −I I/ 3 couple (0.354 V), and similar to I−/I2−% (0.937 V) (Boschloo and Hagfeldt, 2009). These findings indicate the possibility to obtain VOC values in the range of those for − −I I/ 3 (0.6–0.8 V), although the actual VOC values in the solar device can deviate given that the mid-redox potential may not necessarily define the average level of the redox pair, as they are associated with sepa- rated processes of reduction and oxidation, each one featuring different pathways. Further studies will be necessary to better understand the kinetic processes. The photovoltaic performance of DSSCs containing organic dye Y123, PEDOT as CE and electrolytes containing organic redox pairs HQ/BQ, DTHQ/DTBQ, ThymHQ/ThymBQ and PhHQ/PhBQ were measured under AM 1.5 G simulated solar light at a light intensity of 100mW cm−2 with a mask of 0.16 cm2 and were compared with DSSC using Co2+/3+ complex as a reference electrolyte. In our case, PEDOT was chosen as CE because its high activity with quinones (Park et al., 2014) and since several attempts with Pt or graphene were unsuccessful as CE. The chemical composition of the studied organic electrolytes is made up of the hydroquinone derivative (0.4 M), benzoquinone coun- terpart (0.02M), LiTFSI (0.1M) and TBP (0.5M) in dry acetonitrile. The current density–voltage (J–V) curves for the best performing cells for each redox mediator are displayed in Fig. 5. The photovoltage for the organic redox pairs clearly differs from the expected values based on the “redox potential” from the electrochemical analysis (see Table 1). This difference is probably due to kinetic issues regarding reduction and oxidation processes and a possible shift of the conduction band of TiO2 toward more negative values, which might account for the low VOC of such organic shuttles. In addition, the redox potentials of hydroquinones are affected by variables such as solvent polarity, pre- sence of proton donors or acceptors, intra- and intermolecular hydrogen bonding, water content, among others (Bhat, 2012; Gamboa-Valero et al., 2016). In our case, the electrolytes contain TBP and LiTFSI as additives that could affect the redox potential of the derivatives shifting them to lower values, thus decreasing the difference between the CB of the TiO2 and lowering the VOC. As seen from J-V curves (Fig. 5), the performance for the synthe- sized derivatives is better than that of HQ/BQ redox pair, so there is probably a favorable effect in the photovoltaic results of DSSCs as dif- ferent substituent groups are introduced in the hydroquinone molecule, although in the case of PhHQ/PhBQ, the effect is less. Table 3 sum- marizes the photovoltaic results of the electrolytes with the organic mediators. High JSC data were attained for DTHQ and ThymHQ (average values: 12.6 mA cm−2 and 10mA cm−2, respectively) comparable to the best cell performance for Co2+/3+ (14.4 mA cm−2) as electrolyte reference, with DTHQ reaching the highest conversion efficiency (2.5%). However, low fill factors could indicate the existence of alternative paths for the electrons, decreasing the recollected amount at the photoelectrode, thus compromising the efficiency of the solar cells (Fig. 5). To gain insight on the internal processes at DSSCs as to the effect of the different electrolytes containing the organic redox pairs in the charge transfer resistance and chemical capacitance, electrochemical impedance spectroscopy (EIS) was performed. A frequency scan from 7MHz to 0.1 Hz at room temperature in dark conditions were set for a voltage range for all samples with an alternating current (AC) ampli- tude of 10mV. EIS data were fitted with the typical three channel model (Bisquert et al., 2006) using ZView software from Scribner Inc. EIS analysis for the series resistance (RS) and counter electrode re- sistance (RCE) for all the electrolytes show low values (6.2–8.5Ω and 3–9Ω, respectively), an indication that PEDOT is suitable for the redox reaction of the quinone derivatives (Table S1 in Supporting Info). The relation between potential and the density of states (DOS) as- sociated with the chemical capacitance in the devices clearly show a shift of the conduction band of TiO2 toward lower values for all the organic electrolytes compared with the cobalt reference (Fig. 6). To some extent, these measurements likely explain the low VOC values obtained for the organic couples, given that once the conduction band is lowered, the VOC values defined by this level will be then di- minished. Again, some interactions of these derivatives with all the components in the electrolyte (additives and solvent) could decrease the “mid-potential”, which further decrease the VOC. To compare the hydroquinone derivatives with the cobalt reference shuttle under the same electron occupation, taking into consideration the downward shift in the conduction band of DSSCs with the organic electrolytes, EIS measurements were carried out to set recombination resistance and transport resistance as a function of DOS (Fig. 7). The recombination resistance (Rr) for all the organic mediators is lower relative to Co2+/3+ electrolyte by about two orders of magni- tude. These findings account for the loss in photocurrent observed in the sharp slope when going through high potentials extended to VOC in the J-V curve (associated with low FF values). Therefore, the electrons injected from the lowest unoccupied molecular orbital (LUMO) of the dye into the conduction band (CB) of TiO2 easily recombine with the organic redox mediators decreasing the rate of the injected electrons out of the circuit, subsequently affecting FF and decreasing the con- version efficiency of DSSCs. In this context, it has been previously reported that 1,4-benzoqui- nones can adsorb onto the surface of CdSe nanoparticles and other semiconductor surfaces (Uematsu et al., 2016). As such, one cannot rule out the possibility of adsorption of the hydroquinone derivatives on the surface of TiO2 as a plausible explanation for the loss in photocurrent and low recombination resistance. This can increase the recombination of electrons from the TiO2 conduction band to the quinones, while di- minishing the injected photocurrent rate. Indeed, the electron injection rate is governed by the difference in Rr and Rt which enables a low enough resistance to the transport of electrons in the TiO2 and a high resistance toward recombination, particularly when Rt is about 102 orders of magnitude lower than Rr in a larger range of voltage. For the cobalt electrolyte, this behavior is seen, and it renders characteristic photovoltaic parameters as opposed to those found for the organic mediators. In terms of the effect of sub- stituents on the organic mediators, DTHQ/DTBQ exhibits the larger difference between Rr and Rt (Figs. 7 and S7 in Supporting Info), which agrees with higher efficiency (Table 1). In addition, the derivatives DTHQ and ThymHQ achieved lower transport resistances than HQ, which can be ascribed to faster electron transport in the semiconductor layer and faster injection of electrons to the external circuit; these derivatives feature better performances and conversion efficiencies. 0.0 0.2 0.4 0.6 0.8 0 3 6 9 12 15 C ur re nt de ns ity /m A cm -2 Potential / V Co2+/3+ DTHQ/DTBQ ThymHQ/ThymBQ PhHQ/PhBQ HQ/BQ Fig. 5. J-V curves for the best performing cells of the different organic redox pairs sensitized with Y123. N. Flores-Díaz et al. Solar Energy 167 (2018) 76–83 80 As the photovoltaic performance of DTHQ and ThymHQ revealed higher short-circuit currents, they were further tested with another ef- ficient organic sensitizer (LEG4). The photovoltaic results are shown in Table 4, and the corresponding J-V curves are displayed in Fig. 8. The obtained short-circuit currents compare well with cobalt and copper- based electrolytes (11.00 mA cm−2 and 12.07mA cm−2, respectively) (Freitag et al., 2016), and with ruthenium-based sensitizers (12.17 mA cm−2–13.89mA cm−2) also with cobalt electrolytes (Zhang et al., 2016). Again, low VOC and fill factor values had a detrimental effect in the conversion efficiencies with LEG4. Interestingly, the use of a hybrid counter electrode with PEDOT and graphene brings about an improvement in their photovoltaic para- meters and might involve a favorable mechanism in the redox reaction/ charge transfer (Fig. 8). As such, the combination of ThymHQ/ThymBQ with LEG4 and hybrid CE with PEDOT and graphene (regardless of the light intensity) yields the highest conversion efficiencies (3.34% under 12.3% sun, 3.29% under 51.8% sun and 3.19% efficiency under 98% sun). Photocurrent-dynamics as a function of light intensity for DSSCs with Y123 dye and all the organic redox pairs as mediators were also measured (Figs. S9–S13 in Supporting Info). The results for cobalt as reference and the ThymHQ/ThymBQ derivative are shown in Figs. S9 and S10 in Supporting Info. Regardless of the organic redox pairs, comparing the photocurrent data of ThymHQ/ThymBQ with that of Co2+/3+, current dynamics gave the same behavior; seemingly there are no diffusion limitations as the photogenerated current at all light intensities remains constant for each time pulse, which is in agreement with the obtained values for diffusion coefficients of the organic med- iators (Table 2). This behavior can be attributed to the small size of the hydroquinone derivatives compared with bulky cobalt complexes, which do have diffusion limitations, slowing the rates of dye re- generation. This characteristic also indicates that diffusion of the or- ganic species in the electrolyte is not an issue for the current loss and limitation of photovoltages observed for the organic mediators; in fact, it seems they have good diffusion rates that could promote the process of regeneration of the oxidized dye molecules. 4. Conclusions In summary, the facile synthesis and photovoltaic performance of organic mediator couples based on hydroquinone/benzoquinone deri- vatives as redox mediators in DSSCs were demonstrated. DTHQ and ThymHQ feature high values of JSC and can be useful if further mole- cular designs are proposed to improve mainly FF and VOC. Moreover, the derivatives DTHQ, ThymHQ, and PhHQ with their quinone counterparts as redox mediators highlight how, by fine-tuning the steric hindrance of bulky substituents, the performance is enhanced relative to unsubstituted HQ/BQ. ThymHQ/ThymBQ with LEG4 dye and a hybrid counterelectrode (PEDOT and graphene) are highly ver- satile under different light intensities, and yield the highest conversion efficiencies among the different derivatives. Importantly, the loss in photocurrent and low FF values can be ex- plained by low recombination resistances, as suggested by the EIS analysis. As well, the recombination losses can be rationalized by ad- sorption of the quinones on the semiconductor, unfavorable interac- tions of the additives (TBP) with the redox species, and the dynamics of a possible mechanism that enables recombination over regeneration of the dyes. Our investigation will be continued in the future to examine the effect of the kinetics on the efficiency, the changes in the additives on the electrolytes, and especially the synthesis of quinone derivatives bearing electron-donating groups on the overall DSSC performance. Table 3 Photovoltaic parameters of DSSCs obtained with the organic mediator pairs and cobalt redox shuttle as reference. The dye Y123 was used as sensitizer, and PEDOT as catalyst in the CE. Redox pair VOC (mV) JSC (mA cm−2) Wmax (mW cm−2) FF (%) Co2+/3+ 879 ± 8 14.4 ± 0.4 8.6 ± 0.3 0.68 ± 0.01 8.6 ± 0.2 HQ/BQ 533 ± 8 6.5 ± 0.5 1.05 ± 0.04 0.30 ± 0.01 1.08 ± 0.04 PhHQ/PhBQ 528 ± 37 6.3 ± 0.8 1.3 ± 0.3 0.39 ± 0.01 1.3 ± 0.1 DTHQ/DTBQ 542 ± 3 12.6 ± 0.6 2.5 ± 0.4 0.36 ± 0.04 2.5 ± 0.4 ThymHQ/ThymBQ 455 ± 61 10 ± 2 2.0 ± 0.6 0.44 ± 0.01 2.0 ± 0.7 J–V data represents the mean of two devices each, with standard deviation. 1x1019 2x1019 3x1019 4x1019 5x1019 6x1019 7x1019 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Po te nt ia l/ V DOS / cm-3 DTHQ/DTBQ HQ/BQ Co2+/3+ ThymHQ/ThymBQ PhHQ/PhBQ Fig. 6. Corrected potential versus density of states (DOS) under dark condi- tions, DOS for DSSCs with Y123 dye in the range of medium to high potentials for the organic electrolytes (278–611mV). Fig. 7. Recombination resistance (Rr) and transport resistance (Rt) versus DOS for DSSCs with Y123 dye using the organic mediator pairs. DOS are the equivalent to the range of medium to high potentials for the organic electro- lytes. Measurements were performed under dark conditions. N. Flores-Díaz et al. Solar Energy 167 (2018) 76–83 81 Acknowledgements The authors are thankful to Centro de Electroquímica y Energía Química (CELEQ), Centro de Investigación en Productos Naturales (CIPRONA), and Vicerrectoría de Investigación, Universidad de Costa Rica (project number 804-B5-271) for financial support. N. F. greatly acknowledges an Orlando Bravo scholarship for master degree by CELEQ, Sistema de Estudios de Posgrado (SEP), Universidad de Costa Rica, and Prof. Anders Hagfeldt, for a research stay at École Polytechnique Fédérale de Lausanne (EPFL), Switzerland. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.solener.2018.03.084. References Aghazada, S., Gao, P., Yella, A., Marotta, G., Moehl, T., Teuscher, J., Moser, J.-E.E., De Angelis, F., Grätzel, M., Nazeeruddin, M.K., 2016. Ligand engineering for the efficient dye-sensitized solar cells with ruthenium sensitizers and cobalt electrolytes. Inorg. Chem. 55, 6653–6659. http://dx.doi.org/10.1021/acs.inorgchem.6b00842. Alligrant, T.M., Hackett, J.C., Alvarez, J.C., 2010. Acid/base and hydrogen bonding ef- fects on the proton-coupled electron transfer of quinones and hydroquinones in acetonitrile: Mechanistic investigation by voltammetry, 1H NMR and computation. Electrochim. Acta 55, 6507–6516. http://dx.doi.org/10.1016/j.electacta.2010.06. 029. Astudillo, P.D., Tiburcio, J., González, F.J., 2007. The role of acids and bases on the electrochemical oxidation of hydroquinone: Hydrogen bonding interactions in acet- onitrile. J. Electroanal. Chem. 604, 57–64. http://dx.doi.org/10.1016/j.jelechem. 2007.02.031. Bella, F., Galliano, S., Gerbaldi, C., Viscardi, G., 2016. Cobalt-based electrolytes for dye- sensitized solar cells: recent advances towards stable devices. Energies 9, 384. http:// dx.doi.org/10.3390/en9050384. Bella, F., Galliano, S., Falco, M., Viscardi, G., Barolo, C., Grätzel, M., Gerbaldi, C., 2017. Approaching truly sustainable solar cells by the use of water and cellulose deriva- tives. Green Chem. http://dx.doi.org/10.1039/C6GC02625G. Bhat, M.A., 2012. Mechanistic, kinetic and electroanalytical aspects of quinone-hydro- quinone redox system in N-alkylimidazolium based room temperature ionic liquids. Electrochim. Acta 81, 275–282. http://dx.doi.org/10.1016/j.electacta.2012.07.059. Bisquert, J., Grätzel, M., Wang, Q., Fabregat-Santiago, F., 2006. Three-channel trans- mission line impedance model for mesoscopic oxide electrodes functionalized with a conductive coating. J. Phys. Chem. B 110, 11284–11290. http://dx.doi.org/10.1021/ jp0611727. Boschloo, G., Hagfeldt, A., 2009. Characteristics of the iodide/triiodide redox mediator in dye-sensitized solar cells. Acc. Chem. Res. 42, 1819–1826. http://dx.doi.org/10. 1021/ar900138m. Cheng, M., Yang, X., Zhang, F., Zhao, J., Sun, L., 2012. Efficient dye-sensitized solar cells based on hydroquinone/benzoquinone as a bioinspired redox couple. Angew. Chem. Int. Ed. Engl. 51, 9896–9899. http://dx.doi.org/10.1002/anie.201205529. Cheng, M., Yang, X.C., Chen, C., Zhao, J.H., Zhang, F.G., Sun, L.C., 2013. Dye-sensitized solar cells based on hydroquinone/benzoquinone as bio-inspired redox couple with different counter electrodes. Phys. Chem. Chem. Phys. 15, 15146–15152. http://dx. doi.org/10.1039/c3cp51980e. Ellis, H., Eriksson, S.K., Feldt, S.M., Gabrielsson, E., Lohse, P.W., Lindblad, R., Sun, L., Rensmo, H., Boschloo, G., Hagfeldt, A., 2013. Linker unit modification of tripheny- lamine-based organic dyes for efficient cobalt mediated dye-sensitized solar cells. J. Phys. Chem. C 117, 21029–21036. http://dx.doi.org/10.1021/jp403619c. Feldt, S.M., Gibson, E.A., Gabrielsson, E., Sun, L., Boschloo, G., Hagfeldt, A., 2010. Design of organic dyes and cobalt polypyridine redox mediators for high-efficiency dye- sensitized solar cells. J. Am. Chem. Soc. 132, 16714–16724. http://dx.doi.org/10. 1021/ja1088869. Freitag, M., Giordano, F., Yang, W., Pazoki, M., Hao, Y., Zietz, B., Grätzel, M., Hagfeldt, A., Boschloo, G., 2016. Copper phenanthroline as a fast and high-performance redox mediator for dye-sensitized solar cells. J. Phys. Chem. C 120, 9595–9603. http://dx. doi.org/10.1021/acs.jpcc.6b01658. Gamboa-Valero, N., Astudillo, P.D., González-Fuentes, M.A., Leyva, M.A., Rosales-Hoz, M.D.J., González, F.J., 2016. Hydrogen bonding complexes in the quinone-hydro- quinone system and the transition to a reversible two-electron transfer mechanism. Electrochim. Acta 188, 602–610. http://dx.doi.org/10.1016/j.electacta.2015.12. 060. Girault, H.H., 2005. Analytical and Physical Electrochemistry. TrAC Trends in Analytical Chemistry, first ed. EPFL Press, Lausanne. http://dx.doi.org/10.1016/j.trac.2005.07. 002. Guin, P.S., Das, S., Mandal, P.C., 2011. Electrochemical reduction of quinones in different media: a review. Int. J. Electrochem. 2011, 1–22. http://dx.doi.org/10.4061/2011/ 816202. Gupta, N., Linschitz, H., 1997. Hydrogen-bonding and protonation effects in electro- chemistry of quinones in aprotic solvents. J. Am. Chem. Soc. 119, 6384–6391. http:// dx.doi.org/10.1021/ja970028j. Ito, S., Liska, P., Comte, P., Charvet, R., Péchy, P., Bach, U., Schmidt-Mende, L., Zakeeruddin, S.M., Kay, A., Nazeeruddin, M.K., Grätzel, M., 2005. Control of dark current in photoelectrochemical (TiO2/I−–I3−)) and dye-sensitized solar cells. Chem. Commun. (Camb) 4351–4353. http://dx.doi.org/10.1039/b505718c. Kalyanasundaram, K., Hagfeldt, A., Boschloo, G., Sun, L., Kloo, L., Pettersson, H., 2010. Dye-sensitized solar cells. Chem. Rev. 110, 6595–6663. http://dx.doi.org/10.1021/ cr900356p. Monge-Romero, I.C., Suárez-Herrera, M.F., 2013. Electrocatalysis of the hydroquinone/ benzoquinone redox couple at platinum electrodes covered by a thin film of poly(3,4- ethylenedioxythiophene). Synth. Met. 175, 36–41. http://dx.doi.org/10.1016/j. synthmet.2013.04.027. Nair, J.R., Porcarelli, L., Bella, F., Gerbaldi, C., 2015. Newly elaborated multipurpose polymer electrolyte encompassing RTILs for smart energy-efficient devices. ACS Appl. Mater. Interfaces 7, 12961–12971. http://dx.doi.org/10.1021/acsami. 5b02729. O’Reagan, B., Grätzel, M., 1991. A low-cost, high-efficiency solar cell based on dye-sen- sitized colloidal TiO2 films. Nature 353, 737–740. http://dx.doi.org/10.1038/ 353737a0. Park, B.W., Pazoki, M., Aitola, K., Jeong, S., Johansson, E.M.J., Hagfeldt, A., Boschloo, G., 2014. Understanding interfacial charge transfer between metallic PEDOT counter electrodes and a cobalt redox shuttle in dye-sensitized solar cells. ACS Appl. Mater. Interfaces 6, 2074–2079. http://dx.doi.org/10.1021/am405108d. Pavlishchuk, V.V., Addison, A.W., 2000. Conversion constants for redox potentials mea- sured versus different reference electrodes in acetonitrile solutions at 25 °C. Inorg. Chim. Acta 298, 97–102. http://dx.doi.org/10.1016/S0020-1693(99)00407-7. Safdari, M., Lohse, P.W., Häggman, L., Frykstrand, S., Högberg, D., Rutland, M., Álvarez, R., Gardner, J., Kloo, L., Hagfeldt, A., Boschloo, G., 2016. Investigation of cobalt Table 4 Photovoltaic parameters of DSSCs obtained with the organic mediator pairs. The dye LEG4 was used as the sensitizer, and PEDOT or PEDOT+graphene as the catalysts in the CE. Redox pair Dye VOC (mV) JSC (mA cm2) Wmax (mW cm2) FF η (%) DTHQ/DTBQ LEG4 524 ± 3 12 ± 2 1.3 ± 0.3 0.29 ± 0.05 1.3 ± 0.3 ThymHQ/ThymBQ LEG4 510 ± 18 12.02 ± 0.08 2.7 ± 0.2 0.44 ± 0.04 2.7 0.2 DTHQ/DTBQa LEG4 474 ± 51 13.4 ± 0.3 2.63 ± 0.06 0.42 ± 0.04 2.69 ± 0.05 ThymHQ/ThymBQa LEG4 510 ± 5 12.3 ± 0.4 2.9 ± 0.3 0.47 ± 0.02 3.19 ± 0.3 J–V data represents the mean of two devices each, with standard deviation. a Cells using CE with PEDOT+graphene. 0.0 0.1 0.2 0.3 0.4 0.5 0 2 4 6 8 10 12 14 C ur re nt de ns ity /m A cm -2 Potential / V DTHQ/DTBQ P DTHQ/DTBQ P+G ThymHQ/ThymBQ P ThymHQ/ThymBQ P+G Fig. 8. J-V curves for DSSCs with organic electrolytes using derivatives ThymHQ and DTHQ sensitized with LEG4 and PEDOT as CE (dashed lines) and hybrid CE with PEDOT and graphene (solid lines). N. Flores-Díaz et al. Solar Energy 167 (2018) 76–83 82 http://dx.doi.org/10.1016/j.solener.2018.03.084 http://dx.doi.org/10.1021/acs.inorgchem.6b00842 http://dx.doi.org/10.1016/j.electacta.2010.06.029 http://dx.doi.org/10.1016/j.electacta.2010.06.029 http://dx.doi.org/10.1016/j.jelechem.2007.02.031 http://dx.doi.org/10.1016/j.jelechem.2007.02.031 http://dx.doi.org/10.3390/en9050384 http://dx.doi.org/10.3390/en9050384 http://dx.doi.org/10.1039/C6GC02625G http://dx.doi.org/10.1016/j.electacta.2012.07.059 http://dx.doi.org/10.1021/jp0611727 http://dx.doi.org/10.1021/jp0611727 http://dx.doi.org/10.1021/ar900138m http://dx.doi.org/10.1021/ar900138m http://dx.doi.org/10.1002/anie.201205529 http://dx.doi.org/10.1039/c3cp51980e http://dx.doi.org/10.1039/c3cp51980e http://dx.doi.org/10.1021/jp403619c http://dx.doi.org/10.1021/ja1088869 http://dx.doi.org/10.1021/ja1088869 http://dx.doi.org/10.1021/acs.jpcc.6b01658 http://dx.doi.org/10.1021/acs.jpcc.6b01658 http://dx.doi.org/10.1016/j.electacta.2015.12.060 http://dx.doi.org/10.1016/j.electacta.2015.12.060 http://dx.doi.org/10.1016/j.trac.2005.07.002 http://dx.doi.org/10.1016/j.trac.2005.07.002 http://dx.doi.org/10.4061/2011/816202 http://dx.doi.org/10.4061/2011/816202 http://dx.doi.org/10.1021/ja970028j http://dx.doi.org/10.1021/ja970028j http://dx.doi.org/10.1039/b505718c http://dx.doi.org/10.1021/cr900356p http://dx.doi.org/10.1021/cr900356p http://dx.doi.org/10.1016/j.synthmet.2013.04.027 http://dx.doi.org/10.1016/j.synthmet.2013.04.027 http://dx.doi.org/10.1021/acsami.5b02729 http://dx.doi.org/10.1021/acsami.5b02729 http://dx.doi.org/10.1038/353737a0 http://dx.doi.org/10.1038/353737a0 http://dx.doi.org/10.1021/am405108d http://dx.doi.org/10.1016/S0020-1693(99)00407-7 redox mediators and effects of TiO2 film topology in dye-sensitized solar cells. RSC Adv. 6, 56580–56588. http://dx.doi.org/10.1039/C6RA07107D. Shaidarova, L.G., Gedmina, A.V., Budnikov, G.K., 2003. Voltammetry of a benzoquinone- hydroquinone redox couple at electrodes modified with a polyvinylpyridine film doped with cobalt phthalocyanine. J. Anal. Chem. 58, 171–175. http://dx.doi.org/ 10.1023/A:1022366307246. Shanti, R., Bella, F., Salim, Y.S., Chee, S.Y., Ramesh, S., Ramesh, K., 2016. Poly(methyl methacrylate-co-butyl acrylate-co-acrylic acid): physico-chemical characterization and targeted dye sensitized solar cell application. Mater. Des. 108, 560–569. http:// dx.doi.org/10.1016/j.matdes.2016.07.021. Staley, P.A., Newell, C.M., Pullman, D.P., Smith, D.K., 2014. The effect of glassy carbon surface oxides in non-aqueous voltammetry: The case of quinones in acetonitrile. Anal. Chem. 86, 10917–10924. http://dx.doi.org/10.1021/ac503176d. Sun, Z., Liang, M., Chen, J., 2015. Kinetics of iodine-free redox shuttles in dye-sensitized solar cells: interfacial recombination and dye regeneration. Acc. Chem. Res. 48, 1541–1550. http://dx.doi.org/10.1021/ar500337g. Teuscher, J., Marchioro, A., Andrès, J., Roch, L.M., Xu, M., Zakeeruddin, S.M., Wang, P., Grätzel, M., Moser, J.E., 2014. Kinetics of the regeneration by iodide of dye sensiti- zers adsorbed on mesoporous Titania. J. Phys. Chem. C 118, 17108–17115. http:// dx.doi.org/10.1021/jp501481c. Tsao, H.N., Burschka, J., Yi, C., Kessler, F., Nazeeruddin, M.K., Grätzel, M., 2011. Influence of the interfacial charge-transfer resistance at the counter electrode in dye- sensitized solar cells employing cobalt redox shuttles. Energy Environ. Sci. 4, 4921. http://dx.doi.org/10.1039/c1ee02389f. Uchimiya, M., Stone, A.T., 2009. Reversible redox chemistry of quinones: Impact on biogeochemical cycles. Chemosphere 77, 451–458. http://dx.doi.org/10.1016/j. chemosphere.2009.07.025. Uematsu, T., Shimomura, E., Torimoto, T., Kuwabata, S., 2016. Evaluation of surface ligands on semiconductor nanoparticle surfaces using electron transfer to redox species. J. Phys. Chem. C. http://dx.doi.org/10.1021/acs.jpcc.5b12698. acs.jpcc. 5b12698. Wang, J.J., Xie, H., Jin, B.K., 2013. Investigation on electrochemical redox of hydro- quinone-fourier transform infrared spectroelectrochemistry techniques. Fenxi Huaxue/ Chinese J. Anal. Chem. 41, 1006–1012. http://dx.doi.org/10.1016/S1872- 2040(13)60667-2. Wang, M., Chamberland, N., Breau, L., Moser, J.-E., Humphry-Baker, R., Marsan, B., Zakeeruddin, S.M., Grätzel, M., 2010. An organic redox electrolyte to rival triiodide/ iodide in dye-sensitized solar cells. Nat. Chem. 2, 385–389. http://dx.doi.org/10. 1038/nchem.610. Wu, K.-L., Huckaba, A.J., Clifford, J.N., Yang, Y.-W., Yella, A., Palomares, E., Grätzel, M., Chi, Y., Nazeeruddin, M.K., 2016. Molecularly engineered Ru(II) sensitizers compa- tible with cobalt(II/III) redox mediators for dye-sensitized solar cells. Inorg. Chem. 55, 7388–7395. http://dx.doi.org/10.1021/acs.inorgchem.6b00427. Yu, Z., Bu, C., Zhou, Z., Liu, Y., Huang, N., Bai, S., Fu, H., Guo, S., Zhao, X., 2013. Effect of HAc treatment on an open-environment prepared organic redox couple based on hydroquinone/benzoquinone and its application in dye-sensitized solar cells. Electrochim. Acta 107, 695–700. http://dx.doi.org/10.1016/j.electacta.2013.06. 125. Yum, J.-H., Baranoff, E., Kessler, F., Moehl, T., Ahmad, S., Bessho, T., Marchioro, A., Ghadiri, E., Moser, J.-E., Yi, C., Nazeeruddin, M.K., Grätzel, M., 2012. A cobalt complex redox shuttle for dye-sensitized solar cells with high open-circuit potentials. Nat. Commun. 3, 631. http://dx.doi.org/10.1038/ncomms1655. Zhang, X., Xu, Y., Giordano, F., Schreier, M.R., Pellet, N., Hu, Y., Yi, C., Robertson, N., Hua, J., Zakeeruddin, S.M., Tian, H., Grätzel, M., 2016. Molecular engineering of potent sensitizers for very efficient light harvesting in thin film solid state dye sen- sitized solar cells. J. Am. Chem. Soc. 138, 10742–10745. http://dx.doi.org/10.1021/ jacs.6b05281. Zhu, X.Q., Wang, C.H., Liang, H., 2010. Scales of oxidation potentials, pKa, and BDE of various hydroquinones and catechols in DMSO. J. Org. Chem. 75, 7240–7257. http:// dx.doi.org/10.1021/jo101455m. N. Flores-Díaz et al. Solar Energy 167 (2018) 76–83 83 http://dx.doi.org/10.1039/C6RA07107D http://dx.doi.org/10.1023/A:1022366307246 http://dx.doi.org/10.1023/A:1022366307246 http://dx.doi.org/10.1016/j.matdes.2016.07.021 http://dx.doi.org/10.1016/j.matdes.2016.07.021 http://dx.doi.org/10.1021/ac503176d http://dx.doi.org/10.1021/ar500337g http://dx.doi.org/10.1021/jp501481c http://dx.doi.org/10.1021/jp501481c http://dx.doi.org/10.1039/c1ee02389f http://dx.doi.org/10.1016/j.chemosphere.2009.07.025 http://dx.doi.org/10.1016/j.chemosphere.2009.07.025 http://dx.doi.org/10.1021/acs.jpcc.5b12698 http://dx.doi.org/10.1021/acs.jpcc.5b12698 http://dx.doi.org/10.1016/S1872-2040(13)60667-2 http://dx.doi.org/10.1016/S1872-2040(13)60667-2 http://dx.doi.org/10.1038/nchem.610 http://dx.doi.org/10.1038/nchem.610 http://dx.doi.org/10.1021/acs.inorgchem.6b00427 http://dx.doi.org/10.1016/j.electacta.2013.06.125 http://dx.doi.org/10.1016/j.electacta.2013.06.125 http://dx.doi.org/10.1038/ncomms1655 http://dx.doi.org/10.1021/jacs.6b05281 http://dx.doi.org/10.1021/jacs.6b05281 http://dx.doi.org/10.1021/jo101455m http://dx.doi.org/10.1021/jo101455m Neutral organic redox pairs based on sterically hindered hydroquinone/benzoquinone derivatives for dye-sensitized solar cells Introduction Experimental Materials Solar cell preparation Electrochemical measurements Solar cell characterization Electrochemical impedance spectroscopy (EIS) Results and discussion Conclusions Acknowledgements Supplementary material References