Cyclopentadithiophene-functionalized Ru(II)-bipyridine sensitizers for dye-sensitized solar cells

Accepted Manuscript Cyclopentadithiophene- Functionalized Ru(II)-Bipyridine Sensitizers for DyeSensitized Solar Cells Maxence Urbani, María Medel, Sangeeta Amit Kumar, Aravind Kumar Chandiran, David González-Rodríguez, Michael Grätzel, Md.K. Nazeeruddin, Tomás Torres PII: DOI: Reference:

S0277-5387(14)00362-3 http://dx.doi.org/10.1016/j.poly.2014.05.045 POLY 10760

To appear in:

Polyhedron

Received Date: Accepted Date:

14 April 2014 20 May 2014

Please cite this article as: M. Urbani, M. Medel, S.A. Kumar, A.K. Chandiran, D. González-Rodríguez, M. Grätzel, Md.K. Nazeeruddin, T. Torres, Cyclopentadithiophene- Functionalized Ru(II)-Bipyridine Sensitizers for DyeSensitized Solar Cells, Polyhedron (2014), doi: http://dx.doi.org/10.1016/j.poly.2014.05.045

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

1

Cyclopentadithiophene- Functionalized Ru(II)-Bipyridine Sensitizers for Dye-Sensitized Solar Cells Maxence Urbani

a,c,┴

, María Medel

a,┴

, Sangeeta Amit Kumar b, Aravind Kumar Chandiran b,

David González-Rodríguez a, Michael Grätzel b, Md. K. Nazeeruddin b,*, Tomás Torres a,c,* a

Universidad Autónoma de Madrid, Departamento de Química Orgánica, Cantoblanco, 28049 Madrid, Spain.

b

Laboratory of Photonics and Interfaces, Institute of Chemical Sciences and Engineering, Swiss Federal Institute of Technology

(EPFL), Station 6, CH 1015 – Lausanne, Switzerland. c

Instituto Madrileño de Estudios Avanzados (IMDEA)-Nanociencia, c/ Faraday,9, Cantoblanco, 28049 Madrid (Spain).

┴

Both authors contributed equally to this work.

AUTHOR INFORMATION Corresponding Authors *Email: [email protected] *Email: [email protected] Author Contributions ┴

M. Urbani and M. Medel contributed equally to this work.

2

ABSTRACT Two ruthenium sensitizers TT-230 and TT-232 based on cyclopenta[2,1-b:3,4b']dithiophene (CDT) were tested in dye-sensitized solar cells (DSSCs) using a tri-iodide/iodide redox couple, and their performances compared to the benchmark Z907. Both dyes of general formula Ru(LL’)(NCS)2 feature the same 4,4’-dicarboxylic acid 2,2’-bipyridine moiety acting as the anchoring group (L ) to attach the dye on the TiO2 surface, and two CDT subunits as ancillary groups (L’) either directly-linked (TT-230) or vinylene-linked (TT-232) to the bipyridine moiety. In comparison with TT-230, the π-extended conjugation in TT-232 caused significant redshift and broadening of the absorption bands. However, despite better light-harvesting properties, the DSSC device sensitized with TT-232 strikingly underperformed compared to TT-230 under same conditions (PCE= 3.2% and 6.1%, respectively). The use of co-adsorbent (CHENO) or surfactants did not help to improve the photovoltaic properties of TT-230, and resulted in a degradation of the overall efficiency of the cell.

TOC GRAPHIC

KEYWORDS: Dye-sensitized solar cells, Ruthenium dyes, Cyclopentadithiophene

3

1. Introduction New emerging classes of electroactive π-conjugated materials based on thiophene-, bithiophene- or polythiophene- fused ring, have recently attracted growing attention, such as thieno[3,2-b]thiophene (ADT),

(TT),

naphthodithiophenes

dithieno[3,2-b:20,30-d]thiophene (NDTs),

benzo

[1,2-b:4,5-b’]

(DTT),

anthradithiophenes

dithiophene

(BDT)

or

cyclopentadithiophene (CDT), to cite few examples. They rose research interests in many different fields as conducting materials in organic electronics, organic field-effect transistors (OFETs) [1-3], small molecules and solution-processable π-conjugated polymers or co-polymers [4,5] in bulk heterojunction solar cells [6-16]. Extremely interesting and versatile building blocks, their rigid conjugation, tunable optical and electronic properties, make them promising candidates in the field of photovoltaic. Remarkably, the ease of introduction of solubilizing chains during their synthesis allows avoiding aggregation and the preparation of solution processable dyes. Cyclopentadithiophene (CDT) [17-19] is one these emerging classes of molecules that have been successfully used for photovoltaic applications in BHJ [20], polymer solar cell [21,22] and in DSSC as metal-free organic [23-25] or ruthenium dyes [26,27], showing relatively high values of conversions efficiencies over 8%. In our previous work [27], we reported a CDT-based ruthenium dye (TT-230), which in conjunction with a cobalt-based electrolyte, attained high open circuit voltage (VOC) in DSSC. However, despite high VOC, a poor short-circuit current (JSC) was obtained, which resulted in a PCE of only 1.8% under AM 1.5 standard conditions. The main reason for this poor JSC was ascribed to an inefficient dyeregeneration due to a too small overpotential between the cobalt redox couple [Co(bipy)]3+/2+ and Dye+/Dye. We report herein the performances of the dye TT-230 in DSSC using a conventional tri-iodine/iodine-based liquid electrolyte, achieving much higher solar-to-electric power conversion (PCE) over 6%, together with those of a new analogue TT-232 that contain vinylene linkages between CDT units and the bipyridine ruthenium complex.

4

Figure 1. Molecular structures of ruthenium dyes studied in this work: 1) TT-230, 2) TT-232, and 3) benchmark Z907.

2. Experimental

2.1. Reagents and material.

Synthetic procedures were carried out under an inert argon atmosphere, in dry solvent unless otherwise noted. All reagents and solvents were reagent grade and used as received without further purification unless otherwise specified. 4,4’-dicarboxylic acid 2,2’-bipyridine (“dcapy”) and dichloro(p-cymene)-ruthenium(II) dimer were purchased at TCI, 4,4’-dimethyl2,2’-bipyridine at Fluka, ammonium thiocyanate and lithium diisopropyl amide (2M solution in heptane/THF ) at Aldrich. The synthesis and characterization of compound 1 and dye TT-230 were previously reported by us elsewhere in the literature [27]. THF was freshly distilled from sodium benzophenoneketyl under argon prior to use. Chromatographic purifications were performed using silica gel 60 SDS (particle size 0.040-0.063 mm) and/or size-exclusion chromatography (Bio-rad Bio-beds, SX-1). Analytical thin-layer chromatography was performed using Merck TLC silica gel 60 F254. 1H (300 MHz) and 1H-decoupled

13

C NMR (75 MHz)

spectra were obtained on Bruker TopSpin AV-300 spectrometers. Chemical shifts are reported in parts per million (ppm) from the residual signal peak of the deuterated solvent: δ = 7.27 ppm (CDCl3) and δ = 3.31 ppm (MeOD-[d4]) for 1H NMR, and δ = 77.0 ppm (CDCl3) for 13C NMR. Multiplicities are given as: s (singlet), d (doublet), t (triplet), q (quartet), dd (doublet of doublets), m (multiplet), and the coupling constants, J, are given in Hz. UV/Vis measurements were carried

5

out on a Shimadzu UV 3600 spectrophotometer. For extinction coefficient determination, solutions of different concentration were prepared in CH2Cl2 with absorption between 0.5-1, and measured using a 1 cm cuvette (QS). 2.1.1 Synthesis of compound 2.

POCl3 (0.22 ml, 3.65 mmol) was added with a syringe under argon at 0°C to a solution of 1 (580 mg, 1.26 mmol) and DMF (0.27 ml, 2.40 mmol) in DCE (20 mL). Afterwards, the mixture was let to warm to RT, and then heated to reflux overnight. After cooling to RT, a saturated solution of AcOK in H2O was poured into the mixture. After 30 min of stirring, the mixture was extracted with CH2Cl2. After phase separation, the aqueous layer was extracted twice more with CH2Cl2. The combined organic layers were washed with water, dried over MgSO4, filtered, and then the solvents evaporated to dryness. The resulting crude brown-reddish oil was purified by chromatography column (SiO2; Hexane/AcOEt mixtures: gradient eluent from 10:2 to 8:1) affording 2 (470 mg, 0.96 mmol) in 73% yield as an orange solid. UV/Vis (CH2Cl2): λmax/nm (ε/dm3 mol–1cm–1): 399 (199 500). 1H NMR (300 MHz, CDCl3, 25°C) δ (ppm): 9.77 (s, 1H, CHO), 7.51 (s, 1H, ArH), 6.68 (s, 1H, ArH), 2.81 (t, 2H, 3J= 7 Hz, CH2(α) hexyl chain), 1.951.73 (m, 4H, 2CH2(α) ethylhexyl chain), 1.72-1.54 (m, 2H, 2 C*H ethylhexyl chain), 1.41-1.14 (m, 6H, 3 CH2), 1.03-0.47 (m, 33H, 9CH2 + 5CH3). 13C NMR (75 MHz, CDCl3, 25°C) δ (ppm): 182.0 (CHO), 161.8 (3 signals), 156.3 (3s), 151.2, 148.7, 141.9, 133.1, 130.6, 119.9, 53.6, 42.9, 35.0, 34.1, 33.9, 31.6, 31.4, 30.9, 28.5, 28.4, 28.3, 27.3, 27.1, 22.6, 22.5, 22.4, 13.9, 10.5 (2s). MS (FAB+) m/z (%): 515.3 (100) [M+H]+. HRMS (FAB+) m/z: calcd for C32H51OS2: 515.3381 [M+H]+; found: 515.3369. 2.1.2 Synthesis of compound 3 (racemic mixtures of alcohols).

Commercially available 4,4’-dimethyl-2,2’-bipyridine (50 mg, 0.27 mmol) was introduced in an oven dried 50 mL Schlenk flask, and then three cycles of vacuum-argon were realized. Afterwards, THF (5 mL) was introduced under argon with a syringe, and the resulting solution stirred at room temperature until the complete dissolution of the material. Next, the mixture was cooled at –78°C for 10 min, and then a 2 M commercial solution of LDA in heptane/THF (0.27 ml, 0.54 mmol) was added with a syringe to the colorless solution, which

6

turned brown immediately. The solution was let to reach -60°C within 30 min, and then warmed at -10°C for additional 25 min (during this period of time the solution progressively turned from brown to deep dark black-reddish). The solution was cooled again at -78°C, and then a solution of aldehyde 2 (351 mg, 0.68mmol) in THF (5 mL) was transferred under argon in the Schlenk flask with a cannula (the solution progressively turned bright orange). At the end of the addition, the mixture was stirred at -60°C for 2h, and then warmed at -10°C for additional 30 min. Finally, few drops of an aqueous saturated NH4Cl solution were added to quench the reaction. After 10 min of stirring, the reaction mixture was diluted in CH2Cl2 (40 mL), successively dried over Na2SO4, filtered through a small plug of Celite, and then the solvents evaporated to dryness under reduced pressure. The crude orange oil was purified by flash chromatography column on SiO2 using first CH2Cl2 100% as eluent until the disappearance of staring material (aldehyde 2), then CH2Cl2/MeOH/Et3N (100:5:1) to afford 3 in 89% yield (288 mg, 0.24 mmol) as a bright yellowish viscous solid. As 3 was obtained as complex racemic mixtures of alcohols, characterization was done on the next compound, see next step (compound 4). 2.1.3 Synthesis of ligand 4.

A solution of 3 (288 mg, 0.24 mmol), PPTS (50 mg, 0.20 mmol) and a catalytic amount of I2 (~2 mol%) in dry toluene (20 mL) was refluxed in a Dean-Stark apparatus for 36h under Ar. After cooling to room temperature, Et3N (≈ 0.5 mL) was added to the solution, and then the solvents evaporated to dryness under reduced pressure. The crude was purified by successive chromatography columns (SiO2; eluent: CH2Cl2/MeOH/Et3N, 100:0.5:0.1), followed by gel permeation chromatography (eluent: toluene). Purification was achieved by trituration in MeOH (twice) to afford 4 in 50% yield (141 mg, 0.12 mmol) as a dark brown-yellowish viscous solid. UV/Vis (CH2Cl2): λmax/nm (ε/dm3 mol–1cm–1): 282 (25 860), 431 (60 650). 1H NMR (300 MHz, CDCl3, 25°C) δ (ppm): 8.63 (d, 3J= 5 Hz, 2H, 2ArH bipy), 8.50 (s, 2H, 2ArH bipy), 7.59 (d, 3

Jtrans= 16 Hz, 2H, 2C=CH), 7.31 (d, 3J= 5 Hz, 2H, 2ArH bipy), 7.00 (s, 2H, 2ArH CDT), 6.84 (d,

3

Jtrans= 16 Hz, 2H, 2C=CH), 6.66 (s, 2H, 2ArH CDT), 2.84 (t, 3J= 7 Hz, 4H, 2CH2(α) hexyl

chains), 1.94-1.78 (m, 8H, 4CH2(α) ethylhexyl chains), 1.75-1.61 (m, 4H, 4C*H ethylhexyl chains), 1.47-0.52 (m, 78H, 24CH2 + 10CH3). 13C NMR (75 MHz, CDCl3, 25°C) δ (ppm): 158.4 (2 signals), 156.6 (3s), 156.4, 149.4, 147.5(3s), 145.9, 140.6 (3s), 139.0 (2s), 133.9, 127.7(3s),

7

123.4, 122.2 (2s), 120.5, 119.8, 117.4, 53.7, 43.2(3s), 35.1, 34.3, 34.2, 31.9, 31.6, 30.9 (2s) 29.7, 28.7, 28.6, 28.5, 27.4, 27.3, 22.8, 22.8, 22.6, 14.1, 10.7. MS (MALDI-TOF): m/z (%): 1177.8 (100) [M+H]+. HRMS (MALDI-TOF) m/z: calcd for C76H109N2S4: 1177.7468 [M+H]+; found: 1177.7465. 2.1.4 Synthesis of ruthenium dye TT-232.

TT-232 was synthesized according to general procedures described in the literature by us [27] or others [28], from ligand 4 (19.0 mg, 16.1 µmol), dichloro(p-cymene)ruthenium(II) dimer (5.94 mg, 9.70 µmol), dcapy (5.67 mg, 23.2 µmol), and NH4NCS (50 mg, 657 µmol). After removal of DMF under high-vacuum distillation, the remaining pasty solid was triturated in a copious amount of an aqueous 0.1 M HCl solution. The remaining fine black suspension was successively filtered-off, air-dried, and then washed with a copious amount of an aqueous 0.1 M HCl solution, followed by a minimum amount of cold MeOH. The filtrates were discarded, and the remaining solid was recovered by dissolution in CH2Cl2/MeOH mixtures. After evaporation of solvents to dryness, the crude complex was purified on gel permeation chromatography column (eluent: THF/MeOH mixtures). Purification was achieved by trituration in hexanes to afford TT-232 (18.6 mg, 11.3 µmol) in 70% yield as a dark black powder. UV/Vis (CH2Cl2): λmax/nm (ε/dm3mol–1cm–1): 318 (26 000), 473 (47 700), 582 (br sh, 24 100). 1H NMR (300 MHz, CDCl3 +MeOD [d4], 25°C) δ (ppm): 9.71 (br d, 3J~ 5 Hz, 1H, ArH bipy), 9.23 (d, 3J= 6 Hz, 1H, ArH bipy), 8.85 (s, 1H, ArH bipy), 8.70 (s, 1H, ArH bipy), 8.25 (s, 1H, ArH bipy), 8.15 (br s, 1H, ArH bipy), 8.10 (s, 1H, ArH bipy), 7.81-7.65 (m, 2H, 2ArH bipy), 7.65-7.49 (m, 3H, ArH bipy+2ArH CDT), 7.27-7.07 (m, 3H, ArH bipy+2C=CH), 6.92-6.50 (m, 5H, ArH bipy+2ArH CDT+2C=CH) , 2.93-2.73 (m, 4H, 2CH2(α) hexyl chains), 2.08-1.77 (m, 8H, 4CH2(α) ethylhexyl chains), 1.77-1.61 (br s, 4H, 4C*H ethylhexyl chains), 1.45-0.46 (m, 78H, 24CH2+10CH3). MS (MALDI-TOF, DCTB) m/z (%): found 1581.6 (40) [M-(NCS)]+, 1638.6 (100) [M]+. HRMS (MALDI-TOF) m/z (%): calc for RuC90H116N6O4S6: 1638.6441 (100) [M]+; found 1638.6402 (100) .

8

2.2. Photoanode preparation

A transparent fluorine-doped tin oxide conducting glass (NSG10, Japan) was cleaned using ethanol and water followed by an ultrasonic cleaning in DeconnexTM solution for 30 min. The electrodes were then washed with water and ethanol. To remove the organics, a further thermal treatment was done at 500 °C for 30 min. The clean FTO glass was treated twice with TiCl4 (40 mM, 30 min, 75 °C). Two different TiO2 pastes (transparent layer and scattering layer) were screen printed on to the TiCl4 pretreated electrode and followed a series sintering step as described elsewhere in the literature [29]. A 9 µm thickness for the transparent layer (20 nm particles) and another 5 µm for the scattering layer (400 nm particles) were estimated. The photoanodes were further treated with TiCl4 following the steps described above. The TiCl4 post treatment was done to ensure better electronic contact between nanoparticles.

2.3. Device fabrication

The TiO2 electrodes were fired at 500°C for 30 min prior to the sensitization in the dye solution (acetonitrile / t-butanol 1:1 mixture). Surfactants or co-adsorbent (CHENO) were added to ethanolic solutions of the dye to study their effect. The films were then immersed into a 300 mM solution of the dye in ethanol. After 16 hours of dipping, the sensitized electrodes were washed with acetonitrile and ethanol for 30 min to remove the loosely-bounded dye molecules, and then air-dried. The counter electrodes were made by drop casting of ethanolic solution of H2PtCl6 (10 mM) onto a pre-cleaned FTO glass (TEC7, Solaronix, Switzerland), and then the electrodes were fired at 410 °C for 20 min. The sensitized photoanode and counter electrode were melt sealed using a 25 µm polymeric spacer (SurlynTM, Dupont, USA). The electrolyte in acetonitrile was then injected by the vacuum backfilling process through a hole drilled at the side of the counter electrode. The cell fabrication was completed by melt sealing the hole with a glass and soldering the metal solder to make contacts. For these solar cells, the electrolyte used consisted of 1 M dimethyl-methyl imidazolium iodide (DMII), 0.05 M lithium iodide, 0.03 M iodine, 0.05 M guanidinium thiocyanate (GNCS) and 0.5 M of 4-tert-butylpyridine, in a mixture of acetonitrile/butylnitrile.

9

2.4. Photovoltaic characterization.

A 450W xenon lamp (Oriel, USA) was used as a light source. The spectral output of the lamp was filtered using a Schott K113 Tempax sunlight filter (Präzisions Glas & Optik GmbH, Germany) to reduce the mismatch between the simulated and actual solar spectrum to less than 2%. The current-voltage characteristics of the cell were recorded with a Keithley model 2400 digital source meter (Keithley, USA). The photo-active area of 0.159 cm2 was defined by a black metal mask. The amount of the dyes adsorbed on TiO2 films were estimated by measuring the absorbance of desorbed dye solution in DMF and calibrated absorption spectra of each dye. The electron recombination in the mesoporous film was studied by transient photo-voltage measurements [30]. The white light was generated by an array of LED’s while a pulsed red light (0.05 s square pulse width) was controlled by a fast solid-state switch to ascertain rapid submillisecond rise of light perturbation. The voltage decay was recorded on a mac-interfaced Keithley 2602 source meter [31].

10

3. Results and Discussion

3.1. Synthesis The synthesis and characterization of compound 1 and TT-230 were previously reported by us elsewhere in the literature [27]. The synthetic approach to prepare the ruthenium dye TT232 is shown in Scheme 1. The first step is a Mc Murry formylation of 1 using an excess of DMF and POCl3 at reflux in 1,2-dichloro ethane (DCE) overnight to afford, after hydrolysis, aldehyde 2. The second step is a double deprotonation/lithiation of dimethyl 2,2’-bipyridine (1eq) with LDA (2eq), in THF at low temperature. Subsequent quenching of the in-situ generated bis-anion with an excess of aldehyde 2 afforded, after hydrolysis, the functionalized (bis)hydroxy bipyridine 3 in 88% yield, obtained as racemic mixtures of alcohols. Finally, bipyridine 4 was obtained under smooth conditions using pyridinium p-toluene sulfonate salt (PPTS) as a dehydrating agent, in refluxing toluene for 48h. The presence of a catalytic amount of iodine was found to be necessary to ensure the formation of all-trans type double bonds. We finally obtained the ruthenium heteroleptic complex TT-232 in a one-pot, three-step reaction, according to general procedures described in the literature [28], and as previously described by us [27] for the synthesis of TT-230.

11 Scheme 1. Synthesis of TT-232 (DMF = N,N-dimethylformamide, dcapy = 4,4’-dicarboxylic acid 2,2’-bipyridine, PPTS= pyridinium p-toluene sulfonate salt).

3.2. Spectral properties The UV-visible spectra of dyes TT230 and TT-232 in solution (CH2Cl2) are depicted in Figure 2, and absorption data summed up in Table 1. Both dyes display there distinct bands: the first one in the UV region centered at around 320 nm was attributed to the intraligand chargetransfer transition of the dcapy moiety (π→π∗ transition, ILCT1); the main band centered at 439 nm (TT-230) or 473 nm (TT-232) was attributed to the overlap between the intraligand chargetransfer band of the ancillary group bipy(CDT)2 moiety (π→π∗ transition, ILCT2) with major contribution, and the first metal-to-ligand charge-transfer transition (MLCT) of Ru(II) heteroleptic complex with minor contribution. The third band (shoulder) at around 580 nm was attributed to the second MLCT transition. Both dyes display strong absorption, especially in the 400-700 nm visible range with relatively similar and high absorption coefficients up to 50 000 mol-1.cm-1.L. Remarkably, the extension of π-conjugation between the two CDT moieties and bipyridine unit trough the vinylene linkage in TT-232 causes important changes of the main absorption band (ILCT2+MLCT1), with a maximum redshifted by 34 nm and extended to the red region of about 80 nm , with respect to TT-230.

Figure 2. UV-visible spectra of TT-230 (grey line) and TT-232 (black line and circles) in CH2Cl2 solutions.

12

Table 1. Electronic absorption data of dyes in CH2Cl2 solution Dye

Band

λmax (nm)

ε (dm3 mol–1 cm–1)

TT-230

ILCT1 ILCT2a+MLCT1 MLCT2

317 439 581 (br sh)

25 500 51 100 19 500

TT-232

ILCT1 ILCT2a+MLCT1 MLCT2

318 473 582 (br sh)

26 000 47 700 24 100

a

Main contribution is coming from the ILCT2 band ILCT1= π−π* dcapy intraligand charge-transfer transition ILCT2= π−π* bipy(CDT)2 intraligand charge-transfer transition MLCT= Metal-to-Ligand Charge-Transfer transition (dcapy= 4,4’-dicarboxylic acid 2,2’-bipyridine) (CDT= cyclopentadithiophene)

3.3. Photovoltaic properties

The photovoltaic properties and dye-loading for each DSSC devices made with sensitizers TT-230, TT-232, or benchmark Z907 are summed up in Table 2. The J-V characteristics of TT230 DSSC under 10%, 50% and full-sun are depicted in Figure 3. Both TT-230 and TT-232 reached similar dye-coverage on the TiO2 surface. Hence, considering better optical properties for TT-232 it can be assumed a superior light-harvesting efficiency (LHE) than for TT-230. However, all individual photovoltaic parameters (JSC, VOC, FF) and hence the overall PCE were found to be much lower for TT-232 than TT-230. Assuming superior LHE for TT-232 than for TT-230 and dye-loading surface coverage likewise similar for both dyes, the difference in the JSC (7.4 vs 12.3 mA.cm-2, respectively) can be directly ascribed to a much lower adsorbed photon-tocurrent efficiency for TT-232 (APCE = φing ⨉ ηc = φing ⨉ ηcoll ⨉ ηreg ; with φing: electroninjection efficiency; ηc: electron-collection efficiency at the counter electrode, being dependent upon both the efficiency of the photo-generated charge carriers (ηcoll) and efficiency of dyeregeneration (ηreg)). The presence of the double-bond linkages of the bipy(CDT)2 moiety (ancillary group) in TT-232 should perturb and downshift significantly the LUMO levels of the dye, which should disfavor the electron-injection of the excited dye in the TiO2 conduction band. As the HOMO level is mostly delocalized over the Ru center and NCS moieties in N3-type analogues [32], we expect only few difference in regeneration of the two dyes TT-230 and TT232. Moreover, given the same bulky substitution for TT-230 and TT-232, we expect none or very little dye aggregation, which should disable any dye-dye* deactivation channel. The very

13

low tendency of these dyes to undergo aggregation was also observed experimentally: no aggregation phenomena were observed in solution even at high concentration of dye, and neither detected by spectroscopic methods (RMN and UV-VIS spectroscopies). As illustrated in Figure 4, the IPCE values are slightly lower for TT-230 compared to those of Z907 despite superior light-harvesting efficiency for the former. In comparison with the benchmark Z907, the better light-harvesting efficiency of TT-230 must compensate lower electron-injection efficiency, resulting in similar JSC. Considering JSC and fill factor most likely similar, the 44 mV loss in the VOC, from 699 mV for benchmark Z907 to 655 mV for TT-230, mostly account for the difference observed in their overall PCE (η= 6.1% and η= 7.0%, respectively). This seems to point out that, despite having a bulkier substitution, electron recombination rates between photoinjected electrons and tri-iodine ions occur faster for TT-230 than for Z907, most probably due to a less efficient dye-packing of the bulkier dyes on the TiO2 surface. The use of chenodeoxycholic acid “CHENO” as co-adsorbent33 into the dye solutions during the adsorption of the photoanode did not help to improve the performances of the cells, but on the contrary, strong decreases were observed in both cases. This witness that dyeaggregation should be marginal in the DSSCs, and the use of CHENO only results in lower dyeloading that decrease the absolute amount of photo-injected electrons, and hence the JSC. Next, we examined the effect of two surfactants CTAB (cationic) or Triton (neutral) on the performances of TT-230 sensitized DSSC [34]. The role of these surfactants is to increase the surface energy and prevent aggregation. The CTAB cationic surfactant did not change the JSC, but decreased the VOC. On the contrary, the neutral Triton (neutral) did not change significantly the VOC but decreased the JSC. Addition of both types of surfactants resulted in a decrease of the performance of DSSCs, and also witness that dye-aggregation should be negligible for TT-230. Transition photovoltage experiments [30] (Figure 5) shows the evolution of recombination rates and evidenced deep trap states for TT-230 cell. The recombination rate is quite low for the studied sensitizer, which can be ascribed to the steric hindrance of the bulky group in TT-230 groups that should hamper the approach of tri-iodine ions to the TiO2 surface. However, our results tend to confirm that despite having a bulkier substitution, electron recombination between photoinjected electrons and tri-iodine ions is larger for TT-230 (as well as TT-232) than for Z907, most probably because of a less efficient dye-packing of the bulkier dyes on the TiO2 surface, resulting in lower VOC. Thus, lower electron-injection efficiency and larger

14

recombination than the benchmark Z907 are both responsible of the lower performances achieved by TT-230 and TT-232 DSSCs. Table 2. Photovoltaic parametersa of DSSCs measured under AM1.5G solar irradiance (100 mW/cm2) and amount of adsorbed dye on the TiO2 surface. Dye

Coadsorbent

Dye-loading -2 (⨉108 mol. cm )

JSC (mA/cm2)

VOC (mV)

F.F. (%)

Ƞ (%)

TT-230

None CHENO

9.51 (N.A.)

12.3 9.0

654.8 619.2

76 78

6.1 4.3

TT-232

None CHENO

8.78 (N.A.)

7.4 2.7

603.2 505.1

72 76

3.2 1.0

Z-907

None

(N.A.)

12.8

698.7

78

7.0

a The values reported in Table 2 were obtained for the best devices in each configuration. Two cells were made for each condition, and the power conversion efficiency errors are within 5% of the presented values.

Figure 3. Current−voltage curves of TT-230 dye-sensitized solar cell under 10% (lower trace), 50% (middle trace) and full sun (upper trace) illumination (AM 1.5 G solar irradiance, 100 mW/cm2 photon flux).

15

Table 3. Effect of the surfactant on the photovoltaic performances of dyes TT-230 and Z907 measured at AM1.5G solar irradiance (100 mW/cm2) under different conditions. Dye

TT-230

Z907

Surfactant

JSC (mA/cm2)

VOC (mV)

F.F. (%)

Ƞ (%)

None

12.4

654.8

76

6.1

(+) Ctab

12.4

622.0

79

4.7

(0) Triton

11.4

648.8

78

5.7

None

12.8

698.7

78

6.8

(+) Ctab

11.9

686.9

77

6.2

(0) Triton

12.1

690.6

78

6.5

Figure 4. IPCE action spectra of the DSSCs sensitized with dyes TT-230 (▲), TT-232 (●), or benchmark Z907 (o).

(a)

(b)

Figure 5. Transition photovoltage experiments for TT-230 DSSC: (a) recombination rate of the photogenerated electrons measured as a function of the open circuit voltage; (b) open-circuit voltage measured as a function of density of states.

16

4. Conclusion

Ruthenium sensitizers TT-230 and TT-232 based on CDT were tested in DSSCs using a standard tri-iodide/iodide-based electrolyte, and their performances compared to those of the benchmark Z907. Both dyes differ in the type of linkages of their ancillary group, either directlylinked (TT-230) or vinylene-linked (TT-232). TT-230 sensitized cell achieved a maximum solarto-electric power conversion (PCE) of 6.1 % under standard AM 1.5 G standard conditions, while TT-232 achieved a PCE of only 3.2% under same experimental conditions despite improved light-harvesting properties; namely, broader and redshift absorption. Lower electron-injection efficiency and larger recombination at the electrolyte/TiO2 interface, are believed to be the main factors responsible of the decrease in PCE of the DSSCs in the order Z907 (η= 7.0%) > TT-230 (η= 6.1%) >> TT-232 (η= 3.2%). The use co-adsorbent (CHENO) or surfactants did not help to improve the photovoltaic properties of TT-230, and resulted in a degradation of the overall efficiency of the cell.

AUTHOR INFORMATION Corresponding Author *Email: [email protected] *Email: [email protected] Author Contributions ┴ M. Urbani and M. Medel contributed equally to this work. Notes The authors declare no competing financial interests.

ACKNOWLEDGMENT Financial support is acknowledged from European Union within the FP7-ENERGY-2012-1 framework, GLOBALSOL project, Proposal No 309194-2, the “ORION” grant no. NMP229036, the Spanish MEC and MICINN (CTQ2011-24187/BQU, CTQ2011-23659) and PRIPIBUS-2011-1128).

17 References [1] M. Zhang, H. N. Tsao, W. Pisula, C. Yang, A. K. Mishra, K. Müllen, J. Am. Chem. Soc. 129 (2007) 3472. [2] S. Shinamura, I. Osaka, E. Miyazaki, A. Nakao, M. Yamagishi, J. Takeya , K. Takimiya, J. Am. Chem. Soc. 133 (2011) 5024. [3] Z. Chen, M. J. Lee, R. Shahid Ashraf, Y. Gu, S. AlbertSeifried, M. Meedom Nielsen, B. Schroeder, T. D. Anthopoulos, M. Heeney, I. Mc Culloch, H. Sirringhaus, Adv. Mat. 24 (2012) 647. [4] L. Huo, J. Hou, S. Zhang, H.-Y. Chen, Y. Yang, Angew. Chem., Int. Ed. 49 (2010) 1500. [5] Z. B. Henson, G. C. Welch, T. van der Poll, G. C. Bazan, J. Am. Chem. Soc. 134 (2012) 3766. [6] Y.-J. Cheng, S.-H. Yang, C.-S. Hsu, Chem. Rev. 109 (2009) 5868. [7] P. M. Beaujuge, J. M. J. Fréchet, J. Am. Chem. Soc. 133 (2011) 20009. [8] H. Zhou, L. Yang, A. C. Stuart, S. C. Price, S. Liu, W. You, Angew. Chem., Int. Ed. 50 (2011) 2995. [9] P. P. Khlyabich, B. Burkhart, B. C. Thompson, J. Am. Chem. Soc. 134 (2012) 9074. [10] Y. Liang, D. Feng, Y. Wu, S.-T. Tsai, G. Li, C. Ray, L. Yu, J. Am. Chem. Soc. 131 (2009) 7792. [11] L. Huo, Y. Huang, B. Fan, X. Guo, Y. Jing, M. Zhang, Y. Li, J. Hou, Chem. Commun. 48 (2012) 3318. [12] Y.-J. Cheng, S.-W. Cheng, C.-Y. Chang, W.-S. Kao, M.-H. Liao, C.-S. Hsu, Chem. Commun. 48 (2012) 3203. [13] D. Lee, S. W. Stone, J. P. Ferraris, Chem. Commun. 47 (2011) 10987. [14] I. Osaka, T. Abe, S. Shinamura, K. Takimiya, J. Am. Chem. Soc. 133 (2011) 6852. [15] J. Hou, M.-H. Park, S. Zhang, Y. Yao, L.-M. Chen, J.-H. Li, Y. Yang, Macromolecules 41 (2008) 6012. [16] H. J. Son, W. Wang, T. Xu, Y. Liang, Y. Wu, G. Li, L. Yu, J. Am. Chem. Soc. 133 (2011) 1885. [17] P. Coppo, M. L. Turner, J. Mat. Chem. 15 (2005) 1123. [18] P. Coppo, D. C. Cupertino, S. G. Yeates, M. L. Turner, Macromolecules 36 (2003) 2705. [19] P. Coppo, H. Adams, D. C. Cupertino, S. G. Yeates, M. L. Turner, Chem. Commun. (2003) 2548. [20] A. J. Moulé, A. Tsami, T. W. Bünnagell, M. Forster, N. M. Kronenberg, M. Scharber, M. Koppe, M. Morana, C. J. Brabec, K. Meerholz, U. Scherf, Chem. Mat. 20 (2008) 4045. [21] K.-C. Li, J.-H. Huang, Y.-C. Hsu, P.-J. Huang, C.-W. Chu, J.T. s. Lin, K.-C. Ho, K.-H. Wei, H.-C. Lin, Macromolecules 42 (2009) 3681. [22] Z. Zhu, D. Waller, R. Gaudiana, M. Morana, D. Mühlbacher, M. Scharber, C. Brabec, Macromolecules 40 (2007) 1981. [23] H. N. Tsao , C. Yi , T. Moehl, J.-H. Yum, S. M. Zakeeruddin, M. K. Nazeeruddin, M. Grätzel, ChemSusChem 4 (2011) 591. [24] D. Zhou, N. Cai, H. Long, M. Zhang, Y. Wang, P. Wang, J. Phys. Chem. C 115 (2011) 3163. [25] X. Cheng, S. Sun, M. Liang, Y. Shi, Z. Sun, S. Xue, Dyes Pigm. 92 (2012) 1292. [26] J.-J. Kim, K. Lim, H. Choi, S. Fan, M.-S. Kang, G. Gao, H. S. Kang, J. Ko, Inorg. Chem. 49 (2010) 8351.

[27] S. Amit Kumar, M. Urbani, M. Medel, M. Ince, D. GonzalezRodriguez, A. K. Chandiran, A. N. Bhaskarwar, T. Torres, M. K. Nazeeruddin, M. Grätzel, J. Phys. Chem. Lett. 5 (2014) 501. [28] K. Willinger, K. Fischer, R. Kisselev, M. Thelakkat, J. Mat. Chem. 19 (2009) 5364. [29] S. Ito, T. N. Murakami, P. Comte, P. Liska, C. Grätzel, M. K. Nazeeruddin, M. Grätzel, Thin Solid Films 516 (2008) 4613. [30] M. Wang, J. Bai, F. Le Formal, S.-J. Moon, L. Cevey-Ha, R. Humphry-Baker, C. Grätzel, S. M. Zakeeruddin, M. Grätzel, J. Phys. Chem. C 116 (2012) 3266. [31] L. M. Peter, Phys. Chem. Chem. Phys. 9 (2007) 2630. [32] A. Reynal, A. Forneli, E. Martinez-Ferrero, A. Sánchez-Díaz, A. Vidal-Ferran, B. C. O'Regan, E. Palomares, J. Am. Chem. Soc. 130 (2008) 13558. [33] J.-H. Yum, S.-r. Jang, R. Humphry-Baker, M. Grätzel, J.-J. Cid, T. Torres, M. K. Nazeeruddin, Langmuir 24 (2008) 5636. [34] A. I. Maldonado-Valdivia, E. G. Galindo, M. J. Ariza, M. J. García-Salinas, Solar Energy 91 (2013) 263.