Scholarly article on topic 'Succinonitrile-based solid-state electrolytes for dye-sensitised solar cells'

Succinonitrile-based solid-state electrolytes for dye-sensitised solar cells Academic research paper on "Materials engineering"

0
0
Share paper
Academic journal
Prog. Photovolt: Res. Appl.
OECD Field of science
Keywords
{""}

Academic research paper on topic "Succinonitrile-based solid-state electrolytes for dye-sensitised solar cells"

Progress in

P H OTOVO LTAI CS

PROGRESS IN PHOTOVOLTAICS: RESEARCH AND APPLICATIONS

Prog. Photovolt: Res. Appl. (2013) Published online in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/pip.2441

RESEARCH ARTICLE

Succinonitrile-based solid-state electrolytes for dye-sensitised solar cells

Owen Byrne, Aoife Coughlan, Praveen K. Surolia and K. Ravindranathan Thampi*

Schoolof Chemical& Bioprocess Engineering, University College Dublin, Belfield, Dublin 4, Ireland

ABSTRACT

Succinonitrile (SCN), a solid ion conductor (10~4 to 10~3 S/cm) in solid form at room temperature, is mixed with either 1,2-dimethyl-3-propylimidazoliuum iodide or 1-butyl-3-methyl imidazolium iodide ionic liquids for forming a solid plastic phase electrolyte for use in dye-sensitised solar cell (DSSC). Cells containing these two electrolytes showed best energy conversion efficiencies of 6.3% and 5.6%, respectively. The commonly used DSSC electrolyte additives inhibit the formation of the SCN plastic phase. However, for the first time, an SCN-additive (additive = guanidinium thiocyanate) electrolyte composition is reported here, which remains as a solid at room temperatures. By using these new solid electrolytes, a simple and rapid single-step filling procedure for making solid-state DSSC is outlined. This process, which reduces the required manufacturing steps from four to one, is most suitable for continuous, high-throughput, commercial DSSC manufacturing lines. These new electrolytes have been tested under low incident light levels (200 lx) to investigate their suitability for indoor DSSC applications. Copyright © 2013 John Wiley & Sons, Ltd.

KEYWORDS

DSSC; DSC; dye-sensitised solar cell; solid electrolyte; succinonitrile ^Correspondence

K. Ravindranathan Thampi, Schoolof Chemical& Bioprocess Engineering, University College Dublin, Belfield, Dublin 4, Ireland. E-mail: ravindranathan.thampi@ucd.ie

Received 17 December 2012; Revised 30 August 2013; Accepted 17 October 2013

1. INTRODUCTION

A typical dye-sensitised solar cell (DSSC) consists of a TiO2 layer electrode chemisorbed with a monolayer of dye molecules to absorb the visible light spectrum. Incoming sunlight is absorbed by the dye generating excited electrons, which are then injected into the conduction band of the TiO2. The electrons are routed through an external circuit and a counter electrode to an electrolyte containing a suitable redox species (typically, iodide/tri-iodide couple). The residual oxidised dye cations are instantaneously regenerated by the redox species. Thus, the electrolyte is an essential component performing charge transport between the two electrodes and timely dye regeneration as outlined in Figure 1 [1,2].

The best performing liquid electrolytes to date, achieve >11% efficiency with the iodide/tri-iodide redox couple [3,4] in acetonitrile/valeronitrile solvents. The present DSSC record of 12.3% efficiency involves a Co(II/III)tris (bipyridyl) tetracyanoborate complex as the redox couple, also in acetonitrile [5]. Recently, water-based Co-complex electrolytes have been once again proposed after years of work trying to move away from the traditional water-based iodide electrolytes [6]. A major reason for DSSC

operational failure in practical applications is due to leakages of liquid electrolyte caused by seal rupture. Liquid-based electrolytes not only leak through sealant materials and cause seal failure but also freeze at low temperatures as to be expected with water and some other solvents. Furthermore, sealant failure can be caused due to the large vapour pressures exerted by volatile organic solvents such as acetonitrile or by the anomalous expansion of H2O within the cell. For the advancement of DSSC lifetime guarantee, both of these issues need to be resolved. In view of this, various solvent-free, gel/quasi-solid and solid-state electrolytes have been tested as DSSC electrolytes, including: solvent-free eutectic melt of ionic liquids (conversion efficiency, n = 8.2%), poly(acrylonitrile-co-vinyl acetate) based gels (n = 9.03% and 9.46%), quasi-solid polyvinylidene difluoride based gel (n = 6.7%), solid-state hole transport materials such as the p-type inorganic hole conductors CsSnI3 (n = 3.72%), copper iodide (n = 4.7%) and the organic hole conductor 2,2'7,7'-tetrakis-(N,N-di-p-methoxyphenyl-amine)-9,9'-spirobifluorene (spiro-MeOTAD) (n = 6%), and p-type conductive polymers such as poly(3,4-e thylenedioxythiophene) (PEDOT) (n = 6.1%) as listed in Table I [3-5,7-16]. Very recently [17], a new record efficiency of 15.0% was reported by replacing the dye

and poly(acrylonitrile-co-vinyl acetate) gels], difficult handling conditions, requirement of spin-coating techniques, which involve significant material wastage (e.g. spiro-MeOTAD), extra process steps [additional electrode treatment (e.g. CuI)] and improper pore filling of the mesoporous TiO2 layers. The fluorine doped CsSnI2 95F0 05 hole conductor mixed with SnF2 has been reported to achieve 9.28% efficiency, but the procedure requires both additional pre-treatments of the TiO2 electrode with fluorine plasma and drying of the cell following infiltration of the solid electrolyte dissolved in an organic solvent. A guaranteed removal of solvent from devices introduces lengthy manufacturing and quality control procedures.

A solid-state material, which solves the problems of electrolyte leakage and seal degradation by highly volatile solvents, is succinonitrile (SCN). SCN is a white-clear plastic crystal possessing characteristics of good conductivity and flexibility. It is a solid at room temperature enabling easy handling under ambient conditions. Plastic crystalline materials such as SCN exhibit rotational disorder, displaying multiple phase transitions, some of

Table I. Some of the best dye-sensitised solar cellperformances reported for cells containing liquid, solid and quasi-solid

phase electrolytes.*

Electrolyte Sensitiser n (%) Voc (mV) jsc (mA/cm2) FF ref

Liquid AY1:[Co"(bpy)3][B(CN)4]2 (0.165 M), [Com(bpy)3HB(CN)4]3 (0.045 M), TBP (0.8 M), LiClO4 (0.1 M) in ACN solvent YD2-o-C8 + Y123 12.3 935 17.66 74 [5]

DMPII (0.6 M), LiI (0.1 M), I2 (0.05 M), Black dye 11.1 736 20.9 72.2 [3]

TBP (0.5 M) in ACN

E1: BMII (0.6M), I2 (0.03M), GSCN N719 9.82 820 16.0 74.5 [4]

(0.1 M), TBP (0.5 M) in ACN : valeronitrile 12.2 832 20.1 73.1

(volratio: 85:15) [no mask]

Solvent-free: DMII/EMII/EMITCB/I2/NBB/GSCN Z907Na 8.2 741 14.26 77.4 [7]

(Molar ratio: 12;12;16;1.67;3.33;0.67)

Gel/quasi LiI (0.1 M), I2 (0.05 M), TBP (0.5 M), DMPII N719 9.03 797 15.44 73 [8]

solid (0.5 M) in ACN + poly(acrylonitrile-co-vinyl acetate)

LiI (0.1 M), I2 (0.05 M), TBP (0.5 M), DMPII N719 9.46 794 16.23 73 [8]

(0.5 M) in ACN + poly(acrylonitrile-

co-vinyl acetate)

+ TiO2 filler (10wt%)

PMII (0.6 M), I2 (0.1 M), NMBI (0.45 M) in MPN Z907 6.7 749 13.1 68.1 [9]

+ PVDF-HFP (5wt%)

Solid Spiro-MeOTAD doped with Co(III) complex CH3NH3PbI3 15.0 993 20.0 73 [17]

Pi,4I/I2/succinonitrile (5:1:100) (Plasma-treated N719 9.37 763 17.1 71.4 [10]

TiO2 electrode, two-step filling procedure)

CsSnI2.95F0.05 doped with 5% SnF2 N719 9.28 730 17.4 72.9 [11]

(plasma-treated TiO2 electrode)

PEDOT + Li salt/propylene carbonate solution D149 6.1 860 9.3 75 [12]

PEDOT doped with LiTFSI, MPII, TBP N719 5.4 640 14.2 60 [13]

Spiro-MeOTAD (0.17 M), TBP (0.11 mM), C220 6.08 860 10.90 69 [14]

LiN(CF3SO2)2 (0.21 mM ) in chlorobenzene

Spiro-OMeTAD Z907 4 860 9.1 51 [15]

CuI (MgO-coated TiO2 electrode) N3 4.7 620 13.0 58 [16]

*See text and references for explanation of abbreviations.

eV -2-10 1

Liquid Electrolyte RED OX

RED -« ► OX e" iSw OR —

T o_o_<\o_o_o_

HOMO HoPP'"» Regeneration Qye Solid Electrolyte

Working Electrode

Pt Counter Electrode

Figure 1. Operationaloutline of dye-sensitised solarcell. Undesired recombination processes are displayed as red dashed arrows.

molecule light harvester with an organic-inorganic hybrid perovskite sensitiser (CH3NH3PbI3) coupled with spiro-MeOTAD hole transporter doped with a Co (III) complex as 'electrolyte'. However, many of these materials show disadvantages such as the presence of volatile organic solvents [e.g. polyvinylidene difluoride

which lead to local rotatory motions in the crystal lattice. These rotatory motions cause lattice defect formation, which are thought to facilitate high molecular/ionic mobility [18]. SCN exhibits a body-centred cubic plastic phase from -40°C to its melting point at 58°C [19]. In this phase, molecules are rotationally disordered in a mixture of isomers (two gauche isomers and one trans isomer) inverted around the central C-C bond [19]. The availability of the three conformers in succinonitrile creates mono-vacancies in the lattice where the trans isomers act as 'impurities' and leads to high molecular diffusivity. It has been shown that plastic crystals mixed with lithium compounds can result in room temperature waxy solids, which act as a solid-state solvent or 'matrix' for the Li ions exhibiting conductivities as high as 2x10—4 S/cm at 60°C, making these materials very attractive for battery applications [20]. SCN is a non-ionic and highly polar plastic-crystalline organic molecule. High polarity enables it to dissolve various types of salts showing ionic conductivity that originates solely from the guest salt in an otherwise non-ionic matrix [21]. Room temperature conductivities up to 3.4 x 10—4 S/cm are obtained with 5 wt% lithium bis(trifluoromethanesulfonylamide) in SCN [22], and a proton-conducting solid electrolyte with conductivities above 10—3 S/cm was prepared by incorporating a non-aqueous liquid electrolyte into the plastic crystalline phase of SCN [23]. A series of organic cations containing quaternary ammonium were tested as replacements for metal cations [18]. SCN is reported as a suitable matrix for quaternary ammonium based iodides (tetraethylammonium iodide, tetrabutylammonium iodide, N-dimethyl, N-propyl, N-butylammonium iodide, ethyl methyl imidazolium iodide, dimethylimidazolium iodide and ethyl methyl pyrrolidinium iodide) and iodine-yielding waxy solids in the temperature range —40 to 60°C. The resultant materials showed that high I— and I— transport is possible in the solid-state with ionic conductivities up to 3 x 10—3 S/cm. The composition of the electrolyte, both in terms of total iodide/iodine and the relative ratio of these, has a major effect on the final diffusivities and ionic conductivities [18].

Succinonitrile possesses characteristics similar to those that make acetonitrile and glutaronitrile suitable solvents for DSSC [18]. An electrolyte formed by mixing SCN with N-methyl-N-butyl pyrrolidinium iodide (P1j4I) and iodine was reported in DSSC with efficiencies exceeding 6.7% (at 52mW/cm2), which dropped to approximately 5% under full sunlight [24]. At 25°C, this plastic crystal electrolyte has a conductivity of 3.3 x 10—3 S/cm. The DSSC performance of the same electrolyte was recently improved to 6.54% in full sunlight and 7.93% at 30 mW/ cm2 using electrospun hierarchically structured TiO2 nanofiber electrodes [25]. A detailed study employing the same electrolyte (P14I/I2/SCN) also achieved efficiency of 6.54% under full sunlight using a new two-step electrolyte infiltration process, where first the liquid-state electrolyte (E1) containing 0.03 M iodine, 0.1 M guanidinium thiocyanate, 0.5 M 4-ieri-butylpyridine and

0.6 M 1-butyl-3-methyl imidazolium iodide (BMII) in acetonitrile: valeronitrile solution (85:15 by volume) was injected into the TiO2 film and dried at 80°C for 12 h. The high boiling point additives remained entrapped in the TiO2 pores while most of the solvent was evaporated. After that, the plastic SCN electrolyte was injected into the cell at 80°C. The cell was then cooled to room temperature to obtain a waxy solid electrolyte with the presence of additives showing performance of 6.54%. Cell performance was further improved by fluorine plasma treatment of the TiO2 electrode before cell fabrication. This treatment was found to greatly improve infiltration of the SCN electrolyte by increasing the size of nanopores and nanochannels in the TiO2 electrode and reducing the recombination rate as a result of surface passivation of the TiO2 nanoparticle surface. Coupled with the two-step electrolyte infiltration process, efficiencies of 9.37% are possible; the best reported for an SCN-based solid electrolyte. Addition of 3-D photonic crystals behind the DSSC to redirect light, not absorbed by the cell in the first pass, back through the active layers improves efficiency to 10.5%. These additional steps make the cell fabrication procedures lengthy and expensive.

In order for DSSC to compete commercially with other PV technologies, manufacturing line speeds of 2 to >20 m/ min are thought to be necessary [26]. Even though, DSSC has become a highly active area of research and development as evidenced by the fast growth of scientific publications and patent applications over the last 15-20 years [27], manufacturing of small laboratory cells is commonly performed by hand without any concern for time or cost optimization. Simple procedures such as dyeing and electrolyte filling are time consuming and at the cell level all steps must be done rapidly and efficiently for producing DSSC modules at commercial scales [28]. Accordingly, industrial processes involving rapid dyeing (for example, in boiling or hot dye baths) are now available to the industry. However, complicated additional steps such as lengthy waiting periods and electrode plasma treatment will again increase DSSC processing times and reduce the cost advantages associated with high speed manufacturing, as envisaged in roll-to-roll manufacturing steps. Prototypes of glass facade elements (70x200 cm2) consisting of several serially interconnected DSSC modules each of size 30 x 30 cm2 showcased a few typical facade applications of DSSC such as decorative semitransparent glass panels [29]. Recently, the Fraunhofer Institute for Solar Energy Systems developed and fully up-scaled a glass frit-sealed DSSC fabrication procedure for manufacturing modules of size 60 x 100 cm2 on a single substrate. Cell and panel manufacturing up-scaled to such attractive areas for the building-integrated photovoltaics market is vital for DSSC industrialization [30].

To the best of our knowledge, iodides other than P14I and lithium iodide (LiI) have not been reported in SCN-based DSSC cells, even though a range of ammonium iodides and imidazolium iodides are known to show high I— and I— transport in solid-state matrices [18].

The use of common DSSC electrolyte components such as the imidazolium salts, namely 1,2-dimethyl-3-pro-pyl imidazolium iodide (DMPII) and 1-butyl-3-methyl imidazolium iodide (BMII), in the presence of SCN is another option requiring immediate investigation. Similarly, the effects of common electrolyte additives [4-tert-butylpyridine (TBP), LiI, N-methyl benzimidazole (NMBI) and guanidinium thiocyanate (GSCN)], which are beneficial to the overall performance of the cell, on the performance of the SCN electrolyte host require clarity.

Therefore, the focus of our current study is threefold:

• Investigate performance of solid SCN electrolytes incorporating iodide salts other than previously reported P1,4I and LiI.

• Study the effect of common DSSC liquid electrolyte additives (TBP, NMBI, LiI and GSCN) on the performance of solid-state SCN electrolytes (open-circuit voltage, short-circuit current and inhibition of the SCN plastic phase)

• Reduce the required DSSC fabrication steps to allow ready application of SCN-based electrolytes in industrial manufacturing process lines.

All electrolyte performances are generally tested under standard 1 sun (1000 W/m2), AM1.5 illumination to simulate outdoor conditions. We additionally report here for the first time testing of SCN electrolytes in low, fluorescent light (200 lx) conditions for applications of solid-state DSSC in indoor and sensor device applications. Finally, we propose a quick and simple one-step electrolyte filling process ideal for use of these solid electrolytes in high-throughput manufacturing process lines.

2. EXPERIMENTAL

TiO2 electrodes were manufactured using the screen-printing method [32], which is based on a layer-by-layer deposition of TiO2 on a fluorine doped tin oxide (FTO) conducting transparent glass substrate. In all cases, a non-porous dense blocking under-layer of TiO2 was first deposited on the FTO substrate via TiCl4 treatment [33] in order to reduce charge recombination. TiO2 paste was then printed on the TiCl4 treated glass using a TIFLEX Ltd manual screen-printer and involved several cycles. After each TiO2 paste deposition, the films were kept in an ethanol chamber for 6min followed by drying at 125°C for 6min, while the final sintering involved gradual heating in an oven at 325°C (5min) followed successively by 375°C (5min), 450°C (15 min) and finally at 500°C for 30min. After this sintering step, the electrodes were again treated with TiCl4 followed by one more sintering at 500°C for 30 min. The TiO2 active area was 0.28 cm2 (consisting of a 6 mm diameter circular spot). The sintered electrodes were placed in a dye bath of N719 [di-tetra butyl ammonium cis-bis (isothiocyanato)bis(2,2'-bipyridyl-4,4'-dicarboxylato)ruthe-nium(II)] dissolved in acetonitrile: tert-butylalcohol: THF

(vol 4.5:4.5:1) for 16-20 h. Alternatively, a rapid dyeing procedure in hot dye bath is possible as practised in industries. N719 dye was procured from a commercial supplier (Dyesol Ltd) and used without further purification. The counter electrode was prepared with a thin film of Pt catalyst deposited via a drop of H2PtCl6 solution (2 mg Pt in 1 mL ethanol) and heat treated at 400°C for 15 min to remove the solvent. The dye-coated TiO2 electrode and Pt-coated counter electrode were sandwiched together and sealed using a Bynel polymer gasket (50 micron thick). Electrolyte was filled into the space between the two electrodes through a hole in the counter electrode via vacuum backfilling. The back hole was sealed with a thin piece (0.1 mm thick) of glass, heat sealed with Bynel.

TiO2 printing pastes employed were a 'transparent' paste containing TiO2 particles of an average size of 20 nm, formed from P25 powder using a standard fabrication procedure [34]. Ethyl cellulose (Fluka, #46080 and #46070) and anhydrous Terpineol (Sigma-Aldrich, 86480) were used as received for making this paste. A 'scattering' paste of particle size 150-250 nm was purchased from DyeSol Ltd (WER 2-0). Both pastes were screen-printed with a 90 T mesh to yield two electrode configurations used in this study:

• 7 + 2 (7 transparent layers+ 2 scattering layers, total thickness = 15-16 micron)

• 6 + 0 (6 transparent layers + 0 scattering layers, total thickness = 10 micron)

The optimum reported liquid electrolyte used is labelled E1 and composed of 0.03 M I2, 0.1 M GSCN, 0.5 M TBP and 0.6 M BMII in acetonitrile : valeronitrile solution (85:15 by volume) [34]. SCN-based electrolytes were made by heating the SCN host on a hot plate until it becomes a liquid at approximately 80°C. At this temperature, the following iodide salts: DMPII, BMII and LiI were added with stirring. I2 was then added to yield three solidstate compositions in the molar ratio 5:1:100 (X-I:I2:SCN). Electrolytes were vacuum backfilled into the DSSC cell in their liquid state at 80°C. The following materials were used as received; BMII (Merck 4.90187.0100), DMPII (Merck 4.94800.0025), GSCN (Merck 8.20613.0250), TBP (Sigma-Aldrich 142379-25G), NMBI (Sigma-Aldrich 399353-25G), LiI (Sigma-Aldrich 518018-10G), acetonitrile (Sigma-Aldrich 271004-1L), valeronitrile (Sigma-Aldrich 155098-100G) and SCN (Sigma-Aldrich 160962-25G).

Electro-optical characterization including current/voltage (I-V) curves, open circuit voltage (VOC), short circuit current density (JSC) and fill factor (FF) were studied using a Newport 91195A-1000 solar simulator and Newport 69920 Arc Lamp Power Supply (Newport Oriel, Stratford, Connecticut, CT, USA) and recorded with a GAMRY Instruments Potentiostat (Gamry Instruments, Warminster, Pennsylvania, PA, USA). A Newport 81088A Air Mass Filter was placed before the output of the solar simulator to simulate AM 1.5 spectrum with irradiance powers of 1000 W/m2. Low intensity I-V curves were also measured

with a fluorescent bulb as light source with an illuminance of 200 lx to simulate indoor lighting conditions.

3. RESULTS AND DISCUSSION

3.1. Succinonitrile-based solid electrolytes

The SCN plastic crystal host was infiltrated with DMPII, BMII or Lil and iodine in the molar ratio 5:1:100 (X-I:I2: SCN), as described in the experimental section. All three compositions are solid at room temperature. The performance of these electrolytes was tested in fully fabricated DSSC cells using our 7 + 2 working electrode configuration. The results from DSSCs fabricated with standard acetonitrile/valeronitrile-based liquid electrolyte (E1) are also presented for comparison in Figure 2 and Table II.

0.0 0.2 0.4 0.6 0.8

Photovoltage (V)

Figure 2. J-V curves for dye-sensitised solar cellusing different electrolyte materials measured under standard AM1.5 conditions.

Table II. Photovoltaic parameters of dye-sensitised solar cell devices made with the listed electrolytes at 1 sun (1000 Wm— ) incident intensity of AM1.5 simulated solar light.

Electrolyte Jsc (mA/cm2) Voc (V) FF n (%)

Liquid (E1) 18.1 0.815 57 8.47

DMPII:I2:SCN 15.6 0.707 57 6.33

BMII:I2:SCN 13.4 0.729 58 5.62

LiI:I2:SCN 14.6 0.489 50 3.58

DMPII, 1,2-dimethyl-3-propyl imidazolium iodide; I2, iodine; SCN, succinonitrile;

BMII, 1-butyl-3-methyl imidazolium iodide; LiI, lithium iodide.

*Electrolyte compositions: E1 is liquid phase and composed of 0.03 M I2, 0.1 M guanidinium thiocyanate, 0.5 M 4-tert-butylpyridine and 0.6 M BMII in acetonitrile : valeronitrile solution (85:15 by volume). Solid-state electrolytes consist of DMPII, BMII or LiI dispersed with I2 in SCN in the molar ratio 5:1:100 (X-I:I2:SCN).

The liquid-based electrolyte achieves an efficiency of 8.47%, JSC =18.1 mA/cm2 and a particularly attractive VOC = 815mV. The DSSC formed using the SCN-based electrolyte with DMPII showed the best result among the solid electrolytes tested, achieving a best efficiency of 6.33% [average of four devices, n = 5.31% (ESI Table S1)]. This best result compares quite favourably to the P1j4I doped SCN performance of 6.54% [10], exhibiting almost the same VOC values (707 vs 711 mV), slightly better JSC (15.6 vs 14.2 mA/cm2) but with a lower FF (57 vs 64.7%). We believe that FF can be improved by rigorous optimization such as the introduction of extra process steps, e.g. fluorine plasma treatment of the TiO2 electrode[10], but our aim in this paper is to keep process steps to a minimum (Section 3.3). In a large manufacturing line, such additional steps may become practical after appropriate machines and methods are developed.

A solid electrolyte based on BMII was also tested. This is a more suitable comparison to the liquid electrolyte, which also contains the same imidazolium salt. Average performance for four devices was n = 5.33% (ESI Table S1). Similar shape and fill factor are observed when compared with the liquid electrolyte, but the best performing cell exhibited much reduced VOC and JSC values, compared with the liquid E1 electrolyte, resulting in a lower efficiency of 5.62%. The incident photon-to-current conversion efficiency (IPCE) spectra of these best performing DMPII and BMII-based SCN electrolytes are compared with E1 in ESI (Figure S1). All three show similar shapes with a maximum response of 69% (E1), 58% (BMII) and 63% (DMPII) observed at 525 nm. Much like their J-V curves, the shape of the BMII:I2:SCN and BMII containing liquid E1 electrolyte are almost exactly the same [ESI Figure S1(B)], although the IPCE at Xmax is much lower for the solid-state system, which leads to the observed lower photocurrent. Solid-state DMPII:I2:SCN, on the other hand, shows slightly lesser IPCE response in the ~350-450nm region [ESI Figure S1(B)], but its overall greater IPCE contribution compared with BMII:I2:SCN leads to the observed larger photocurrent values.

Finally, a LiI-based solid electrolyte was also fabricated, and efficiencies of 3.58% were measured. This low performance is due to a significant drop in cell VOC. The VOC recorded was 489 mV, which is far below the voltages obtained in the reference liquid electrolyte. The adsorption of the Li cation causes a positive shift in the conduction band energy of the titanium oxide, decreasing VOC [31]. The size of the shift is attributed to the charge-to radius ratio of the cation, i.e. the larger counter-cation, the larger the VOC, if the charge remains the same [35,36]. As Li ions have a relatively small radius, the VOC is decreased significantly. Furthermore, the small Li+ ions can intercalculate into the dye-sensitised TiO2 films thereby increasing recombination and further lowering VOC [31]. However, a relatively good performance in JSC of 14.6 mA/cm2 is observed.

Previous LiI-based SCN electrolytes are reported to achieve an efficiency of 3.92% in a DSSC with the use

of TBP additive (JSC = 8.76mA/cm2, VOC = 611mV, FF = 73, n =3.92%). However, the TBP inhibited the formation of SCN plastic phase, and additional silica nanoparticles were required to solidify the electrolyte for achieving a reported 3.81% efficiency in solid-state [37] (Jsc = 8.51 mA/cm2, VOC = 633 mV, FF = 71).

3.2. Succinonitrile electrolyte with additives

In an effort to compare the E1 liquid electrolyte containing BMII, I2, GSCN and TBP with a similar SCN-based solid electrolyte, the effects of various additives on the performance of the BMII:I2:SCN solid were compared. Additives frequently employed in liquid electrolytes to improve the different aspects of DSSC performance are LiI, GSCN, TBP and NMBI. Their effects are summarised in Table III [31]. For this set of testing, a 6 + 0 working electrode configuration was employed. Additives were added to the BMII: I2:SCN electrolyte with stirring at 80°C to yield a molar ratio of 5:1:5:100 (BMII:I2: Additive:SCN).

All additives inhibited formation of the plastic phase of SCN by reducing its melting point and forming room temperature liquids in all cases except for the BMII:I2: GSCN:SCN combination, which remains as a solid at room temperature (Table IV). It is known that the melting point of SCN is lowered by the addition of salts [21], and hence the results here are not surprising. A very low residual energy is required for the plastic-crystal phase to enter a more disordered state. Vacancies could account for several percent of the SCN's volume, but as the concentration of solutes increase, a tendency towards a partial loss of plastic-crystal local order occurs and eventually results in a complete inhibition of the plastic phase formation itself [21]. Decreased melting temperatures are noted with TBP, NMBI and

LiI, which resulted in room temperature liquid formation. The addition of GSCN also exhibits a lowering of melting point as expected, but it still yields a room temperature solid. To the best of our knowledge, this is the first SCN-based electrolyte to remain solid upon the addition of an extra additive. SCN consists of two CH2CN moieties both involving a C triple bonded to N connected by a central C-C bond [NCCH2-CH2CN]. When additives are added to the system, a tendency towards inhibition of the plastic phase is observed due to the disordering of the SCN molecules in the plastic phase. The thiocyanate anion exhibits structural similarities to the SCN molecule, containing a C triple bonded to N. It is possible that the nitrogen lone pair of the thiocyanate anion is interacting with the SCN host molecules, inducing some stabilisation of the crystal order, and thereby comprising a slightly higher melting temperature than the other additive materials.

The additives TBP and NMBI performed as expected, increasing the cell voltage from 682 to 747 mV and 718 mV, respectively as displayed in Table IV and Figure 3a. However, this increase in voltage is associated with a decrease in cell current to JSC values of 9.09 and 8.28 mA/ cm2, respectively. Of these two additives, TBP performed better, which results in a higher VOC and JSC. Although VOC is increased in both cases, due to the decrease in cell current, a drop in overall efficiencies is observed from 4.84% (no additive) to 4.35% (TBP) and 3.57% (NMBI).

The LiI additive yielded a remarkable increase in photocurrent. A JSC value of 16.0 mA/cm2 was observed, much greater than the BMII analogue with no additives. This large current increase can be due to Li cations being co-adsorbed onto TiO2 surface, which increases electron injection yield and thereby the cell current. However, a low voltage of 536 mV is observed. Coupled

Table III. Summary of the effects of different electrolyte additives on dye-sensitised solar cellperformance[31].

4-tert-butylpyridine (TBP) and

Lithium iodide Guanidinium thiocyanate N-methylbenz-imidazole

Additive (LiI) (GSCN) (NMBI)

Increase /sc Increase /sc

Effect Reduce VOC Increase VOC Increase VOC

Comment Li+ adsorbed onto TiO2 surface: Guanidinium cations Negative shift of TiO2 conduction band:

adsorbed onto TiO2 surface:

^ increase ^ increase electron injection ^ increase VOC

electron injection yield (increase current) Longer electron lifetimes in CB

yield (increase current)

^ positive shift of TiO2 ^ self assembly of compact ^ slower electron

conduction band dye layer - slower electron recombination rate

recombination rate

(decrease VOC) (increase VOC) (increase VOC)

Li ions intercalculate into

dye-sensitised TiO2 films

^ increase recombination

(decrease VOC)

Table IV. Photovoltaic parameters of dye-sensitised solar celldevices made with 1-butyl-3-methylimidazolium iodide : iodine : succinonitrile electrolyte and our 6 + 0 working electrode configuration with different additives at 1 sun (1000 Wm~ ) incident intensity of AM1.5 simulated solar light. Their

phase at room temperature and 10°C is also listed.

Electrolyte Jsc (mA/cm2) Voc (V) FF n(%) Phase at room temp. Phase at 10°C

Liquid (E1) 11.8 0.720 72 6.24 Liquid Liquid

BMII:I2:SCN 12.2 0.682 58 4.84 Solid Solid

BMII:I2:LiI:SCN 16.0 0.536 44 3.75 Liquid Solid

BMII:I2:GSCN:SCN 12.8 0.758 54 5.22 Solid Solid

BMII:I2:TBP:SCN 9.09 0.747 64 4.35 Liquid Solid

BMII:I2:NMBI:SCN 8.28 0.718 60 3.57 Liquid Solid

BMII, 1-butyl-3-methyl imidazolium iodide; I2, iodine; SCN, succinonitrile; LiI, lithium iodide; GSCN, guanidinium thiocyanate; TBP, 4-tert-butylpyridine; NMBI, N-methyl benzimidazole.

*Electrolyte compositions: E1 is liquid phase and composed of 0.03 M I2, 0.1 M GSCN, 0.5 M TBP and 0.6 M BMII in acetonitrile:valeronitrile solution (85:15 by volume). SCN-based electrolytes consist of: BMII dispersed with I2 in SCN with the listed additive (LiI, GSCN, TBP or NMBI) in the molar ratio 5:1:5:100 (BMII : I2 : Additive : SCN).

with a poor FF, a decrease in overall cell performance is noticed with an energy conversion efficiency of 3.75%. Li ions can cause a positive shift of TiO2 conduction band, decreasing VOC [31]. Also, the small Li ions can intercalculate into the dye-sensitised TiO2 films increasing recombination also lowering VOC. Furthermore, the presence of I— in the LiI additive can change the concentration of I—/I2 in the electrolyte (compared with non-iodide containing TBP, NMBI and GSCN additives). This can change the poly-iodide distribution in the cell and alter the redox potential leading to higher or lower VOC values. IPCE spectra were measured, and it did not show any significant change in the iodide absorption spectral region (ESI Figure S2), but this effect cannot be ruled out entirely, and it likely also affects the VOC. In any case, the LiI addition showed poor cell performance as it resulted in a room temperature liquid due to inhibition of the SCN plastic phase formation. So, it was not pursued any further.

The GSCN additive resulted in a room temperature solid (Table IV). To the best of our knowledge, this is the first SCN/additive combination reported to form a room temperature solid without any additional solid-ifiers. The guanidinium cations act much like the Li cation in that they are co-adsorbed onto TiO2 surface and increase the electron injection yield. Furthermore, guanidinium cations co-adsorbed onto the TiO2 surface alongside the dye anions allow for screening of the lateral Coulombic repulsion of the sensitiser [38]. As a result, a compact dye monolayer can be formed without molecular aggregation, and the dark current systematically reduced. This causes a considerable increase in the VOC. Consequently, compared with the non-additive BMII analogue, an increase in VOC from 682 to 758 mV is observed. Unlike TBP and NMBI, there is no associated decrease in current. The trend of decreasing photocurrent as one progress from LiI to GSCN to TBP to NMBI additives (Table IV) is in agreement with the measured IPCE spectra (ESI Figure S2). The magnitude of IPCE is significantly lower for the

TBP and NMBI additives, which is in agreement with their observed lower photocurrents. Although exhibiting

e 0) Q

■С CL

£ 1— 3

■Lil(1) GSCN(2) No Additive(3) TBP(4) NMBI(5)

0.2 0.4 0.6

Photovoltage (V)

Solid (BMII)(3) Solid (BMII+GSCN)(2) Liquid (BMII+GSCN+TBP)(1)\

"o2~ 0.4 0.6

Photovoltage (V)

Figure 3. (a) J-V curves for dye-sensitised solar cellusing BMII: I2:SCN electrolyte with different additives. (b) J-V curves for dye-sensitised solar cellusing BMII: I2 based electrolytes and our best TiO2 electrode configuration.

Table V. Photovoltaic parameters of the devices made with 1-butyl-3-methyl imidazolium iodide : iodine electrolyte with different additives and/or solvents and our best 7 + 2 working electrode configuration at 1 sun (1000Wm" ) incident intensity of AM1.5

simulated solar light.

Electrolyte Jsc (mA/cm2) Voc (V) FF n (%)

Liquid (E1) (BMII:I2:GSCN:TBP) 18.1 0.815 57 8.47

Solid (BMII:I2:GSCN:SCN) 15.5 0.804 48 5.94

Solid (BMII:I2:SCN) 13.4 0.729 58 5.62

BMII, 1-butyl-3-methyl imidazolium iodide; I2, iodine; GSCN, guanidinium thiocyanate; TBP, 4-tert-butylpyridine; SCN, succinonitrile. *Electrolyte compositions: E1 is liquid phase and composed of 0.03 M I2, 0.1 M GSCN, 0.5 M TBP and 0.6 M BMII in acetonitrile:valeronitrile solution (85:15 by volume). SCN-based electrolytes consist of: BMII dispersed with I2 in SCN in the molar ratio 5:1:100 (BMII: I2 : SCN) and with GSCN additive in the molar ratio 5:1:5:100 (BMII: I2 : GSCN : SCN).

a poor FF of 54% with GSCN additive cell efficiency increases considerably from 4.84% to 5.22%.

The best combination of additive (GSCN) with BMII doped SCN was tested using our best 7 + 2 working electrode. Compared with the BMII:I2:SCN electrolyte with no additive, cell VOC improves considerably from 729 to 804 mV and is now comparable with that achieved with the liquid electrolyte (815 mV) as displayed in Table V and Figure 3b. There is no associated decrease in JSC density, either. However, the 15.5mA/cm2 JSC is still lower than that of the liquid electrolyte (18.1 mA/cm2). GSCN produces promising photocurrents and photovoltages, although the poor FF (48%) translates them into a lower efficiency of 5.94%. The low FF may be inherent to this solid electrolyte and could be related to poor pore filling of the thicker (15 micron) TiO2 electrode. Improvements of this parameter would increase cell performance considerably. For example, with a relatively conservative FF of 58% and the present VOC and JSC values (804 mV, 15.5 mA/cm2), a remarkable efficiency of 7.2% would be expected. Greater pore infiltration can be achieved by fluorine plasma treatment of the TiO2 electrode [10], which increases the size of nanopores in the TiO2 electrode; however, this would involve an additional fabrication step and was not pursued here.

Finally, of utmost interest is the cell durability/stability results of these electrolytes. Because of the low melting point of the SCN solid host (mp = 58°C), they are expected to be most suitable for indoor DSSC applications (Section 3.4). In this regard, we present the long term stability of the best performing cells after a period of 122 days stored in indoor, ambient, room temperature conditions in the dark (ESI Figure S3). 16 days after cell fabrication, the cells containing BMII:I2:SCN, DMPII:I2:SCN, BMII:I2: GSCN:SCN and E1 electrolytes show performances of 95%, 88%, 81% and 93%, respectively, compared with their maximum efficiency values. BMII:I2:SCN indeed exhibits quite a promising performance. The liquid E1 cell shows 79% of its maximum performance after 122 days, and the BMII:I2:SCN solid electrolyte maintains a promising 88%. Both the DMPII:I2:SCN and BMII:I2:GSCN:SCN electrolytes continue to exhibit the poor stability, which was observed already within the initial 16 days itself, now showing only 56% (DMPII:I2:SCN) and 43% (BMII:I2: GSCN:SCN) after 122days.

3.3. Manufacturing advantages

The design of modules and integration of processing steps are critical to achieving high-throughput and low-cost DSSC production. Procedures such as electrolyte filling that are time consuming at the cell level must be done rapidly and efficiently when producing DSSC modules at commercial scales [28]. SCN-based solid electrolytes offer several advantages in terms of DSSC manufacturing processes. Currently liquid phase electrolytes are introduced into the DSSC via a vacuum backfilling process performed after cell sealing, which involves several steps including

• drilling hole in counter electrode (before cell fabrication)

• sandwich seal counter electrode and working electrode together

• vacuum backfill electrolyte

• seal back hole in counter electrode

On the basis of this study, we propose that these steps can be reduced to one quick manufacturing line process suitable for rapid, low-cost, high-throughput continuous processing as outlined in Figure 4. A small bead of solid SCN electrolyte is dropped onto the dye-coated TiO2 at ambient temperature. The counter electrode is then brought towards the working electrode, sandwiching sealant material between the two electrodes. The cell is

Figure 4. Electrolyte filling and dye-sensitised solar cellsealing step using solid-state succinonitrile-based electrolytes.

Table VI. Photovoltaic parameters of dye-sensitised solar celldevices made with different electrolyte materials at 200 lx incident

intensity of fluorescent light.

Electrolyte Jsc (^A/cm2) Voc (V) FF (%) Pmax (^W)

Liquid (E1) 23.7 0.654 61 2.65

DMPII:I2:SCN 16.5 0.379 48 0.85

BMII:I2:SCN 17.2 0.419 50 1.01

BMII:I2:GSCN:SCN 16.6 0.542 65 1.63

DMPII, 1,2-dimethyl-3-propyl imidazolium iodide; I2, iodine; SCN, succinonitrile; BMII, 1-butyl-3-methyl imidazolium iodide; GSCN, guanidinium thiocyanate. 'Electrolyte compositions: E1 is liquid phase and composed of 0.03 M I2, 0.1 M GSCN, 0.5 MTBPand 0.6 M BMII in acetonitrile:valeronitrile solution (85:15 by volume). SCN-based electrolytes consist of: DMPII dispersed with I2 in SCN in the molar ratio 5:1:100 (DMPII: I2 : SCN), BMII dispersed with I2 in SCN in the molar ratio 5:1:100 (BMII: I2 :SCN) and with GSCN additive in the molar ratio 5:1:5:100 (BMII: I2: GSCN : SCN).

heated in order to cure the sealant material thereby sealing the cell. (Sealants currently employed in DSSC manufacturing line production generally involve curing conditions of UV irradiation with gentle heating at temperatures up to 80°C. This temperature range is perfect for use with low melting point SCN electrolyte and is amenable to the requirement that high temperature processing is avoided after dyeing as dye molecules are sensitive to temperature above 100°C [28]). Simultaneously, the SCN electrolyte will begin to melt and slowly intercalculate into the pores of the TiO2 layer. The cell is then allowed to cool gradually to room temperature resulting in a sealed DSSC cell that is electrolyte-filled and ready for use without the need for any tricky and time consuming vacuum electrolyte filling and back hole drilling and sealing steps. This procedure makes SCN very suitable for use in low-cost, high-throughput, continuous processing either in roll-to-roll manufacture on flexible materials or on rigid glass substrates.

3.4. Low light (200 lx) measurement - indoor applications

Finally, considering that the melting point of the SCN host is in the region of 60°C, these electrolytes may be considered more suitable for use in indoor conditions where temperatures should not exceed 40°C. In this regard, the performance of the best cells was measured under low light conditions of a fluorescent bulb at an intensity of 200 lx to simulate indoor lighting conditions as displayed in Table VI. The liquid electrolyte produced a maximum cell power of 2.65 ^W. Both the DMPII:I2:SCN and BMII:I2: SCN solids achieved less than half the performance of the liquid electrolyte. However, BMII:I2:GSCN:SCN performed quite well achieving a power maximum of 1.63 ^W and exhibiting a much better FF under low light conditions compared with full sun conditions. Although not optimised, these results suggest that these solid electrolytes will perform under low light conditions and further optimisation of the electrolyte composition for use in low light (e.g. lower iodine/iodide concentrations) could yield promising potential for real world application of these SCN-based materials.

4. CONCLUSION

Succinonitrile doped with iodine and common imidazolium salts, DMPII and BMII, is possible to be used as a solid electrolyte in DSSC exhibiting best efficiencies of 6.3% and 5.6%, respectively. This result is achieved with a simple electrolyte filling process, without the use of any additives or additional TiO2 electrode treatment.

The effect of common liquid electrolyte additives on their solid-state analogue of BMII:I2:SCN is now better understood. TBP and NMBI operate as expected, improving cell voltages but reducing cell currents. LiI increases cell current but causes considerable reduction in cell voltages and fill factor. GSCN improves cell VOC significantly with no drop in cell current. Cell performance can be improved from 5.6% to 5.94% when using the GSCN additive. Additives are found to inhibit formation of the SCN plastic phase. For the first time, an SCN/additive (GSCN) electrolyte composition is reported as a room temperature solid although its long term stability was found to be poor. The non-additive containing BMII:I2:SCN electrolyte shows most promising stability of the solid electrolytes tested.

The solid nature of the SCN host allows for easy processing in DSSC manufacturing lines. We outline a simple, ambient condition, single-step procedure for DSSC electrolyte filling. Required manufacturing steps are reduced from four to one, allowing ready application of these solid electrolytes into DSSC manufacturing lines and enhancing DSSC's ability to compete with other PV technologies in terms of cost and manufacturing line speeds. Finally, these new electrolytes when tested under low light conditions at 200 lx showed excellent promise for its use in DSSC meant for indoor and sensor applications.

ACKNOWLEDGEMENTS

K. R. T. acknowledges the SFI-Stokes Professorship grant support and award [S07/EN/E013].O. B. received support from the European Commission's FP7 SMARTOP project under the grant agreement number: 265769. P. K. S. acknowledges the funding support received from IRCSET and SolarPrint under the EMPOWER Industry partnership research funding programme. The authors would like to thank SolarPrint Ltd for 200 lx measurements.

REFERENCES

1. Wu J, Lan Z, Hao S, Li P, Lin J, Huang M, Fang L, Huang Y. Progress on the electrolytes for dye-sensitized solar cells. Pure and Applied Chemistry 2008; 80: 2241-2258. DOI: 10.1351/pac200880112241

2. Chou C-S, Chou F-C, Kang J-Y. Preparation of ZnO-coated TiO2 electrodes using dip coating and their applications in dye-sensitized solar cells. Powder Technology 2012; 215-216: 38-45. DOI: 10.1016/j. powtec.2011.09.003

3. Chiba Y, Islam A, Watanabe Y, Komiya R, Koide N, Han L. Dye-sensitized solar cells with conversion efficiency of 11.1%. Japanese Journal of Applied Physics 2006; 45: L638-L640. DOI: 10.1143/JJAP.45.L638

4. Ito S, Nazeeruddin MK, Liska P, Comte P, Charvet R, Pechy P, Jirousek M, Kay A, Zakeeruddin SM, Grätzel M. Photovoltaic characterization of dye-sensitized solar cells: effect of device masking on conversion efficiency. Progress in Photovoltaics: Research and Applications 2006; 14: 589-601. DOI: 10.1002/pip.683

5. Yella A, Lee H-W, Tsao HN, Yi C, Chandiran AK, Nazeeruddin MK, Diau EW-G, Yeh C-Y, Zakeeruddin SM, Grätzel M. Porphyrin-sensitized solar cells with cobalt (n/III)-based redox electrolyte exceed 12 percent efficiency. Science 2011; 334: 629-634. DOI: 10.1126/science.1209688

6. Spiccia L. Aqueous dye-sensitized solar cell electrolytes based on the cobalt(II)/(III) tris(bipyridine) redox couple. Energy & Environmental Science 2012. DOI: 10.1039/C2EE23317G

7. Bai Y, Cao Y, Zhang J, Wang M, Li R, Wang P, Zakeeruddin SM, Gratzel M. High-performance dye-sensitized solar cells based on solvent-free electrolytes produced from eutectic melts. Nature Materials 2008; 7: 626-630. DOI: 10.1038/nmat2224

8. Chen C-L, Teng H, Lee Y-L. In situ gelation of electrolytes for highly efficient gel-state dye-sensitized solar cells. Advanced Materials 2011; 23: 4199-4204. DOI: 10.1002/adma.201101448

9. Wang P, Zakeeruddin SM, Grätzel M. Solidifying liquid electrolytes with fluorine polymer and silica nanoparticles for quasi-solid dye-sensitized solar cells. Journal of Fluorine Chemistry 2004; 125: 1241-1245. DOI: 10.1016/j.jfluchem.2004.05.010

10. Lee B, Buchholz DB, Guo P, Hwang D-K, Chang RPH. Optimizing the performance of a plastic dye-sensitized solar cell. The Journal of Physical Chemis-tryC 2011; 115: 9787-9796. DOI: 10.1021/jp201555n

11. Chung I, Lee B, He J, Chang RPH, Kanatzidis MG. All-solid-state dye-sensitized solar cells with high efficiency. Nature 2012; 485: 486-489. DOI: 10.1038/nature11067

12. Liu X, Zhang W, Uchida S, Cai L, Liu B, Ramakrishna S. An efficient organic-dye-sensitized solar cell with in situ polymerized poly(3,4-ethylenedioxythiophene) as a hole-transporting material. Advanced Materials 2010; 22: E150-E155. DOI: 10.1002/adma.200904168

13. Koh JK, Kim J, Kim B, Kim JH, Kim E. Highly efficient, iodine-free dye-sensitized solar cells with solid-state synthesis of conducting polymers. Advanced Materials 2011; 23: 1641-1646. DOI: 10.1002/adma.201004715

14. Cai N, Moon S-J, Cevey-Ha L, Moehl T, Humphry-Baker R, Wang P, Zakeeruddin SM, Grätzel M. An organic D-n-A dye for record efficiency solid-state sensitized heterojunction solar cells. Nano Letters 2011; 11: 1452-1456. DOI: 10.1021/nl104034e

15. Fabregat-Santiago F, Bisquert J, Cevey L, Chen P, Wang M, Zakeeruddin SM, Grätzel M. Electron transport and recombination in solid-state dye solar cell with spiro-OMeTAD as hole conductor. Journal of the American Chemical Society 2009; 131: 558-562. DOI: 10.1021/ja805850q

16. Kumara GRA, Okuya M, Murakami K, Kaneko S, Jayaweera VV , Tennakone K. Dye-sensitized solidstate solar cells made from magnesium oxide-coated nanocrystalline titanium dioxide films: enhancement of the efficiency. Journal of Photochemistry and Photobiology A: Chemistry 2004; 164: 183-185. DOI: 10.1016/j.jphotochem.2003.11.020

17. Burschka J, Pellet N, Moon S-J, Humphry-Baker R, Gao P, Nazeeruddin MK, Gratzel M. Sequential deposition as a route to high-performance perovskite-sensitized solar cells. Nature 2013; 499: 316-319. DOI: 10.1038/nature12340

18. Dai Q, MacFarlane DR, Forsyth M. High mobility ITHs redox couple in a molecular plastic crystal: A potential new generation of electrolyte for solidstate photoelectrochemical cells. Solid State Ionics 2006; 177: 395-401. DOI: 10.1016/j.ssi.2005.11.004

19. Hawthorne HM, Sherwood JN. Lattice defects in plastic organic solids. Part 2.-anomalous self-diffusion in succinonitrile. Transactions of the Faraday Society 1970; 66: 1792-1798. DOI: 10.1039/TF9706601792

20. MacFarlane DR, Huang J, Forsyth M. Lithium-doped plastic crystal electrolytes exhibiting fast ion conduction for secondary batteries. Nature 1999; 402: 792-794. DOI: 10.1038/45514

21. Alarco P-J, Abu-Lebdeh Y, Abouimrane A, Armand M. The plastic-crystalline phase of succinonitrile as a universal matrix for solid-state ionic conductors. Nature Materials 2004; 3: 476-481. DOI: 10.1038/nmat1158

22. Long S, MacFarlane DR, Forsyth M. Fast ion conduction in molecular plastic crystals. Solid State Ionics 2003; 161: 105. DOI: 10.1016/S0167-2738 (03)00208-X

23. Abu-Lebdeh Y, Abouimrane A, Alarco PJ, Hammami A, Ionescu-Vasii L, Armand M. Ambient temperature proton conducting plastic crystal electrolytes. Electrochemistry Communications 2004; 6: 432. DOI: 10.1016/j.elecom.2004.02.015

24. Wang P, Dai Q, Zakeeruddin SM, Forsyth M, MacFarlane DR, Grätzel M. Ambient temperature plastic crystal electrolyte for efficient, all-solid-state dye-sensitized solar cell. Journal of the American Chemical Society 2004; 126: 13590-13591. DOI: 10.1021/ja045013h

25. Hwang D, Jo SM, Kim DY, Armel V, MacFarlane DR, Jang S-Y. High-efficiency, solid-state, dye-sensitized solar cells using hierarchically structured TiO2 nanofibers. ACS Applied Materials & Interfaces 2011; 3: 1521-1527. DOI: 10.1021/am200092j

26. Desilvestro H, Berotz M, Tulloch S, Tulloch GE. Packaging, Scale-up, and Commercialization of Dye Solar Cells'. In Kalyanasundaram K. (ed.), Chapter 7, Dye Sensitised Solar Cells, EPFL Press: Lausanne (Switzerland), 2010.

27. Pettersson H, Nonomura K, Kloo L, Hagfeldt A. Trends in patent applications for dye-sensitized solar cells. Energy & Environmental Science 2012; 5: 7376-7380.

28. Baxter JB. Commercialization of dye sensitized solar cells: present status and future research needs to improve efficiency, stability, and manufacturing. Journal of Vacuum Science & Technology, A: Vacuum, Surfaces, and Films 2012; 30: 020801. DOI: 10.1116/ 1.3676433

29. Hinsch A, Brandt H, Veurman W, Hemming S, Nittel M, Würfel U, Putyra P, Lang-Koetz C, Stabe M, Beucker S, Fichter K. Dye solar modules for facade applications: recent results from project ColorSol. Solar Energy Materials and Solar Cells 2009; 93: 820-824. DOI: http:// dx.doi.org/10.1016/j.solmat.2008.09.049

30. Hinsch A, Veurman W, Brandt H, Loayza Aguirre R, Bialecka K , Flarup Jensen K. Worldwide first fully up-scaled fabrication of 60 x 100 cm2 dye solar module prototypes. Progress in Photovoltaics: Research and Applications 2012; 20: 698-710. DOI: 10.1002/ pip.1213

31. Yu Z, Vlachopoulos N, Gorlov M, Kloo L. Liquid electrolytes for dye-sensitized solar cells. Dalton Transactions 2011; 40: 10289-10303. DOI: 10.1039/C1DT11023C

32. Kroon JM, Bakker NJ, Smit HJP, Liska P, Thampi KR, Wang P, Zakeeruddin SM, Grätzel M, Hinsch A, Hore S, Würfel U, Sastrawan R, Durrant JR, Palomares E, Pettersson H, Gruszecki T, Walter J, Skupien K, Tulloch GE. Nanocrystalline dye-sensitized solar cells having maximum performance. Progress in Photovoltaics: Research and Applications 2007; 15: 1-18. DOI: 10.1002/pip.707

33. Ito S, Murakami TN, Comte P, Liska P, Grätzel C, Nazeeruddin MK, Grätzel M. Fabrication of thin film dye sensitized solar cells with solar to electric power conversion efficiency over 10%. Thin Solid Films 2008; 516: 4613-4619. DOI: 10.1016/j.tsf.2007.05.090

34. Ito S, Chen P, Comte P, Nazeeruddin MK, Liska P, Pechy P, Grätzel M. Fabrication of screen-printing pastes from TiO2 powders for dye-sensitised solar cells. Progress in Photovoltaics: Research and Applications 2007; 15: 603-612. DOI: 10.1002/pip.768

35. Sapp SA, Elliott CM, Contado C, Caramori S, Bignozzi CA. Substituted polypyridine complexes of cobalt(ii/iii) as efficient electron-transfer mediators in dye-sensitized solar cells. Journal of the American Chemical Society 2002; 124: 11215-11222. DOI: 10.1021/ja027355y

36. Liu Y, Hagfeldt A, Xiao X-R, Lindquist S-E. Investigation of influence of redox species on the interfacial energetics of a dye-sensitized nanoporous TiO2 solar cell. Solar Energy Materials and Solar Cells 1998; 55: 267-281. DOI: 10.1016/s0927-0248(98)00111-1

37. Li D, Qin D, Deng M, Luo Y, Meng Q. Optimization the solid-state electrolytes for dye-sensitized solar cells. Energy & Environmental Science 2009; 2: 283-291. DOI: 10.1039/B813378F

38. Grätzel M. Conversion of sunlight to electric power by nanocrystalline dye-sensitized solar cells. Journal of Photochemistry and Photobiology A: Chemistry 2004; 164: 3-14. DOI: 10.1016/j.jphotochem.2004.02.023

SUPPORTING INFORMATION

Additional supporting information may be found in the

online version of this article at the publisher's web site.