Effects of Introducing Methoxy Groups into the Ancillary Ligands in Bis(diimine) Copper(I) Dyes for Dye-Sensitized Solar Cells

A systematic investigation of four heteroleptic bis(diimine) copper(I) dyes in n-type Dye-Sensitized Solar Cells (DSSCs) is presented. The dyes are assembled using a stepwise, on-surface assembly. The dyes contain a phosphonic acid-functionalized 2,2′-bipyridine (bpy) anchoring domain (5) and ancillary bpy ligands that bear peripheral phenyl (1), 4-methoxyphenyl (2), 3,5-dimethoxyphenyl (3), or 3,4,5-trimethoxyphenyl (4) substituents. In masked DSSCs, the best overall photoconversion efficiency was obtained with the dye [Cu(5)(4)]+ (1.96% versus 5.79% for N719). Values of JSC for both [Cu(5)(2)]+ (in which the 4-MeO group is electron releasing) and [Cu(5)(4)]+ (which combines electron-releasing and electron-withdrawing effects of the 4and 3,5-substituents) and are enhanced with respect to [Cu(5)(1)]+. DSSCs with [Cu(5)(3)]+ show the lowest JSC. Solid-state absorption spectra and external quantum efficiency spectra reveal that [Cu(5)(4)]+ benefits from an extended spectral range at higher energies. Values of VOC are in the order [Cu(5)(4)]+ > [Cu(5)(1)]+ > [Cu(5)(2)]+ > [Cu(5)(3)]+. Density functional theory calculations suggest that methoxyphenyl character in MOs within the HOMO manifold in [Cu(5)(2)]+ and [Cu(5)(4)]+ may contribute to the enhanced performances of these dyes with respect to [Cu(5)(1)]+.


Introduction
Dye-sensitized solar cells (DSSCs) [1] convert solar to electrical energy using an optically transparent, wide-band gap semiconductor functionalized with a surface-bound dye which extends the absorption range into the visible spectrum [2][3][4].The semiconductor is commonly mesoporous TiO 2 and the sensitizer is typically a ruthenium(II) complex such as the standard reference dye N719 (Scheme 1) or an organic dye in DSSCs based on n-type semiconductors.Photoconversion efficiencies (η) of up to 11-14% have been recorded using ruthenium-based, organic, or zinc(II) porphyrin-based sensitizers [5][6][7][8][9][10].However, the scarcity of ruthenium in the Earth's crust and its associated high cost have motivated us and others to investigate the use of inorganic dyes containing Earth abundant metals.Copper(I) [11][12][13] and iron(II) [14,15] complexes are of primary interest.For copper-sensitized DSSCs, reported values of η in the range 3-5% [16][17][18] confirm the potential of DSSCs containing copper(I) dyes.When comparing these lower photoconversion efficiencies with the values obtained for state-of-the-art ruthenium(II) dyes, it is important to recognize that the dye structures and dye/electrolyte combinations in the ruthenium-based systems have been optimized for over a quarter of a century.In contrast, copper-based DSSCs are still in their infancy and, with systematic tuning of dye and electrolyte components [16,[19][20][21][22] and the use of co-sensitization [18], enhanced performances are gradually being achieved.Most copper(I) sensitizers are bis(diimine) copper(I) complexes and "push-pull" dyes which facilitate electron injection necessarily require the use of heteroleptic complexes.To overcome the problems of rapid ligand redistribution in solution (Equation ( 1)), Odobel and coworkers have used the HETPHEN approach [17], in which sterically demanding groups in the 6,6 -positions of a 2,2 -bipyridine (bpy) or 2,9-positions in a 1,10-phenanthroline ligand stabilize the heteroleptic complex.Our own strategy is a "surfaces-as-ligands, surfaces as complexes" (SALSAC) approach in which the heteroleptic dye is assembled on the glass FTO/TiO 2 electrode in a stepwise manner [11].For performance enhancement, heteroleptic dyes are essential and the SALSAC approach remains the most adaptable.It also has an advantage of allowing dye regeneration [23,24].
For the anchoring domain in a dye in DSSCs, carboxylic acid (or carboxylate) anchors are commonly employed [25].However, for heteroleptic copper(I) dyes, we have found that phosphonic acids are superior to carboxylic acids [11].This is consistent with the strong binding of phosphonic acids and phosphonates to TiO 2 [26] and the applications of phosphonic acid anchors in some ruthenium dyes [27].
We have investigated the use of ancillary ligands with hole-transporting dendrons [28], but such ligands are synthetically time-intensive, and we have generally found that dyes containing structurally simple ancillary ligands [16,18,29] perform as well, if not better, than those with more elaborate architecture.In accordance with the maxim "small is beautiful" [30], we now present an investigation of the effects on DSSC performance of introducing peripheral methoxy substituents into the ancillary ligand in bis(diimine) copper(I) dyes.The ancillary ligands 1-4 and anchoring ligand 5 (Scheme 2) each contain methyl substituents in the 6,6 -positions of the bpy unit to stabilize the copper(I) complexes.Copper(I) prefers a tetrahedral coordination geometry, and substituents close to the copper(I) centre prevent flattening to the square-planar geometry favoured by copper(II) [31].The ancillary ligands 1, 2, 3, and 4 contain peripheral phenyl, 4-methoxyphenyl, 3,5-dimethoxyphenyl, and 3,4,5-trimethoxyphenyl substituents, respectively.We were interested to know how the overall electron-releasing behaviour of the para-OMe or electron-withdrawing nature of the meta-OMe substituents [32][33][34] would influence the DSSC performances of the dyes [Cu(5)(L ancillary )] + (L ancillary = 1, 2, 3, or 4).Such substituent effects have been used by Wu et al. to tune properties of hole-transporting materials for perovskite solar cells [35].

Synthesis and Compound Characterization
Compounds 1-3 and 5 were prepared as previously reported [36][37][38][39].Ligand 4 was synthesized using Kröhnke [40] methodology as shown in Scheme 3. The 1 H and 13 C NMR spectra of 4 are displayed in Figures S1 and S2 and were assigned by COSY, NOESY, HMQC, and HMBC methods.A cross peak in the NOESY spectrum between H Me and H A5 distinguished the signals for H A3 and H A5 .The methyl groups of the methoxy substituents give rise to 1 H NMR resonances at δ 3.97 and 3.92 ppm (relative integrals of 2:1) while the signal assigned to the protons of the 6,6 -dimethyl groups appears at δ 2.72 ppm.nuclei C B3 and C B4 are characterized by 13 C NMR signals at δ 154.2 and 139.9 ppm, respectively.In [Cu(4) 2 ][PF 6 ], the 13 C NMR resonances for C A2 and C A6 could not be resolved in the 1D spectrum (Figure S8), and the HMBC spectrum was used to locate them at δ 157.4 and 152.1 ppm (Figure 2).In the respective electrospray mass spectrums of [Cu(1) 2 ][PF 6 ], [Cu(3) 2 ][PF 6 ], and [Cu(4) 2 ][PF 6 ], the base peak corresponded to the [M-PF 6 ] + ion at major isotopomer (m/z) = 735.24,975.31, and 1095.30,respectively.Each peak envelope showed the predicted isotope pattern (Figure S9).

Solution and Solid-State Absorption Spectra
The solution absorption spectra of the homoleptic complexes are presented in Figure 3 and the absorption maxima and intensity data are given in Table 1.We include the spectrum of [Cu(2) 2 ][PF 6 ] for comparison.The previously reported data was for an MeCN solution [37] whereas, here, the data is for CH 2 Cl 2 solutions.Intense absorption bands below 390 nm arise from π*←π transitions in [Cu(1) 2 ][PF 6 ] and from π*←π and π*←n transitions in the compounds containing methoxy groups.Each compound exhibits a metal-to-ligand charge transfer (MLCT) band in the range 482-486 nm.] has previously been reported [37].

DSSC Performances
Working electrodes with a scattering layer were functionalized with the heteroleptic dyes [Cu(5)(L ancillary )] + dyes (L ancillary = 1, 2, 3, or 4) as summarized in Figure 4, and duplicate masked DSSCs for each dye were assembled (see Materials and Methods).Reference DSSCs containing N719 (Scheme 1) were also fabricated (see Materials and Methods), and the plots of current density (J) against potential (V) in Figure 6 confirm the reproducibility of their performances.We have observed [16] that copper-dye functionalized DSSCs may improve in performance after initial fabrication.The J-V measurements for the DSSCs were, therefore, made on the day of sealing the cells (day 0) and repeated seven days later.The performance parameters for these masked DSSCs are given in Table 2. DSSCs with [Cu(5)(1)] + and [Cu(5)(3)] + were referenced to N719 cell 1, with [Cu(5)(2)] + to N719 cell 2, and with [Cu(5)(4)] + to N719 cell 3 (Table 2).The final column in Table 2 gives relative values of the efficiency η with respect to N719 set at 100%, and (on the day that the DSSCs were made) values range from 24.2% and 25.4% for [Cu(5)(3)] + to 30.2% and 33.9% for [Cu(5)(4)] + .Comparisons of the DSSC parameters for days 0 and 7 indicate that the devices are stable over this period but that there are no ripening effects [41,42] leading to significantly enhanced performance.

Dye
On the Day of DSSC Fabrication  Table 2 and the J-V curves in Figure 7 reveal that the highest short-circuit current densities (J SC ) are obtained for the dyes [Cu(5)(2)] + and [Cu(5)(4)] + with the 4-MeO and 3,4,5-(MeO) 3 substitution patterns in each phenyl ring (Scheme 2).While differences in J SC are small for these two dyes, values of the open-circuit voltages (V OC ) are larger for [Cu(5)(4)] + than [Cu(5)(2)] + .The replacement of the phenyl substituents in [Cu(5)(1)] + by 4-methoxyphenyl or 3,4,5-trimethoxyphenyl groups in [Cu(5)(2)] + and [Cu(5)(4)] + , respectively, enhances J SC but leads to a fall in V OC for [Cu(5)(2)] + , as opposed to a gain in V OC for [Cu(5)(4)] + .While Figure 7 shows this for the best performing cells, the scatter plots in Figure 8 confirm that the trend is true for the duplicate cells.The presence of the 4-MeO group is critical.The dye [Cu(5)(3)] + containing 3,5-dimethoxyphenyl groups exhibits the lowest values of V OC and J SC of all four dyes (Figures 7 and 8).The trends are consistent with the competitive mesomeric (+M) and inductive (−I) effects of the methoxy substituents.Electrons are released by the +M effect but are withdrawn by the −I effect.On going from 1 to 2, the electron-releasing 4-methoxy groups [32][33][34] enhance the "push-pull" characteristics of the dye, resulting in higher J SC .In contrast, introducing the electron-withdrawing 3,5-dimethoxy substituents [32][33][34] upon replacing ancillary ligand 1 by 3 has a detrimental effect on DSSC performance.Ancillary ligand 4 combines the effects of both ligands 2 and 3 and, interestingly, this results in the best-performing dye.DSSCs with [Cu(5)(4)] + exhibit both the highest V OC and J SC (Figures 7 and 8) and photoconversion efficiences up to 33.9% relative to N719 set at 100%.The trends in J SC are reflected in the external quantum efficiency (EQE) spectra displayed in Figure 9. Table 3 gives values of EQE max and λ max , and values of λ max are similar to the absorption maxima in the solid-state spectra in Figure 5.The slightly lower EQE max for [Cu(5)(4)] + versus [Cu(5)(2)] + is offset by the broader spectral range extending to higher energies; this is also observed in the solid-state absorption spectrum (Figure 5).

Characteristics of the HOMO and LUMO Manifolds
Ground state density functional theory (DFT) calculations were used to examine how methoxy-substitution affects the characteristics of the molecular orbitals in the HOMO and LUMO manifolds of the dyes.In an earlier study of bis(diimine) copper(I) complexes, we demonstrated that use of different atomic orbital basis sets (6-311++G** basis set on all atoms, 6-311++G** on Cu, 6-31G* on C, H and N, or 6-31G* on all atoms) strongly influences the calculated absorption spectra but had no significant effect upon the characteristics of the MOs lying in the HOMO and LUMO manifolds [43].In the present investigation, we therefore employed a 6-31G* basis set on all atoms to reduce the computational burden, and we focus mainly on orbital composition.
The characters of the MOs from HOMO−3 to LUMO+1 and energy levels are shown in Figures 10  and 11.For all sensitizers, the LUMO is localized on the anchoring ligand while the LUMO+1 is localized on the bpy unit of the ancillary ligand.The energy gap between the LUMO and LUMO+1 is similar in each complex (Figure 11).Inspection of the character of the HOMO in each complex is instructive.The HOMO is essentially centered on copper.Although, in each of the methoxy-functionalized dyes there are additional contributions from the ancillary ligand extending over the peripheral aryl groups.This is most apparent for [Cu(5)(2)] + .The appearance of dominant methoxyphenyl character in HOMO−3 in [Cu(5)(2)] + and [Cu(5)(4)] + , as well in HOMO−2 and HOMO−3 in [Cu(5)(3)] + , may aid hole transport over the ancillary ligand.This is reminiscent of [Cu(5)(7)] + (7 = 4,4 -bis(4-iodophenyl)-6,6-dimethyl-2,2 -bipyridine), for which we have suggested that the improved performance of DSSCs containing [Cu(5)( 7)] + with respect to other halo-analogues may be associated with better electron transfer from the electrolyte over the 4-IC 6 H 4 substituent [16].

General
1 H and 13 C NMR spectra were recorded on a Bruker Avance III-500 NMR spectrometer (Bruker BioSpin AG, Fällanden, Switzerland), and 1 H and 13 C chemical shifts were referenced to residual solvent peaks with respect to δ(TMS) = 0 ppm.Solution and solid-state absorption spectra were recorded on a Cary 5000 spectrophotometer, (Agilent Technologies Inc., Santa Clara, CA, United States), respectively.Electrospray ionization (ESI) mass spectra were recorded on a Shimadzu LCMS-2020 instrument (Shimadzu Schweiz GmbH, Roemerstr, Switzerland), and high resolution ESI mass spectra were recorded on a Bruker maXis 4G QTOF instrument (Bruker Daltonics GmbH, Faellanden, Switzerland).

DSSC Fabrication
TiO 2 electrodes (Solaronix Test Cell Titania Electrodes, Solaronix SA, Aubonne, Switzerland) were washed with milliQ water and HPLC grade EtOH, heated at 450 • C for 30 min, then cooled to ca. 80 • C. The thicknesses of the transparent and scattering layers were ~9 and 3 µm, respectively, by SEM [45].The electrodes were then placed in a DMSO solution of 5 (1.0 mM) for 24 h at room temperature.The electrodes were removed from the solution, washed with DMSO and EtOH, and dried in an N 2 stream.Each electrode was then immersed in a CH 2 Cl 2 solution of , or [Cu(4) 2 ][PF 6 ] (0.1 mM) for 3 days at room temperature.After removal from the dye-bath, the electrodes were washed with CH 2 Cl 2 and dried in an N 2 stream.For N719 (Solaronix SA, Aubonne, Switzerland), TiO 2 electrodes were soaked in a solution of N719 (EtOH, 0.3 mM) for 3 days.The electrodes were taken out of the dye-bath, washed with EtOH, and dried in an N 2 stream.Commercial counter electrodes (Solaronix Test Cell Platinum Electrodes, Solaronix SA, Aubonne, Switzerland) were washed with EtOH and then heated at 450 • C for 30 min to remove volatile organics.

Electrodes for Solid-State Absorption Spectroscopy
Dye-functionalized electrodes were assembled using the immersion procedure above but using Solaronix Test Cell Titania Electrodes Transparent (Solaronix SA, Aubonne, Switzerland).

DSSC and External Quantum Efficiency (EQE) Measurements
The DSSCs were masked.The mask was made from a black-coloured copper sheet with an aperture (ca.0.06 cm 2 , each mask accurately calibrated) smaller than the surface area of TiO 2 .Black tape was used to complete the top and side masking of each DSSC.Performance measurements were made by irradiating the DSSC from behind with a LOT Quantum Design LS0811 instrument (LOT-QuantumDesign GmbH, Darmstadt, Germany, 100 mW cm −2 = 1 sun, AM1.5 G conditions) and the simulated light power was calibrated with a silicon reference cell.EQE measurements were made using a Spe Quest quantum efficiency setup (Rera Systems, Nijmegen, The Netherlands) operating with a 100 W halogen lamp (QTH) and a lambda 300 grating monochromator (L.O.T.-Oriel GmbH & Co. KG, Darmstadt, Germany).The monochromatic light was modulated to 3 Hz using a chopper wheel (ThorLabs Inc., Newton, NJ, USA), and the cell response was amplified with a large dynamic range IV converter (Melles Griot B.V., Didam, The Netherlands) and measured with a SR830 DSP Lock-In amplifier (Stanford Research Systems Inc., Sunnyvale, CA, USA).

DFT Calculations
Ground state density functional theory (DFT) calculations were carried out using Spartan 16 (v.2.0.9)[46] at the B3LYP level with a 6-31G* basis set in vacuum.Initial energy optimization was carried out at a semi-empirical (PM3) level.

Scheme 2 .
Scheme 2. Structures of ligands and atom labelling for NMR assignments.

Table 2 .
Performance parameters (under 1 sun illumination) of duplicate, masked DSSCs (mask aperture calibrated and ca.0.06 cm 2 ) with [Cu(5)(L ancillary )] + (L ancillary = 1, 2, 3, or 4) on the day of sealing the cells, and after seven days.The data is compared to a DSSC containing N719 and relative η values are with respect to N719 set at 100% a .