Neutral copper(i) dipyrrin complexes and their use as sensitizers in dye-sensitized solar cells

Article abstract of DOI:10.1039/c3dt53334d

Heteroleptic neutral copper(i) dipyrrin complexes have been synthesised with the general formula [Cu(4,4′-(R)-6,6′-(CH₃)- bipyridine)(dipyrrin)], R = CH₃ or CO₂Et, and H-dipyrrin is either 1,3,7,9-tetramethyldipyrromethene

Full text of DOI:10.1039/c3dt53334d

Dalton
Transactions
View Article Online
View Journal
PAPER
Neutral copper(I) dipyrrin complexes and their use
as sensitizers in dye-sensitized solar cells†
Cite this: DOI: 10.1039/c3dt53334d
Tracy E. Hewat, Lesley J. Yellowlees and Neil Robertson*
Heteroleptic neutral copper(I) dipyrrin complexes have been synthesised with the general formula
[Cu(4,4’-(R)-6,6’-(CH3)-bipyridine)(dipyrrin)], R = CH3 or CO2Et, and H-dipyrrin is either 1,3,7,9-tetra-
methyldipyrromethene (HL1), 1,13-diphenyl-6,8-diisoindolemethene (HL2), or 1,13-diphenyl-3,11-di(tri-
fluoromethyl)-6,8-diisoindolemethene (HL3). Improved stability was observed across the series L1–L3,
likely due to better steric constraints between the ligands. Visible light absorption has also been enhanced
with a red-shift in absorption from 450 nm to 600 nm. Complex 3 shows photoluminescence lifetime in
the order of nanoseconds suggesting singlet fluorescence which is supported by theoretical calculations.
Study of the complexes as sensitisers in dye-sensitised solar cells was achieved by assembling the dye
in situ on the surface of TiO2 in a series of steps (anchoring ligand followed by ancillary ligand and
[Cu(CH3CN)4][BF4]. The highest efficiency achieved was 0.41% for the dye with HL3, attributed to better
dye regeneration due to a more favourable oxidation potential.
Received 26th November 2013,
Accepted 19th January 2014
DOI: 10.1039/c3dt53334d
www.rsc.org/dalton
the d10 configuration, which avoids low-lying metal d–d states
that might cause non-radiative decay. However, conformational
Introduction
Copper(I)-polypyridyl complexes have attracted a lot of recent change of the complex following MLCT excitation may lead to
interest as alternatives to Ru(II)-polypyridyls in photophysical rapid non-radiative decay,19,20 and similarly, conformational
and photoelectrochemical applications, particularly due to change following Cu(I)/Cu(II) oxidation leads to slow electron
their ease of synthesis, their low toxicity, and the high abun- transfer. To achieve Cu(I) complexes suitable for photoelectro-
dance of copper in comparison with 2nd or 3rd row transition chemical applications, this must be prevented by ligand
metals. This has led to use of such species in organic light design.21 Typically addition of bulky substituents in the 6,6′-
emitting diodes (OLEDs),1–3 light-emitting electrochemical positions of bipyridine or 2,9-positions of phenanthroline will
cells (LECs)4,5 and dye sensitised solar cells (DSSCs).6–18
minimise the conformational change and maintain the
Copper(I)-polypyridyl complexes can display long-lived pseudo-tetrahedral geometry. Another method is to have a
metal-to-ligand charge transfer (MLCT) excited states due to large ancillary ligand to enhance steric rigidity.6–9,11,20,22
For solar energy applications, chromophoric ligands are
attractive to enhance light harvesting. The BODIPY dye family
has been widely studied as they are highly fluorescent; soluble
in most organic solvents; they have a narrow emission band-
width; are highly photostable; and have a relatively long-lived
excited state.23–25 This makes them ideal candidates in a wide
range of applications such as biofluorescent probes, photo-
sensitisers or pigments for dye lasers.24–26
In this context, dipyrromethene (hereafter dipyrrin) species,
that form the delocalised framework of BODIPYs, have been
used as ligands for a range of metals such as Pt(II),27 Cu(II),28
Ru(II),29 Ir(III),30 Zn(II)26 and more recently Cu(I).31 The unsub-
stituted parent compound is unstable,32,33 however stability is
enhanced by adding groups at the meso-position and 1,9-posi-
tions to block nucleophilic or electrophilic attack.
School of Chemistry and EaStChem, University of Edinburgh, King’s Buildings,
Edinburgh, UK. E-mail: neil.robertson@ed.ac.uk; Fax: +44 (0)131 650 4743;
Tel: +44 (0)131 650 4755
†Electronic supplementary information (ESI) available: Synthetic procedures for
the synthesis of HL3 and dmdecbpy; Fig. S1: Reaction scheme for HL3; Fig. S2:
Stability tests of 1; Fig. S3: Stability tests of 3; Fig. S4: 2D COSY and 2D NOESY
spectra of 3; Fig. S5: 19F NMR of HL3; Fig. S6: 1H NMR spectra of HL3 and 5;
Fig. S7: (a) Absorbance spectrum of HL2 in varying solvents, (b) Absorbance
spectrum of 3 in varying solvents; Fig. S8: Excitation and emission spectra of
1–4; Fig. S9: Molecular orbital images of 1; Fig. S10: Molecular orbital images of
2; Fig. S11: Molecular orbital images of 3; Fig. S12: Differential pulse voltammo-
grams of 1–5; Table S1: Percentage contributions from component parts of 1 to
selected molecular orbitals; Table S2: TD-DFT calculated visible absorption wave-
lengths of 1; Table S3: Percentage contributions from component parts of 2 to
selected molecular orbitals; Table S4: TD-DFT calculated visible absorption wave-
lengths of 2; Table S5: Percentage contributions from component parts of 3 to
selected molecular orbitals; Table S6: TD-DFT calculated visible absorption wave-
lengths of 3; Table S7: Reagents and solvents used for in situ solar studies;
Table S8: JV data for in situ solar studies. See DOI: 10.1039/c3dt53334d
Generally, dipyrrins have been used to form divalent or tri-
valent homoleptic complexes, with limited reports of hetero-
leptic species. A recent report described the first synthesis of
This journal is © The Royal Society of Chemistry 2014
Dalton Trans.

View Article Online
Paper
Dalton Transactions
heteroleptic Cu(I) dipyrrin complexes with phosphine ancillary room temperature and 77 K in ethanol, using a Fluoromax2
ligands, and concluded that the low energy absorption and fluorimeter controlled by the ISAMain software. Electrochemi-
emission bands were intraligand charge transfer and thus the cal studies were recorded using the General Purpose Electro-
complex displayed fluorescence rather than phosphor- chemical System (GPES) software that was connected to an
escence.31 Teets et al. reported the first neutral Cu(I) complex Autolab system containing a PGSTAT 30. The technique com-
of the related azadipyrromethene complex,34 which was weakly prised a Pt disk working electrode, a Pt rod counter electrode,
luminescent at room temperature.
and an Ag/AgCl reference electrode. The data collected were
Given the emerging importance of Cu(I) polypyridyl com- standardised against ferrocene/ferrocenium (measured at
plexes in optoelectronic fields, such species that also incorpor- +0.507 V),43 and adjusted for NHE (addition of 0.23 V to Ag/
ate a dipyrrin ligand would form an attractive and tunable AgCl values). The supporting electrolyte was 0.1 M tetrabutyl-
family of functional dyes. Accordingly, this work focuses on ammonium tetrafluoroborate in acetonitrile.
the synthesis and analysis of the first examples of copper(I)
Theoretical calculations were performed using the func-
complexes that feature a dipyrrin ligand and a functionalised tional B3LYP, and the basis set LANL2DZ, to describe the
2,2′-bipyridyl ligand. A first assessment is made of the poten- copper atom and all other atoms were described by
tial of this complex type to be applied as sensitisers to dye-sen- 6-31G*.44–47 The initial structure was computed using Avoga-
sitised solar cells, in which context increased light harvesting, dro software48 and frequency calculations were generated from
tunable redox potential and
a long-lived charge-transfer the optimised geometry. Time-dependent density functional
excited state are all key features to optimise. We have used theory (TD-DFT) calculations were carried out using an aceto-
three different dipyrrin ligands (Fig. 1), with varying degrees of nitrile polarisable continuum model and seventy singlet-
π-conjugation and substituents, to explore the effect on the singlet transitions were calculated using the optimised
photophysical and electrochemical properties of the dye. The geometry.
bipyridine ligands used have substituents in 6,6′-positions to
Synthesis
of
Cu(L1)(tmbpy)
(1). [Cu(CH3CN)4][BF4]
maintain a pseudo-tetrahedral geometry upon excitation. An (39.6 mg, 0.13 mmol), HL1·HCl (29.8 mg, 0.13 mmol) and 1 M
in situ dye assembly method on the TiO2 surface was adopted NaOH (0.25 mL) were added to degassed acetone (10 mL) and
to establish that such species can function as sensitisers.12,13
CH2Cl2 (5 mL). The solution was stirred at room temperature,
under nitrogen, for 15 min. Tmbpy (26.7 mg, 0.13 mmol) was
added to the mixture and the solution was stirred for a further
45 min. The mixture was filtered and the solvent was removed
under reduced pressure. The resulting yellow/brown solid was
stirred in water for 30 min and then filtered. Yield: 46.2 mg,
77%. 1H NMR (DMSO, 400 MHz): δ 8.00 (s, 2H, H-bpy), 7.11 (s,
2H, H-bpy), 6.87 (s, 1H, CH-meso), 5.99 (s, 2H, H-pyrrolic), 2.50
(s, 6H, CH3-bpy), 2.36 (s, 6H, CH3-bpy), 2.28 (s, 6H, CH3-L3),
2.18 (s, 6H, CH3-L3). +ESI/MS: m/z 475.4 (M+). Anal. Calc. for
C27H31CuN4: C 68.26, H 6.58, N 11.79. Found: C 68.19, H 6.53,
N 11.64.
Experimental
General procedures
The synthesis of [Cu(CH3CN)4][BF4],35 4,4′-6,6′-dimethyl-2,2′-
bipyridine (tmbpy),36 4,4′-dicarboxy-6,6′-dimethyl-2,2′-bipyri-
dine
(HL1·HCl),39
(dmdcbpy),9,37,38
and
1,3,7,9-tetramethyldipyrromethene
1,13-diphenyl-6,8-diisoindolemethene
(HL2),40 were carried out according to literature procedures.
4,4′-(CO2Et)-6,6′-dimethyl-2,2′-bipyridine (dmdecbpy) was syn-
thesised by adapted literature procedures.8 1,13-diphenyl-3,11-
di(trifluoromethyl)-6,8-diisoindolemethene (HL3) is a novel
ligand adapted from literature methods for related species (see
ESI† for procedure)40–42 All other chemicals were purchased
from Sigma Aldrich, VWR, Fischer Scientific, Merck and used
as received, except for 2,4-lutidine which was purified further
by vacuum distillation.
Synthesis of Cu(L2)(tmbpy) (2). [Cu(CH3CN)4][BF4] (28 mg,
0.09 mmol), HL2 (35 mg, 0.09 mmol) and 1 M NaOH
(0.09 mL) were added to degassed acetone (10 mL) and stirred
at room temperature, under nitrogen, for 15 min. Tmbpy
(19 mg, 0.09 mmol) was added and the solution was stirred for
a further 45 min. The mixture was filtered and the solvent was
removed under reduced pressure. The purple solid was washed
1
with water and dried. Yield: 47.6 mg, 78.3%. H NMR (DMSO,
All UV/Vis spectra were recorded using a JASCO V-670 series
spectrophotometer and the data collected using Spectra
Manager™ II software. Emission spectra were recorded at
400 MHz): δ 8.30 (s, 1H, CH), 8.25 (d, 2H, J = 7.9 Hz, H-L4),
7.84 (s, 2H, H5-bpy), 7.68 (d, 2H, J = 7.9 Hz, H-L4), 7.33 (t, 2H,
J = 7.6 Hz, H-L4), 7.31 (d, 4H, J = 7.4 Hz, Ph), 7.13 (t, 2H, J =
7.4 Hz, H-L4), 6.96 (s, 2H, H3-bpy), 6.89 (t, 2H, J = 7.3 Hz, Ph),
6
6.72 (t, 4H, J = 7.6 Hz, Ph), 2.31 (s, 6H, CH3 -bpy), 2.01 (s, 6H,
4
CH3 -bpy). +ESI/MS: m/z 457.2 (CuL4). Anal. Calc. for
C43H35CuN4: C 76.93, H 5.26, N 8.35. Found C 76.82, H 5.20,
N 8.26.
Synthesis Cu(L2)(dmdecbpy) (3). [Cu(CH3CN)4][BF4] (24 mg,
0.08 mmol), HL2 (30 mg, 0.08 mmol) and 1 M NaOH
(0.08 mL) were added to degassed acetone (10 mL) and stirred
at room temperature, under nitrogen, for 10 min. Dmdecbpy
Fig. 1 Ligands used in this work.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
(25 mg, 0.08 mmol) was added to the solution and stirred for a allowed to cool (80–100 °C) and immersed into a DMSO solu-
further 1.5 h. The mixture was filtered and the solvent was tion of dmdcbpy (3 mM) overnight. The colourless slide was
removed under reduced pressure. The purple solid was washed removed from the solution and washed with DMSO followed
with water and dried. A sephadex column (MeOH, then by ethanol and dried. The electrodes were then immersed into
acetone) was used for further purification. Yield: 48.8 mg, a acetonitrile solution of [Cu(CH3CN)4][BF4] (1 mM) and
81.3%. 1H NMR (DMSO, 500 MHz): δ 8.35 (s, 2H, H5-bpy), 8.34 HL1·HCl, HL2, or HL3 (1 mM) for 24 h. The electrodes were
(s, 1H, CH), 8.28 (d, 2H, H-L4), 7.66 (s, 2H, H3-bpy), 7.64 (s, removed and washed with acetonitrile and dried. Counter elec-
2H, H-L4), 7.36 (t, 2H, H-L4), 7.20 (d, 4H, Ph), 7.15 (t, 2H, trodes were platinized using 10 µL of H2PtCl6 solution (5 mM)
H-L4), 6.83 (t, 2H, Ph), 6.70 (t, 4H, Ph), 4.45 (q, 4H, O-CH2), in isopropyl alcohol and heated to 400 °C for 15 minutes. The
6
2.18 (s, 6H, CH3 -bpy), 1.41 (t, 6H, CH3). +ESI/MS: m/z 804 solar cells were assembled by sandwiching the two electrodes
[M + NH4]. Anal. Calc. for C47H39CuN4O4: C 71.69, H 4.99, together using a sealant (Bynel, 25 µm), and heated to 220 °C,
N 7.12. Found: C 71.79, H 4.92, N 6.96.
applying even pressure to ensure the Bynel melted evenly.
Synthesis of Cu(L3)(tmbpy) (4). [Cu(CH3CN)4][BF4] (23 mg, Electrolyte was vacuum filled into the hole and the hole was
0.07 mmol), HL3 (40 mg, 0.07 mmol) and 1 M NaOH sealed by Bynel and a cover glass. Current–voltage behaviour
(0.07 mL) were added to degassed acetone (10 mL) and stirred of the cells was determined using a Sciencetech solar simu-
at room temperature, under nitrogen, for 15 min. Tmbpy lator equipped with an AM 1.5G filter, calibrated to 100 mW
(16 mg, 0.07 mmol) was added and the solution was stirred for cm−1 using a silicon photodiode. The electrolyte used was
a further 1.5 hours. The mixture was filtered and the solvent 0.4 M lithium iodide (LiI), 0.03 M iodine (I2), 0.1 M guanidi-
was removed under reduced pressure. The purple solid was nium thiocyanate (GuNCS), 0.5 M tert-butylpyridine (TBP), 1 M
washed with water and dried. The solid was purified by 1,3-dimethylimidazolium iodide (DMII) in acetonitrile–valero-
washing with toluene to filter off the impurity (Cu(tmbpy)2BF4) nitrile (MeCN–VN) 85 : 15.
and then running a Sephadex column in DMF to remove any
further free ligands. Solvent was removed from the filtrate to
yield the pure product. Yield: 15 mg, 25%. 1H NMR (DMSO,
Results and discussion
Syntheses
500 MHz): δ 8.62 (s, 1H, CH), 8.53 (d, 2H, H-L4), 7.91 (s, 4H,
H-L4), 7.84 (s, 2H, H5-bpy), 7.63 (d, 2H, H-L4), 7.22 (d, 4H, Ph),
6.99 (s, 2H, H3-bpy), 6.96 (t, 2H, Ph), 6.77 (t, 4H, Ph), 2.36 (s,
Copper(I) dipyrrin complexes were synthesised at room temp-
6
6H, CH3 -bpy), 1.99 (s, 6H, CH36-bpy). 19F NMR (DMSO,
erature via a simple one-pot method (Fig. 2) using stoichio-
metric quantities of ligands and base (to deprotonate the free
dipyrrin ligand). The reaction was carried out under nitrogen
to minimise oxidation to copper(II). Further purification was
376 MHz): δ −59.17, −59.41.
Synthesis of Cu(L3)(dmdecbpy) (5). [Cu(CH3CN)4][BF4]
(20 mg, 0.07 mmol), HL3 (35 mg, 0.07 mmol) and 1 M NaOH
(0.07 mL) were added to degassed acetone (10 mL) and stirred
at room temperature, under nitrogen, for 15 min. Dmdecbpy
(21 mg, 0.07 mmol) was added and the solution was stirred for
a further 1.5 hours. The mixture was filtered and the solvent
was removed under reduced pressure. The purple solid was
achieved by Sephadex column where necessary (3–5). The com-
plexes were characterised by 1H NMR, ESI-MS and elemental
analysis, confirming the proposed structures. In the case of 4
and 5, only NMR characterisation was possible due to difficul-
ties in obtaining sufficient yield (4) and residual ligand
washed with water and dried. The solid was purified by
washing with toluene to filter off the impurity (Cu(dmdecb-
py)2BF4) and then running a Sephadex column in DMF to
remove any further free ligands, but some free ligands still
remained. Yield: 6 mg, 10%. 1H NMR (DMSO, 500 MHz):
δ 8.67 (s, 1H, CH), 8.63 (s, 2H, H5-bpy), 8.56 (d, 2H, H-L4), 7.89
(s, 2H, H-L4), 7.81 (s, 2H, H3-bpy), 7.66 (d, 2H, H-L4), 7.22 (d,
4H, Ph), 6.89 (t, 2H, Ph), 6.75 (t, 4H, Ph), 4.41 (q, 4H, O-CH2),
6
2.69 (s, 6H, CH3 -bpy), 1.37 (t, 6H, CH3). 19F NMR (DMSO,
376 MHz): δ −59.26, −59.51.
Preparation of solar cells
Titanium dioxide paste (Dyesol, DSL-18NR-T) was deposited
onto pre-treated (TiCl4, 40 mM, 75 °C for 30 min) fluorine-
doped tin oxide conductive glass (3 mm thick, 8 Ω sheet resist-
ance, Solaronix) by doctor-blading. The film thickness was
approx 11 µm. The TiO2 electrodes (active area = 4 × 4 mm2)
were heated to 500 °C for 30 min, cooled to room temperature,
treated with a dilute solution of TiCl4 (40 mM, 75 °C for
30 min) and sintered at 500 °C for 30 min. The electrodes were Fig. 2 Reaction scheme of neutral copper(I) complexes.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans.

View Article Online
Paper
Dalton Transactions
impurities (5). Clearly, complex formation with the new ligand
L3 has proven difficult, however we include 4 and 5 in the
study as they have the greatest relevance to DSSC application
and the in situ dye assembly described later.
Stability and reproducibility
Copper(I) complexes are extremely labile which accounts for
the common reproducibility problems for heteroleptic
species.12,49,50 During the reaction, ligands will continually
exchange with other ligands in solution, giving rise to a
mixture of products, unless one is more favoured due to steric
constraints or binding energies.
The stability of complexes 1–3 in solution was explored via
1H NMR at varying time intervals; 0, 7 and 17 days. The stabi-
lity of 2 is shown in Fig. 3.
Fig. 4 Absorption spectra of ligands (HL1·HCl, HL2 and HL3) and com-
plexes (1–4) in acetonitrile.
After seven days, there is a slight change in the intensity of
the peak at 7.31 ppm, representing the phenyl group on the
dipyrrin ligand (red box), which became more noticeably
different after 17 days. After 17 days there are also new peaks
(purple circles) arising which suggests [Cu(tmbpy)2][BF4]
(8.36 ppm and 7.42 ppm) is forming in solution with the
breakdown of the complex. The long-term instability in solu-
tion is the likely reason that precluded the growth of single
crystals for diffraction. This is similarly observed for complexes
1 and 3 (ESI†).
Generally, increasing conjugation will reduce the HOMO–
LUMO gap of the ligand hence lower the energy of the π–π*
transition. Adding electron-withdrawing groups (CF3) to the
dipyrrin (HL3) resulted in a slight blue shift in comparison to
HL2.
Complexes 2 and 3 show very similar spectroscopic pro-
perties, with the main visible band around 600 nm, a slight
red-shift from the free ligand. A slight red-shift and intensity
difference between 2 and 3 is due to a change in substituents
on bipyridine. Theoretically, increasing the electron-withdraw-
ing character (methyl to an ester) will shift the MLCT absorp-
tion band to lower energies, and can also enhance the charge
transfer character of the low-energy transitions where these
involve orbitals on the bipyridine ligand. Complex 4 shows
similar absorption values to HL3 (570 nm), although the inten-
sity is greater (Fig. 5).
UV-Vis absorption spectroscopy
The spectra of all ligands and complexes display the character-
istic intense visible absorption typical of dipyrrins. The notice-
able difference between the protonated ligand (HL1·HCl) and
complex 1 is the considerable decrease in intensity and broad-
ening of the main band (432 nm) when the dipyrrin is co-
ordinated to copper. Within porphyrin chemistry this is
usually a result of loss in degeneracy of the LUMO51 and a
similar effect may be occurring here. Complex 1 also shows a
second absorption at longer wavelengths (498 nm) (Fig. 4).
Increasing the conjugation of the dipyrrin backbone for
HL2 results in a red shift of the π–π* transition from 464 nm
(HL1·HCl) to 584 nm, and a decrease in the intensity and
broadening of the band, possibly due to phenyl rotation.
Photoluminescence
Complexes 1–4 are emissive in ethanol at room temperature
and 77 K. The results are summarised in Table 1 and com-
pared with data for the free dipyrrin ligands. All of the com-
plexes emit at a comparable energy to their respective free
dipyrrin ligands and at room temperature, the emission
maxima and the shape of the band for complexes 2, 3 and HL3
are all very similar. These observations suggest that the copper
or bipyridine orbitals are not significantly involved in the tran-
sition. This is often seen in literature for dipyrrin complexes,
or other highly chromophoric, conjugated ligands,31,52 where
the photophysical behaviour derives solely from the ligand
orbitals. The excitation data for complex 4 correlate well with
the absorption and are very similar to the uncomplexed ligand
HL3, which further supports the conclusion that the emissive
behaviour derives from the dipyrrin. Complexes 1–3 show λmax
(excitation) values at higher energy than the maxima in their
respective absorption spectra. Complexes 2 and 3 both display
higher-energy emission when frozen, suggesting a significant
degree of structural relaxation of the molecule after excitation
that is prevented in the rigid matrix (77 K). In addition to this,
the change in solvent (ethanol, acetonitrile and DCM) does
not have a significant effect on the position and shape of the
Fig. 3 Stability testing of 2 in d6-DMSO by NMR.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
absorption and emission spectra, only in the intensity (ESI†).
This implies that there is weak polarisation between the
ground and excited states, supporting the assignment of the
transition as ligand centred rather than charge transfer.40
Time-resolved fluorescence was used to measure the life-
time of complex 3 and ligand HL2, selected as representative
samples that gave sufficient signal with the excitation laser
wavelength available (shortest = 486 nm). The average excited-
state lifetime observed for HL2 and 3 are similar, with that of
the complex being slightly longer. In keeping with the steady-
state measurements, this similarity and also the nanosecond
timescale, suggest that emission can be assigned as 1(π–π*)
fluorescence from the dipyrrin centre and not 1MLCT or
3MLCT. Dipyrrin centred intraligand charge transition was
also reported in literature for copper(I) complexes with phos-
phine coligands.31
Electrochemistry
The redox potentials for free ligands and complexes are sum-
marised in Table 2. Although the poor purity of 5 prevented its
inclusion in the spectroscopic study, it has been included here
since minor impurity peaks in the electrochemical data can be
readily identified and neglected.
The reduction potentials reported for each complex can be
assigned as bipyridine ligand based, although possibly also
with some dipyrrin character in the case of 1 as suggested by
computational work (vide infra). Complex 2 and complex 3
Table 2 Electrochemical data for ligands HL1·HCl, HL2, HL3 and
complexes 1–5 in 0.1 M TBABF4–MeCN vs. NHE
Compound
Eox/V
Ered/V
HClL1
HL2
HL3
1
2
3
4
5
1.55a, 1.68
−0.45a
0.84b, 1.02b
—
0.97b, 1.08b
—
0.97a, 1.25
−1.44a
0.45a, 0.64b, 0.86
0.55a, 0.68b, 0.90
0.62a, 0.96b, 1.12b
0.76b, 0.97, 1.13b
−1.26a
−1.18a
Fig. 5 Photoluminescence of (top) HL1·HCl and 1, (centre) HL2, 2 and
3 in Ethanol at 77 K (complex 1 recorded at room temperature) and
(bottom) HL3 and 4 in acetonitrile at 77 K.
In solvent window
−1.24a
a Irreversible. b Pseudo-reversible.
Table 1 Photophysical properties of the ligands HL1·HCl, HL2, HL3 and complexes 1–4
Room temperatureb
77 Kb
Compound λabsa/nm (εmax × 103 M−1 cm−1
)
λmax/nm Excitation λmax/nm Emission λmax/nm Excitation λmax/nm Emission <τ>b/ns
HL1·HCl
HL2
HL3
1
464 (145)
584 (44.5)
571 (29.6)
432 (53.2)
498 (6.4)
600 (33.2)
608 (38)
485
600
590
390
492
650
650
490
470
605
580
370
490
630
660
490
Not recorded
0.25
Not recorded
Not recorded
2
3
4
530
540
570
665
665
645
550
560
585a
635
640
645a
Not recorded
0.34
Not recorded
572 (44.2)
a Measured in acetonitrile. b Measured in ethanol.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans.

View Article Online
Paper
Dalton Transactions
both display a similar oxidation potential attributed to the Computational and electronic properties of molecules
Cu(dipyrrin) fragment which remains unchanged between
them. The irreversible character of these precludes interpret- The experimental and calculated UV-Vis spectra for 1–3 are
ation of the subsequent oxidation processes. In the context of shown in Fig. 6. The calculated results are qualitatively com-
DSSC applications, these complexes show insufficiently positive parable to experimental data, although with some differences
oxidation potential to be used with the common redox electro- in the exact energy of the visible transitions. For 1, the lowest
lyte I−/I3− 53 as the dye would not be rapidly regenerated.
energy transition calculated at 581 nm represents MLCT from
Complex 1 is not as easily oxidised as 2 and 3. The reason copper centre to either bipy or dipyrrin ligand, which is typical
for this is due to less conjugation of the dipyrrin structure, for heteroleptic copper(I) complexes.33 The LUMO and sub-
also reflected in comparing HL1·HCl and HL2. This supports sequent levels are localised on dipyrrin and bipy as both
the assignment of the first oxidation for 2 and 3 as dipyrrin ligands show similar electron-donating and withdrawing abil-
based, consistent with the electronic spectroscopy results, ities and therefore one is not predominantly favoured over the
where the lowest excitation is localised on the dipyrrin.
Generally the redox processes for copper(I) complexes are Cu character in contrast to 2 and 3 where the HOMO is calcu-
not chemically reversible as observed here.54
lated to be almost entirely dipyrrin based. The dominant
other. Notably, the HOMO of 1 is calculated to have dominant
Increasing the conjugation of the dipyrrin ligand red shifts absorption feature in the visible spectrum, calculated at
the absorption, improving light harvesting. However, the 399 nm, is shown to be predominantly ligand-centred on
complex becomes too easily oxidised. One method to resolve dipyrrin.
this is to add electron-withdrawing groups to dipyrrin struc-
Complex 2, which features the same bipyridine moiety as 1,
ture, such as CF3, CN, or NO2 groups. This was the strategy but more conjugated dipyrrin, showed a similar LUMO. The
behind the synthesis of HL3, 4 and 5 and indeed we observed minimal electron-withdrawing functionality on the bipyridine
a shift in oxidation potentials to more positive values.
positions these orbitals higher in energy such that the LUMO
Fig. 6 Theoretical UV-Vis spectra for 1–3. Black line represents the experimental spectrum in acetonitrile, red dotted line represents the calculated
spectrum, and red solid line represents the calculated electronic transition.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
is localised on dipyrrin whereas the LUMO+1 is localised on electrochemical properties of dyes (3 and 5) appropriate for
bipy. TD-DFT data suggest the low-energy visible transition has use as DSSC sensitisers. For solar cell fabrication however,
a high contribution of ligand-centred character instead of dyes with acid, rather than ester functionality, are required
charge transfer, suggesting that the metal centre does not sig- and we experienced synthetic difficulty in the isolation of het-
nificantly affect the absorption properties of the complex. This eroleptic complexes consisting of an acid-bipy and dipyrrin
supports experimental data such as fluorescence with a short ligand, due to the ease of deprotonation and reprotonation of
lifetime instead of phosphorescence and the ease of oxidation dipyrrin. In situ methods, based on the procedure of the Con-
compared with 1.
stable group,12,13 were therefore adopted to enable solar cell
For 3, the main visible absorption band shows a high con- performance of these dyes to be assessed without the need to
tribution of copper-to-bipy (MLCT) transition and a small con- isolate the acid complexes.
tribution localised on dipyrrin ligand (27%). A weak MLCT
The electrochemical properties indicate that 3 is too easily
band is predicted at 744 nm, however with very low oscillator oxidised, such that it is poorly designed for use in DSSC with
−
strength such that this band not observed in the experimental the I−/I3 electrolyte. Complex 4 was designed to contain
data. Table 3 indicates that the LUMO is now localised on the additional electron-withdrawing groups (CF3) which results in
bipyridine ligand and the LUMO+1 is localised on the dipyrrin a 0.07 V shift in oxidation potential to more positive values
instead of the other way around for complex 2. This is due to and 5 shows a further 0.14 V shift in oxidation potential. This
electron-withdrawing ester groups on bipyridine (Table 4).
suggests dye regeneration may be possible if the copper sensi-
tiser features a dipyrrin with CF3 groups. The method is shown
in Fig. 7. Table 5 reports the efficiency data of the best cells.
The working electrodes were first immersed into a DMSO
solution of dmdcbpy ligand for 24 hours. The electrodes were
rinsed and dried and then immersed into a solution of [Cu
(CH3CN)4][BF4] and dipyrrin for 24 hours. The colour change
is indicative of coordination occurring on the surface of TiO2.
For comparison, a reference cell was also prepared using the
known homoleptic copper(I) dye (Fig. 8).8
In situ solar studies
The above studies included bipy ligands with carboxylic ester
groups as these give information on the spectroscopic and
Table 3 Percentage contributions from component parts of 1–3 to
selected molecular orbitals. Also quoted are the calculated energies for
these molecular orbitals
% Contribution from
Dipyrrin- Bipy-
MO
Cu-based based
based
Complex MO
energy/eV orbitals
orbitals
orbitals
1
HOMO−1 −4.80
66.58
13.32
20.82
2.19
97.52
99.64
96.0
20.1
11.0
94.55
0.99
0.20
1.70
94.10
9.44
17.25
0.18
89.28
2.12
HOMO
LUMO
−4.72
−1.94
−1.79
−4.73
−2.24
−1.95
68.18
3.26
1.49
0.16
2.30
2.49
76.61
63.58
0.12
4.99
2.06
0.53
LUMO+1
HOMO
LUMO
LUMO+1
HOMO−3 −5.98
2
3
Fig. 7 In situ protocol for producing a copper complex on the surface
of TiO2 working electrode.
3.41
13.95
19.17
99.70
5.73
95.82
3.60
HOMO−1 −5.19
HOMO
−4.78
−2.74
−2.31
−2.29
Table 5 DSSC efficiency data of the best cells. The active layer was
0.16 cm2 and the cells were measured under 1 sun
LUMO
LUMO+1
LUMO+2
95.87
Sample
Voc/mV
Jsc/mA cm−2
ff
η/%
[Cu(dmdcbpy)2][BF4]
530
520
510
520
2.33
0.40
0.91
1.21
0.68
0.62
0.68
0.64
0.83
0.13
0.31
0.41
a
b
c
Table 4 TD-DFT calculated visible absorption wavelengths for 1–3,
illustrating the molecular orbitals involved and their relative contribution
to the absorption
Main charge
transitions
Relative
contribution
(%)
Main visible
absorbance/nm
Complex
MO from
MO to
1
581
HOMO−1
HOMO−1
HOMO
HOMO
HOMO
HOMO−3
HOMO−1
HOMO
LUMO
LUMO+1
LUMO
LUMO
LUMO+1
LUMO
23
36
41
66
34
13
60
27
2
3
541
546
LUMO+2
LUMO+1
Fig. 8 Homoleptic Cu(I) complex [Cu(dmdcbpy)2][BF4].8
This journal is © The Royal Society of Chemistry 2014
Dalton Trans.

View Article Online
Paper
Dalton Transactions
The electrolyte used in this work consisted of 0.4 M LiI,
0.03 M I2, 0.1 M GuNCS, 0.5 M TBP, 1 M DMII in acetonitrile–
Conclusions
valeronitrile (85 : 15).
This work demonstrates the first examples of heteroleptic
The solar cell construction was explored (ESI†) via changes copper(I) complexes, consisting of a bipyridine ligand and a
in (i) the solvent system, (ii) introducing either a [Cu(II)(dipyr- dipyrrin ligand. A series of three dyes has been synthesised
rin)2] homoleptic dye or Cu(I) salt as the copper source, and and fully characterised by electrochemical, photophysical and
(iii) addition of base to deprotonate the dipyrrin prior to computational techniques, with two other complexes obtained
coordination. Changes in solvent system can affect the overall but difficult to fully study due to poor stability and low yield.
efficiency of the solar cell. For instance, dichloromethane or Copper(I) materials are highly labile towards ligand exchange
acetonitrile were used to dissolve the copper salt and dipyrrin. which inevitably poses problems with stability, especially
If dichloromethane were used in the dye bath the photocurrent when they are in solution for long periods of time. Interest-
was lower and the photovoltage was higher than if acetonitrile ingly, the complexes did show improved stability with
were used. This could suggest the energy of the acceptor states enhanced steric constraint caused by the bulkiness of the
in the TiO2 is higher for the DCM cell, affecting charge injec- dipyrrin ligand.
tion and reducing the photocurrent of the cell.
Increasing the π-conjugation of the dipyrrin ligand shifts
One method adopted to instigate ligand exchange whilst in the visible bands to lower energies. Complex 1 also shows a
solution used a homoleptic [Cu(II)(dipyrrin)2] species. Several small band at 498 nm corresponding to MLCT transition,
attempts were made to reduce the copper in situ; Zn/Hg, NaBH4 according to theoretical calculations. Complex 2 and 3 showed
or methanol as the reducing agent. The most successful very similar absorption properties to the free dipyrrin ligand.
attempt used methanol. The colour change on the surface was Lifetime measurements suggest that complex 3 shows fluo-
either deep red or deep yellow, which is indicative of surface rescence rather than phosphorescence, with a nanosecond
bound dye. However, the photocurrent voltage results were timescale.
lower than using a copper(I) species as the starting reagent.
According to electrochemical data, complex 2 and 3 are
Table 5 represents the highest efficiencies achieved in this easily oxidised, which may cause a problem for dye regener-
work. Sample a used the addition of [Cu(CH3CN)4][BF4] and ation within a DSSC device. Decreasing the conjugation
HL1·HCl to the bound acid-bipy anode in dichloromethane, b (complex 1) resulted in an increase in HOMO–LUMO gap,
used the addition of [Cu(CH3CN)4][BF4] and HL2 in aceto- which ultimately blue-shifts the charge transfer transition and
nitrile, and c used the addition of [Cu(CH3CN)4][BF4] and HL3 reduces light harvesting. Complex 4 features CF3 groups which
in acetonitrile.
shift the oxidation potential to more positive values.
The charge injection should not greatly change between
Despite the low oxidation potentials in situ solar studies did
samples a, b, and c as the LUMO should be localised on the show that the complexes assembled function as sensitisers,
acid-bipy ligand. The improved photocurrent could be a result of with efficiencies of 0.13–0.41%. The study shows that
improved dye regeneration which would be expected for L3; increasing the conjugation of dipyrrin ligand (L1 to L2)
addition of electron-withdrawing groups (CF3) to the dipyrrin improved the efficiency. It is likely that this is due to better
ligand should reduce the HOMO energy level, causing the light harvesting. The highest efficiency was obtained for
complex to be less easily oxidised and therefore improve dye the copper complex with L3 ligand. The electron-withdrawing
regeneration. Sample c has indeed shown an increase in the trifluoromethyl groups increased the oxidation potential,
cell’s efficiency in comparison to samples a and b, through both which possibly improved dye regeneration. Although the
an improved photocurrent and open-circuit voltage (Fig. 9).
absolute value of these efficiencies remains low, we have
demonstrated the concept of using a chromophoric ancillary
ligand on a heteroleptic Cu(I) complex to enhance light
harvesting. We hope this will stimulate further molecular
design in the active field of DSSC sensitisers based on earth-
abundant elements.18
Acknowledgements
We thank EPSRC (Supergen Excitonic Solar Cells) for financial
support. We also thank the COSMIC facility at Edinburgh Uni-
versity for carrying out lifetime measurements. This work has
made use of the resources provided by the EaStChem Research
Computing Facility (http://www.eastchem.ac.uk/rcf). This facil-
ity is partially supported by the eDIKT initiative (http://www.
edikt.org).
Fig. 9 Photocurrent voltage curves of in situ studies. The cells were all
tested under the same conditions and the best cells are reported.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
22 N. Armaroli, G. Accorsi, F. Cardinali and A. Listorti, Photo-
chemistry and Photophysics of Coordination Compounds I,
Springer, 2007, p. 69.
Notes and references
1 Z. Liu, M. F. Qayyum, C. Wu, M. Whited, P. I. Djurovich,
K. O. Hodqson, B. Hedman, E. I. Solomon and 23 R. Sakamoto, S. Kusaka, Y. Kitagawa, M. Kishida,
M. E. Thompson, J. Am. Chem. Soc., 2011, 133, 3700–3703.
2 I. Satoshi, H. Masashi, K. Isao, Y. Masataka, H. Mikio and
O. Masahisa, J. Mater. Chem. C, 2013, 1, 542–551.
3 H. Masashi, I. Satoshi, Y. Masataka, K. Isao, H. Mikio
and O. Masahisa, J. Am. Chem. Soc., 2011, 133, 10348–
10351.
4 R. D. Costa, D. Tordera, E. Ortí, H. J. Bolink, J. Schönle,
S. Graber, C. E. Housecroft, E. C. Constable and
J. A. Zampese, J. Mater. Chem., 2011, 21, 16108–16118.
5 Y.-M. Wang, F. Teng, Y.-B. Hou, Z. Xu and Y.-S. Wang, Appl.
Phys. Lett., 2005, 87, 233512–233515.
M. Hayashi, Y. Takara, M. Tsuchiya, J. Kakinuma,
T. Takeda, K. Hirata, T. Ogino, K. Kawahara, T. Yagi,
S. Ikehira, T. Nakamura, M. Isomura, M. Toyama,
S. Ichikawa, M. Okumura and H. Nishihara, Dalton Trans.,
2012, 41, 14035–14037.
24 F. López Arbeloa, J. Bañuelos, V. Martinez, T. Arbeloa
and I. López Arbeloa, Int. Rev. Phys. Chem., 2005, 24,
339–374.
25 T. A. Fayed, Reviews in Fluorescence, Springer, 2009, ch. 4,
pp. 75–111.
26 R. Sakamoto, S. Kusaka, Y. Kitagawa, M. Kishida,
M. Hayashi, Y. Takara, M. Tsuchiya, J. Kakinuma,
T. Takeda, K. Hirata, T. Ogino, K. Kawahara, T. Yagi,
S. Ikehira, T. Nakamura, M. Isomura, M. Toyama,
S. Ichikawa, M. Okumura and H. Nishihara, Dalton Trans.,
2012, 41, 14035–14037.
6 N. Alonso-Vante, J.-F. Nierengarten and J.-P. Sauvage,
J. Chem. Soc., Dalton Trans., 1994, 1649–1654.
7 S. Sakaki, T. Kuroki and T. Hamada, J. Chem. Soc., Dalton
Trans., 2002, 840–842.
8 T. Bessho, E. C. Constable, M. Graetzel, A. H. Redondo,
C. E. Housecroft, W. Kylberg, Md. K. Nazeeruddin, 27 C. Bronner, S. A. Baudron, M. W. Hosseini, C. A. Strassert,
M. Neuburger and S. Schaffner, Chem. Commun., 2008,
3717–3719.
9 E. C. Constable, A. H. Redondo, C. E. Housecroft,
A. Guenet and L. De Cola, Dalton Trans., 2010, 39, 180–184.
28 S. R. Halper, M. R. Malachowski, H. M. Delaney and
S. M. Cohen, Inorg. Chem., 2004, 43, 1242–1249.
M. Neuburger and S. Schaffner, Dalton Trans., 2009, 6634– 29 G. Li, P. G. Bomben, K. C. D. Robson, S. I. Gorelsky,
6644.
C. P. Berlinguette and M. Shatruk, Chem. Commun., 2012,
48, 8790–8792.
10 B. Bozic-Weber, E. C. Constable, C. E. Housecroft,
M. Neuburger and J. R. Price, Dalton Trans., 2010, 39, 30 K. Hanson, A. Tamayo, V. V. Diev, M. T. Whited,
3585–3594.
P. I. Djurovich and M. E. Thompson, Inorg. Chem., 2010,
49, 6077–6084.
11 C. L. Linfoot, P. Richardson, T. E. Hewat, O. Moudam,
M. M. Forde, A. Collins, F. White and N. Robertson, Dalton 31 X. Liu, H. Nan, W. Sun, Q. Zhang, M. Zhan, L. Zou, Z. Xie,
Trans., 2010, 39, 8945–8956.
X. Li, C. Lu and Y. Cheng, Dalton Trans., 2012, 41, 10199–
12 B. Bozic-Weber, E. C. Constable, C. E. Housecroft,
10210.
P. Kopecky, M. Neuburger and J. A. Zampese, Dalton Trans., 32 J. B. Paine, The Porphyrins, D. Dolphin, Academic Press,
2011, 40, 12584–12594. New York, 1978, vol. 1.
13 B. Bozic-Weber, V. Chaurin, E. C. Constable, 33 P. Bamfield, A. W. Johnson and J. Leng, J. Chem. Soc., 1965,
C. E. Housecroft, M. Meuwly, M. Neuburger, J. A. Rudd, 7001–7005.
E. Schönhofer and L. Siegfried, Dalton Trans., 2012, 41, 34 T. S. Teets, D. V. Partyka, A. J. Esswein, J. B. Updegraff III,
14157–14169.
M. Zeller, A. D. Hunter and T. G. Gray, Inorg. Chem.
Commun., 2007, 46, 6218–6220.
35 O. Moudam, in LCC-Toulouse, LCC Toulouse, 2007.
14 Y.-J. Yuan, Z.-T. Yu, J.-Y. Zhang and Z.-G. Zou, Dalton
Trans., 2012, 41, 9594–9597.
15 B. Bozic-Weber, S. Y. Brauchli, E. C. Constable, S. O. Fürer, 36 N. R. Kelly, S. Goetz, S. R. Batten and P. E. Kruger, CrystEng-
C. E. Housecroft, F. J. Malzner, I. A. Wright and
J. A. Zampese, Dalton Trans., 2013, 42, 12293.
16 N. Robertson, ChemSusChem, 2008, 977.
Comm, 2008, 10, 68–78.
37 Y. Takada, K. Nomura and S. Matsubara, Org. Lett., 2010,
12, 5204–5205.
17 M. Sandroni, M. Kayanuma, A. Planchat, N. Szuwarski, 38 V.-M. Mukkala and J. J. Kankare, Helv. Chim. Acta, 1992, 75,
E. Blart, Y. Pellegrin, C. Daniel, M. Boujtita and F. Odobel,
Dalton Trans., 2013, 42, 10818.
18 B. Bozic-Weber, E. C. Constable and C. E. Housecroft,
Coord. Chem. Rev., 2013, 257, 3089.
1578–1592.
39 J. L. Sessler, L. R. Eller, W.-S. Cho, S. Nicolaou, A. Aguilar,
J. T. Lee, V. M. Lynch and D. J. Magda, Angew. Chem., Int.
Ed., 2005, 44, 5989–5992.
19 D. V. Scaltrito, D. W. Thompson, J. A. O’Callaghan and 40 G. Ulrich, S. b. Goeb, A. De Nicola, P. Retailleau and
G. J. Meyer, Coord. Chem. Rev., 2000, 208, 243–266.
20 N. Amaroli, Chem. Soc. Rev., 2001, 30, 113–119.
21 R. M. Williams, L. De Cola, F. Hartl, J.-J. Lagref,
R. Ziessel, J. Org. Chem., 2011, 76, 4489–4505.
41 J. M. Mellor, A. H. El-Sagheer, E.-S. H. El-Tamany and
R. N. Metwally, Tetrahedron, 2000, 56, 10067–10074.
J.-M. Planeix, A. De Cian and M. W. Hosseini, Coord. Chem. 42 N. Zanatta, R. Barichello, H. G. Bonacorso and
Rev., 2002, 230, 253–261.
M. A. P. Martins, Synthesis, 1999, 765–768.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans.

View Article Online
Paper
Dalton Transactions
43 I. Noviandri, K. N. Brown, D. S. Fleming, P. T. Gulyas, 50 E. C. Constable, C. E. Housecroft, P. Kopecky,
P. A. Lay, A. F. Masters and L. Phillips, J. Phys. Chem. B,
1999, 103, 6713–6722.
E. Schönhofer and J. A. Zampese, CrystEngComm, 2011, 13,
2742.
44 P. J. Hay and W. R. Wadt, J. Chem. Phys., 1985, 82, 270–283.
45 W. R. Wadt and P. J. Hay, J. Chem. Phys., 1985, 82, 284–
298.
46 P. J. Hay and W. R. Wadt, J. Chem. Phys., 1985, 82, 299–310.
47 A. D. Becke, J. Chem. Phys., 1993, 98, 5648–5652.
48 Avogadro 1.1.0, http://sourceforge.net/projects/avogadro/
51 K. M. Smith, Porphyrin and metalloporphyrin, Elsevier Scien-
tific Publishing, New York, 1975.
52 Y. Kajiwara, A. Nagai and Y. Chujo, Bull. Chem. Soc. Jpn.,
2011, 84, 471–481.
53 G. Boschloo, E. A. Gibson and A. Hagfeldt, J. Phys. Chem.
Lett., 2011, 2, 3016–3020.
49 A. Hernandez Redondo, E. C. Constable and 54 G. F. Manbeck, W. W. Brennessel and R. Eisenberg, Inorg.
C. E. Housecroft, Chimia, 2009, 63, 205. Chem., 2011, 50, 3431–3441.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2014