Substituted [Cu(I)(POP)(bipyridyl)] and related complexes: Synthesis, structure, properties and applications to dye-sensitised solar cells

Article abstract of DOI:10.1039/c0dt00190b

The synthesis and subsequent spectroscopic, electrochemical, photophysical and computational characterisation of a series of heteroleptic Cu(i) complexes of general formula: [CuPOP{4,4′(R)-bipyridyl}][BF₄] and [CuPOP{4,4′,6,6′(R)-bipyridyl}][BF₄] is described (POP = bis{2-(diphenylphosphanyl)phenyl} ether; R = Me, CO₂H, CO ₂Et. The steric constraint imposed by the POP ligand can impede distortion towards square planar geometry upon MLCT excitation or oxidation and this is explored in the context of varying substituents on the bipyridyl ligand. The insight gained opens new avenues for design of functional Cu(i) systems suitable for photophysical and photoelectrochemical applications such as sensitisers for dye-sensitised solar cells (DSSCs). The Royal Society of Chemistry 2010.

Full text of DOI:10.1039/c0dt00190b

View Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
Substituted [Cu(I)(POP)(bipyridyl)] and related complexes: Synthesis,
structure, properties and applications to dye-sensitised solar cells†‡
Charlotte L. Linfoot, Patricia Richardson, Tracy E. Hewat, Omar Moudam, Michael M. Forde, Anna Collins,
Fraser White and Neil Robertson*
Received 23rd March 2010, Accepted 8th July 2010
DOI: 10.1039/c0dt00190b
The synthesis and subsequent spectroscopic, electrochemical, photophysical and computational
characterisation of a series of heteroleptic Cu(I) complexes of general formula:
[CuPOP{4,4¢(R)-bipyridyl}][BF4] and [CuPOP{4,4¢,6,6¢(R)-bipyridyl}][BF4] is described (POP =
bis{2-(diphenylphosphanyl)phenyl} ether; R = Me, CO2H, CO2Et. The steric constraint imposed by the
POP ligand can impede distortion towards square planar geometry upon MLCT excitation or
oxidation and this is explored in the context of varying substituents on the bipyridyl ligand. The insight
gained opens new avenues for design of functional Cu(I) systems suitable for photophysical and
photoelectrochemical applications such as sensitisers for dye-sensitised solar cells (DSSCs).
could present detrimental practical implications in the long term.
Introduction
Research into the field of first row transition metals has been
Recent research into the electrochemical and photophysical prop-
limited due to their often extremely short-lived excited states
erties of phenanthroline- and bipyridine-based Cu(I) complexes1–4
compared to 2nd and 3rd row transition metal complexes, arising
has revealed their potential for application in dye-sensitised solar
from the low lying metal d-d excited states which allow non-
cells (DSSCs),2,3,5–7 organic light emitting diodes (OLEDs)4,8 and
radiative decay of the MLCT excited state. This is a limiting factor
light-emitting electrochemical cells (LECs).4,8 The most successful
in terms of DSSC functionality as charge injection from the dye is
not efficiently achieved.13,14
DSSCs developed to date are based on Ru(II) containing sensitiser
molecules,9 some of which are being commercially developed for
niche markets.10 Lower efficiency DSSCs have also been con-
structed using transition metal complexes of Os(II), Pt(II), Re(I),
Cu(I), Fe(II)11 and Ni(II).12 The use of third row transition metal
sensitisers has been particularly successful due to their ability to
display long-lived excited states following MLCT absorption in
the visible. However low abundance, and toxicity considerations
To overcome this problem, complexes with a strong crystal
field have been studied, aimed at raising the energy of any d-d
excited states. This strategy has been explored by Ferrere13,15,16 with
work on Fe(II) sensitiser molecules analogous to known Ru(II)
sensitisers that exhibit strong crystal-field splitting.9 Solar cell
function was observed but it was significantly reduced compared
to the corresponding Ru(II) sensitiser. Similarly low efficiencies
using Ni(II)(dithiolate)(diimine) complexes as DSSC sensitiser
molecules has recently been presented by our group, with limited
solar cell function being observed,12 also attributed to short excited
state lifetimes.
School of Chemistry and EaStChem, University of Edinburgh, King’s
Buildings, Edinburgh, UK. E-mail: neil.robertson@ed.ac.uk; Fax: +44 131
6504743; Tel: +44 131 650 4755
†
‡ Electronic supplementary information (ESI) available: Figure
A. Schematic of the POP chelate ring; Figure B. Mercury plots
of [Cu(dmbpy)2Cl][BF4] demonstrating p–stacking interactions; Figure
C. Mercury plots of 1 demonstrating no p–stacking interactions; Figure D.
(a) Absorbance Spectra for 1 and 3; (b) Absorbance Spectra for 4, 6, 7 and
9; Figure E. Excitation and Emission Spectra; Figure F. Selected Molecular
Orbital Images for 4; Figure G. Selected Molecular Orbital Images for 5;
Figure H. Selected Molecular Orbital Images for 7; Figure I. Selected
Molecular Orbital Images for 9; Figure J. Molecular Orbital Diagram
for 4; Figure K. Molecular Orbital Diagram for 7; Figure L. Molecular
Orbital Diagram for 5; Figure M. I–V Curve for 5; Figure N. I–V
Curve for 5 with Cheno treatment; Figure O. I–V Curve for 5 with
TiCl4 treatment; Table A. Crystallographic data for complexes 6, 9 and
10; Table B. Percentage contributions from component parts of 4 to
selected molecular orbitals; Table C. TDDFT calculated visible absorption
wavelengths for 4; Table D. Percentage contributions from component
parts of 5 to selected molecular orbitals; Table E. TDDFT calculated
visible absorption wavelengths for 5; Table F. Percentage contributions
from component parts of 7 to selected molecular orbitals; Table G. TDDFT
calculated visible absorption wavelengths for 7; Table H. Percentage
contributions from component parts of 9 to selected molecular orbitals;
Table I. TDDFT calculated visible absorption wavelengths for 9. CCDC
reference numbers 771439–771443. For ESI and crystallographic data in
CIF or other electronic format see DOI: 10.1039/c0dt00190b
In contrast, Cu(I) complexes can display longer excited-state
lifetimes due to their d10 configuration, leading to a power-
conversion efficiency of 2.6%.17 To date however, Cu(I) DSSC
research has been limited to homoleptic sensitisers,2,3,5,7 and in
this work we extend this to an initial investigation of heteroleptic
Cu(I) complexes. In analogous Ru(II) dyes, the heteroleptic ligands
control directionality of the MLCT transition to minimise recom-
bination losses and also allow tuning of the HOMO energy.18
During the redox process; Cu(I)/Cu(II),19,20 a conformational
geometry change typically occurs from distorted tetrahedral to
tetragonally-flattened, respectively.21 Research investigating Cu(I)
bis-phenanthroline complexes1,8,20–22 has revealed that substituents
in the 2- and 9-positions of the phenanthroline ligand acts as
“blocking groups” to sterically constrain the molecule.1,20,21
A
similar approach has been utilised for DSSC Cu(I) sensitisers
by including substituents groups in the 6,6¢positions of the
bipyridine, however this leads to increased synthetic complexity
of the bipyridine ligands which must also be 4,4¢-substituted
with acid groups to bind to TiO2.2,3,5–7 In this work, we
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 8945–8956 | 8945
©

View Online
explore the use of a geometrically inflexible co-ligand bis{2-
(diphenylphosphanyl)phenyl} ether (POP), enabling the simpler
4,4¢-(CO2H)2-bipyridine to be used as the TiO2-binding ligand,
without requiring 6,6¢-substitution and therefore simplifying the
intensive bipyridine synthesis required for previous Cu(I)-based
sensitisers. POP has been used effectively in OLED studies
due to this rigid and inflexible geometry it can impose on the
complex.13,23,24
The work outlined here reports the synthesis and subsequent
electrochemical and photophysical characterisation of a series of
homo- and heteroleptic Cu(I)(bipyridine) complexes (Fig. 1 and
2), enabling a comparison of structural rigidity imposed by the
POP and variously-substituted bipyridine ligands. Included is the
first example of a heteroleptic Cu(I) complex investigated as a
DSSC sensitiser.
bipyridine (tmbpy) and bis{2-(diphenylphosphanyl)phenyl}
ether (POP), and the starting complexes [Cu(MeCN)4][BF4] and
[Cu(POP)(MeCN)2][BF4] were synthesised using modified litera-
ture methods as described in the Experimental section. 4,4¢,6,6¢-
tetracarboxy-2,2¢-bipyridine (tcbpy) and 4,4¢,6,6¢-tetra(CO2Et)-
2,2¢-bipyridine (tecbpy) are novel ligands that were synthesised
via oxidation of the tmbpy ligand. The reaction schemes
shown in Fig. 1 and 2 demonstrate how the Cu(I) complexes
were synthesised by simple addition, with stirring, of the
relevant starting materials in stoichiometric amounts, to the
appropriate solvent at room temperature. It should be noted
that [Cu(tcbpy)2][BF4] 2, and [Cu(POP)(tcbpy)][BF4] 8 were
1
synthesised and characterised by H NMR under nitrogen, but
were not air stable and so were not further investigated. The
methyl and ester analogues have been synthesised alongside the
acids, despite not being directly applicable for use in a DSSC, as
they can give insight into the properties of the acid complexes
while being more convenient for detailed characterisation due
to better solubility. The Cu(I) complexes 1, 3–7 and 9 were
Results and discussion
Syntheses
1
characterised by H NMR, ESI-MS, +FAB-MS and elemental
The ligands 4,4¢-dicarboxy-2,2¢-bipyridine (dcbpy), 4,4¢-
di(CO2Et)-2,2¢-bipyridine (decbpy), 4,4¢,6,6¢-tetramethyl-2,2¢-
analysis, verifying molecular structures.
Fig. 1 Synthetic reaction scheme for homoleptic complexes reported in this work. All procedures carried out under a nitrogen atmosphere. (i) 1 degassed
chloroform; 2 dry, degassed MeOH; 3 dry, degassed DCM.
Fig. 2 Synthetic reaction scheme for heteroleptic complexes reported in this work. (1) 4, 5, 6, 8 acetone; 7 anhydrous acetonitrile (ii) DCM (iii) acetone.
8946 | Dalton Trans., 2010, 39, 8945–8956
This journal is
The Royal Society of Chemistry 2010
©

View Online
˚
Table 1 Selected bond lengths (A) and bond angles (deg) of complexes
[Cu(dmbpy)2Cl][BF4] and 1
[Cu(dmbpy)2Cl][BF4]
1
Cu–N(1A)
Cu–N(1B)
Cu–N(2A)
Cu–N(2B)
N(1A)–Cu–N(1B)
N(2A)–Cu–N(2B)
N(1A)–Cu–N(2B)
N(2A)–Cu–N(1B)
N(1A)–Cu–N(2A)
N(1B)–Cu–N(2B)
2.154(2)
2.082(2)
2.001(2)
1.981(2)
100.97(8)
169.34(8)
91.78(8)
96.60(8)
79.07(8)
79.64(8)
2.055(2)
2.056(2)
2.055(2)
2.056(2)
117.63(9)
117.63(9)
134.37(9)
134.37(9)
80.60(13)
80.63(13)
Structure analysis
The single crystal structures of [Cu(II)(dmbpy)2Cl][BF4] and
compounds 1, 6, 9 and 10 have been determined allowing insight
into the structural features of these complexes (Table 1 and 2), in
particular the blocking group effects. All crystals were grown from
solutions open to air. Despite obtaining a crystal structure of 10,
we were unable to isolate the complex in a pure form and therefore
further characterisation was not carried out. In attempting to
crystallise the 4,4¢-dimethyl-2,2¢-bipyridine analogue of 1, the
complex oxidised in air along with coordination of chlorine to
give [Cu(II)(dmbpy)2Cl][BF4] (Fig. 3a). In contrast, the structure
of complex 1 (Fig. 3b) illustrates the stabilisation of the Cu(I)
oxidation state by addition of methyl groups in the 6,6¢-positions of
the bipyridine. The angle between the planes of the two bipyridyls
is 72.18◦ giving a pseudo-tetrahedral geometry and suggesting
considerable flexibility of this geometric parameter.
The crystal structures of the heterolpetic complexes (6, 9 and
10) shown in Fig. 4 illustrate the role of POP in stabilisation
in air of a Cu(I) oxidation state over Cu(II) as the complexes
are forced to adopt distorted tetrahedral conformations with
reduced structural flexibility. The angles between the bipyridyl
plane and the P–Cu–P plane for these complexes range from 84.05
to 88.61◦ indicating a geometry closer to tetrahedral than that of 1
(Table 2). Additionally, the bite angle of POP in each case remains
comparatively close to that expected for an ideal tetrahedron
(Table 2). The blocking functionality of POP arises as one of
Fig. 3 Mercury plots of (a) [Cu(dmbpy)2Cl][BF4]; and (b) 1. Ther-
mal ellipsoids are drawn at the 50% probability level. [BF4] counter
ion omitted for clarity. Selected bond lengths and bond angles for
˚
[Cu(dmbpy)2Cl][BF4] involving the chloride atom: Cu–Cl: 2.2770(7) A;
Cl–Cu–N(1A) 123.54(6)◦; Cl–Cu–N(2A) 93.08(6)◦; Cl–Cu–N(1B)
135.48(6)◦; Cl–Cu–N–(2B) 96.57 (6)◦.
the phenyl rings attached to the phosphorus atoms is constrained
to be over the bipyridyl ligand resulting in H-phenyl-to-bipyridyl
˚
˚
˚
plane distances of 2.891 A, 2.941 A, 2.790 A for 6, 9 and 10
respectively that prevent structural distortion. This geometrical
arrangement is caused by the rigid nature of the chelating ring
(Cu–P–C–C–O–C–C–P). Among the three structures presented
here (see supplementary information) and also those reported
˚
Table 2 Selected bond lengths (A) and bond angles (deg) from single crystal X-ray crystallography for 6, 9 and 10.; and from solvated DFT calculations
for 9
X
6
9: single crystal X-ray structure
9: DFT geometry optimisation
10
Cu–N(1)
Cu–N(2)
Cu–P(1)
Cu–P(2)
N(1)–Cu–N(2)
N(1)–Cu–P(1)
N(2)–Cu–P(1)
N(1)–Cu–P(2)
N(2)–Cu–P(2)
P(1)–Cu–P(2)
bpy plane–PCuP planea
Cu–P(2)–Cb
2.037 (2)
2.0621(19)
2.2213(6)
2.2465 (6)
80.30(8)
128.87(6)
111.47(5)
109.74(6)
112.26(5)
110.40(2)
84.05
2.068(2)
2.093(2)
2.2482(7)
2.2756(7)
79.74(8)
121.08(6)
118.49(6)
108.47(6)
113.65(6)
111.77(2)
85.77
2.105
2.107
2.400
2.411
2.103(2)
2.070(2)
2.2158(7)
2.2851(7)
80.16(8)
128.91(6)
119.24(6)
102.14(6)
102.75(6)
116.30(3)
88.61
79.64
121.96
115.30
106.85
112.25
114.50
N/A
114.38(8)
115.58(8)
115.89
107.54(9)
a for 6 C C17; for 9 C C9; for 10 C C35.
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 8945–8956 | 8947
©

View Online
Fig. 4 Mercury plots of (a) 6; (b) 9 and (c) 10. Thermal ellipsoids are drawn at the 50% probability level. 10 lies on an inversion centre. The [BF4] counter
ions are omitted for clarity.
previously, there is little variation between the angles within the
chelate ring providing strong support for the lack of flexibility
in the POP coordination geometry and hence the rigidity of the
whole molecule. In general, the structures shows little geometric
deviation compared with eleven POP containing Cu(I)-structures
described in the literature.14,24–28 Furthermore, the structures of 6
and 9 clearly demonstrate that in a heteroleptic complex with an
appropriate co-ligand, 6,6¢-substituted bipyridine ligands are not
required to maintain an air-stable Cu(I)-complex. This enables the
use in Cu(I) DSSC sensitisers of the straightforward dcbpy ligand
widely exploited in Ru sensitisers and bypassing the need for more
complex multiply substituted bipyridyl ligands.
The nature of each electronic transition was determined by
visual inspection (Argus Lab34) of each contributing molecular
orbital as an isosurface map. Many of the calculated occupied
and virtual orbitals were seen to be of mixed character, with
electron density simultaneously present on both the ligand and
metal. However, the electronic transitions could broadly be
classified into three distinct types; metal-to-ligand-charge-
transfer (MLCT), where the occupied orbital is predominantly
metal based and the virtual orbital ligand based, either on the POP
or bpy; ligand-ligand (LL), where the transition involves occupied
and virtual orbitals on the same ligand; and ligand-to-ligand-
charge-transfer (LLCT), where the transition involves occupied
and virtual orbitals on different ligands. These are discussed
further in the context of the experimental results below.
DFT and TDDFT calculations
UV/Vis absorption spectroscopy
All calculations were carried out using Gaussian 0329 with the
B3LYP/LANL2DZ30–33 functional and basis set with induced
solvent effects. This level of theory has been used successfully in
the literature to calculate the structural and electronic properties
of a number of coordination complexes.25 Calculation details
are outlined in full in the Experimental section. The optimised
geometry for 9 is in very good agreement with the structure
obtained crystallographically (Table 2), but as expected the DFT
calculated solvated bond lengths are longer than those in the
crystal structure. Packing forces present in the calculated liquid-
phase are weaker than those in the solid-phase. The DFT
calculated gas-phase and solution-phase structures for 4, 5, 7 and 9
produce identical overall structure energies and comparable bond
distances and angles.
Time dependent density functional theory (TDDFT), also
at the B3LYP/LANL2DZ level of theory, was used to probe
the electronic transitions that give rise to the visible and near-
UV absorption of each complex. Although current TDDFT
implementations have a tendency to underestimate the energy
of charge-transfer interactions, the method has been shown to
provide useful qualitative information on the electronic structure
and photophysical behaviour of similar coordination complexes.25
Seventy singlet-singlet transitions were calculated with solvent,
for each complex. Analysis of the components of the TDDFT
expansion shows that the majority of predicted transitions are
not well described by a single electron-promotion, but involve
transitions between several different orbital pairs.
The experimental and calculated UV-Vis absorption spectra are
shown in Fig. 5, with the main absorption bands and related
extinction coefficients given in Table 3. For each complex the
experimental spectra show two distinct groupings; very strong
absorption in the near uv and weaker absorption bands in the
visible that are sensitive to the nature of the biypridine substituents.
From previous work,1 it is known that the absorption peak of the
p→p* intra-bipyridine transition occurs between 300–320 nm.
Each complex in the homoleptic series exhibits an absorption peak
in this range (Table 3), however this absorption peak cannot be
resolved for the heteroleptic series as it is believed to be masked by
the stronger POP ligand UV absorption band. In the absorbance
spectra collected for 4 and 7 the shoulder present at 320 nm may
possibly be attributed to the bipryidyl absorption.
The TDDFT calculations reproduce the experimental spectra
reasonably well in the UV region, predicting a large number of
transitions with significant oscillator strength (Fig. 5). Examina-
tion of the molecular orbitals involved in each transition shows
that the UV absorption bands are dominated by p→p* transitions
on both the bipyridine and the POP ligands (see supplementary
information‡).
The longer wavelength bands have been assigned as Cu(I) to
bipyridyl MLCT transfer bands based upon comparison with
literature. This assignment is supported by the calculated energies
and deduced characters of the frontier orbitals (Table 4) and the
percentage contribution to the lowest energy MLCT bands, of
8948 | Dalton Trans., 2010, 39, 8945–8956
This journal is
The Royal Society of Chemistry 2010
©

View Online
Table 3 Absorbance measurements for all complexes. 1 and 4 - 9 carried out in MeCN; 3 carried out in DCM
Absorption/lmax/nm (emax(¥103M-1cm-1))
Complex
Intraligand transitions
MLCT
[Cu(tmbpy)2]
[Cu(tecbpy)2]
1
3
300 (25.8)
315 (32.0)
358 (1.4)
477 (4.9)
449 (3.5)
594 (7.4)
[CuPOP(MeCN)2]
[CuPOP(dmbpy)]
[CuPOP(dcbpy)]
[CuPOP(decbpy)]
[CuPOP(tmbpy)]
[CuPOP(bpym)]
—
4
272 (14.2)
282 (22.6)
276
290 (19.7)
288 (25.5)
294 (16.1)
248 (25.1)
369 (3.2)
416
424 (2.4)
358 (2.2)
403 (2.0)
5
6
246 (26.6)
250 (26.3)
247 (35.5)
7
460 (0.2)
9
Fig. 5 Theoretical UV/Vis absorption spectra against experimental absorption spectra for (a) 4; (b) 5; (c) 7; (d) 9.Where red solid columns = calculated
electronic transition; red dotted line = calculated spectra; black solid line = solution spectrum of (a) 4, (b) 6, (c) 7 and (d) 9 in MeCN.
each orbital pair (Table 5) in the TDDFT expansion. It must be
noted that TDDFT has a tendency to over stabilise, and indeed the
calculated MLCT absorptions are consistently predicted at lower
energy than those seen experimentally.35
orbital images for the LUMO of 4 and 5. 5 displays increased
delocalisation and therefore a lower energy LUMO (Table 4)
which subsequently reduces the required MLCT energy, causing
the absorption to take place at longer wavelength.
For the 4,4¢-disubstituted complexes the TDDFT calculations
agree with experimental data in placing the MLCT absorption
for 5 at lower energy than that of 4. The proposed explanation
for this, based upon the previous experimental data, is that 5
has a lower lying LUMO as a result of the electron-withdrawing
acid substituents. This proposal can be further confirmed upon
inspection of the TDDFT results with Fig. 6 showing molecular
The MO diagram for 4 looks almost as expected for a pseudo-
tetrahedral complex (see supplementary information‡). Three
MOs are close in energy (HOMO,
HOMO-1 and HOMO-2) and approximate as t2 (dxy, dxz and
dyz), and another two MOs at lower energy (HOMO-4 and
HOMO-5) which approximate as e (dx2–y2 and dz2). The MLCT
transition can therefore be said to be t2 to p* (bipyridine) in
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 8945–8956 | 8949
©

View Online
Table 4 Percentage contributions from component parts of 4, 5, 7 and 9 to selected molecular orbitals. Also quoted are the calculated energies for these
molecular orbitals
% Contribution from
Complex
MO
MO character
MO energy/eV
Cu-based orbitals
POP-based orbitals
Bpy-based orbitals
4
HOMO-1
HOMO
LUMO
HOMO-3
HOMO-2
HOMO
dxz or dyz
dxz,or dyz
p*-bpy
-5.99
-5.87
-2.27
-6.60
-6.42
-6.08
-3.16
-2.60
-6.27
-6.03
-2.12
-6.33
-6.07
-2.92
-2.27
63.78
44.34
1.48
14.09
38.48
36.98
2.57
8.80
50.38
5.68
83.47
54.01
58.41
4.25
27.42
5.28
92.84
2.44
7.51
4.61
93.18
99.07
9.27
19.92
94.32
28.13
5.51
5
LUMO
p*-bpy
p*-bpy
dxy
dxz,or dyz
p*-bpy
LUMO+1
HOMO-2
HOMO-1
LUMO
HOMO-1
HOMO
0.40
0.53
7
9
51.93
61.10
1.44
59.87
38.12
1.47
38.80
18.98
4.24
12.00
56.37
3.80
LUMO
LUMO+1
p*-bpy
p*-bpy
94.73
94.62
1.97
3.41
Table 5 TD-DFT calculated visible absorption wavelengths for 4, 5, 7 and
9, indicating the molecular orbitals involved and their relative contribution
to the absorption
interactions with the ligand-based orbitals. Inspection of Table 5
demonstrates that the LUMO and LUMO+1 orbitals for all four
of the complexes are bpy-based, and that the HOMO to HOMO-3
orbitals are Cu(I)- and POP-based, as predicted.
This is of importance as the main charge-transfer transitions
occur between these orbitals (see Table 5), indicating MLCT across
the molecule in the desired direction for DSSC function of the acid
analogous: towards the acid groups bound to the TiO2.
Main charge
transitions
Main visible
Complex absorbance/nm Mo from
Relative
contribution
MO to
4
5
416
HOMO-1
HOMO
HOMO
HOMO-3
HOMO-2
HOMO
LUMO
LUMO
LUMO
LUMO
LUMO
16%
84%
100%
29%
10%
Emission spectroscopy
509
420
Excitation and emission spectra were investigated for 4, 7 and
9 in ethanol at room temperature and in a frozen glass at
77 K. For comparison, data were also collected for the uncom-
plexed dmbpy and tmbpy ligands, and the precursor complex
[CuPOP(MeCN)2][BF4] (Table 6).Maxima values are given in
Table 6 with full spectra given in the supplementary information.
Complex 9 was found to be emissive only at 77 K and not at room
temperature. Complexes 4 and 7 both showed room-temperature
emission with a significant increase in intensity upon freezing
accompanied by some shifting of the absorption and emission
maxima (Fig. 7). This behaviour is consistent with emission from
a triplet MLCT state in keeping with characteristic behaviour of
related Cu(I) polypyridyl complexes.8
LUMO+1 61%
7
9
402
481
383
HOMO-1
HOMO-2
HOMO-1
HOMO
HOMO-1
HOMO
LUMO
LUMO
LUMO
LUMO
LUMO+1 46%
LUMO+1 54%
59%
41%
37%
63%
Fig. 8 shows a comparison of the emission intensity for 4 and
7, where the excitation wavelengths used for both complexes have
comparable molar absorption coefficients. The role of the 6,6¢-
methyl groups on the bipyridine ligand is apparent in providing ad-
ditional steric constraint to the complex leading to the significantly
greater emission intensity of 7. Nevertheless, the observation of
emission from 4, 7 and 9, all of which lack 6,6¢-substituents on
the bipyridyl, demonstrates the role of the ancilliary POP ligand
in providing structural rigidity to the complex. To function as a
sensitiser in a DSSC, a complex must possess a sufficiently long
excited-state lifetime to enable charge injection into the semicon-
ductor condution band and the observation of photoluminescence
is generally taken as an indication that the excited-state lifetime will
be sufficiently long. Alonside the structural studies, the observed
emission from 4, 7 and 9 demonstrates the suitability of using the
dcbpy ligand in conjuction with a sterically-blocking co-ligand
Fig. 6 Isosurface images generated from DFT calculations of (a) LUMO
for 4; (b) LUMO for 5.
nature. Despite the lowest energy MLCT transition taking place
in 5, 7 and 9 being of the same character, their MO diagrams vary
more from the ideal pseudo-tetrahedral model due to increased
8950 | Dalton Trans., 2010, 39, 8945–8956
This journal is
The Royal Society of Chemistry 2010
©

View Online
Table 6 Excitation and emission measurements for dmby, tmbpy, [CuPOP(MeCN)2][BF4], 4, 7, and 9
Room-temperature EtOH
77 K rigid matrix (EtOH)
Compound
Excitation maxima lmax/nm
Emission maxima lmax/nm
Excitation maxima lmax/nm
Emission maxima lmax/nm
dmbpy
tmbpy
CuPOP(MeCN)2
4
7
9
300, 330
264, 304
324
408
369
410
373
457
618
554
Not observed
Not recorded
Not recorded
Not recorded
340
369
328, 356
Not recorded
Not recorded
Not recorded
592
527
515
Not observed
Fig. 7 Absorbance and emission spectra for heteroleptic complexes 4
and 7. Absorbance spectrum carried out in MeCN; fluorescence emission
spectrum recorded in a rigid matrix of EtOH.
Fig. 8 Emission spectra for 4 and 7 recorded in EtOH at 300 K. The
spectra have been normalised against concentration of solution used.
at more positive potentials, around +1.4 V (Table 7). This increase
in Cu(I) oxidation potential is due to the orbital interaction with
the electron withdrawing POP ligand.8 The oxidation potentials
vary little across 4 – 7, suggesting the HOMO is largely Cu(I)/POP
in character as this feature is invariant across the series. This is also
consistent with the TDDFT results shown in Tables 4 and 5 and is
appropriate for use of 5 as a DSSC sensitiser since positive charge
density on the oxidised sensitiser will be located away from the
semiconductor. For long-term DSSC operation, it would clearly be
preferable to develop sensitisers that exhibit chemically reversible
oxidations, however the acid analogue, 5, is still viable for study
as model sensitiser since cyclic voltammetry experiments occur on
a far longer timescale relative to those taking place in a DSSC
device.
to design Cu(I) DSSC sensitisers. Due to the qualitatively large
photoluminescence observed for 7, the photophysical properties
of complexes 4 and 7 are currently under further investigation
in a variety of media and a range of temperatures along with
computational studies of structural rigidity in the excited state.
This work will be reported elsewhere.
Electrochemistry
The redox potentials for all complexes are shown in Table 7.
The first oxidation potential for the homoleptic complexes occurs
within the range +0.75 to +0.91 V and can be assigned as
an irreversible Cu(I)/Cu(II) redox process. The corresponding
Cu(I)/Cu(II) oxidation for the heteroleptic complexes (4 – 7) occurs
Table 7 Oxidation and reduction potentials for all complexes; 1, 4–7 and 9 = 0.3M TBABF4/MeCN; 3 = 0.3M TBABF4/DCM
1
2
E
(V vs. Ag/AgCl)
Complex
ERed.4
ERed.3
ERed.2
ERed.1
EOx.1
EOx.2
EOx.3
EOx.4
[Cu(tmbpy)2]
[Cu(tecbpy)2]
[CuPOP(dmbpy)]
[CuPOP(dcbpy)]
[CuPOP(decbpy)]
[CuPOP(tmbpy)]
[CuPOP(bpym)]
1
3
4
5
6
7
9
-1.68
-1.27
-1.34
-0.87
-1.60
-1.40
-0.91
-1.66
-0.78*
+0.77
+0.91*
+1.44
+1.41
+1.42
+1.41
+0.55*
+1.73
+1.64
+1.68
+1.65
+1.24
-1.61
-1.52
-1.63
-1.44
-1.08
+1.44
+1.75
All potentials vs. Ag/AgCl (saturated solution); recorded at 298 K; Peaks are chemically and electrochemically irreversible and values shown represent
1
peak potential, except; * = Peaks are chemically reversible and electrochemically irreversible, and values shown represent E .
2
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 8945–8956 | 8951
©

View Online
Table 8 I–V Characterisation data for 5. Dye Bath diethyl ether. Adsorption time 24 h. VOC = open circuit potential, ISC = short circuit current, ff = fill
factor, h = power conversion efficiency. N719 measure ISC = 7.96 mA; VOC = 753 mV; ff = 0.51 and h = 3.05%
Dye concentration/mM
Cheno concentration/mM
Post-treated with TiCl4
I
SC/mA
V
OC/mV
ff
h/%
1
2
1
2
1
2
N/A
N/A
1
N
N
N
N
Y
Y
0.176
0.233
0.198
0.300
0.210
0.230
361
347
333
373
388
345
0.51
0.65
0.48
0.45
0.47
0.58
0.032
0.053
0.030
0.050
0.038
0.046
2
N/A
N/A
Upon comparison with literature,1 the reductions exhibited by
all the complexes can be assigned as bipyridyl-ligand based, which
is consistent with the predictions from TDDFT. The variation
in reduction potential seen for the complexes is consistent with
their calculated LUMO energies, with the acid and ester analogues
requiring a less negative reduction potential due to the extended
conjugation. These electrochemical results support the expectation
that the charge transfer transition will have directionality in the
desired direction; towards the carboxylate groups (on the bipyridyl
ligand) that bind to TiO2.
circuit currents produced in some systems36–40 by acting as a spacer
group on the TiO2 surface, limiting the extent of dye aggregation.
A range of dye:Cheno concentration ratios were tried for 5
with representative examples in Table 8. Improved photocurrents
for the dye:Cheno cells over the non-Cheno cells was observed
in every case, with an optimum ratio of 1 : 1, although no
improvement to the photovoltage was seen. Secondly, a TiCl4 post-
treatment of the TiO2 film prior to adsorption of the sensitiser was
applied as a potential method for improving photocurrents.9,40 No
improvement to photocurrent was recorded upon the introduction
of the TiCl4 post-treatment, and photovoltages remained within
the same range seen throughout this work (Table 8).
Solar cell measurements
Despite the low power-conversion efficiency of cells sensitised
with 5, the key feature of the work is the demonstration of sensitiser
function for a dye of this design. The POP ligand was chosen for
its known steric rigidity and successfully demonstrates that simple
dcbpy ligands can be used in heteroleptic Cu(I) DSSC sensitisers
due to the chemical stability and sufficiently-long excited-state
lifetime imposed by the POP. The low power-conversion efficiency
is largely the result of poor harvesting of the solar spectrum by
the dye leading to low ISC. Our future work is now focused on the
development of sterically-constraining co-ligands that also impart
more optimised spectroscopic characteristics on heteroleptic Cu(I)
DSSC sensitisers.
I–V characterisation was carried out on solar cells made using
5 adsorbed onto TiO2 films. Cell construction and treatments
are outlined in the Experimental section and power-conversion
efficiency was determined while irradiating under AM 1.5 light
(100 mWcm-2). Cell efficiencies (h) were calculated using the
resulting values of ISC = short circuit current (mA), VOC = open
circuit voltage (mV) and ff = fill factor.
Adsorption studies revealed the dye bath conditions yielding
the optimum TiO2 surface coverage of 5 to be a sensitiser
concentration of 2 mM in diethyl ether. The results for 1 mM
diethyl ether dye baths have also been quoted for completeness.
Approximate relative TiO2 surface coverage was determined using
UV/Vis spectroscopy. This method also allowed confirmation
that the dye bound without degradation by comparison with the
absorption spectra of 5 in solution. Results for a selected range of
these cells are shown in Table 8 and for comparative reasons,
N719 DSSCs were constructed and tested following the same
procedures.
The observed photovoltages for cells using complex 5 were
consistently around 350 mV, lower than that observed for N719
(753 mV), and lower than the best homoleptic Cu(I) sensitisers
quoted in the literature (around 550 mV),2,3,5–7 The photocurrents
were also significantly lower than those recorded for our N719
cells and for existing homoleptic
Cu(I) sensitisers (up to 0.3 mA compared to 8.0 and 5.0 mA
respectively). This however is as expected due to the blue shift
of the MLCT absorption caused by the electron-accepting POP
ligand such that the complex only just absorbs in the visible region
(lmax = 394 nm). The low photocurrents may also play a role in the
reduced photovoltages displayed using this sensitiser.
Conclusions
Shown here is the synthesis of a series of homoleptic com-
plexes of general formula [Cu(I)(diimine)2][BF4] (1–3), and
a series of heteroleptic complexes of the general formula
[Cu(I)(POP)(diimine)][BF4] (4–9). Electrochemical, spectroscopic
and computational methods were used to understand the elec-
tronic characteristics of these complexes. In particular, this work
demonstrates the first example of a heteroleptic Cu(I) complex;
[Cu(I)(POP)(dcbpy)][BF4] (5), that functions as a sensitiser in a
dye-sensitised solar cell. The photocurrents and photovoltages
achievable by 5 are lower than existing Cu(I)-sensitisers7 but this
is a result of the dye absorbing weakly in the visible region (lmax
=
394 nm and emax = 3.3 ¥ 103 M-1cm-1).
A key point of this work is that stable heteroleptic Cu(I)-
sensitisers can be achieved with simple 4,4¢(CO2H)-bipyridine,
synthetically less intensive to produce than the 4,4¢,6,6¢-substituted
analogues. Work is now moving towards replacing the POP ligand
with a more suitable second ligand that can have a beneficial
impact on the absorption properties of the sensitiser as well as
providing the required steric constraint.
The studies reported in Table 8 include two additional DSSC
fabrication procedures outlined in the literature. Firstly, the
inclusion of Chenodeoxycholic acid (Cheno) in the dye baths was
carried out as it has been shown to dramatically improve the short
8952 | Dalton Trans., 2010, 39, 8945–8956
This journal is
The Royal Society of Chemistry 2010
©

View Online
introduced into the cell through the two holes which were drilled
in the counter electrode. The holes were subsequently sealed using
Kapton tape. The default electrolyte used was 0.6 M LiI, 0.03 M
I2, 0.1 M Guanidinium thiocyanate, 0.5 M 4-tert-butylpyridine,
15 : 85 Valeronitrile:Acetonitrile. The active area of the cell was
1 cm2. The current–voltage characteristics of the cells were
measured under simulated AM 1.5 illumination (100 mWcm-2)
provided by a solar simulator (1 kW Xe with AM 1.5 filter, Mu¨ller)
calibrated using a Digital Solar Power Meter. The Cheno additive
was dissolved in the Dye Baths at the same time as the sensitiser.
The TiCl4 post-treatment of the TiO2 films was carried out prior
to dye adsorption. The TiO2 coated FTO cells were coated with
a 40mM aqueous TiCl4 solution and placed into a steam bath at
70 ◦C for 30 min. The cells were removed, washed with deionised
water, dried at 100 ◦C for 15 min and then sintered at 500 ◦C for
30 min. Dye adsorption was then carried out as described above.
Experimental
General procedures
The synthesis of bis{2-(diphenylphosphanyl)phenyl} ether
(POP),8 4,4¢-dicarboxy-2,2¢-bipyridine (dcbpy),41 4,4¢-di(CO2Et)-
2,2¢-bipyridine (decbpy),42 4,4¢,6,6¢-tetramethyl-2,2¢-bipyridine
(tmbpy),43 [Cu(MeCN)4][BF4]44 and [Cu(I)POP(MeCN)2][BF4]8
were carried out according to literature procedures. 4,4¢,6,6¢-
tetracarboxy-2,2¢-bipyridine (tcbpy) and 4,4¢,6,6¢-tetra(CO2Et)-
2,2¢-bipyridine (tecbpy) are novel complexes and have been
synthesized using adapted literature methods.42
4,4¢-dimethyl-2,2¢-bipyridine (dmbpy), bipyrimidine (bpym)
and all other chemicals were purchased from Aldrich and used
as received.
Electrochemistry was carried out using a Pt working electrode,
Pt rod counter electrode and Ag/AgCl reference electrode. All
electrochemical experiments were carried out in either acetonitrile
or DCM and the supporting electrolyte used was TBABF4 (0.3M).
After each experiment the reference electrode was calibrated
against the ferrocene/ferrocenium couple which was found to be at
0.55 V. The absorption spectra were recorded using a PerkinElmer
Lambda 9 spectrophotometer controlled using the UV/Winlab
software. Emission spectra were recorded at room temperature
and using frozen samples, with ethanolic solutions of 4, 7, 9,
[CuPOP(MeCN)2][BF4], dmbpy and tmbpy (0.1–0.6 mM), using
a Fluoromax2 fluorimeter controlled by the ISAMain software.
Density functional theory calculations were performed using
the Gaussian 03 program29 with the starting structures for 4,
5 and 7 inputted using the builder program Arguslab and the
crystal structure coordinates used as the starting point for 9. All
calculations were carried out using the Becke’s three parameter
exchange functional with the Lee–Yang–Parr for the correlation
functional (B3LYP),30 using the Los Alamos National Laboratory
basis sets, known as LANL2DZ (developed by Hay and Wadt),31–33
which comprises ECP + double zeta for copper, and the all-
electron valence double zeta basis sets developed by Dunning
(D95V) for light atoms.45 Frequency calculations were carried
out to ensure that optimised geometries were minima on the
potential energy surface. Solvent effects were included via the
self-consistent reaction field (SCRF) method using the polarised
continuum model (PCM),46 with slight modifications to the default
cavity parameters to aid convergence. Time-dependent density
functional theory (TDDFT) was performed in an acetonitrile
polarizable continuum model, with the first 70 singlet transitions
calculated. ArgusLab34 was used to generate the orbital isosurface
maps.
Cu(tmbpy)2][BF4] (1)
[Cu(MeCN)4][BF4] (140 mg, 0.45 mmol) and tmbpy (200 mg,
0.89 mmol) were stirred in 20 mL degassed chloroform under
nitrogen for 30 min. The solvent was removed under reduced
pressure and the solid obtained triturated in hexane. The product
was obtained by filtration. Yield = 75%, 193 mg. dH (DMSO,
250 MHz)/ppm = 8.36 (s, 4H, H-bpy), 7.42 (s, 4H, H-bpy), 2.45
(s, 12H, CH3), 2.11 (s, 12H, CH3). MS (positive ESI); m/z: 487.10
(M-BF4 )+. Elemental analysis: calculated for C28H32BCuF4N4: C
-
58.49, H 5.61, N 9.74. Found: C 57.78, H 5.46, N 9.57.
[Cu(tcbpy)2][BF4] (2)
[Cu(MeCN)4][BF4] (20.5 mg, 0.07 mmol) and tcbpy (42.3 mg,
0.13 mmol) were stirred in 5 mL degassed, dry MeOH under
nitrogen for 2 h. A white solid (uncomplexed tcbpy) was removed
by filtration. The pure product was then obtained by precipitation
with isopropyl alcohol overnight. Yield = 32%, 17 mg. dH(DMSO,
250 MHz)/ppm: 8.84 (s, 4H, H-bpy), 8.22 (s, 4H, H-bpy). MS
(positive ESI): m/z = 725.31 (M-BF4 )+. Elemental analysis:
-
calculated for C28H16BCuF4N4O16: C 41.27, H 1.98, N 6.88. Found:
C 41.44, H 2.24, N 7.96.
[Cu(tecbpy)2][BF4] (3)
[Cu(MeCN)4][BF4] (14.2 mg, 0.05 mmol) and tecbpy (40 mg,
0.09 mmol) were stirred in 50 mL degassed, dry DCM under
nitrogen for 30 min. The pure product was obtained by precip-
itation with isopropyl alcohol. Yield = 53%, 28 mg. dH (CDCl3,
250 MHz)/ppm = 9.15 (s, 4H, H-bpy), 8.69 (s, 4H, H-bpy), 4.54
(q, J = 7.10 Hz, 8H, –CH2–CH3), 3.88 (q, J = 7.14 Hz, 8H, –
CH2–CH3), 1.47 (t, J = 7.18 Hz, 12H, –CH2–CH3), 1.03 (t, J =
To make the DSSCs titanium dioxide paste (Dyesol, DSL-
18NR-T) was deposited onto cleaned fluorine doped tin oxide
conductive glass (TEC 8, Pilkington, UK) by doctor-blading.
The film was dried at 100 ◦C for 15 min and then sintered
at 450 ◦C for 30 min to remove the organics and to form a
mesoporous film structure. The thickness of the film was about
18 mm. The films were sensitized with 5 using a solution of
the dye in diethyl ether. The platinized counter electrode was
fabricated following the previously reported procedure.47 The
cell was completed by sealing the dye coated TiO2 electrode
and Pt electrode of a cell together by a thermal plastics spacer
(Surlyn 1702, 25 mm, Solaronix) at 120 ◦C. The electrolyte was
-
7.11 Hz, 12H, CH2–CH3). MS (positive ESI); m/z: 950 (M-BF4 )+.
Elemental analysis: calculated for C44H48N4O16CuBF4·CH2Cl2: C
48.08, H 4.48, N, 4.98. Found: C 47.60, H 4.33, N 4.79.
[Cu(POP)(dmbpy)][BF4] (4)
[Cu(I)POP(MeCN)2][BF4] (201 mg, 0.26 mmol) and dmbpy
(51 mg, 0.06 mmol) were stirred in 5 mL acetone for 3 h resulting
in an orange solution. The solvent was removed under reduced
pressure. The pure product was then obtained by reprecipitaiton
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 8945–8956 | 8953
©

View Online
Table 9 X-Ray crystallography data
[Cu(dmbpy)2Cl] [BF4]
1
6
9
10
CCDC deposition number
Empirical Formula
771442
(C24H24N4CuCl)2.
771443
C28H32N4Cu·BF4
771439
C52H44N2CuO5P2
771440
C44H34N4OP2Cu·BF4
771441
C80H62N4O2P4Cu2·
2[BF4]·2[C3H6O]
(BF4)2·(CH2Cl2)
1193.47
150(2)
BF4·C2H3N
1030.24
100(2)
Formula weight
T/K
574.93
150(2)
Orange
847.08
150(2)
Orange
1652.15
150(2)
Crystal colour
Crystal dimensions
Crystal system
Space group
Cyan
Orange
Orange-Yellow
0.15 ¥ 0.21 ¥ 0.25
Triclinic
0.50 ¥ 0.40 ¥ 0.30
0.90 ¥ 0.59 ¥ 0.43
Monoclinic
C2/c
13.8531(14)
16.4771(14)
13.5039(19)
90
119.730(7)
90
2676.7(5)
4
0.15 ¥ 0.15 ¥ 0.08
0.22 ¥ 0.33 ¥ 0.36
Monoclinic
P21/c
9.5405(2)
19.8580(4)
20.4442(4)
90
95.4810(10)
90
3855.55(14)
4
Triclinic
Triclinic
¯
¯
¯
P1
P1
P1
˚
a/A
10.8528(3)
14.4072(4)
17.6085(5)
83.9868(16)
74.9744(14)
73.3850(14)
2546.69(12)
4
12.3345(3)
12.6111(3)
18.0200(4)
98.4411(13)
100.4738(12)
113.2549(12)
2457.18(10)
2
11.3321(3)
13.2821(3)
14.2188(3)
108.665(1)
105.605(1)
97.186(1)
1899.81(7)
1
˚
b/A
˚
c/A
a (◦)
b (◦)
g (◦)
3
˚
V/A
Z
Dc/Mg m-3
1.556
8998
1.427
2747
1.392
10570
1.459
10239
1.440
10287
Independent reflections
[Rint = 0.0411]
36169/0/666
1.119
[Rint = 0.0492]
13995/184/190
0.869
[Rint = 0.0606]
49789/63/680
0.577
[Rint = 0.038]
31367/0/514
0.711
[Rint = 0.037]
27104/0/496
0.719
All data/restraints/parameters
Absorption correction/mm-1
R1, wR2 (observed data:
F2>2s(F2))
0.0366/0.0933
0.0526/0.1470
0.0422/0.0937
0.0485/0.1071
0.0537/0.1160
from acetonitrile upon addition of isopropyl alcohol and leaving
overnight. Yield = 67%, 153 mg. dH(DMSO, 250 MHz)/ppm:. 8.51
(s, 2H, H-bpy), 8.37 (d, 2H, H-bpy), 7.47-7.15 (m, 18H, H-POP),
7.09 (t, 2H, H-bpy), 6.98 (m, 8H, H-POP), 6.61 (m, 2H, H-POP),
[Cu(POP)(tmbpy)][BF4] (7)
[Cu(I)POP(MeCN)2][BF4] (101 mg, 0.13 mmol) and tmbpy (25 mg,
0.03 mmol) were stirred in 10 mL anhydrous acetonitrile under N2
for 3 h resulting in an orange solution. The volume of reaction
mixture was reduced by half under reduced pressure. The pure
product was then obtained upon addition of isopropyl alcohol and
leaving overnight. The desired product remains in the filtrate and is
obtained by solvent removal under reduced pressure. Yield = 75%,
90 mg. dH(DMSO, 250 MHz)/ppm: 8.09 (s, 2H, H-bpy), 6.85-
7.34 (m, 30H, H-bpy and POP), 2.37 (s, 6H, CH3), 2.31 (s, 6H,
CH3). MS (positive FAB): m/z (%) = 812.6 (M-BF-)+. Elemental
analysis: calculated for C50H44BCuF4N2OP2: C 66.64, H 4.92, N
3.11. Found C 66.40, H 4.76, N 3.00.
2.47 (s, 6H, CH3-bpy). MS (positive FAB): m/z = 784.7 (M-BF4 )+.
-
Elemental analysis: calculated for C48H40BCuF4N2OP2: C 66.03,
H 4.62, N 3.21. Found: C 65.58, H 4.42, N 3.90.
[Cu(POP)(dcbpy)][BF4] (5)
[Cu(I)POP(MeCN)2][BF4] (100 mg, 0.13 mmol) and dcbpy (28 mg,
0.03 mmol) were stirred in 5 mL acetone for 3 h resulting in a yellow
solution. The pure product was then obtained by reprecipitation
upon addition of isopropyl alcohol and leaving overnight. Yield =
26%, 96 mg. dH(DMSO, 250 MHz)/ppm: 9.17 (s, 2H, H-bpy),
8.91 (d, 2H, H-bpy), 8.07 (d, 2H, H-bpy), 7.4 (m, 28H, H-POP).
MS (positive ESI): m/z = 844.80 (M-BF-)+. Elemental analysis:
calculated for C48H36CuN2O5P2.BF4: C 61.73, H 3.89, N 3.00.
Found C 62.06, H 4.20, N 2.87.
[Cu(POP)(tcbpy)][BF4] (8)
[Cu(I)POP(MeCN)2][BF4] (100 mg, 0.13 mmol) and tcbpy (45 mg,
0.04 mmol) were stirred in 5 mL acetone for 3 h resulting in a dark
purple solution and white solid. The white solid was removed by
filtration. The pure product was then obtained by reprecipitation
upon addition of isopropyl alcohol and leaving overnight. The
desired product remains in the filtrate and was obtained by solvent
removal under reduced pressure. Yield = 53%, 72 mg. dH(DMSO,
250 MHz)/ppm: 9.32(s, 2H, H-bpy), 8.72(s, 2H, H-bpy), 6.87-7.67
(m, 28H, POP). MS (positive FAB): m/z (%) = 617.3 (CuPOP).
Elemental analysis: calculated for C50H36BCuF4N2O9P2: C 58.81,
H 3.55, N 2.74. Found C 59.10, H 3.88, N 2.86.
[Cu(POP)(decbpy)][BF4] (6)
[Cu(I)POP(MeCN)2][BF4] (201 mg, 0.26 mmol) and decbpy
(51 mg, 0.05 mmol) were stirred in 5 mL acetone for 3 h resulting
in an orange solution. The solvent was removed under reduced
pressure. The pure product was then obtained by reprecipitation
upon addition of isopropyl alcohol and leaving overnight. Yield =
87%. 231 mg, dH(DMSO, 250 MHz)/ppm: 8.79 (s, 2H, H-bpy),
8.75 (d, J = 4.74 Hz, 2H, H,-bpy), 7.98 (dd, J = 1.32 Hz, J =
5.42 Hz, 2H, H-bpy), 7.10 (m, 28H, H-POP), 4.53 (dd, J = 7.07 Hz,
4H, CH2CH3), 1.49 (t, J = 7.13 Hz, 6H, CH2CH3). MS (positive
ESI): m/z = 784.87 (M-BF-)+. Elemental analysis: calculated for
C52H44CuN2O5P2.BF4: C 63.14, H 4.48, N 2.83. Found C 62.71, H
4.36, N 2.11.
[Cu(POP)(bpym)][BF4] (9)
[Cu(I)POP(MeCN)2][BF4] (486 mg, 0.63 mmol) and bpym
(100 mg, 0.63 mmol) were stirred in 20 mL DCM. The
product was obtained by concentration of the reaction mix-
ture then precipitation by the addition of ether. Yield = 64%.
8954 | Dalton Trans., 2010, 39, 8945–8956
This journal is
The Royal Society of Chemistry 2010
©

View Online
dH(CD3OD, 250 MHz)/ppm: 8.90 (s, 4H, H-bpym), 7.59–6.77
(m, 30H, H-bpym and H-POP). Elemental analysis: calculated for
C44H34BCuF4N4OP2: C 62.39, H 4.05, N 6.61. Found C 61.56, H
3.43, N 6.34
8 N. Armaroli, G. Accorsi, M. Holler, O. Moudam, J.-F. Nierengarten,
Z. Zhou, R. T. Wegh and R. Welter, Adv. Mater., 2006, 18, 1313.
9 M. K. Nazeeruddin, A. Kay, I. Rodicio, R. Humphry-Baker, E. Mueller,
P. Liska, N. Vlachopoulos and M. Gra¨tzel, J. Am. Chem. Soc., 1993,
115, 6382.
10 http://www.g24i.com/.
11 N. Robertson, Angew. Chem., Int. Ed., 2006, 45, 2338.
12 C. L. Linfoot, P. Richardson, K. L. McCall, J. R. Durrant, A.
Morandeira and N. Robertson, Solar Energy, in press.
13 S. Ferrere and B. Gregg, J. Am. Chem. Soc., 1998, 120, 843.
14 S.-M. Kuang, D. G. Cuttell, D. R. Mcmillin, P. E. Fanwick and R. A.
Walton, Inorg. Chem., 2002, 41, 3313.
[Cu2(POP)2(g4-bpym)][(BF4)2] (10)
[Cu(I)POP(MeCN)2][BF4] (33 mg, 0.04 mmol) and 11 (34 mg,
0.04 mmol) were stirred in 5 mL acetone for 30 h. The product was
obtained by concentration of the reaction mixture then precipita-
tion by the addition of ether. Despite obtaining a crystal structure
we were unable to obtain a pure product with satisfactory ele-
mental analysis. Yield = 48%, 32 mg. dH(DMSO, 250 MHz)/ppm:
9.01 (s, 4H, H-bpym), 7.85-6.58 (m, 58H, H-bpym and H-POP).
Elemental analysis: calculated for C80H62B2Cu2F8N4O2P4: C 62.56,
H 4.07, N 3.65. Found C 59.06, H 3.96, N 3.01.
15 S. Ferrere, Chem. Mater., 2000, 12, 1083.
16 S. Ferrere, Inorg. Chim. Acta, 2002, 329, 79.
17 N. Robertson, ChemSusChem, 2008, 1, 977.
18 L. Spiccia, G. B. Deacon and C. M. Kepert, Coord. Chem. Rev., 2004,
248, 1329.
19 R. M. Williams, L. D. Cola, F. Hartl, J.-J. Lagref, J.-M. Planeix, A. D.
Cian and M. W. Hosseini, Coord. Chem. Rev., 2002, 230, 253.
20 N. Armaroli, G. Accorsi, F. Cardinali and A. Listorti, Photochemistry
and Photophysics of Coordination Compounds I, 69, Springer, 2007.
21 M. Schmitted, C. Michel, S. -X. Liu, D. Schildbach and D. Fenske,
Eur. J. Inorg. Chem., 2005, 5, 1155.
X-ray crystallography
22 M. Ruthkosky, C. A. Kelly, F. N. Castellano and G. J. Meyer, Coord.
Chem. Rev., 1998, 171, 309.
23 O. Moudam, A. Kaeser, B. Delavaux-Nicot, C. Duhayon, M. Holler,
G. Accorsi, N. Armaroli, I. Se´guy, J. Navarro, P. Destruel and J.-F.
Nierengarten, Chem. Commun., 2007, 3077.
24 T. McCormick, W. L. Jia and S. Wang, Inorg. Chem., 2006, 45,
147.
25 L. Yang, J.-K. Feng, A.-M. Ren, M. Zhang, Y.-G. Ma and X.-D. Liu,
Eur. J. Inorg. Chem., 2005, 1867.
26 Y.-M. Xie and J.-H. Wu, Inorg. Chem. Commun., 2007, 10,
1561.
27 U. Monkowius, Y. N. Svartsov, T. Fischer, M. Zabel and H. Yersin,
Inorg. Chem. Commun., 2007, 10, 1473.
28 Q. Zhang, J. Ding, Y. Cheng, L. Wang, Z. Xie, X. Jing and F. Wang,
Adv. Funct. Mater., 2007, 17, 2983.
Crystals of [Cu(dmbpy)2][BF4] were grown by slow diffusion of
diethyl ether into a saturated DCM solution of 4. Crystals of
1 and 9 were grown by slow diffusion of diethyl ether into a
saturated DCM solution of 1 or 9 respectively. Crystals of 6
were grown by slow diffusion of diethyl ether into a saturated
solution of 6 in acetonitrile. Crystals of 10 was grown by slow
diffusion of hexane into a saturated solution of 10 in acetone.
Single crystal X-ray diffraction data are given in Table 9 and
˚
were collected using Mo-Ka radiation (l = 0.71073 A) on a
Smart APEX CCD diffractometer equipped with an Oxford
Cryosystems low-temperature device operating at 150 K. An
absorption correction was applied using the multi-scan procedure
SADABS.48 The structures were solved by Direct methods (Shelx49
for [Cu(dmbpy)2][BF4], 1 and 6; and SIR9250 for 9 and 10) and
refined by full-matrix least squares against |F|2 using all data
(Shelx49 for [Cu(dmbpy)2][BF4], 1 and 6; and CRYSTALS51 for 9
and 10). Figures were prepared using the programme Mercury.52
All non-H atoms were refined with anisotropic displacement
parameters.
29 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb,
J. R. Cheeseman, J. A. Montgomery Jr., T. Vreven, K. N. Kudin, J. C.
Burant, J. M. Millam, S. S. Iyengar, J. Tomasi, V. Barone, B. Mennucci,
M. Cossi, G. Scalmani, N. Rega, G. A. Petersson, H. Nakatsuji, M.
Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T.
Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E.
Knox, H. P. Hratchian, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo,
R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi,
C. Pomelli, J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth,
P. Salvador, J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D.
Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K.
Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S.
Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz,
I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y.
Peng, A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson,
W. Chen, M. W. Wong, C. Gonzalez and J. A. Pople, 2004, Gaussian
03, Revision C.02, Gaussian, Inc., Wallingford, CT.
Acknowledgements
We thank EaSTChem and the EPSRC Supergen Excitonic Solar
Cell Consortium for funding. This work has made use of
the resources provided by the EaStChem Research Computing
Facility (http://www.eastchem.ac.uk/rcf). This facility is partially
supported by the eDIKT initiative (heep://www.edikt.org).
30 A. D. Becke, J. Chem. Phys., 1993, 98, 5648.
31 W. R. Wadt and P. J. Hay, J. Chem. Phys., 1985, 82, 270.
32 W. R. Wadt and P. J. Hay, J. Chem. Phys., 1985, 82, 284.
33 W. R. Wadt and P. J. Hay, J. Chem. Phys., 1985, 82, 299.
34 Arguslab 4.0, M. A. Thompson, Planaria, Software LLC, Seattle,
http://www.arguslab.com.
Notes and references
35 K. L. McCall, J. R. Jennings, H. Wang, A. Morandeira, L. M. Peter,
J. R. Durrant, L. J. Yellowlees, J. D. Woollins and N. Robertson,
J. Photochem. Photobiol. A–Chem., 2000, 202, 196.
36 S. Ito, H. Miura, S. Uchida, M. Takata, K. Sumioka, P. Liska, P. Comte,
P. Pe´chy and M. Gra¨tzel, Chem. Commun, 2008, 5194.
37 R. Chen, X. Yang, H. Tian, X. Wang, A. Hagfeldt and L. Sun, Chem.
Mater., 2007, 19, 4007.
38 H. Choi, J. K. Lee, K. H. Song, K. Song, S. O. Kanga and J. Koa,
Tetrahedron, 2007, 63, 1553.
39 K. Sayama, S. Tsukagoshi, T. Mori, K. Hara, Y. Ohga, A. Shinpou, Y.
Abe, S. Suga and H. Arakawa, Sol. Energy Mat. Sol. C., 2003, 80, 47.
40 J. H. Yum, S. Moon, R. Humphry-Baker, P. Walter, T. Geiger, F.
Nu¨esch, M. Gra¨tzel and M. K. Nazeeruddin, Nanotechnology, 2008,
19, 424005.
1 N. Armaroli, Chem. Soc. Rev., 2001, 30, 113.
2 N. Alonso-Vante, J.-F. Nierengarten and J.-P. Sauvage, J. Chem. Soc.,
Dalton Trans., 1994, 1649.
3 S. Sakaki, T. Kuroki and T. Hamada, T., J. Chem. Soc., Dalton Trans.,
2002, 840.
4 A. Barbieri, G. Accorsi and N. Armaroli, Chem. Commun., 2008, 2185.
5 T. Bessho, E. C. Constable, M. Graetzel, A. H. Redondo, C. E.
Housecroft, W. Kylberg, Md. K. Nazeeruddin, M. Neuburger and S.
Schaffner, Chem. Commun., 2008, 3717.
6 A. Hernandez Redondo, E. C. Constable and C. E. Housecroft, Chimia,
2009, 63, 205.
7 E. C. Constable, A. H. Redondo, C. E. Housecroft, M. Neuburger and
S. Schaffner, Dalton Trans., 2009, 6634.
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 8945–8956 | 8955
©

View Online
41 P. G. Hoertz, A. Staniszewski, A. Marton, G. T. Higgins, C. D.
Incarvito, A. L. Rheingold and G. J. Meyer, J. Am. Chem. Soc., 2006,
128, 8234.
42 F. H. Case, J. Am. Chem. Soc., 1946, 68, 2574.
43 P. Ghosh and T. G. Spiro, J. Am. Chem. Soc., 1980, 102,
5543.
44 O. Moudam, Thesis, LCC-Toulouse, France, July 2007.
45 T. H. Dunning Jr. and P. J. Hay, in Modern Theoretical Chem-
istry, Ed. H. F. Schaefer III, Vol. 3 (Plenum, New York, 1976) 1–
28.
47 N. Papageorgiou, W. F. Maier and M. Gra¨tzel, J. Electrochem. Soc.,
1997, 144, 876.
48 G. M. Sheldrick, SADABS, Version 2004/1, University of Gottingen,
Germany.
49 G. M. Sheldrick, Acta Cryst., 2008, A64, 112.
50 A. Altomane, G. Cascarano, C. Giacovazzo, A. Guagliardi, M. C.
Burla, G. Polidori and M. Carnalli, J. Appl. Cryst, 1994, 27, 435.
51 P. W. Betteridge, J. R. Carruthers, R. I. Cooper, K. Prout and D. J.
Watkin, J. Appl. Cryst, 2003, 36, 1487.
52 C. F. Macrae, P. R. Edgington, P. McCabe, E. Pidcock, G. P. Shields, R.
Taylor, M. Towler and J. van de Streek, J. Appl. Cryst., 2006, 39, 453.
46 A. Vleck and S. Zalis, Coord. Chem. Rev., 2007, 251, 258.
8956 | Dalton Trans., 2010, 39, 8945–8956
This journal is
The Royal Society of Chemistry 2010
©