Spectroscopic, electrochemical and computational study of Pt-diimine-dithiolene complexes: Rationalising the properties of solar cell dyes

Article abstract of DOI:10.1039/b719789f

A series of [Pt(ii)(diimine)(dithiolate)] complexes of general formula [Pt{X,X′-(CO₂R)₂-bpy}(mnt)] (where X = 3, 4 or 5; R = H or Et, bpy = 2,2′-bipyridyl and mnt = maleonitriledithiolate), have been spectroscopically, electrochemically and computationally characterised and compared with the precursors [Pt{X,X′-(CO₂R) ₂-bpy}Cl₂] and X,X′-(CO₂R) ₂-bpy. The study includes cyclic voltammetry, in situ EPR spectroelectrochemical studies of fluid solution and frozen solution samples, UV/Vis/NIR spectroelectrochemistry, hyrid DFT and TD-DFT calculations. The effect of changing the position of the bpy substituents from 3,3′ to 4,4′ and 5,5′ is discussed with reference to electronic changes seen within the different members of the family of molecules. The performance of the mnt complexes in dye-sensitised solar cells has been previously described and the superior performance of [Pt{3,3′-(CO₂R)₂-bpy} (mnt)] is now explained in terms of decreased electronic delocalisation through twisting of the bipyridyl ligand as supported by the EPR and computational results.

Full text of DOI:10.1039/b719789f

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
Spectroscopic, electrochemical and computational study of
Pt–diimine–dithiolene complexes: rationalising the properties
of solar cell dyes†
Elaine A. M. Geary,a Keri L. McCall,a Andrew Turner,a Paul R. Murray,a Eric J. L. McInnes,b Lorna A. Jack,a
Lesley J. Yellowleesa and Neil Robertson*a
Received 2nd January 2008, Accepted 22nd April 2008
First published as an Advance Article on the web 29th May 2008
DOI: 10.1039/b719789f
A series of [Pt(II)(diimine)(dithiolate)] complexes of general formula [Pt{X,Xꢀ-(CO2R)2-bpy}(mnt)]
(where X = 3, 4 or 5; R = H or Et, bpy = 2,2ꢀ-bipyridyl and mnt = maleonitriledithiolate), have been
spectroscopically, electrochemically and computationally characterised and compared with the
precursors [Pt{X,Xꢀ-(CO2R)2-bpy}Cl2] and X,Xꢀ-(CO2R)2-bpy. The study includes cyclic voltammetry,
in situ EPR spectroelectrochemical studies of fluid solution and frozen solution samples, UV/Vis/NIR
spectroelectrochemistry, hyrid DFT and TD-DFT calculations. The effect of changing the position of
the bpy substituents from 3,3ꢀ to 4,4ꢀ and 5,5ꢀ is discussed with reference to electronic changes seen
within the different members of the family of molecules. The performance of the mnt complexes in
dye-sensitised solar cells has been previously described and the superior performance of
[Pt{3,3ꢀ-(CO2R)2-bpy}(mnt)] is now explained in terms of decreased electronic delocalisation through
twisting of the bipyridyl ligand as supported by the EPR and computational results.
The dye molecule in a DSSC requires a linker moiety which
attaches the dye to the surface of the semiconductor9 and in
Introduction
the large majority of studies this has been the (4,4ꢀ-CO2H)2-bpy
ligand (bpy = 2,2ꢀ-bipyridyl). The 4,4ꢀ-substituted carboxylic acid
groups on the bpy bind to the surface of the semiconductor
allowing electron transfer from the excited dye molecule to the
semiconductor. This electron transfer reaction is facilitated by the
interaction of the p* orbital on the substituted bpy with the 3d
orbital manifold of the conduction band of the TiO2.
An attractive feature of [Pt(diimine)(dithiolene)] complexes is
the ease with which the HOMO and LUMO of the molecule can be
separately tuned. Over the past two decades, Eisenberg et al. have
systematically studied the electrochemistry and luminescence of a
range of these molecules by sequentially changing the substituents
on the diimine and dithiolene components.2 This has shown that
the excited state energies can be tuned by up to 1 eV. Such studies
have demonstrated that the LUMO on [Pt(II)(diimine)(dithiolate)]
complexes is located on the diimine functionality and the HOMO
is a mixture of Pt(d) and S(p) orbital character. The visible charge-
transfer transition for these molecules is therefore assigned as a
MMLLꢀCT (mixed metal–ligand to ligand charge transfer) from
the Pt/S HOMO to the diimine based LUMO.1,2,10 These previous
studies have focussed almost exclusively on complexes with 4,4ꢀ-
disubstitued diimine ligands.1,2,10
Complexes of Pt(II) with polypyridyl ligands have been widely
studied due to properties including redox activity, solva-
tochromism and luminescence.1,2 In most cases, ligands on the
Pt centre can be readily varied, allowing for manipulation of the
molecular and electronic properties of these complexes making
them useful as model systems for the study of many physical
properties. In recent years square-planar Pt(II) coordination
complexes have also been studied for their potential use in the area
of dye-sensitised solar cells (DSSCs)3–6 and in solar to chemical
energy conversion.7 In a DSSC, a dye molecule is adsorbed on to
the surface of a semiconductor (usually TiO2). Upon illumination
with visible light, the photo-excited dye injects an electron into the
conduction band of the semiconductor, which is supported on a
transparent electrode. The dye is re-reduced by a redox couple
(typically I−/I3−), which in turn is regenerated at the counter
electrode, thus completing the circuit.8
aSchool of Chemistry, University of Edinburgh, King’s Buildings, Edinburgh,
UK EH9 3JJ. E-mail: neil.robertson@ed.ac.uk; Fax: +44 1316504743;
Tel: +44 1316504755
bSchool of Chemistry, University of Manchester, Oxford Road, Manchester,
UK M13 9 PL
We have recently reported the synthesis and study in DSSCs
of a series of [Pt(II)(diimine)(dithiolate)] complexes, that bind
to the semiconductor surface through a bpy ligand substituted
with carboxylic acid groups at the positions, 3,3ꢀ, 4,4ꢀ and
5,5ꢀ respectively. The solar-energy conversion efficiencies of the
resulting complexes were studied and interestingly showed the
3,3ꢀ-disubstituted bpy molecule to be the most efficient sensitiser of
those tested.6 In marked contrast, studies of Ru–bpy dyes (the most
efficient class of sensitiser) in DSSCs have focussed almost entirely
† Electronic supplementary information (ESI) available: Table S1: UV/Vis
of oxidised and reduced species; Fig. S1–S4: UV/Vis spectra of 2a0/1−
,
2b0/1−/2−, 2c0/1−/2−, 2c0/1+; Fig. S5–S7: EPR of 1b1−, 2c1−, 1b1− (frozen);
Tables S2 and S3: frontier orbital energies of 2a, 3a; Table S4: calculated
contribution of ring C and N atoms to the LUMO of 1c, 2c, 3c;
Table S5: calculated frontier orbital energies of 1b, 1c, 2c, 3c; Fig. S8–
S17 representations of frontier orbitals of 2a, 3a, 2c, 3c, 1b; Table S6:
comparison of calculated and experimental structure of 1c; Table S7:
calculated energies and % contributions of HOMO and LUMO for 1b,
1c, 2c, and 3c; Tables S8–S10: TD-DFT results for 1c, 2c, 3c; Fig. S10:
TD-DFT results for 2c, 3c. See DOI: 10.1039/b719789f
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 3701–3708 | 3701
©

View Article Online
on 4,4ꢀ-substitution. We are only aware of one Ru dye involving
3,3ꢀ-(CO2H)2-bpy ligands that has been studied in DSSCs and this
showed slightly poorer results than the analogous Ru complex
with 4,4ꢀ-(CO2H)2-bpy ligands.11 Presumably, the slightly poorer
performance is the reason for the lack of further interest, however
our results on Pt complexes suggest that 3,3ꢀ-(CO2H)2-bpy ligands
are worth revisiting.
Following this discovery, the aim of the current work is to
greatly extend our understanding of the electronic properties of
[Pt{X,Xꢀ-(CO2Et)2-bpy}(mnt)] (X = 3, 4 or 5; mnt = maleoni-
triledithiolate) by systematic electronic characterisation of each
member of this family and comparison with related precursors,
(Fig. 1) using UV/Vis/NIR spectroelectrochemistry, in situ EPR
spectroelectrochemistry and DFT calculations. In particular, this
allows insight into the bpy based LUMO which influences not only
the photophysical and electronic properties of the molecules, but
also the coupling to TiO2 and consequent electronic properties of
the composite system. We have reported some initial work on the
UV/Vis/NIR spectroelectrochemistry of 1c–3c6 in our original
study and this is incorporated here in the discussion.
Results and discussion
(i) Electrochemistry
The electrochemistry of the [Pt{X,Xꢀ-(CO2Et)2-bpy}(mnt)] fam-
ily (1c–3c), the related [Pt{X,Xꢀ-(CO2Et)2-bpy}Cl2] precursor
molecules (1b–3b) and the corresponding uncoordinated bpy
ligands (1a–3a) was studied by cyclic voltammetry in a solution
of 0.1 M TBABF4 in DMF at 293 K. Redox potentials for all
processes are listed in Table 1.
The electrochemical properties show various trends which are
common within each of the 3,3ꢀ, 4,4ꢀ and 5,5ꢀ motifs. In each case
the free bpy ligand shows either one or two reductions within
the solvent window. On binding to the Pt centre, the reduction
potentials significantly shift to more positive values due to the
stabilisation of the LUMO after complexation of the free ligand
to the Pt metal. On substitution of the two chloride ligands with
the mnt ligand, each [Pt{X,Xꢀ-(CO2Et)2-bpy}(mnt)] complex (1c,
2c and 3c) shows two fully reversible reductions, the potentials
of which remain similar to those of their respective [Pt{X,Xꢀ-
(CO2Et)2-bpy}Cl2] precursors (Fig. 2), particularly for the first
reduction. This shows that there is only a small effect on the
LUMO of the complex after substitution of the chloride ligands
for mnt, consistent with assignment of the LUMO as bpy based.2
Fig. 2 Cyclic voltammogram showing reduction processes for 2a (dash–
dot green), 2b (solid black) and 2c (dashed red) in 0.1 M TBABF4–DMF
Fig. 1 Synthesis of members of the family of [Pt{X,Xꢀ-(CO2R)2-bpy}-
(mnt)]. (i) KMnO4(aq.), 120 ◦C/HCl; (ii) EtOH/H2SO4, 115 ◦C; (iii)
K2PtCl4, 125 ◦C; (iv) Na2(mnt), RT (140 ◦C for 2c). The principle axis
system for the EPR analysis is also shown with the z-axis perpendicular to
the molecular plane.
at 293 K and scan rate 0.1 V s−1
.
In addition, each of the [Pt{X,Xꢀ-(CO2Et)2-bpy}(mnt)] com-
plexes (1c, 2c and 3c) shows one irreversible oxidation at a similar
Table 1 Redox potentials of the [Pt{X,Xꢀ-(CO2Et)2-bpy}(mnt)] family and related precursor molecules and ligandsa
No.
Compound
E2/V
E1/V
Eox/V
1a
1b
1c
1d
2a
2b
2c
2d
3a
3b
3c
3d
3,3ꢀ-(CO2Et)2-bpy
—
−1.84
−0.61
−0.59
−0.69∗
−1.53
−0.67
−0.65
−0.82∗
−1.23
−0.53
−0.52
−0.69∗
—
—
1.35∗
1.27∗
—
[Pt{3,3ꢀ-(CO2Et)2-bpy}Cl2]
[Pt{3,3ꢀ-(CO2Et)2-bpy}(mnt)]
[Pt{3,3ꢀ-(CO2H)2-bpy}(mnt)]
4,4ꢀ-(CO2Et)2-bpy
−1.27
−1.20
−1.05∗
—
−1.25
−1.20
−1.19
−1.61
−1.03
−0.99
−1.16
[Pt{4,4ꢀ-(CO2Et)2-bpy}Cl2]
[Pt{4,4ꢀ-(CO2Et)2-bpy}(mnt)]
[Pt{4,4ꢀ-(CO2H)2-bpy}(mnt)]
5,5ꢀ-(CO2Et)2-bpy
—
1.39∗
1.27∗
—
[Pt{5,5ꢀ-(CO2Et)2-bpy}Cl2]
[Pt{5,5ꢀ-(CO2Et)2-bpy}(mnt)]
[Pt{5,5ꢀ-(CO2H)2-bpy}(mnt)]
—
1.41∗
1.27∗
a * irreversible, § quasi reversible reduction.
3702 | Dalton Trans., 2008, 3701–3708
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
potential to each other, consistent with location of the HOMO on
the dithiolate moiety.
show similar features (e.g. Fig. S4†). On generation of the 1c1+,
2c1+ and 3c1+, the peaks associated with the p–p* transition of
the coordinated bpy shift to lower energy. This has previously
been shown in a study of [Ru(bpy)3]3+ and [Ir(bpy)3]3+ where the
p–p* transition of the coordinated bpy shifted to lower energy
as the positive charge on the complex was increased.16,17 In the
monooxidised species, the decrease in intensity of the peak at
19 000 cm−1, which has already been assigned as a MMLLꢀCT
peak from the mnt moiety, is consistent with the site of oxidation
as based at least in part on the mnt motif.
Reduction of the free ligand becomes harder in the order
{5,5ꢀ-(CO2Et)2-bpy}, {4,4ꢀ-(CO2Et)2-bpy}, {3,3ꢀ-(CO2Et)2-bpy},
since the electron-withdrawing groups at the 5,5ꢀ position of
the bpy provide greater stabilisation of the monoanion than
substituents at the 4,4ꢀ or 3,3ꢀ positions.12 The electrochemistry
of [Pt{5,5ꢀ-(CO2Et)2-bpy}(mnt)] (3c) shows both reductions at
the least negative potential of the series 1c–3c, as expected,12
however the first reduction potential (E1) of [Pt{3,3ꢀ-(CO2Et)2-
bpy}(mnt)] (1c) occurs at less negative potential than that of
[Pt{4,4ꢀ-(CO2Et)2-bpy}(mnt)] (2c) which is in contrast with the
reduction potentials of the free ligands 1a and 2a. The crystal
structure of [Pt{3,3ꢀ-(CO2Et)2-bpy}(mnt)] previously reported by
us6 shows a significant torsion angle (30.7◦) between the two
pyridine rings of the bpy moiety while [Pt(II)(bpy)(dithiolate)]
complexes with substituents at the 4,4ꢀ positions have much
smaller torsion angles e.g. 0.68◦ for a di-Me substituted bpy,13
and 12.69◦ for a di-tBu substituted bpy14 (the biggest for any 4,4ꢀ-
disubstituted bpy). This suggests that the structural differences
between the 3,3ꢀ and 4,4ꢀ disubstituted molecules influence their
relative electrochemical responses and this results in the first
reduction of 1c being intermediate between that of 2c and 3c.
For each of the molecules studied, the separation of the first and
second reduction is between 0.36 V and 0.66 V. This is associated
with the spin pairing energy of the two added electrons in the same
orbital (vide infra).
(iii) EPR in fluid solution
In situ EPR spectroelectrochemistry was performed on each
member of the [Pt{X,Xꢀ-(CO2Et)2-bpy}(mnt)] series (1c, 2c, 3c)
and comparison made where possible with the precursor species
1a,b, 2a,b and 3a,b. Data observed were broadly comparable to
previous studies of related complexes.18
(a) Reduced 3,3ꢀ-(CO2Et)2-bpy ligand and complexes (1a1−,
1b1−, 1c1−)
Reduction of 1c in a solution of 0.1 M TBABF4 in DMF at 233K
gives an EPR active solution of 1c1−. The coupling to the ligand
nuclei is best resolved at 293 K (Fig. 3) and the coupling constants
for 1c1− are presented, alongside those of 1b1− (Fig. S5†) in Table 2.
Coupling is observed to the Pt centre, two ring N atoms and two
ring H atoms. These are most likely to be the hydrogens in the
5,5ꢀ positions as this has previously been shown to be the most
electronically important position.12 This assignment is supported
by DFT calculations (vide infra) that show the LUMO of the
neutral complex to have a much higher contribution from the ring
carbons in the 5-position than any of the other carbon atoms of
The electrochemistry of the carboxylic acid substituted com-
plexes (1d, 2d and 3d) shows broadly similar oxidation and
reduction potentials to those of the ester-substituted complexes
(1c, 2c and 3c), suggesting that the more-soluble ester species used
for convenience in the further studies below are good models for
the acid derivatives used in DSSCs.
(ii) Spectroelectrochemistry
UV/Vis/NIR spectroelectrochemistry of reduced 1c, 2c and 3c
was reported in our prior work6 and in this work we have also
studied the spectroelectrochemistry of the corresponding species
1a,b, 2a,b and 3a,b as well as oxidations of 1c, 2c and 3c, which
were observed to be chemically reversible at −60 ◦C (observed
absorptions are tabulated in the ESI, Table S1, Fig. S1–S4†).
Electrochemical reduction to the monoanionic and dianionic
species for 1a–c, 2a–c and 3a–c show similar trends. For the
complexes 1–3b and 1–3c, similar bands appeared compared
with the free ligands 1–3a consistent with the reduction electron
locating on the bpy ligand. Spectral features of the reduced,
substituted bpy ligands in turn were similar to the unsubstituted
bpy1−.15 Upon oxidation of 1c, 2c and 3c, each of the spectra
Fig. 3 Left: 2nd derivative EPR spectrum of 1c1− in 0.1 M TBABF4–DMF
at 293 K. Right: simulated spectrum using parameters given in Table 2.
Table 2 EPR data for 1b1− and 1c1− in 0.1 M TBABF4 in DMF at 293 K. All hyperfine coupling constants given in G with 10−4 cm−1 in parenthesis. D =
linewidth
1b1−
3.7 (3.4)
3.7 (3.4)
45.5 (42.2)
4.1
1c1−
4.0 (3.7)
4.0 (3.7)
42.0 (39.2)
4.3
2b1−
2c1−
3b1−
3c1−
aiso (H × 2)
aiso (N × 2)
aiso (Pt)
D
g
59.5 (54.9)
1.988
56.0 (52.1)
1.994
39
36
1.997
2.001
1.994
1.967
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 3701–3708 | 3703
©

View Article Online
the ring (Table S4†). On introduction of another electron to each
complex and formation of 1b2− and 1c2− respectively, the signal
collapses as the second electron enters the same orbital as the first
and they spin pair. The similarities between the spectra of 1b1−
and 1c1− are consistent with the reduction electron entering an
orbital primarily located on the derivatised bpy motif analogous
to the results found in the UV/Vis/NIR spectroelectrochemistry
experiments. We were unable to obtain the spectrum for 1a1− due
to poor stability in the reduced form.
positions give much the largest contribution to the LUMO of the
neutral compound compared with the other ring carbons (Table 3).
The solution EPR of 2b1− at 233 K in solution of 0.1 M TBABF4
in DMF shows a broad line with two 195Pt satellites and any
coupling to the ligand nuclei for this species remains unresolved
at a range of temperatures. This is analogous to the in situ EPR
spectroelectrochemistry results for the related complex 2c1−. As
was the case for 1b2− and 1c2−, on introduction of another electron
to form both 2b2− and 2c2− the signals collapse as the second
electron enters the same orbital as the first and they spin pair
forming a diamagnetic species.
(b) Reduced 4,4ꢀ-(CO2Et)2-bpy ligand and complexes (2a1−,
2b1−, 2c1−)
(b) Reduced 5,5ꢀ-(CO2Et)2-bpy ligand and complexes (3a1−,
3b1−, 3c1−)
In situ reduction of the ligand (2a) gives an EPR active solution
(Fig. 4) with coupling constants listed in Table 3. The reduction
electron couples strongly to the ring nitrogen and to one ring
hydrogen on each pyridine ring. Again, the ring hydrogen atoms
that show strong coupling can be assigned as those in the 5-
position through DFT calculations that show the carbons in these
1−
The in situ EPR spectrum of the {5,5ꢀ-(CO2Et)2-bpy} ligand
(3a1−) shows coupling of the reduction electron to the two ring
nitrogen atoms, and two sets of two ring hydrogen atoms assigned
as shown in Table 3. It is of note in this case that no large coupling
to ring hydrogens occurs since the most electronically important
5-positions contain the ester substituents rather than H-atoms.
The solution EPR of 3b1− at 273 K in solution of 0.1 M TBABF4
in DMF shows coupling of the reduction electron to the 195Pt
nucleus. Any coupling to the ligand nuclei for this species remains
unresolved at a range of temperatures similar to the results found
for 2b1−. Again in this case the EPR spectrum of 3c1− is analogous
to the in situ EPR spectrum for the related dichloride complex 3b1−
which shows a similar magnitude of the coupling of the reduction
electron to the Pt nucleus.
(iv) EPR in frozen solutions
On cooling to 173 K rhombic X-band EPR spectra were
obtained for 1b1−, 1c1− and 2c1− and, following spectral simulation,
the parameters obtained are listed in Table 4. For all compounds,
the average of g1 + g2 + g3 is in good agreement with giso.
The small shift in giso from the free electron value, ge of 2.0023
suggests that there is only a small admixture of metal orbitals
in the SOMO and that the reduction electron is therefore based
mainly on the bpy ligand. These results reiterate the findings of the
Fig. 4 Left: EPR spectrum of 2a1− in 0.1 M TBABF4–DMF at 233 K.
Egen = −1.85 V. Right: simulated EPR spectrum of 2a1− using parameters
given in Table 3.
Table 3 EPR coupling constants for 2a1− and 3a1− in 0.1 M TBABF4–DMF at 233 K. All hyperfine coupling constants given in G. D = linewidth. For
comparison, the calculated percentage contribution to the LUMO from the ring carbon and nitrogen atoms for the neutral compounds is given, with the
position of the atom on the bipy ring shown in parentheses
Parameter
2a1−
%LUMO
3a1−
%LUMO
aiso (H × 2)
aiso (H × 2)
aiso (H × 2)
aiso (N × 2)
D
3.45
0.42
0.40
2.90
0.56
2.004
13.41 (5-position)
0.11 (6-position)
0.10 (3-position)
12.75
0.90
—
0.41
1.30
0.69
2.003
5.89 (4-position)
1.88 (6-position)
3.78 (3-position)
5.62
g
Table 4 EPR parameters of compounds in 0.1 M TBABF4–DMF at 173 K. The Aiso values are taken from solution spectra. Nominal A3 values of 20 G
were used in the simulations as these are not resolved. The A3 values in the table are calculated as 3Aiso − (A1 + A2). W1–3 refer to linewidths
Compound
giso
g1
g2
g3
A
iso/cm−1
A1/cm−1
A2/cm−1
A3/cm−1
W1/G
W2/G
W3/G
1b
1c
2c
1.9975
2.0014
1.994
2.038
2.033
2.044
2.012
2.019
2.025
1.972
1.975
1.936
42.2 × 10−4
39.2 × 10−4
52.1 × 10−4
44 × 10−4
40 × 10−4
55 × 10−4
60 × 10−4
49 × 10−4
64 × 10−4
23 × 10−4
29 × 10−4
37 × 10−4
5
5
8
7
7
8
18
18
20
3704 | Dalton Trans., 2008, 3701–3708
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
UV/Vis/NIR spectroelectrochemical study whereby the reduction
electron locates itself on the bpy motif. In each case, coupling was
resolved to the 195Pt nucleus for each of the g-components but no
further splittings to ligand nuclei were resolved. Experimental and
simulated spectra are shown in Fig. 5 and 6 and S7.† Previous
studies have shown that the singly-occupied orbital in related
complexes is of b2 symmetry assuming C2v point symmetry with
metal contributions to the singly occupied orbital consisting of 5dyz
and 6pz contributions. (The principle axes are defined as shown in
Fig. 1.18) The contribution of these orbitals to the SOMO can be
determined using eqn (1)–(3).18,19
Ayy = As − (2/7)Pda2 − (2/5)Ppb2
Azz = As + (2/7)Pda2 + (4/5)Ppb2
The parameters a2 and b2 are respectively the contribution of
5dyz and 6pz to the SOMO, Pd and Pp are the electron nuclear
dipolar coupling parameters for platinum 5d and 6p electrons
respectively and As is the isotropic Fermi contact term. Using
calculated values Pd = 0.0549 cm−1 and Pp = 0.0402 cm−1,20 we
can derive values for a2 and b2 (Table 5).
This allows us to compare the effects of mnt substitution for two
chloride ligands by examination of the values for 1b1− and 1c1−. It is
apparent that a reduction in Pt contribution to the SOMO occurs
in the mnt complex 1c1− and this may be associated with greater
electron delocalisation in the latter consistent with the extended
conjugation of the mnt. Likewise, we can compare the influence
of altering the ester groups from the 4,4ꢀ to the 3,3ꢀ-positions (2c1−
and 1c1−). The former shows a noticably higher Pt contribution to
the SOMO. This may be rationalised in terms of a better overlap
between the bpy and the Pt orbitals for the 4,4ꢀ-derivative due to
greater planarity across the molecules, whereas the twist in the bpy
ligand of the 3,3ꢀ-analogue reduces the orbital overlap of the bpy
ligand with the Pt centre.
(2)
(3)
Axx = As − (4/7)Pda2 − (2/5)Ppb2
(1)
(v) DFT calculations and discussion
To further interpret the electronic character of these complexes, hy-
brid density functional theory (DFT) calculations were performed
on 1b, 1c, 2a, 2c, 3a and 3c (see Experimental section for details).
Plots of the isosurfaces and calculated energies of the HOMO,
LUMO, HOMO − 1 and LUMO + 1 orbitals can be found in
Fig. 7 (1c) and in the ESI (Fig. S8–S17†). A comparison of selected
Fig. 5 EPR spectrum of 1c1− in a solution of 0.1 M TBABF4 in DMF at
173 K. Egen = − 2.0 V. Simulated spectrum shown beneath.
Fig. 6 EPR spectrum of 2c1− in a solution of 0.1 M TBABF4 in DMF at
173 K. Egen = −1.15 V. Simulated spectrum shown beneath.
Fig. 7 (A) Calculated HOMO of 1c. (B) Calculated LUMO of 1c.
(C) Calculated HOMO − 1 of 1c. (D) Calculated LUMO + 1 of 1c.
Table 5 Platinum 5dyz and 6pz admixtures and total contribution to the SOMO of the anions determined by EPR, compared with hybrid-DFT
calculations of the Pt contribution to the LUMO and to the HOMO of the neutral complexes
a2
b2
Total Pt SOMO (EPR)
Total Pt LUMO (calc.)
Total Pt HOMO (calc.)
1b1−
1c1−
2c1−
0.034
0.019
0.019
0.050
0.023
0.073
0.084
0.042
0.092
1b
1c
2c
3c
0.078
0.058
0.083
0.429
0.118
0.106
0.111
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 3701–3708 | 3705
©

View Article Online
angles and bond distances from the calculated structure of 1c and
the XRD structure shows that the calculated structure is very
close to that experimentally observed, Table S6.† All calculations
are consistent with the HOMO based largely on the mnt moiety
and the LUMO largely based on the bpy moiety (Table S7†) as
previously reported, with a larger total contribution of Pt orbitals
to the HOMO than to the LUMO (Table 5) for 1c, 2c and 3c.1,2,10
For the dichloride 1b, the Pt contribution to the HOMO is seen to
be much larger than for the three mnt-complexes in keeping with
the capacity of the dithiolene ligand to enhance delocalisation of
the frontier orbitals in the complex. We also observe that the Pt
contribution to the LUMO in 1b, 1c and 2c matches the trend
for the Pt contribution to the SOMO determined from the EPR
results. In fact, there is also a reasonable match with the absolute
values however, although gratifying, this may be fortuitous and it
is the reproduction of the trend that is most significant.
The noticable difference in the Pt contribution to the
LUMO/SOMO of 1c/1c1− compared with 2c/2c1− is likely to be
due to the non-planar character of the former caused by the steric
hindrance at the 3,3ꢀ-positions and we can therefore surmise that
this causes poorer overlap with the Pt orbitals. Furthermore, the
computational results show clear evidence of the consequences of
this non-planarity in the LUMO and LUMO + 1 orbitals (Fig. 7).
In particular, the latter orbital is localised on one of the pyridyl
rings of the ligand only, in complete contrast to the corresponding
orbital for the planar 2c (Fig. S13†) which is distributed over the
whole bpy ligand.
900 cm−1 from those observed. The relative size of the oscillator
strengths match the relative intensities of the observed spectra
well, with the bpy intraligand transitions being the most intense
and the two CT bands being of a similar magnitude to one another.
Lastly, the observed trend in the energy of the HOMO–LUMO
transitions was reproduced with 1c being the lowest and 2c being
the highest.
In our prior work we determined that 1c showed a better solar-
to-electric power-conversion efficiency than 2c (and much better
than 3c).6 In particular this arose from a higher open-circuit
voltage for a cell using 1c which was explained by the longer
charge recombination time (between electrons in the TiO2 and the
oxidised dye) that was determined for this complex.6 This charge-
recombination is known as a key loss process and its inhibition is
one of the widely-studied approaches to dye design for DSSCs.24
The results presented in this work suggest that the longer-
lived charge-separated state for complex 1c on TiO2 is related
to the non-planar geometry of the complex. We have shown, by
the lower contribution of the Pt orbitals to the SOMO/LUMO
of 1c compared with 2c as shown by EPR/DFT studies, that
this reduces the electronic coupling between the bpy ligand
and the Pt(mnt) fragment of the dye as well as reducing the
electronic coupling across the bpy ligand itself. According to the
electrochemical and spectroelectrochemical results, the positive
charge density on the oxidised dye is located largely on the mnt
ligand and this conclusion is supported by DFT calculation of the
HOMO location on the neutral complexes. The poorer electron
delocalisation across the dye in the case of 1c will then lead to
poorer electronic communication between the cation on the dye
and electrons in the TiO2 giving a longer-lived charge-separated
state.
Time-dependant density functional theory (TD-DFT) calcu-
lations were carried out for 1c, 2c and 3c in the presence of
DMF. This method has been previously used on similar platinum–
diimine–dithiolate systems to great effect with the results show-
ing a remarkable similarity to experimental observations.21,22
A
Interestingly, the observation of significantly poorer photo-
voltaic performance using 3c is also consistent with this analysis.
The electrochemical and EPR results in this work and previous
ENDOR results demonstrate that the 5,5ꢀ-positions on the bpy
are the most strongly electronically coupled to the ring. Thus, the
poor performance of 3c may in part be explained by enhanced
charge recombination via strong electronic communication from
the TiO2 to the positive charge on the oxidised dye via the acid
groups in the 5-position of the bpy.
summary of relevant calculated transitions for each complex can
be found in Tables S8–S10† and figures showing the calculated
transitions relative to the observed absorption spectrum are shown
in Fig. 8 and S18.† The calculated composition of the lowest energy
transition for all three complexes was found to be 70% HOMO–
LUMO and MMLLꢀCT in nature as expected. The calculated
energy was slightly underestimated, by an average of 1660 cm−1,
for all three complexes. It should be noted that the inclusion of
solvent in these calculations is essential, as has been observed
previously in similar systems,23 as the TD-DFT studies carried out
under vacuum vastly underestimated the energy of the MMLLꢀCT
by ∼12000 cm−1. The highest energy transitions were identified as
bpy intraligand transitions and calculated to be no more than
Conclusions
This work contains the first detailed spectroelectrochemical
(UV/Vis/NIR and EPR) and DFT study on a family of
[Pt(diimine)(dithiolate)] dyes and their synthetic precursors. These
molecules are important in their relevance to the area of solar cell
dyes and also more generally for their optical and electronic func-
tionality. The optically transparent thin layer electrode (OTTLE)
and in situ EPR experiments confirm previous investigations which
assign the location of the bpy based LUMO, although they also
demonstrate partial involvement of the metal centre in the LUMO
as indicated by coupling of the reduction electron to the Pt nucleus,
supported by computational results. Trends in the OTTLE study of
the mnt dyes indicate that in general the location of the reduction
electron is similar within the 3,3ꢀ, 4,4ꢀ and 5,5ꢀ motifs.
Most significantly, the EPR and DFT studies provide a rationale
for the superior performance of the 3,3ꢀ-substituted bpy-complex
over the 4,4ꢀ-substituted bpy counterpart in DSSCs. This provides
Fig. 8 Calculated TD-DFT results shown as a bar chart (left axis) relative
to the observed absorption spectrum (right axis) in DMF of 1c.
3706 | Dalton Trans., 2008, 3701–3708
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
important guidance in the optimisation of such solar cell dyes in
future studies and we suggest that the focus in the DSSC field on
4,4ꢀ-bipy ligands with almost no investigation of 3,3ꢀ-bipy may be
a missed opportunity for dye enhancement.
financial support and we thank the EPRSC Multifrequency EPR
National Service.
Notes and references
1 S. D. Cummings and R. Eisenberg, J. Am. Chem. Soc., 1996, 118, 1949.
2 M. Hissler, J. E. McGarrah, W. B. Connick, D. K. Geiger, S. D.
Cummings and R. Eisenberg, Coord. Chem. Rev., 2000, 208, 115.
3 E. A. M. Geary, N. Hirata, J. Clifford, J. R. Durrant, S. Parsons,
A. Dawson, L. J. Yellowlees and N. Robertson, Dalton Trans., 2003,
3757.
4 A. Islam, H. Sugihara, K. Hara, L. Pratap Singh, R. Katoh, M.
Yanagida, Y. Takahashi, S. Murata and H. Arakawa, New J. Chem.,
2000, 24, 343.
5 A. Islam, H. Sugihara, K. Hara, L. P. Singh, R. Katoh, M. Yanagida,
Y. Takahashi, S. Murata, H. Arakawa and G. Fujihashi, Inorg. Chem.,
2001, 40, 5371.
6 E. A. M. Geary, L. J. Yellowlees, L. A. Jack, I. D. H. Oswald, S. Parsons,
N. Hirata, J. R. Durrant and N. Robertson, Inorg. Chem., 2005, 44,
242.
7 J. Zhang, P. Du, J. Schneider, P. Jarosz and R. Eisenberg, J. Am. Chem.
Soc., 2007, 129, 7726.
8 M. K. Nazeeruddin and M. Graetzel, Comprehensive Coordination
Chemistry II, ed. J. A. McCleverty and T. J. Meyer, Elsevier Science,
Amsterdam, 2004, vol. 9, p. 719.
9 E. Galoppini, Coord. Chem. Rev., 2004, 248, 1283; N. Robertson,
Angew. Chem., Int. Ed., 2006, 45, 2338.
10 W. Paw, S. D. Cummings, M. A. Mansour, W. B. Connick, D. K. Geiger
and R. Eisenberg, Coord. Chem. Rev., 1998, 171, 125.
11 P.-H. Xie, Y.-J. Hou, T.-X. Wei, B.-W. Zhang, Y. Cao and C.-H. Huang,
Inorg. Chim. Acta, 2000, 308, 73; Y.-J. Hou, P.-H. Xie, B.-W. Zhang, Y.
Cao, X.-R. Xiao and W.-B. Wang, Inorg. Chem., 1999, 38, 6320.
12 L. Jack, Ph.D. Thesis, University of Edinburgh, 2003.
13 J. M. Bevilacqua and R. Eisenberg, Inorg. Chem., 1994, 33, 2913.
14 B. W. Smucker, J. M. Hudson, M. A. Omary and K. R. Dunbar, Inorg.
Chem., 2003, 42, 4714.
15 E. Koenig and S. Kremer, Chem. Phys. Lett., 1970, 5, 87.
16 V. T. Coombe, G. A. Heath, A. J. MacKenzie and L. J. Yellowlees, Inorg.
Chem., 1984, 23, 3423.
17 G. A. Heath, L. J. Yellowlees and P. S. Braterman, Chem. Commun.,
1981, 287.
18 E. J. L. McInnes, R. D. Farley, C. C. Rowlands, A. J. Welch, L. Rovatti
and L. J. Yellowlees, J. Chem. Soc., Dalton Trans., 1999, 4203.
19 P. H. Rieger, J. Magn. Reson., 1997, 124, 140.
Experimental
The synthesis of each compound 1–3(a–d) has previously been
reported.6 Electrochemical studies were carried out using a
DELL GX110PC with General Purpose Electrochemical System
(GPES), version 4.8, software connected to an autolab system
containing a PGSTAT 20 potentiostat. The technique used a three
electrode configuration, with a 0.5 mm diameter Pt disc working
electrode, a Pt rod counter electrode and an Ag/AgCl (saturated
KCl) reference electrode against which the ferrocenium/ferrocene
couple was measured to be +0.55 V. The supporting electrolyte
was 0.1 M tetrabutylammonium tetrafluoroborate (TBABF4),
solvent was dry, degassed DMF and concentration of compound
approximately 1 mmol.
OTTLE measurements were taken using a quartz cell of 0.5 mm,
a Pt/Rh gauze working electrode, an Ag/AgCl reference electrode
and a Pt wire counter electrode. UV-Vis spectra were recorded
on a Perkin-Elmer Lambda 9 spectrophotometer, controlled by a
Datalink PC, running UV/Winlab software. 0.1 M TBABF4 was
used as the supporting electrolyte in all cases.25
All in situ EPR spectra were recorded on an X-band Bruker
ER200D-SCR spectrometer, connected to a Datalink 486DX
PC running EPR Acquisition System, version 2.42 software. In
situ EPR experiments were electrogenerated using a BAS CV-
27 voltammograph. Variable temperature work was carried out
using a Bruker ER111VT variable temperature unit. In some
cases, solution spectra were recorded at reduced temperature to
ensure long-term stability of the reduced species. All g values
were corrected to 2,2ꢀ-diphenyl-1-picrylhydrazyl with gliterature
2.0036 0.0002. Spectra were simulated by manipulation of
=
parameters until a satisfactory reproduction of experimental
spectra was obtained, assuming couplings to the following nuclei:
20 E. J. L. McInnes, R. D. Farley, S. A. MacGregor, K. J. Taylor, L. J.
Yellowlees and C. C. Rowlands, J. Chem. Soc., Faraday Trans., 1998,
2985.
1
195Pt (I = 2 , 33.8% abundance), H (I = 12 , 100% abundance)
and 14N (I = 1, 100% abundance). Since each of these gives a
different combination of nuclear spin (hence splitting) and natural
abundance, no ambiguity in nuclear types arises in the simulations.
Density functional theory calculations of all complexes were
performed using either the Gaussian 03 program package26 or the
GAMESS-UK package.27 The starting structure was input using
the builder program Arguslab and default convergence conditions
of Gaussian 03 were used. The Becke three parameters hybrid
exchange and the Perdew–Wang 1991 correlation functionals
(B3PW91) were used.28,29 For the platinum atom the Hay–Wadt
VDZ(n+1)ECP was used30 with the other atoms described by 6-
31G*.31 The optimised structures were verified as minima on the
potential energy surface by the absence of negative values in the
frequency calculations. TD-DFT calculations were carried out in
the presence of a polarisable continuum model (PCM) DMF32
solvation field, with the first 25 singlet transitions calculated.
1
21 C. J. Adams, N. Fey, M. Parfitt, S. J. A. Pope and J. A. Weinstein, Dalton
Trans., 2007, 4446–4456.
22 C. Makedonas and C. A. Mitsopoulou, Eur. J. Inorg. Chem., 2006,
2460–2468; C. Makedonas, C. A. Mitsopoulou, F. J. Lahoz and A. I.
Balana, Inorg. Chem., 2003, 42, 8853.
23 A. Vlcˇek Jr. and S. Za´lisˇ, Coord. Chem. Rev., 2007, 251, 258–287.
24 N. Hirata, J.-J. Lagref, E. J. Palomares, J. R. Durrant, M. K.
Nazeeruddin, M. Gra¨tzel and D. Di Censo, Chem.–Eur. J., 2004, 10,
595.
25 E. Alessio, S. Daff, M. Elliot, E. Iengo, L. A. Jack, K. G. Macnamara,
J. M. Pratt and L. J. Yellowlees, Spectroelectrochemical techniques,
Trends Mol. Electrochem., 2004, 339.
26 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. 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. G. Johnson,
Acknowledgements
We thank the University of Edinburgh Christina Miller fund
and the EPSRC Excitonic Solar Cells Supergen Consortium for
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 3701–3708 | 3707
©

View Article Online
W. Chen, M. W. Wong, C. Gonzalez and J. A. Pople, GAUSSIAN 03
(Revision C.02), Gaussian, Inc., Wallingford, CT, 2004.
28 J. P. Perdew, J. A. Chevary, S. H. Vosko, K. A. Jackson, M. R. Pederson,
D. J. Singh and C. Fiolhais, Phys. Rev. B, 1993, 48.
27 GAMESS-UK is a package of ab initio programs. See: http://www.
cfs.dl.ac.uk/gamess-uk/index.shtml:M. F. Guest, I. J. Bush, H. J. J.
van Dam, P. Sherwood, J. M. H. Thomas, J. H. van Lenthe, R. W. A.
Havenith and J. Kendrick, The GAMESS-UK electronic structure
package: algorithms, developments and applications, Mol. Phys., 2005,
103, 719–747.
29 J. P. Perdew, K. Burke and Y. Wang, Phys. Rev. B, 1996, 54, 16533.
30 P. J. Hay and W. R. Wadt, J. Chem. Phys., 1985, 82, 299.
31 M. M. Francl, W. J. Pietro, W. J. Hehre, J. S. Binkley, M. S. Gordon,
D. J. Defrees and J. A. Pople, J. Chem. Phys., 1982, 77, 3654.
32 E. S. Boes, P. R. Livotto and H. Stassen, Chem. Phys., 2006, 331,
142.
3708 | Dalton Trans., 2008, 3701–3708
This journal is
The Royal Society of Chemistry 2008
©