Increased photocurrent in a tandem dye-sensitized solar cell by modifications in push-pull dye-design

Article abstract of DOI:10.1039/c4cc10230d

Donor-π-acceptor photosensitizers for NiO photocathodes that exhibit a broad spectral response across the visible region are presented. These enabled an increase in the photocurrent density of p-type dye-sensitized solar cells to 8.2 mA cm^-2 and a tandem cell to be assembled which generated a photocurrent density of 5.15 mA cm^-2.

Full text of DOI:10.1039/c4cc10230d

View Article Online
View Journal
ChemComm
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: C. J. Wood, G. H.
Summers and E. Gibson, Chem. Commun., 2015, DOI: 10.1039/C4CC10230D.
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
Accepted Manuscripts are published online shortly after
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
Information for Authors.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the Ethical guidelines still
apply. In no event shall the Royal Society of Chemistry be held
responsible for any errors or omissions in this Accepted Manuscript
or any consequences arising from the use of any information it
contains.
www.rsc.org/chemcomm

View Article Online
DOI: 10.1039/C4CC10230D
Page 1 of 3
ChemComm
Journal Name
RSCPublishing
COMMUNICATION
Increased photocurrent in a tandem dye-sensitized
solar cell by modifications in push-pull dye-design.
Cite this: DOI: 10.1039/x0xx00000x
C. J. Wood,a G. H. Summersa,b and E. A. Gibsona,b*
Received 00th January 2012,
Accepted 00th January 2012
DOI: 10.1039/x0xx00000x
www.rsc.org/
and bodipy chromophores.6,8 Appending these groups through a
Donor-π-acceptor photosensitizers for NiO photocathodes
that exhibit a broad spectral response across the visible
region are presented. These enabled an increase in the
photocurrent density of p-type dye-sensitized solar cells to 8.2
mA cm–2 and a tandem cell to be assembled which generated
a photocurrent density of 5.15 mA cm–2.
thiophene π-linker to the 4-(diphenyl amino) benzoic acid
anchor/donor motif gives dyes which absorb longer wavelengths that
complement state-of-the-art photoanodes in tandem cells (Figure 1).
For the photoanode dye we selected D35 (Figure S1 in the ESI), a
well-known push-pull sensitizer for TiO2 that is designed to transfer
Dye-sensitized solar cells (DSCs) based on mesoporous electron density towards the n-type semiconductor via the carboxylic
semiconductor electrodes are often presented as low-cost alternatives acid anchoring group.9 Conversely, P1 pulls electrons away from the
to conventional silicon devices.1 However to date their efficiency has semiconductor surface and is one of the best performing dyes with
only reached 13% compared to 25% for crystalline silicon.2 State-of- NiO (reported IPCE = 63%).10 However, the spectral response for P1
the-art DSCs are based on photoanodes, so the photocurrent overlaps almost entirely with D35 (Figure 1). Qin et al. modified P1
generated when light is absorbed is a result of electron transfer from to absorb longer wavelengths by incorporating additional cyano
the excited dye to the conduction band of the TiO2, an n-type groups on the malonitrile acceptor unit.10 This resulted in a
semiconductor. An efficient dye-sensitized photocathode would significant lowering of the LUMO energy which diminished the
–
enable tandem cells to be constructed, where the platinised counter driving force for dye regeneration‡ by the I3 /I– redox mediator such
electrode found in n-DSCs is replaced with a dye-sensitized p-type that there was a complete loss of photocurrent. The cationic 1-hexyl-
semiconductor.3,4 This enables more light to be converted more 2,3,3-3H-indolium acceptor dye (CAD3) was synthesised by
efficiently, giving a maximum theoretical efficiency of 43% condensing 1-hexyl-2,3,3-trimethyl-3H-indolium PF6 onto 4-
compared to 33% for a single junction device. Despite this potential carboxy-4’,4’’-diformyl-2-thienyl)triphenylamine in the presence of
for a step-change in the photoconversion efficiency of dye-sensitized piperidine (see ESI). The bodipy analogue (GS1) was prepared by
solar cells, a tandem device that is more efficient than TiO2 alone reacting 2-phenyl-1H-pyrrole with 4-carboxy-4’,4’’-diformyl-2-
has not been reported because an efficient p-DSC has not yet been thienyl)triphenylamine in the presence of CF₃CO₂H before
produced. An increasing amount of work has been devoted to
developing p-DSCs, where light absorption by the dye is followed
by rapid electron transfer from the valence band of the
oxidation with chloranil and coordination using BF3.O(C2H5)2.
The ground state geometry and electronic distribution in GS1 and
CAD3 were predicted using hybrid-DFT calculations (See ESI). The
semiconductor, typically NiO, to the dye.3 However, the efficiency HOMO–LUMO transition was confirmed to be the dominant
of p-DSCs, often based on dye-sensitized NiO, remains far lower
(1.4%5) than that obtained for n-DSCs and so tandem devices
generally have a lower efficiency§ than typical n-type devices. This
electronic transition in both dyes by TDDFT (Figure S11). B3LYP is
known to underestimate the energy of photoexcitation for charge
transfer transitions, however the calculated solution phase spectra
effect is further enhanced because of the overlap of the light compare well with the experimental data presented in Figure S11.11
absorption between the n- and p-type dyes.4
The isodensity plots illustrate the “push-pull” nature of the dyes by
Dyes typically used in conventional TiO2-based DSCs convert
predicting that this low energy transition is accompanied by a shift
of electron density from the 4-(diphenyl amino) benzoic acid
photons up to 600 nm with high quantum efficiency.1 Therefore,
dyes used for the photocathode should harvest wavelengths longer anchor/donor unit (HOMO) to the bodipy (LUMO) in GS1 or 1-
than 600 nm. To date there have been very few reports of dyes for
photocathodes that absorb at longer wavelengths and generate
sufficient photocurrents to be used in tandem DSCs.6,7 Noticeably,
hexyl-2,3,3-trimethyl-3H-indolium (LUMO) for CAD3. The optical
and electrochemical properties of CAD3 and GS1 determined
experimentally are summarised in the ESI (Table S1). CAD3 has a
compared to the number of recent publications on p-DSCs, relatively very broad absorption spectrum with λmax well towards the red
few tandem DSCs have been reported.3,4
The aim of this work was to build on our previous success with dyes
incorporating a cationic 1-hexyl-2,3,3-3H-indolium acceptor unit
(Figure S3 and S4). Adsorption of CAD3 onto NiO gave rise to a 48
nm hypsochromic shift indicating strong electronic interaction
between the dye and the NiO. When GS1 was adsorbed on NiO there
This journal is © The Royal Society of Chemistry 2012
J. Name., 2012, 00, 1-3 | 1

View Article Online
ChemComm
DOI: 10.1039/C4CC10P23a0gDe 2 of 3
COMMUNICATION
Journal Name
energy but the orbital energy dictates the estimated driving force for
charge separation (ΔGinj) and dye regeneration (ΔGreg).‡ According
to Liu et al. ΔGinj > –0.8 eV is required for efficient photoinduced
charge separation at the dye/NiO interface.12 We estimate‡
ΔGinj(CAD3) ≈ –0.91 eV and ΔGinj (GS1) ≈ – 0.98 eV, this implies
that these new dye molecules should be able to withdraw electrons
from NiO following light absorption. Charge-recombination at the
dye/semiconductor interface is rapid in NiO DSCs, so ΔGreg of the
dye by the electrolyte must be substantial in order for charge transfer
in the forwards direction to compete with the reverse process.10 We
estimate‡ ΔGreg (CAD3) ≈ –0.17 eV and ΔGreg (GS1) ≈ –0.54 eV
indicating that electron density on either the indolium or bodipy
moiety can be intercepted by the redox electrolyte.13 Less energy is
wasted in these electron transfer processes for CAD3 and GS1 than
for P1(ΔGinj (P1) ≈ –1.21 eV and ΔGreg (P1) ≈ –0.64 eV).
An n-DSC incorporating D35 and three p-DSCs incorporating P1,
GS1 and CAD3 were assembled using platinised conductive glass
–
counter electrodes and infiltrated with I3 /I– electrolyte. Figure 1
shows the spectral response of the p-DSCs compared to the n-DSC
and Table 1 lists the p-DSC characteristics. Shifting the λmax of the
dyes to longer wavelength means more photons are absorbed leading
to a higher photocurrent density from the p-DSC. Of the three dyes,
the CAD3 device had the highest spectral response in the red region
and had the highest efficiency.
The photocurrent density obtained with GS1 was almost twice that
generated with the previous bodipy dye reported by our group,
which was based on 2,4-dimethyl-3-ethylpyrrole.8 By avoiding
methyl substituents at the 2 position of the bodipy, the electron
delocalisation has been extended from the triphenylamine-thiophene
donor motif to the bodipy in GS1, as evidenced by the isodensity
plots in Figure S12. Accordingly, the HOMO-LUMO gap has been
reduced, broadening and red-shifting the absorption by the dye from
λmax = 540 nm previously to λmax = 565 nm for GS1. This change in
structure and optical properties has had a direct impact on the p-DSC
characteristics, increasing the IPCE from 28% previously to 53% for
GS1. Extending the spectral response to the red for CAD3 led to a
photocurrent density of 8.2 mA cm–2. This is more than double the
photocurrent density we previously obtained using asymmetric
indolium cationic electron-acceptor dyes (CAD1: JSC = 3.6 IPCE =
33% and CAD2: JSC = 3.3 mA cm−2 IPCE = 27%).6 The high molar
absorption coefficient of CAD3 is almost four times that of CAD1
and CAD2 and we attribute the improvement in photocurrent to the
increased light harvesting efficiency of the CAD3 electrode.6
Figure 1. Overlaid IPCE plots for a P1 (a), GS1 (b), CAD 3 (c) NiO p-DSC and a D35
TiO2 n-DSC (orange) with the molecular structures adjacent. The structure of D35
is provided in Figure S1 in the ESI.
Table 1. Photovoltaic Performance of p-DSCs and Tandem p/n DSCs.
JSC
VOC FF
η
IPCE
Dye Loading
Dye Cell type
[mA cm-2]a) [mV] [%] [%] [%]b) [10-6 mol cm-2]c)
p-DSC
CAD3
8.21
5.15
5.87
4.54
5.37
3.71
101 31 0.25 50
613 54 1.7
7.36 × 10-06
1.04 × 10-05
1.24 × 10-05
p/n-DSC
p-DSC
GS1
106 31 0.20 53
638 43 1.3
p/n-DSC
p-DSC
P1
89 33 0.16 54
732 38 1.1
p/n-DSC
a)Jsc is the short-circuit current density at the V = 0 intercept, Voc is the open-
circuit voltage at the J = 0 intercept, FF is the device fill factor, η is the
power conversion efficiency; b)ICPE is the monochromatic incident photon-
to-current conversion efficiency; c) number of moles of dye absorbed on a 0.2
cm2 area NiO electrode.
Recently two papers have reported improved photocurrent densities
for p-DSCs using push-pull dyes; Click et al. obtained 7.4 mA cm–2
by incorporating two perylene units per dye molecule (“BH4”),
giving a molar absorption coefficient of 100 000 L mol–1 cm–1,14 Liu
et al. achieved 7.57 mA cm−2 by substituting thiophene units for
fluorene units, extending the π-linker (“zzx-op1–2”).14 However, the
spectral response only extends to 650 nm and therefore if this dye
was incorporated in a tandem cell, both electrodes would be
competing for light. Our modifications to the push-pull design by
substituting the electron acceptor to capture more of the solar
spectrum has led to an even higher photocurrent density. The IPCE
for each dye was c.a. 50%, which is impressive for NiO p-DSCs, but
lower than reported for P1 (63%), PMI-6T-TPA (62%) and zzx-op1–
2 (74%).4,10,15 Since ΔGinj is sufficient for charge separation, we
attribute the lower IPCEs recorded in this work to inefficient dye-
was no shift in the peak maximum but a broadening of the
absorption spectra was observed. The molar absorption coefficient
was lower for GS1 (at λmax= 565 nm, ε = 66 000 L mol–1 cm–1) than
for CAD3 (at λmax= 614 nm, ε = 95 000 L mol–1 cm–1) but higher
than P1.10 GS1 and CAD3 were only mildly emissive in chlorinated
solvents but luminesced sufficiently for us to estimate the lowest
electronic transition energy (E0–0 = 1.78 eV for CAD3 and E0–0
=
2.11 eV for GS1) from the intercept of the normalised absorption
and emission spectra. The electrochemical properties of CAD3 and
GS1 were investigated by cyclic voltammetry and differential pulse
voltammetry, the results are provided in the ESI. GS1 is more easily
reduced than CAD3 and the first reduction process was reversible for
GS1 but not for CAD3.
For the p-DSC to work, the energy levels of the dye must be
carefully positioned so that there is a spontaneous “downhill”
process for the electron transfer from the NiO valence band to the
dye HOMO and from the dye LUMO to the redox electrolyte. The
main challenge associated with dyes that absorb at longer
wavelengths is that the frontier orbitals must be brought closer in
–
regeneration by I3 and/or differences in the NiO preparation
compared to other groups. This gives us confidence that even higher
photocurrent densities are possible, providing that the frontier orbital
energies of the dyes and the quality of the NiO are further optimised.
The VOC for the cells containing CAD3 (101 mV) and GS1 (106 mV)
were slightly better than those obtained with P1 (89 mV) and only
2 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 2012

View Article Online
DOI: 10.1039/C4CC10230D
Page 3 of 3
Journal Name
ChemComm
COMMUNICATION
slightly lower than that reported for zzx-op1−2 (117 mV).10.15 enable better photovoltage to be obtained. However, we believe that
Inspection of the dark current curves (Figure S15) suggests that the the main barrier to providing the promised “step-change” in solar
recombination reaction between the NiO and the electrolyte is cell efficiency is the use of NiO as the p-type semiconductor, and
accelerated for P1, causing the slightly lower photovoltage. The low efforts need to be focused on discovering a semiconductor with an
fill factors are typical for p-DSCs and are due to a combination of energetically lower lying valence band. This would allow us to
detrimental dark and light induced recombination reactions.16
exploit the dyes reported here in a tandem cell so that a VOC of up to
1.5 V is achievable.
EAG thanks the Royal Society for a Dorothy Hodgkin Fellowship
and Research Project and the University of Nottingham for funding.
10
8
6
4
2
0
Notes and references
a
School of Inorganic Chemistry, The University of Nottingham,
University Park, Nottingham, NG7 2PL, United Kingdom.
b Now at the School of Chemistry, Newcastle University, Newcastle upon
Tyne, NE1 7RU, United Kingdom. E-mail: Elizabeth.gibson@ncl.ac.uk
§Nattestad et al. reported p/n DSC with η=1.91%, JSC = 2.4 mA
cm–2, VOC = 1079 mV, FF = 74% when illuminated through the
TiO2; η=2.42% when illuminated through the NiO9)
–
-
‡
ΔGinj = e[EVB(NiO) – ED*/D–]; ΔGreg = e[E(I3 /I2•–) – ED/D–]; E(D*/D )
=
E(D/D-) + E0-0; EVB (NiO) ≈ –0.12 V vs. Fe(Cp)2+/0 15; E°’(I3 /I2•–) = –0.82 V
–
0
100
200
300
400
500
600
700
vs. Fe(Cp)2+/0 in acetonitrile.13
Photovoltage (mV)
Electronic Supplementary Information (ESI) available: experimental
section, absorption and emission spectra, DFT calculations,
Figure 2. Current-voltage plots for a CAD3/NiO p-DSC (triangles), D35/TiO2 n-
DSC (diamonds) and a D35/CAD3 p/n DSC (open triangles).
electrochemistry
DOI: 10.1039/c000000x/
data,
solar
cell
characterisation.
See
The promising photocurrents, and unprecedented current response in
the red region, prompted us to assemble p/n tandem DSCs with P1,
GS1 and CAD3. To achieve current matching between the
photoanode and photocathode, the film thickness of the TiO2
electrode was varied and current-voltage and IPCE measurements
were taken until a current was obtained that matched the current that
would be produced by the photocathode when positioned at the
bottom of the cell (e.g. 5 mA cm–2 for a tandem DSC with CAD3,
Figure S18-S20). This was challenging to achieve for P1 due to the
spectral overlap with D35 which led to very low photocurrents. Very
thin TiO2 layers were used which led to good VOC but the devices
suffered from poor fill factors because we were unable to match the
currents at the two photoelectrodes. Much better results were
achieved with GS1 and CAD3 since the spectral response is red-
shifted relative to P1 and D35. The current-voltage plots for the n-
DSC, p-DSC and tandem DSC for CAD3 are provided in Figure 2
(the equivalent plots for GS1 and P1 are provided in Figure S21 in
the ESI). The photocurrents for the CAD3/D35 and GS1/D35 cells
are the highest reported so far for tandem DSCs. However, the VOC
for each cell, although greater than the VOC of the individual n-DSC
and p-DSC, was lower than typically achieved in the best n-DSCs
(typically >700 mV).1 This is a direct result of the electrolyte chosen
for the devices. A high concentration of lithium ions shifts the
valence band edge of the NiO to lower energy, increasing the VOC in
p-DSCs, but also lowers the conduction band edge of the TiO2
lowering the VOC of the n-DSC.17,18 Attempts to use electrolyte
mixtures optimised for n-DSCs resulted in very poor performances
of the p-DSCs; the electrolyte that worked best in our p-DSCs
caused a large drop in photovoltage at the photoanode (VOC = 500-
550 mV in our n-DSCs compared to 780 mV reported by Jiang et
al.).9 The best tandem cells were obtained with the electrolyte
optimised for p-DSCs (0.1 M I2, 1.0 M LiI).
1
2
A. Hagfeldt, G. Boschloo, L. Sun, L. Kloo, H. Pettersson, Chem. Rev.
2010, 110,6595.
M. A. Green, K. Emery, Y. Hishikawa, W. Warta, E. D. Dunlop, Prog.
Photovolt. 2014, 22, 701.S. Mathew, A. Yella, P. Gao, R. Humphry-
Baker, B. F. E. Curchod, N. Ashari-Astani, I. Tavernelli, U.
Rothlisberger, M. K. Nazeeruddin, M. Grätzel, Nat. Chem. 2014, 6,
242.
3
J. He, H. Lindström, A. Hagfeldt, S.-E. Lindquist, Sol. Energy Mater.
Sol. Cells 2000, 62, 265; F. Odobel, Y. Pellegrin, E. A. Gibson, A.
Hagfeldt, A. L. Smeigh, L. Hammarström, Coord. Chem. Rev. 2012,
256, 2414.
4
5
A. Nattestad, A. J. Mozer, M. K. R. Fischer, Y.-B. Cheng, A. Mishra, P.
Bäuerle, U. Bach, Nat. Mater. 2010, 9, 31.
S. Powar, T. Daeneke, M. T. Ma, D. Fu, N. W. Duffy, G. Götz, M.
Weidelener, A. Mishra, P. Bäuerle, L. Spiccia, U. Bach, Angew. Chem.
2013, 52, 602.
6
C. J. Wood, M. Cheng, C. A. Clark, R. Horvath, I. P. Clark, M. L.
Hamilton, M. Towrie, M. W. George, L. Sun, X. Yang, E. A. Gibson, J.
Phys. Chem. C 2014, 118, 16536.
7
8
9
C.-H. Chang, Y.-C. Chen, C.-Y. Hsu, H.-H. Chou, J. T. Lin, Org. Lett.
2012, 14, 4726.
J.-F. Lefebvre, X.-Z. Sun, J. A. Calladine, M. W. George, E. A. Gibson,
Chem. Commun. 2014, 50, 5258.
X. Jiang, T. Marinado, E. Gabrielsson, D. P. Hagberg, L. Sun, A.
Hagfeldt, J. Phys. Chem. C, 2010, 114, 2799. M. Liang, J. Chen. Chem.
Soc. Rev., 2013,42, 3453
10 P. Qin, J. Wiberg, E. A. Gibson, M. Linder, L. Li, T. Brinck, A.
Hagfeldt, B. Albinsson, L. Sun, J. Phys. Chem. C 2010, 114, 4738.
11 A. Dreuw, M. Head-Gordon, Chem. Rev. 2005, 105, 4009.
12 Z. Liu, D. Xiong, X. Xu, Q. Arooj, H. Wang, L. Yin, W. Li, H. Wu, Z.
Zhao, W. Chen, M. Wang, F. Wang, Y.-B. Cheng, H. He, ACS Appl.
Mater. Interfaces 2014, 6, 3448.
13 E. A. Gibson, L. Le Pleux, J. Fortage, Y. Pellegrin, E. Blart, F. Odobel,
A. Hagfeldt, G. Boschloo, Langmuir 2012, 28, 6485.
14 K. A. Click, D. R. Beauchamp, B. R. Garrett, Z. Huang, C. M. Hadad,
Y. Wu, Phys. Chem. Chem. Phys. 2014, 16, 26103.
15 Z. Liu, W. Li, S. Topa, X. Xu, X. Zeng, Z. Zhao, M. Wang, W. Chen,
F. Wang, Y.-B. Cheng, H. He, ACS Appl. Mater. Interfaces 2014, 6,
10614.
16 D. Dini, Y. Halpin, J. G. Vos, E. A. Gibson, (2015). Coord. Chem.
Rev., 2015, in press CCR-D-14-0.
17 H. Zhu, A. Hagfeldt, G. Boschloo, J. Phys. Chem. C 2007, 111, 17455.
18 Y. Liu, A. Hagfeldt, X. Xiao, S. Lindquist, Sol. Energy Mater. Sol.
Cells 1998, 55, 267.
Conclusions
By re-designing simple donor-π-acceptor dyes to capture the lower
energy portion of visible light and promote photoinduced electron
transfer from NiO to a I3 /I– electrolyte we have increased the
–
photocurrent density in p-DSCs from 5.4 to 8.2 mA cm–2. This has
enabled us to assemble tandem cells with up to 5.2 mA cm–2, which
is substantially higher than any previous tandem DSC. Further
modifications in dye-design will enable us to increase the
photocurrent even further, and alternative redox electrolytes should
This journal is © The Royal Society of Chemistry 2012
J. Name., 2012, 00, 1-3 | 3