You are on page 1of 10

Journal of Power Sources 458 (2020) 228063

Contents lists available at ScienceDirect

Journal of Power Sources


journal homepage: www.elsevier.com/locate/jpowsour

Spiro[fluorene-9,90 -phenanthren]-100 -one as auxiliary acceptor of D-A-π-A


dyes for dye-sensitized solar cells under one sun and indoor light
Rui-Yu Huang , Wen-Hsuan Tsai , Jun-Jie Wen , Yuan Jay Chang *, Tahsin J. Chow **
Department of Chemistry, Tunghai University, No.1727, Sec.4, Taiwan Boulevard, Xitun District, Taichung, 40704, Taiwan

H I G H L I G H T S G R A P H I C A L A B S T R A C T

� Auxiliary acceptor of spiro[fluorene-


9,90 -phenanthren]-100 -one in D-A-π-A
dyes.
� The co-sensitization achieved the high­
est PCE of 27.04% at 2500 lux indoor
light.
� Stability of co-sensition retained 84.49%
of its original PCE after 336 h in air.

A R T I C L E I N F O A B S T R A C T

Keywords: Six novel organic dyes (RY1~RY6) containing spiro[fluorene-9,90 -phenanthren]-100 -one as an auxiliary acceptor
Dyes sensitized solar cells were synthesised and effectively used for the fabrication of D-A-π-A type dye-sensitized solar cells (DSSCs). The
Co-sensitization molecular structures were modified by introducing a novel spiro[fluorene-9,90 -phenanthren]-100 -one auxiliary
Spiro[fluorene-9,90 -phenanthren]-100 -one
acceptor group between the donor and the π-bridge. The molecular rigidity can be enhanced by depressing
High performance under indoor light
intermolecular aggregation and carbonyl group can trapping the Liþ or I3 ions to retard the charge recombi­
nation. The sensitizer RY3 was found to perform remarkable light-harvesting efficiency of 6.30% at AM1.5 solar
condition and 21.67% at TL84 (2500 lux) illuminations without DCA co-deposition. For further improvement, a
higher efficiency can be achieved through a suitable co-sensitization of N719 and RY3, which displayed an
efficiency of 8.55% under one sun (AM 1.5). While operated under indoor light, the efficiency was boosted to
25.57% and 27.04% at 1000 & 2500 lux illuminations, respectively. The high performance of co-sensitization of
N719 and RY3 can be ascribed to a high surface coverage and a broader range of absorption wavelength. Sta­
bility test of the device co-sensitized with N719 and RY3 showed a mild decay of PCE 3.74% after 96 h, while it
retained 84.49% of its original PCE after 336 h in ambient atmosphere without encapsulation.

1. Introduction dioxide emission urged us to investigate and develop renewable and


environmental friendly energy sources. Solar energy is the most abun­
The world limited fossil fuel storage and the increasing carbon dant and promising energy source, despite its intermittent limitation. In

* Corresponding author.
** Corresponding author.
E-mail addresses: jaychang@thu.edu.tw (Y.J. Chang), tjchow@thu.edu.tw (T.J. Chow).

https://doi.org/10.1016/j.jpowsour.2020.228063
Received 31 October 2019; Received in revised form 28 February 2020; Accepted 16 March 2020
Available online 26 March 2020
0378-7753/© 2020 Elsevier B.V. All rights reserved.
R.-Y. Huang et al. Journal of Power Sources 458 (2020) 228063

the past two decades, dye-sensitized solar cells (DSSCs) have received design of hole-transporting materials for high performance perovskite
considerable attention due to their facile fabrication, lightweight, and solar cells [34]. A few examples of the π-spacer in DSSCs with rigid
low cost comparing with the silicon-based photovoltaic cells. The DSSCs spiro-structure are listed in Table 1 [35–40]. The DSSCs made with the
are particularly suitable for indoor applications, therefore can be used in sensitizers FHD5, FHD6, 2, 4, JK-88 and JK-89 showed high Voc values
portable devices under both indoor light and sun light. of 697–760 mV, due to good molecular packing that reduces intermo­
In the development of DSSCs, the most crucial component is the lecular aggregation [38–40]. Most of the sensitizers containing thio­
design of high-performance photo-sensitizers. O’Regan and Gra €tzel have phene moieties showed higher light harvesting efficiency. However,
reported the pioneer work of DSSC by the deposition of ruthenium-based they were usually accompanied by lower values of Voc, due to the strong
sensitizers onto nanoporous TiO2 to achieve an optimal power conver­ interactions between iodine and sulphur atoms that form sensitizer-I2
sion efficiency over 11% [1–3]. However, the ruthenium-based sensi­ and sensitizer-I-3 complexes [41–43].
tizers were expensive due to the high price of rare metal ruthenium. An In this work, we designed six new sensitizers by incorporating spiro
enormous number of metal-free organic sensitizers have been investi­ [fluorene-9,90 -phenanthren-100 -one] as an auxiliary acceptor moiety in
gated and developed in subsequent years. Metal-free organic sensitizers D-A-π-A system (RY1~RY6), and apply them onto DSSCs (Fig. 1). The
have the advantages of environmental friendliness, low cost, higher synthetic procedures are shown in Scheme 1, and all compounds were
structural flexibility and easier purification, etc. For example, the confirmed by spectroscopic analyses.
indenoperylene-based and porphyrin-based sensitizers have achieved a The advantage of inserting spiro[fluorene-9,90 -phenanthren-100 -
maximal efficiency up to 12%. However, the synthesis of sensitizers was one] as an auxiliary acceptor moiety into the dye structure can be evi­
sometimes hampered by the multi-step synthetic procedures and the denced by the following rationales. Compared to MK-160 and MK-162,
lower stability during long-time operations [4,5]. Yano and Hanaya used the analogous compounds RY3 and RY4 exhibited significantly higher
a method of co-sensitization of an alkoxysilyl anchored dye (ADEKA-1) Voc, Jsc and PCE values under an AM1.5 solar condition (100 mW cm 2).
and a metal-free organic dye LEG-4 to achieve a record-high perfor­ The bulky moiety of spiro[fluorene-9,90 -phenanthren-100 -one] increases
mance of 14.3% [6]. Most metal-free organic dyes have a linear shaped the rigidity of molecular structure. It inhibits intermolecular aggrega­
structure consisting of an electron donor (D) and an electron acceptor tion, and prevents the dye from lying down on TiO2 surface. The
(A), which are connected by a π-bridge to form a dipolar D-π-A system. insertion of an auxiliary acceptor onto RY-sensitizers helps them achieve
Recently, Zhu developed a new type “D-A-π-A00 system of organic a better molecular packing on the top of TiO2, both reducing the inter­
sensitizer by inserting an auxiliary acceptor between the D group and molecular aggregation and trapping the Liþ or I3 ions, therefore reduces
the π-bridge, which displayed better photovoltaic performance by the charge recombination and enhances the efficiency of light harvest­
broadening the absorption spectra [7–9]. Various auxiliary groups have ing. The effect was apparent in the experiment of co-deposition with
been reported in the literature, such as benzothiadiazole [10–13], ben­ deoxycholic acid (DCA), while the photocurrent and PCEs were
zotriazole [8,14,15], diketopyrrolopyrrole [16,17], fluoro-substitent decreased due to the reduction of dye loading amount. The sensitizer of
[18,19], isoindgo [20,21], quinoxaline [9,22–26], and triazine [27] to RY3 displayed a high conversion efficiency of 6.30% with short-circuit
tailor the absorption energy levels. Another effective strategy that has current (Jsc) 13.10 mA cm 2, open-circuit voltage (Voc) 697 mV and fill
been reported in the literature is to add Liþ trapping groups to block the factor (ff) 0.69 without adding DCA at AM1.5 solar condition.
surface of TiO2, such as using 12-crown-4 ether, long alkyl/alkoxy chain, A panchromatic effect is investigated by the co-sensitization of N719
or [2.2]paracyclophane moiety [28–30]. It can enhance the conduction and RY3. The absorption intensity increased due to an improvement of
band energy level of TiO2 and reduce the degree of charge recombina­ dye-coverage on TiO2 surface (Fig. 5(c and d)) [9,32,33,44,45]. The best
tion. As a result, the Voc value of DSSC is increased. The improvement of performance of the device with co-sensitization of N719 and RY3 (8hr
Jsc by co-sensitization can be realized by enhancing the panchromatic and 4hr) displayed a short-circuit current (Jsc) 17.3 mA cm 2, an
effect, broadening the absorption spectrum, and matching energy levels open-circuit voltage (Voc) 727 mV, and a fill factor (ff) 0.66, corre­
simultaneously [31–33]. sponding to an overall conversion efficiency 8.55% under AM1.5 solar
In our previous report, we have shown an effective use of spiro condition.
[fluorene-9,90 -phenanthren-100 -one] as an acceptor moiety in a D-A-D It would be beneficial if indoor light can be harvested to provide the

Table 1
Photochemical and electrochemical parameters of the sensitizers.
dye λmaxa (nm)/(ε (M 1cm 1
)) λmax (TiO2) Eoxb (V) E0-0 (eV) Eredc (V) Jsc (mA∙cm 2
) Voc (mV) FF ηd (%) Ref.

RY1 440(43400) 402 0.80 2.44 1.64 12.30 678 0.68 5.67 this
RY2 397(41700) 367 0.80 2.66 1.86 9.01 697 0.66 4.14 work
RY3 450(45000) 426 0.57 2.34 1.77 13.10 697 0.69 6.30
RY4 410(37700) 363 0.57 2.52 1.95 10.15 698 0.67 4.74
RY5 444(45100) 403 0.59 2.41 1.82 12.90 698 0.66 5.94
RY6 402(28300) 359 0.59 2.61 2.02 9.18 703 0.69 4.24
N719 535(16300) – – – – 16.07 701 0.67 7.54
MK-160 370(44000) – 0.71 1.93 1.22 4.46 530 0.62 1.45 15a
MK-162 370(53000) – 0.65 1.78 1.13 4.92 520 0.59 1.52 15a
SSD1 392(47000) 422 1.10 2.75 1.65 8.9 630 0.67 3.75 15b
SD8 450(33000) 456 1.13 2.38 1.25 9.12 630 0.72 4.12 15c
FHD5 485(21000) 547 1.11 2.03 0.92 8.16 650 0.69 3.73 15d
FHD6 443(15000) 482 1.11 2.26 1.15 7.47 700 0.71 3.69 15d
2 452(17100) 461 0.63 2.41 1.78 12.8 705 0.66 5.96 15e
4 492(16800) 476 0.36 2.38 2.02 12.2 697 0.72 6.12 15e
JK-88 410(28000) – 1.12 2.47 1.36 9.28 760 0.75 5.28 15f
JK-89 428(34000) – 1.13 2.43 1.31 13.02 750 0.70 6.83 15f
a
Absorption are in CH2Cl2.
b
Oxidation potential in THF (10 3 M) containing 0.1 M (n-C4H9)4NPF6 with a scan rate 50 mV s 1 (vs. NHE).
c
Ered calculated by Eox E0-0.
d
Best performance of DSSCs measured in a 0.28 cm2 working area on a FTO (8Ω/square) substrate. Electrolyte: LiI (0.5 M), I2 (0.05 M), and TBP (0.5 M) in MeCN.
The TiO2 electrode was dipped in methene chloride of sensitizer with 3 � 10 4 M for 12hr.

2
R.-Y. Huang et al. Journal of Power Sources 458 (2020) 228063

Fig. 1. Chemical structure of RY-sensitizers.

Scheme 1. Synthetic conditions of organic dye sensitizers: (i) NBS, DMF, 0 � C; (ii) Pd(dppf)Cl2⋅CH2Cl2, AcOK, bis(pinacolato)diboron, dioxane/toluene, 90 � C; (iii)
Pd(PPh3)4, 2,7-dibromo-100 H-spiro[fluorene-9,90 -phenanthren]-100 -one, THF/toluene/K2CO3(aq), 90 � C; (iv) (5-(1,3-dioxolan-2-yl)thiophen-2-yl)tributylstannane or
(4-(1,3-dioxolan-2-yl)phenyl)tributylstannane, Pd(PPh3)2Cl2, DMF, 90 � C, and then 3 M HCl/THF, rt; (v) Cyanoacetic acid, piperidine, CHCl3, 60 � C.

electric power of certain indoor applications, which would be unaffected The indoor light is commonly used to brighten our offices, factories, first
by weather or the intermittent restriction of sunlight. In this regard, aid rooms, hospitals, stores, laboratories, libraries, production halls, and
DSSCs are recognized to be a more efficient device than silicon-based showrooms, etc., mostly generated by luminescent devices such as
cells, because they performed better in the longer wavelength region. fluorescent LED lamps, ranging typically in 300–1500 lux.

3
R.-Y. Huang et al. Journal of Power Sources 458 (2020) 228063

Encouragingly, many effective organic sensitizers have been developed dyes exhibited two broad, high-intensity absorption bands in the range
by optimizing the spectral matching with indoor light sources (e.g., of 330–450 nm. The short wavelength absorptions (330–358 nm;
TL84 and CWF). Gra €tzel and Hagfeldt’s group has demonstrated that 2.41–4.69 � 104 M 1cm 1) are attributed to localized π–π* transitions,
using co-sensitized XY1 and D35 dyes with a copper-based electrolyte, whereas the long wavelength ones (396–450 nm; 2.83–4.51 � 104
the PCE can be promoted to a remarkably high level of 28.9% under M 1cm 1) are assigned to the characteristic intramolecular charge
1000 lux of fluorescent light [46]. Wei and Yeh’s group reported that an transfer (ICT) transitions between the donor group and acceptor cya­
acetylene-based anthracene sensitizer TY6 yielded 28.56% PCE under noacrylic acid moiety. Compared with phenylene π-spacers RY2, RY4
6000 lux T5 fluorescent light [47]. Lin and Wei used and RY6, the ICT bands of RY1, RY3 and RY5 are red-shifted by
pyrazine-incorporated panchromatic sensitizer MD7 that reached PCE approximately 30–54 nm due to the presence of the thiophene group.
values of 18.85% and 27.17%, respectively, under 300 lux and 6000 lux The ICT absorptions of solid films on TiO2 surface exhibit broader band-
of T5 fluorescent light [48]. In our pervious report, a co-sensitization of shapes due to a stronger interaction between the sensitizers and TiO2
MM-3 and MM-6 achieved PCE values of 27.76%, 28.74% and 30.45% surface. The ICT bands of RY-sensitizers also exhibit a slight blue-shift
at 600 lux, 1000 lux and 2500 lux under TL84 fluorescent light, comparing to those in solutions, which may be ascribed either to the
respectively [26]. In this work, through co-sensitization of N719 and formation of H-aggregation or to the deprotonation of cyanoacrylic acid
RY3, we achieved PCEs values of 25.57% and 27.04% at 1000 lux and moiety [2,50]. The bulky auxiliary acceptor in RY-sensitizers may
2500 lux under TL84 fluorescent light, respectively. reduce the degree of intermolecular J-aggregation. Such an effect can be
verified by a DCA test, which will be discussed in the following para­
2. Results and discussion graphs. The zero-zero energy (energy gap) of all sensitizers can be
estimated by the offset of absorption band and found to be in the ranges
2.1. Synthesis of sensitizers of 2.34–2.67 eV.

The synthesis of sensitizers RY1~RY6 started from compound 1 (1a,


1b and 1c) (Scheme 1). Three types of arylamine (1) were treated with 2.3. Electrochemical properties
bromination reaction to produce the bromo-substituted compounds 2.
The dibromo-substitued auxiliary acceptor was constructed by a To examine the electrochemical-stability and the HOMO and LUMO
convenient process developed previously by us in high yield (3 steps, energies of RY-sensitizers, cyclic voltammetry experiments (CV) were
~72%) [34]. Compound 2 was then subjected to a Suzuki-Miyaura performed. All RY-sensitizers displayed nearly reversible waves on CV
coupling reaction to yield the boronic ester precursor 3. The three de­ (Fig. S19). The absence of alkoxy substituent on the donor moieties of
rivatives of 3 (3a, 3b and 3c) were coupled with 2,7-dibromo-100 H-spiro RY1 and RY2 reduces their HOMO levels, thus increases their oxidation
[fluorene-9,90 -phenanthren]-100 -one through a Suzuki-Miyaura reaction potentials. The HOMO energy levels of RY1 and RY2 are lower than
to yield the corresponding derivatives of 4 (4a, 4b and 4c) [49]. These those of RY3~RY6 in an amount of 0.11–0.13 eV. It leads to broader
compounds exist as mixtures of optical isomers, which were not HOMO–LUMO energy gaps in the range of 0.02–0.33 eV (Fig. 3). The
resolved. The structures of 4 were elongated by coupling with either introduction of thiophene into the π-spacer of sensitizers induces a
(5-(1,3-dioxolan-2-yl)thiophen-2-yl)tributylstannane or (4-(1,3-dioxo­ smaller HOMO–LUMO gap with respect to the corresponding phenylene
lan-2-yl)phenyl)tributylstannane through a Stille reaction to yield analogues. The smaller thiophene group reduces the steric hindrance
compounds 5 and 6 in 78–91% yield. Finally, all the aldehyde de­ between adjacent aryl rings, therefore increases the planarity of mo­
rivatives of 5 and 6 were subjected to Knoevenagel condensation re­ lecular structure and elevates the degree of π-resonance. This result is
actions to yield the target compounds. The details of all synthesis can be consistent with the red-shift of ICT bands in their absorption spectra, and
obtained in ESIy. can be verifies by the difference of dipole moments estimated by theo­
retical computations (Table S1). The narrower HOMO–LUMO energy
gaps of RY1, RY3 and RY5, compared with those of RY2, RY4 and RY6
2.2. Photophysical properties are apparently induced by the presence of a thiophene unit. The LUMO
levels of the dyes were determined by measuring the Eox and the band
The ultraviolet–visible absorption spectra of all sensitizers, measured gaps that were obtained by the zero–zero energy at the offset of the
in 3.0 � 10 5 M CH2Cl2 solution, were shown in Fig. 2. The RY-series absorption spectra (Fig. 2(a) and Table 1). The estimated HOMO and

Fig. 2. (a) Absorption spectra of dyes RY1~RY6 in CH2Cl2 and N719 in MeCN/tBuOH(1/1) (3.0 � 10 4
M); (b) Absorption spectra of dyes RY1~RY6 on TiO2
thin film.

4
R.-Y. Huang et al. Journal of Power Sources 458 (2020) 228063

trend is also consistent with the ICT transition of experimental data and
the dipole moments of computation result. Charge distribution from
HOMO to LUMO can be clearly verified by the calculated dipole mo­
ments, i.e. RY3 (14.00 D) � RY5 (13.99 D) > RY1 (10.66 D) and RY4
(11.30 D) > RY6 (10.60 D) > RY2 (9.02 D). Therefore, the ICT bands in
the dyes RY1, RY3 and RY5 exhibit a bathochromic shift comparing
with those of the dyes RY2, RY4 and RY6. The physical and theoretical
parameters and the ICT transition dipole moments of the RY-sensitizers
are summarized in Table S1.

2.5. Photovoltaic performance of DSSCs

The designed RY-photosensitizers were used in the DSSCs devices


fabrication, while they were absorbed on the np-TiO2 surface according
to the standard procedure. The electrolyte used in the devices was a
solution contained I /I3 that was made of LiI (0.5 M) and I2 (0.05 M).
DCA was used as a co-absorbent in the experiments, and N719 was
served as the control reference sensitizer. Co-sensitization system of
Fig. 3. HOMO-LUMO energy levels of RY-sensitizers.
N719 and RY3 were used in the experiments with various deposition
time. The incident photocurrent conversion efficiency (IPCEs) and
LUMO levels are illustrated in Fig. 3. The variation of energy levels was photocurrent–voltage (J–V) curve of all dyes were measured either
analysed by using theoretical computations as discussed in the following under AM 1.5 solar light (100 mW cm 2) or under indoor light using a
section. TL84 (4100 K, European shop fluorescent), or a CWF (4150 K, cool white
fluorescent, shop lighting) lamp. The short-circuit current (Jsc), open-
2.4. Computational analysis circuit voltage (Voc), fill factor (FF), and solar-to-electrical photocur­
rent density (η) are summarized in Table 1.
The optimized structures of RY-sensitizers were obtained by time- The J–V curve and IPCE of all devices are plotted in Fig. 5. Compared
dependent density functional theory (TD-DFT) calculations using a with RY1 and RY2, a slight improvement of 4–20 mV in Voc was
B3LYP/6-31G(d) basis set in Gaussian 16W. We monitored two dihedral observed for dyes RY3~RY6 that contain long-chain alkoxyl sub­
angles in the sensitizer structures, i.e., the angle of donor/auxiliary stituents. Previous report showed that the inclusion of a thiophene
acceptor and the angle of auxiliary acceptor/π-spacer. The dihedral moiety usually exhibited a good light harvesting effect, yet accompanied
angles between the donor and the auxiliary acceptor were quite close to by a lower Voc, because of the strong interactions between iodine and
each other at around 35.94� –37.66� . Compounds containing thiophene sulphur atoms, which form sensitizer-I2 and sensitizer-I-3 complexes
groups, i.e. RY1, RY3 and RY5, displayed smaller dihedral angles be­ [41–43]. In the present cases this phenomenon was observed only on
tween the auxiliary acceptor and the π-spacer comparing with others RY1 and RY2, while for RY3~RY6 they showed similar high Voc value
that possessing more co-planar thiophene groups (Fig. S20). The elec­ in the range of 697–703 mV. The reason could be attributed to both the
tron density distributions in the frontier orbitals of RY-sensitizers are long-chain alkoxyl substituents and rigid spiro-structure of the latter,
shown in Fig. 4 and Fig. S20. The electron density in the HOMOs is both could reduce charge recombination and lower dark current.
mainly localized around the arylamine donor moieties, while in the Sensitizer RY3 exhibited a high Voc value of 697 mV and a high Jsc value
LUMOs around the cyanoacrylic acid acceptor moieties. Upon photo- of 13.10 mA cm 2 without using DCA co-absorbent, due to the formation
excitation, an electron is promoted from the HOMO to the LUMO and of better film morphology. The device performance of RY3 displayed a
therefore induces a considerable amount of electron density redistri­ short-circuit current (Jsc) 13.10 mA cm 2, an open-circuit voltage (Voc)
bution. The calculated HOMO/LUMO levels were 5.31/ 2.95 eV for 697 mV, a fill factor (ff) 0.69, corresponding to an overall conversion
RY1, 5.28/ 2.90 eV for RY2, 4.96/ 2.89 eV for RY3, 4.93/ 2.85 efficiency 6.30% without DCA treatment under AM1.5 solar condition.
eV for RY4, 4.95/ 2.65 eV for RY5, and ¡4.90/ 2.54 eV for RY6. Good dye morphology on the TiO2 surface plays a crucial factor for
These results are consistent with the observed λmax values in the ab­ reducing the rate of charge recombination [51,52]. Higher dye
sorption spectra, i.e., 577 nm, 568 nm, 661 nm, 632 nm, 591 nm, and morphology quality leads to a better coverage of TiO2 surface, and
571 nm for RY1~RY6, respectively. The trend of absorption λmax prevent contact between Liþ/I3 ions and TiO2. The hydrophobic long
showed RY3 > RY5 > RY1 and RY4 > RY6 > RY2 of the same π-bridge. chain usually plays the role of surface blocking that increases the energy
This trend also consistent with Jsc value (Fig. 5(a)) and dyes loading gap between the Fermi levels of TiO2 and electrolyte, therefore enhances
amount (Table S4). The dyes containing a thiophene π-linkage exhibited the value of Voc [53–55]. For the purpose of improving dye alignment,
red-shifts comparing with those containing a phenylene π-linkage. The DCA was usually added to help the dye molecules standing vertically on

Fig. 4. The frontier HOMO and LUMO orbitals of RY-sensitizers estimated by time-dependent DFT/B3LYP (6-31G* basis set).

5
R.-Y. Huang et al. Journal of Power Sources 458 (2020) 228063

Fig. 5. (a) J-V plots and (b) IPCE of DSSCs devices of MM-1~6 without DCA co-absorbent. (c) J-V plots and (d) IPCE of DSSCs devices of RY3 by adding DCA co-
absorbent and cosensitization system with N719.

the TiO2 surface. better understanding of the photovoltaic properties of the DSSCs de­
Upon adding DCA (10 mM) as a co-absorbent, the Voc of DSSCs made vices. Nyquist and Bode plots were drawn based on a forward bias under
with RY-sensitizers increased slightly in 0.5%–4.0%, but the Jsc a dark condition (Fig. 6). In the Nyquist plots the major semicircles that
decreased in 3.89%–12.66% due to the reduction of loading amount. As observed for each sensitizer are associated with charge recombination
a result the overall power conversion efficiency decreased in 2.28%– resistance (Rrec) between the TiO2/electrolyte/dye interfaces. A larger
10.77% (Table 2). It indicated that the RY-sensitizers are capable of semicircle relates to Rrec at low frequency, indicating a lower tendency
forming good vertical alignment by themselves without DCA. Never­ of charge recombination and thus a longer device electronic lifetime.
theless, the RY-sensitizers exhibited lower Jsc value of 9.01–13.10 mA The radii of the semicircles increase in the order of RY1 (115 Ω) <
cm 2 due to their narrow absorption spectra. RY2 (134 Ω) � RY3 (138 Ω) � RY4 (135 Ω) � RY5 (135 Ω) < RY6 (175
The absorption wavelengths of N719 and RY3 showed comple­ Ω) < N719 (190 Ω) < N719 þ RY3 (236 Ω). This trend is similar to the
mentary coverage of the whole visible spectral range. The effect of co- Voc of the dyes, i.e., RY1 (678 mV) < RY2 (697 mV) � RY3 (697 mV) �
deposition of N719 and RY3 in different ratios was conducted for RY4 (698 mV) � RY5 (698 mV) < RY6 (703 mV) � N719 (701 mV) <
advanced improvements. Sensitizer N719 was deposited on TiO2 first, N719 þ RY3 (727 mV). The high performance of co-deposition of N719
followed by RY3 with different deposition time (N719/RY3: 4hr/8hr; and RY3 is confirmed by a larger arc of the major semicircle (236 Ω) in
6hr/6hr; 8hr/4hr). Both Jsc and Voc were improved and the power Nyquist plot and a longer electron lifetime (79.57 ms). The optimal
conversion efficiency increased by 0.7%–13.42%. The role of sensitizer device power conversion efficiency of 8.55% under one sun AM1.5
RY3 for improving the performance of N719 was not only broadening condition is higher than that using standard N719 (7.54%) dye under the
the spectral converage but also helping the dye pack denser on TO2 same condition.
surface. The most appropriate deposition times for N719 and RY3 were The electron diffusion lifetime (τe) of the device was related to the
8 h and 4 h, respectively. The device exhibited higher Jsc and Voc values value of 1/2πf. A shift to lower frequency (f) in a Bode plot corresponds
than the devices made with either N719 or RY3 alone, i.e., a Jsc value of to a longer electron lifetime. The electron lifetime of the dyes increases
17.09 (�0.21) mA∙cm 2 with broad IPCE and a high Voc of 719 (�8) mV in the order of RY1 (43.60 ms) < RY2 (56.23 ms) � RY3 (60.28 ms) �
(in Table 2 and Fig. 5(c and d)). The best performance of a co-sensitized RY4 (55.45 ms) � RY5 (67.72 ms) < RY6 (71.05 ms) � N719 (68.89
device displayed a short-circuit current (Jsc) 17.3 mA cm 2, an open- ms) < N719 þ RY3 (79.57 ms), where a higher τe indicates a lower
circuit voltage (Voc) 727 mV, a fill factor (ff) 0.68, corresponding to an charge recombination rate (Fig. 6(b) & (d)). The larger charge recom­
overall conversion efficiency 8.55% under AM1.5 solar condition. bination resistance and the longer electron diffusion lifetime may be
Electrochemical impedance spectroscopy (EIS) was examined for a attributed to good compact packing of the sensitizers on TiO2 surface.

6
R.-Y. Huang et al. Journal of Power Sources 458 (2020) 228063

Table 2 observed upon the addition of Liþ ion (Fig. S21). The reduction of both
Photovoltaic parameters of devices made RY-sensitizers with/without DCA and the π-π* and ICT transitions reveals the presence of a significant inter­
co-sensitization system. action between the carbonyl moiety and the Liþ ion. The amount of
Dyea DCA Jsc Voc FF ηb (%) intensity change was proportional to the concentration of Liþ ion. Flu­
(mM) (mA∙cm 2
) (mV)
Averagec best
orene that does not have carbonyl substituent did not show a similar
spectral change in the presence of Liþ ion. These results support a Liþ
RY1 0 11.97 � 665 0.66 5.25 � 5.67
trapping effect by the carbonyl group, and may be responsible for the
0.33 � 13 � 0.42
0.02 high Voc of the RY-sensitizers.
10 10.91 � 692 0.68 5.13 � 5.37
0.16 � 12 � 0.24 2.6. DSSCs performance under indoor light condition
0.01
RY2 0 8.78 � 0.23 682 0.64 3.83 � 4.14
� 15 � 0.31
We have demonstrated that the RY-sensitizers and co-sensitization
0.02 showed promising conversion efficiency in DSSCs under one-sun illu­
10 7.84 � 0.25 686 0.66 3.55 � 3.82 mination (AM1.5). We continued to explore the possibility of using these
�9 � 0.27 sensitizers under indoor light illumination in the intensity range of
0.02
1000–2500 lux by using two kinds of fluorescent lamps (TL84, 4100 K;
RY3 0 12.92 � 686 0.67 5.94 � 6.30
0.18 � 11 � 0.36 CWF, 4150 K). These RY-sensitizers display a high degree of spectral
0.02 overlap with the indoor light emission wavelengths (Fig. S22).
10 11.69 � 679 0.65 5.30 � 5.57 The indoor-light cell efficiency and parameters at 1000 and 2500 lux
0.31 � 14 0.27

(TL84 and CWF) are summarized in Table 3 and Table S2. Compared
0.02
RY4 0 9.87 � 0.28 690 0.64 4.15 � 4.74
with CWF condition, all devices are shown better results under TL84
�8 � 0.59 indoor light source illumination (Fig. S23 and Table S2). The devices
0.03 made with sensitizers RY1, RY3 and RY5 showed cell efficiencies 18%–
10 8.62 � 0.25 715 0.68 4.19 � 4.49 21% under stronger illumination of 2500 lux (TL84), higher than those
�9 0.30
made with RY2, RY4 and RY6 containing phenylene π-spacer. The latter

0.02
RY5 0 12.71 � 688 0.65 5.68 � 5.94 dyes exhibited narrower absorption bandwidths, which is a situation
0.19 � 10 � 0.26 similar to that under AM 1.5. Among all sensitizer, RY3 showed the
0.01 highest performance, i.e., 18.15(�1.00)% under 1000 lux and 20.83
10 11.24 � 686 0.66 5.09 � 5.47
(�0.84)% under 2500 lux. Although the RY-sensitizers exhibited lower
0.27 � 13 � 0.38
0.02
performance than N719 under similar conditions, e.g., 20.88(�1.29)%
RY6 0 8.99 � 0.19 688 0.66 4.16 � 4.45 under 1000 lux and 24.91(�1.12)% under 2500 lux, nevertheless RY-
� 15 � 0.29 sensitizers can be used as good co-absorbent in co-sensitization systems.
0.03 Our current results are highly competitive with the co-sensitization
10 8.64 � 0.17 698 0.66 3.98 � 4.24
system of XY-1 and D35 using Cu(II/I)(tmby) as the redox mediator
� 11 � 0.26
0.02 (25.5% at 200 lux; 28.9% at 1000 lux), which has been developed by
N719 (12hr)c – 15.92 � 691 0.65 7.15 � 7.54 Gra€tzel and Hagfeldt [46]. We have also reported in a previous work, by
0.15 � 10 � 0.39 the co-sensitization of MM-3 and MM-6, remarkable performances of
0.02 27.76% (600 lux), 28.84% (1000 lux) and 30.45% (2500 lux) could be
N719þRY3d – 16.36 � 704 0.65 7.48 � 7.92
(4hrþ8hr) 0.22 �9 � 0.44
achieved. The best PCE was found to be higher than 30% (30.45% at
0.02 2500 lux) under a TL84 fluorescent indoor light source [26].
N719þRY3d – 16.09 � 689 0.65 7.20 � 7.60 Recent reports indicated that the commonly used reference sensitizer
(6hrþ6hr) 0.26 � 16 � 0.40 N719 shows the efficiency of dye-sensitized solar cells in 17.25% (600
0.01
lux) ~ 27.64% (6000 lux) under indoor light [48,56,57]. In this work,
N719þRY3d – 17.09 � 719 0.66 8.11 � 8.55
(8hrþ4hr) 0.21 �8 � 0.44 we also compared the efficiency of DSSCs made with the following three
0.02 different compositions under illumination of indoor light sources CWF
a and TL84: (1) RY3 only; (2) N719 only; and (3) N719 co-deposited with
Concentration of dye is 3 � 10 4 M in CH2Cl2.b Performance of DSSCs
measured in a 0.28 cm2 working area on an FTO (7 Ω/square) substrate under RY3. The results are shown in Fig. S23 and Table S2. The
AM 1.5 condition.c The average was estimated based on 10 devices.c The TiO2 co-sensitization of N719 and RY3 (deposited time 8hr and 4hr) per­
electrode was dipped in methene chloride of sensitizer with 3 � 10 4 M for formed the best after fine-tune the mole ratios to achieve an optimal
12hr.fThe TiO2 electrode was dipped in tBuOH/MeCN of N719 (3 � 10 4 M), morphology on TiO2 surface. The best performance PCE was found to be
rinsed with ACN, then dipped in methene chloride of RY3 (3 � 10 4 M). 26.19 (�0.85)% at 2500 lux under a TL84 fluorescent indoor light
source. The high performance of co-sensitization under indoor light is
The presence of long alkoxy chains in the structure of dyes consistent with that under sunlight.
RY3~RY6 improves the film quality on TiO2 surface, thus results in A stability test on the devices were performed at 25 � C under TL84
higher values of Rrec and τe than RY1. The highly ordered shapes of indoor light and AM1.5 in an ambient environment. The device RY3,
RY3~RY6 induce a more compact molecular packing, therefore block N719 and co-sensitization of N719 and RY3 retained 92.9%, 88.06%
the Liþ or I3 ions in electrolyte effectively from a direct contact with the and 90.37% efficiency, respectively, under 2500 lux after 192 h. The
electrode and therefore reduce the dark current. It is worth noting the plots of PCE decay curves are shown in Fig. S24, and the list of param­
case of RY2, which exhibits high values of Rrec and τe even without the eters is provided in Table S3 [58]. The stability test on device
long-chain substituents. This result indicated that the auxiliary spiro co-sensitized with N719 and RY3 retained 84.5% after 336 h at 2500 lux
[Fluorene-9,90 -phenanthren-100 -one] moiety may have both the effect of under TL84 illumination. These results indicated that our devices
reducing the intermolecular aggregation and the effect of trapping the possess remarkable stability under ambient condition.
Liþ or I3 ions to retard the charge recombination. The Liþ ion trapping
effect can be evidenced by examining the UV/vis spectra of spiro-diBr 3. Conclusion
and 4b, while a significant decrease of absorption intensity was
We demonstrated the successful application of a series of D-A-π-A

7
R.-Y. Huang et al. Journal of Power Sources 458 (2020) 228063

Fig. 6. Impedance spectra of RY-series dyes in CH2Cl2 (a,c) Nyquist plots (b,d) Bode phase plots at 0.73 V bias in the dark. The electron diffusion lifetime τe ¼
1/2πf.

Table 3
Photovoltaic parameters of the RY-series dyes measured using TL84 indoor light.
Sensitizer Irradiance (mW/cm2) Illuminance (lux) Jsc (μA∙cm 2) Voc (mV) FF(%) PCE (%)a

RY1 0.185 1000 78.3(�4.3) 555(�6.5) 70.88(�1.02) 16.64(�1.38)


0.462 2500 195.5(�6.2) 590(�3.4) 71.94(�0.85) 17.94(�0.91)
RY2 0.185 1000 45.3(�2.5) 572(�5.1) 58.77(�0.58) 8.23(�0.62)
0.462 2500 110(�3.7) 610(�3.5) 61.76(�0.68) 9.02(�0.40)
RY3 0.185 1000 85.8(�3.5) 590(�3.6) 66.33(�0.53) 18.15(�1.00)
0.462 2500 222(�5.1) 630(�2.9) 68.83(�0.82) 20.83(�0.84)
RY4 0.185 1000 55.2(�2.5) 573(�4.9) 58.72(�0.39) 10.03(�0.62)
0.462 2500 146(�3.1) 632(�4.1) 59.28(�0.46) 11.76(�0.50)
RY5 0.185 1000 80.7(�3.9) 595(�2.8) 65.27(�0.81) 16.94(�1.12)
0.462 2500 201(�4.7) 650(�3.6) 69.83(�0.63) 19.74(�0.76)
RY6 0.185 1000 51.9(�5.7) 588(�3.7) 56.78(�1.22) 9.36(�1.32)
0.462 2500 137(�6.9) 650(�3.5) 60.94(�0.81) 11.74(�0.83)
N719 (12hr) 0.185 1000 110(�4.2) 550(�4.2) 64.62(�1.51) 21.13(�1.49)
0.462 2500 275(�6.5) 621(�3.6) 67.25(�1.53) 24.85(�1.32)
b
N719þRY3 (8hrþ4hr) 0.185 1000 120(�2.6) 569(�3.7) 66.26(�1.13) 24.45(�1.12)
0.462 2500 278(�4.4) 628(�4.3) 69.32(�0.65) 26.19(�0.85)
a
Voc: open-circuit voltage; Jsc: short-circuit current; FF: fill factor; PCE: power-conversion efficiency. The data are based on three measurements. Experiments were
conducted by using TiO2 photoelectrodes, with approximately 12 mm thickness and 0.28 cm2 working area on FTO (7Ω/square) substrate.
b
The TiO2 electrode was dipped in tBuOH/MeCN of N719 (3 � 10 4 M) for 6 h, rinsed with ACN, then dipped in methene chloride of RY3 (3 � 10 4 M) for 6hr.

sensitizers incorporating spiro[fluorene-9,90 -phenanthren]-100 -one aggregation and trapping the Liþ or I3 ions to retard charge recombi­
moiety as an auxiliary acceptor (RY1~RY6) in high-performance DSSCs nation. All films made of RY-sensitizers show good morphology. Among
under one sun and indoor light illumination (CWF and TL84) conditions. them, the ones containing a thiophene group, i.e., RY1, RY3 and RY5,
The auxiliary moiety showed the effect of reducing intermolecular perform better than those containing a corresponding phenyl group, i.e.,

8
R.-Y. Huang et al. Journal of Power Sources 458 (2020) 228063

RY2, RY4 and RY6. The smaller steric hindrance of a thiophene group, W solar simulator (Oriel Sol3A Class AAA Solar Simulator 9043A,
comparing to a phenyl group, induces a more planar molecular geom­ Newport), and passed through an AM 1.5 filter (Oriel 74110). The light
etry and a more regular alignment on the surface of TiO2. The thiophene intensity was further calibrated using an Oriel reference solar cell (Oriel
group also reduces the HOMO-LUMO energy gap, thus broaden the 91150) and adjusted to 1.0 sun. The monochromatic quantum efficiency
absorption wavelength range. The dye RY3 performed remarkably well was recorded using a monochromator (Oriel 74100) under a short cir­
without using DCA in DSSCs, which showed short-circuit current (Jsc) cuit condition. Electrochemical impedance spectra of DSSCs were
13.10 mA cm 2, open-circuit voltage (Voc) 697 mV, and fill factor (ff) recorded by an Impedance/CIMPS/IVMS analyser (Zahner Ennium).
0.69, corresponding to an overall conversion efficiency of 6.30% under DSSCs device performances under indoor light: The stabilized indoor
AM1.5 solar condition. We also found that a co-deposition of N719 and light sources system (model CMS-PV101, designed by Industrial Tech­
RY3 together can improve not only the dye-coverage on TiO2 surface, nology Research Institute, Taiwan) consisted of TL84 (4100 K, European
but also the wavelength range of absorption. Therefore, the RY3 can be a shop fluorescent) and CWF (4150 K, cool white fluorescent, shop
good candidate as co-absorbent in DSSCs. The best performance of the lighting), which meet the standards of SEMI PV80-0218 and CIE. Both of
device co-deposited with N719 and RY3 displayed short-circuit current light sources are intensity adjustable in a range of 0–2500 lux, with non-
(Jsc) 17.3 mA cm 2, open-circuit voltage (Voc) 727 mV, and fill factor (ff) uniformity less than 2% (within the area of 20 cm � 20 cm) and tem­
0.68, corresponding to an overall conversion efficiency of 8.55% under poral instability less than 2%. A calibrated spectroradiometer (ISM-Lux,
AM1.5 solar condition. The most impressive result was obtained when Isuzu Optics, Japan) was embedded at underlying system. The acquired
N719 and RY3 were used on indoor applications. The devices co- illuminance value was attained by altering LED lamp lifting platform to
sensitized with N719 and RY3 achieved the highest PCE of 27.04% at a moderate position and incessantly confirmed by spectroradiometer
2500 lux under a TL84 fluorescent lamp. It displays high stability that until it reached firmly stable condition. After then, the DSSC device was
retained 84.5% after 336 h at 2500 lux under TL84 illumination. put at the upper site of the spectroradiometer and the J-V curves were
measured via a computer-controlled digital source meter (Keithley
4. Experimental section 2401, USA) under various dim light illumination, followed the re­
quirements of SEMI PV57-1214.
Characterization and reagents: All reactions and manipulations were Quantum Chemistry Computations: Geometry optimization of the dyes
performed under a nitrogen atmosphere, and solvents were freshly were accomplished by Gaussian G16W program (B3LYP/6-31G* hybrid
distilled according to standard procedures. The 1H and 13C NMR were functional). For excited states, the time-dependent density functional
recorded on a Bruker AVIII HD 400. The 1H and 13C NMR chemical shifts theory (TDDFT) and the B3LYP functional were used. The frontier
were reported on δ scale downfield from Me4Si. The coupling constants orbital plots of the highest occupied molecular orbital (HOMO) and the
(J) were given in hertz. Absorption spectra were recorded on a Shimadzu lowest unoccupied molecular orbital (LUMO) were drawn by using
UV-1800 spectrophotometer. Redox potentials were measured using GaussView 06.
cyclic voltammetry on a Zahner Ennium analyser. All measurements
were performed in CH2Cl2 solution containing 0.1 M tetrabutylammo­ Declaration of competing interest
nium hexaflourophosphate as supporting electrolyte under ambient
conditions after purging the solution with N2 for 10 min. The conven­ The authors declare that they have no known competing financial
tional three-electrode configuration was employed that consists of a interests or personal relationships that could have appeared to influence
glassy carbon working electrode, a platinum counter electrode, and an the work reported in this paper.
Ag/Agþ reference electrode calibrated with ferrocene/ferrocenium (Fc/
Fcþ) as an internal reference. Mass spectra were recorded on a Jeol JMS- CRediT authorship contribution statement
700 double-focusing mass spectrometer. Analytical thin layer chroma­
tography (TLC) was performed on Silica gel 60 F 254 Merck. Column Rui-Yu Huang: Data curation, Formal analysis. Wen-Hsuan Tsai:
chromatography was performed using the silica gel from Merck (Kie­ Software, Validation. Jun-Jie Wen: Software, Validation. Yuan Jay
selgel Si 60; 40–63 μm). Solvent THF was distilled with sodium benzo­ Chang: Supervision, Funding acquisition, Writing - original draft.
phenone ketyl. Toluene and methylene chloride were distilled with Tahsin J. Chow: Funding acquisition, Writing - review & editing.
CaH2. All other solvents and reagents were reagent grade, purchased
from Acros, Alfa, Merck, Lancaster, TCI, Sigma-Aldrich, and Showa, Acknowledgments
separately, and used without further purification.
Fabrication and Characterization of DSSCs: The FTO conducting glass This work was financial supported by the Ministry of Science and
(2.2 mm thick, fluorine doped tin oxide over-layer, transmission >90% Technology, Taiwan (MOST 107-2113-M-029-010 and 107-2113-M-
in the visible, sheet resistance 7 Ω/square), titania-oxide pastes of Ti- 029-011). Tunghai University is gratefully acknowledged.
Nanoxide T/SP (adsorption layer) and Ti-Nanoxide R/SP (scattering
layer) were purchased from Solaronix. A thin film of TiO2 (adsorption Appendix A. Supplementary data
layer ~12 μm; scattering layer: ~6 μm) was coated on a 0.28 cm2 FTO
glass substrate. It was immersed in a methylene chloride solution con­ Supplementary data to this article can be found online at https://doi.
taining 3 � 10 4 M dye sensitizers for 12 h, then rinsed with anhydrous org/10.1016/j.jpowsour.2020.228063.
acetonitrile and dried. The other piece of FTO with sputtering 100 nm
thick Pt was used as a counter electrode. The active area was controlled References
at a dimension of 0.28 cm2 by adhering 60 μm thick polyester tape on
[1] M.K. Nazeeruddin, A. Kay, I. Rodicio, R. Humphry-Baker, E. Mϋeller, P. Liska,
the Pt electrode. The photocathode was placed on top of the counter
N. Vlachopoulos, M. Gr€ aetzel, J. Am. Chem. Soc. 115 (1993) 6382–6390.
electrode and was tightly clipped together to form a cell. Electrolyte was [2] M.K. Nazeeruddin, P. P�echy, T. Renouard, S.M. Zakeeruddin, R. Humphry-Baker,
then injected into the seam between two electrodes. An acetonitrile P. Comte, P. Liska, L. Cevey, E. Costa, V. Shklover, L. Spiccia, G.B. Deacon, C.
solution containing LiI (0.5 M), I2 (0.05 M) and 4-tert-butylpyridine A. Bignozzi, M. Gr€
atzel, J. Am. Chem. Soc. 123 (2001) 1613–1624.
[3] G. Maerker, F.H. Case, J. Am. Chem. Soc. 80 (1958) 2745–2748.
(TBP) (0.5 M) was used as the electrolyte. Devices made with com­ [4] A. Yella, H.-W. Lee, H.N. Tsao, C. Yi, A.K. Chandiran, M.K. Nazeeruddin, E.W.-
mercial dye N719 under the same condition (3 � 10 4 M, Solaronix S.A., G. Diau, C.-Y. Yeh, S.M. Zakeeruddin, M. Gr€ atzel, Science 334 (2011) 629.
Switzerland) were used as references. The cell parameters were obtained [5] Z. Yao, M. Zhang, H. Wu, L. Yang, R. Li, P. Wang, J. Am. Chem. Soc. 137 (2015)
3799–3802.
under incident light that had an intensity of 100 mW cm 2 measured
using a thermopile probe (Oriel 71964), which was generated by a 300

9
R.-Y. Huang et al. Journal of Power Sources 458 (2020) 228063

[6] K. Kakiage, Y. Aoyama, T. Yano, K. Oya, J.-i. Fujisawa, M. Hanaya, Chem. [32] J.-J. Cid, J.-H. Yum, S.-R. Jang, M.K. Nazeeruddin, E. Martínez-Ferrero,
Commun. 51 (2015) 15894–15897. E. Palomares, J. Ko, M. Gr€ atzel, T. Torres, Angew. Chem. Int. Ed. 46 (2007)
[7] Y. Wu, W. Zhu, Chem. Soc. Rev. 42 (2013) 2039–2058. 8358–8362.
[8] W. Li, Y. Wu, Q. Zhang, H. Tian, W. Zhu, ACS Appl. Mater. Interfaces 4 (2012) [33] J.-H. Yum, E. Baranoff, S. Wenger, M.K. Nazeeruddin, M. Gr€ atzel, Energy Environ.
1822–1830. Sci. 4 (2011) 842–857.
[9] K. Pei, Y. Wu, H. Li, Z. Geng, H. Tian, W.-H. Zhu, ACS Appl. Mater. Interfaces 7 [34] Y.-C. Chen, S.-K. Huang, S.-S. Li, Y.-Y. Tsai, C.-P. Chen, C.-W. Chen, Y.J. Chang,
(2015) 5296–5304. ChemSusChem 11 (2018) 3225–3233.
[10] W. Zhu, Y. Wu, S. Wang, W. Li, X. Li, J. Chen, Z.-s. Wang, H. Tian, Adv. Funct. [35] M. Can, H. Bilgili, K. Demirak, B. Gultekin, S. Koyuncu, M. Kus, C. Zafer, S. Demic,
Mater. 21 (2011) 756–763. S. Icli, Int. J. Hydrogen Energy 41 (2016) 21293–21299.
[11] S. Haid, M. Marszalek, A. Mishra, M. Wielopolski, J. Teuscher, J.-E. Moser, [36] D. Heredia, J. Natera, M. Gervaldo, L. Otero, F. Fungo, C.-Y. Lin, K.-T. Wong, Org.
R. Humphry-Baker, S.M. Zakeeruddin, M. Gr€ atzel, P. B€
auerle, Adv. Funct. Mater. Lett. 12 (2010) 12–15.
22 (2012) 1291–1302. [37] P. Shen, Y. Tang, S. Jiang, H. Chen, X. Zheng, X. Wang, B. Zhao, S. Tan, Org.
[12] Y. Wu, M. Marszalek, S.M. Zakeeruddin, Q. Zhang, H. Tian, M. Gr€ atzel, W. Zhu, Electron. 12 (2011) 125–135.
Energy Environ. Sci. 5 (2012) 8261–8272. [38] M. Xu, X. Hu, Y. Zhang, X. Bao, A. Pang, J.-K. Fang, ACS Appl. Energy Mater. 1
[13] D. Joly, M. Godfroy, L. Pellej� a, Y. Kervella, P. Maldivi, S. Narbey, F. Oswald, (2018) 2200–2207.
E. Palomares, R. Demadrille, J. Mater. Chem. 5 (2017) 6122–6130. [39] W.-S. Chao, K.-H. Liao, C.-T. Chen, W.-K. Huang, C.-M. Lan, E.W.-G. Diau, Chem.
[14] Y. Cui, Y. Wu, X. Lu, X. Zhang, G. Zhou, F.B. Miapeh, W. Zhu, Z.-S. Wang, Chem. Commun. 48 (2012) 4884–4886.
Mater. 23 (2011) 4394–4401. [40] N. Cho, H. Choi, D. Kim, K. Song, M.-s. Kang, S.O. Kang, J. Ko, Tetrahedron 65
[15] S. Chaurasia, W.-I. Hung, H.-H. Chou, J.T. Lin, Org. Lett. 16 (2014) 3052–3055. (2009) 6236–6243.
[16] S. Qu, C. Qin, A. Islam, Y. Wu, W. Zhu, J. Hua, H. Tian, L. Han, Chem. Commun. 48 [41] M. Miyashita, K. Sunahara, T. Nishikawa, Y. Uemura, N. Koumura, K. Hara,
(2012) 6972–6974. A. Mori, T. Abe, E. Suzuki, S. Mori, J. Am. Chem. Soc. 130 (2008) 17874–17881.
[17] P. Ganesan, A. Yella, T.W. Holcombe, P. Gao, R. Rajalingam, S.A. Al-Muhtaseb, [42] M. Zhang, J. Liu, Y. Wang, D. Zhou, P. Wang, Chem. Sci. 2 (2011) 1401–1406.
M. Gr€atzel, M.K. Nazeeruddin, ACS Sustain. Chem. Eng. 3 (2015) 2389–2396. [43] R.-Y. Huang, Y.-H. Chiu, Y.-H. Chang, K.-Y. Chen, P.-T. Huang, T.-H. Chiang, Y.
[18] D.-Y. Chen, Y.-Y. Hsu, H.-C. Hsu, B.-S. Chen, Y.-T. Lee, H. Fu, M.-W. Chung, S.- J. Chang, New J. Chem. 41 (2017) 8016–8025.
H. Liu, H.-C. Chen, Y. Chi, P.-T. Chou, Chem. Commun. 46 (2010) 5256–5258. [44] S.-Q. Fan, C. Kim, B. Fang, K.-X. Liao, G.-J. Yang, C.-J. Li, J.-J. Kim, J. Ko, J. Phys.
[19] Y.J. Chang, T.J. Chow, J. Mater. Chem. 21 (2011) 9523–9531. Chem. C 115 (2011) 7747–7754.
[20] W. Ying, F. Guo, J. Li, Q. Zhang, W. Wu, H. Tian, J. Hua, ACS Appl. Mater. [45] L. Yu, K. Fan, T. Duan, X. Chen, R. Li, T. Peng, ACS Sustain. Chem. Eng. 2 (2014)
Interfaces 4 (2012) 4215–4224. 718–725.
[21] S.-G. Li, K.-J. Jiang, J.-H. Huang, L.-M. Yang, Y.-L. Song, Chem. Commun. 50 [46] M. Freitag, J. Teuscher, Y. Saygili, X. Zhang, F. Giordano, P. Liska, J. Hua, S.
(2014) 4309–4311. M. Zakeeruddin, J.-E. Moser, M. Gr€ atzel, A. Hagfeldt, Nat. Photon. 11 (2017) 372.
[22] K. Pei, Y. Wu, A. Islam, Q. Zhang, L. Han, H. Tian, W. Zhu, ACS Appl. Mater. [47] Y.S. Tingare, Nguy^en, S.n. Vinh, H.-H. Chou, Y.-C. Liu, Y.-S. Long, T.-C. Wu, T.-
Interfaces 5 (2013) 4986–4995. C. Wei, C.-Y. Yeh, Adv. Energy Mater. 8 (2018), 1802405.
[23] J. Yang, P. Ganesan, J. Teuscher, T. Moehl, Y.J. Kim, C. Yi, P. Comte, K. Pei, T. [48] M.B. Desta, N.S. Vinh, C.H. Pavan Kumar, S. Chaurasia, W.-T. Wu, J.T. Lin, T.-
W. Holcombe, M.K. Nazeeruddin, J. Hua, S.M. Zakeeruddin, H. Tian, M. Gr€ atzel, C. Wei, E. Wei-Guang Diau, J. Mater. Chem. 6 (2018) 13778–13789.
J. Am. Chem. Soc. 136 (2014) 5722–5730. [49] S.R. Chemler, D. Trauner, S.J. Danishefsky, Angew. Chem. Int. Ed. 40 (2001)
[24] Y. Gao, X. Li, Y. Hu, Y. Fan, J. Yuan, N. Robertson, J. Hua, S.R. Marder, J. Mater. 4544–4568.
Chem. 4 (2016) 12865–12877. [50] T. Kitamura, M. Ikeda, K. Shigaki, T. Inoue, N.A. Anderson, X. Ai, T. Lian,
[25] J.-S. Ni, T.-Y. Chiu, W.-S. Kao, H.-J. Chou, C.-c. Su, J.T. Lin, ACS Appl. Mater. S. Yanagida, Chem. Mater. 16 (2004) 1806–1812.
Interfaces 8 (2016) 23066–23073. [51] J. Bisquert, J. Phys. Chem. B 106 (2002) 325–333.
[26] M.L. Jiang, J.-J. Wen, Z.-M. Chen, W.-H. Tsai, T.-C. Lin, T.J. Chow, Y.J. Chang, [52] Q. Wang, J.-E. Moser, M. Gr€ atzel, J. Phys. Chem. B 109 (2005) 14945–14953.
ChemSusChem 12 (2019) 3654–3665. [53] S.A. Haque, E. Palomares, B.M. Cho, A.N.M. Green, N. Hirata, D.R. Klug, J.
[27] J. Liu, K. Wang, F. Xu, Z. Tang, W. Zheng, J. Zhang, C. Li, T. Yu, X. You, R. Durrant, J. Am. Chem. Soc. 127 (2005) 3456–3462.
Tetrahedron Lett. 52 (2011) 6492–6496. [54] G. Boschloo, L. H€ aggman, A. Hagfeldt, J. Phys. Chem. B 110 (2006) 13144–13150.
[28] D. Kuang, C. Klein, H.J. Snaith, J.-E. Moser, R. Humphry-Baker, P. Comte, S. [55] Z. Ning, Y. Fu, H. Tian, Energy Environ. Sci. 3 (2010) 1170–1181.
M. Zakeeruddin, M. Gr€ atzel, Nano Lett. 6 (2006) 769–773. [56] Y.S. Tingare, N.S.n. Vinh, H.-H. Chou, Y.-C. Liu, Y.-S. Long, T.-C. Wu, T.-C. Wei, C.-
[29] Y. Uemura, T.N. Murakami, N. Koumura, J. Phys. Chem. C 118 (2014) Y. Yeh, Adv. Energy Mater. 7 (2017), 1700032.
16749–16759. [57] C.-H. Chen, P.-T. Chou, T.-C. Yin, K.-F. Chen, M.-L. Jiang, Y.J. Chang, C.-K. Tai, B.-
[30] Y.-H. Chiu, M. Shibahara, R.-Y. Huang, M. Watanabe, Z.-S. Wang, Y.-J. Hsiao, B.- C. Wang, Org. Electron. 59 (2018) 69–76.
F. Chang, T.-H. Chiang, Y.J. Chang, Dyes Pigments 136 (2017) 761–772. [58] S. Yun, P.D. Lund, A. Hinsch, Energy Environ. Sci. 8 (2015) 3495–3514.
[31] S.-Q. Fan, C. Kim, B. Fang, K.-X. Liao, G.-J. Yang, C.-J. Li, J.-J. Kim, J. Ko, J. Phys.
Chem. C 115 (2011) 7747–7754.

10

You might also like