Position engineering of cyanoacrylic-acid anchoring group in a dye for DSSC applications

Article abstract of DOI:10.1016/j.dyepig.2020.108470

This study provides insights into the molecular tailoring of DSSC dyes via synthesis/investigation of three new organic dyes with the position engineering of the acceptor moiety. Differences in photophysical, electrochemical and photovoltaic properties are found and interpreted in detail. The excitation-spectrum study shows that the population of the intramolecular charge transfer (ICT) is incident-wavelength dependent; and the ICT excitons originate not only from direct ICT band transition but also from π-π* energy transfer, and the degree of the latter depends much on the position of the anchoring group. The transient photoluminescence discloses that the non-injection quenching of the ortho dye is the most severe possibly due to the instability of the excited state resulting from the hurdling of electron delocalization by the large steric hinderance. Large steric hinderance in the ortho dye also results in the least dye loading on TiO₂ photoanode. Due to broader/stronger absorption, higher dye loading, high injection efficiency and low charge recombination rate, the para-substituted dye DSSC device achieves the best performance with power conversion efficiency of 6.63%; while the ortho dye based device performs the worst with significant lower of photocurrent and photovoltage compared to the para dye. The dye dependent V_OC is explained by combination of dipole moment and dye loading. Particularly, the employing of excitation spectrum as tools to investigate the intrinsic photophysical complicity of dyes with D-π-A structure, which is scarcely reported, is proved to be an effective way.

Full text of DOI:10.1016/j.dyepig.2020.108470

Dyes and Pigments 180 (2020) 108470
Contents lists available at ScienceDirect
Dyes and Pigments
journal homepage: http://www.elsevier.com/locate/dyepig
Position engineering of cyanoacrylic-acid anchoring group in a dye for
DSSC applications
,
*
Shuaishuai Liua, Yunfei Jiaoa, Yongjie Dinga, Xingli Fanb, Juan Songb, Baoxiu Mia
,
,
**
Zhiqiang Gaob
a Key Laboratory for Organic Electronics and Information Displays (KLOEID), Jiangsu National Synergetic Innovation Center for Advanced Materials (SICAM), Institute
of Advanced Materials (IAM), Nanjing University of Posts & Telecommunications, 9 Wenyuan Road, Nanjing, 210023, China
b Jiangsu Engineering Centre for Plate Displays & Solid State Lighting, School of Material Science and Engineering, Nanjing University of Posts & Telecommunications, 9
Wenyuan Road, Nanjing, 210023, China
A R T I C L E I N F O
A B S T R A C T
Keywords:
This study provides insights into the molecular tailoring of DSSC dyes via synthesis/investigation of three new
organic dyes with the position engineering of the acceptor moiety. Differences in photophysical, electrochemical
and photovoltaic properties are found and interpreted in detail. The excitation-spectrum study shows that the
population of the intramolecular charge transfer (ICT) is incident-wavelength dependent; and the ICT excitons
Dye-sensitized solar cell
Organic dyes
Acceptor
Link position
Phenothiazine
originate not only from direct ICT band transition but also from π-π* energy transfer, and the degree of the latter
depends much on the position of the anchoring group. The transient photoluminescence discloses that the non-
injection quenching of the ortho dye is the most severe possibly due to the instability of the excited state resulting
from the hurdling of electron delocalization by the large steric hinderance. Large steric hinderance in the ortho
dye also results in the least dye loading on TiO2 photoanode. Due to broader/stronger absorption, higher dye
loading, high injection efficiency and low charge recombination rate, the para-substituted dye DSSC device
achieves the best performance with power conversion efficiency of 6.63%; while the ortho dye based device
performs the worst with significant lower of photocurrent and photovoltage compared to the para dye. The dye
dependent VOC is explained by combination of dipole moment and dye loading. Particularly, the employing of
excitation spectrum as tools to investigate the intrinsic photophysical complicity of dyes with D-
π-A structure,
which is scarcely reported, is proved to be an effective way.
1. Introduction
replacing conventional metal containing dyes for DSSC. Generally,
organic dyes are designed as donor- -acceptor (D- -A) structures mainly
due to the following material features. Firstly, the pull-push tendency in
these D- -A dyes leads to efficient intramolecular charge transfer (ICT)
from the donor to the acceptor through the -bridge upon light ab-
sorption [9–11], which can redshift the absorption spectrum compared
π
π
Dye-sensitized solar cells (DSSCs) have drawn great attention in
recent couple of decades mainly due to their easy fabrication, high
power conversion efficiency (PCE) and low cost [1–3]. A conventional
DSSC device consists of TiO2 photoanode, dye sensitizer, redox elec-
trolyte and counter electrode. Among them, the dye not only plays the
role of sunlight harvesting, but also relates to other main optoelectronic
processes occurred in the device [4], such as electron injection to TiO2,
as well as dye regeneration with the help of redox in electrolyte.
Therefore, dye design and property investigation for DSSC have been
intensively studied these years. Having the advantages of abundant
π
π
to the typical π-π* transitions in a conjugated organic material, thus
more matching with the solar irradiation profile and beneficial to light
harvesting; and from an electronic standpoint this ICT process controls
the intrinsic photovoltaic potential of a pure organic dye. Secondly, this
D-π-A structure offers great potential for property tuning at electronic
levels. For instance, by changing D-, π-, and A-moieties and/or their
resource, easy synthesis, low cost, high molar extinction coefficients (
ε)
linking modes, dyes with different highest occupied molecular orbitals
(HOMOs) and lowest unoccupied molecular orbitals (LUMOs) can be
and relative low toxicity, organic dyes [5–8] have been gradually
* Corresponding author.
** Corresponding author.
E-mail addresses: iambxmi@njupt.edu.cn (B. Mi), iamzqgao@njupt.edu.cn (Z. Gao).
https://doi.org/10.1016/j.dyepig.2020.108470
Received 22 February 2020; Received in revised form 21 April 2020; Accepted 21 April 2020
Available online 11 May 2020
0143-7208/© 2020 Elsevier Ltd. All rights reserved.

S. Liu et al.
Dyes and Pigments 180 (2020) 108470
Scheme 1. Synthetic routes for CBPTZ-o, CBPTZ-m and CBPTZ-p: (a) KOH, TBAB DMSO, rt, 8 h; (b) CHCl3, 0 �C, 6 h; (c) K2CO3, Pd(PPh3)4, TBAB, THF, 85 �C, 13 h;
(d) K2CO3, Pd(PPh3)4, TBAB, THF, 85 �C, 13 h; (e) HAc, NH4Ac, 90 �C, reflux, 24 h.
feasibly obtained [12,13], making it easy to tune the dyes energetically
favorable both for electron injection to TiO2 and dye regeneration via
feasible for organic dye development [35–38]. Yan-Duo Lin et al. syn-
thesized a series of organic dyes with the triphenylamine moiety linked
to the main chromophore through either para- or meta-position across a
phenyl group. The result shows that the overall properties of the par-
a-isomer are better than that of the meta-isomer [39]. Walid Sharmoukh
et al. reported two organic dyes comprising para- and ortho-linked donor
relative to the anchoring group. The para-dye has wider range of ab-
sorption and lower bandgap, hence higher incident photon-to-current
conversion efficiency (IPCE) as well as better PCE were achieved
compared to the ortho-dye [40]. Our group has synthesized two organic
IÀ /IÀ3 . Moreover, the D-
π-A structure renders donor-localized HOMO--
electrons and acceptor-localized LUMO-electrons, which facilitate
electron injection into TiO2 photoanode.
In the efforts of paving the ways toward good quality organic dyes for
DSSC, a variety of moieties have been investigated, such as electron rich
donors of carbazole (Cz) [14,15], triarylamine [16,17], phenothiazine
(PTZ) [18,19] and indoline [20,21]; electron deficient anchoring ac-
ceptors of cyanoacrylic acid [22] and rhodanine-3-acetic acid [23], as
well as conjugated bridges of oligothiophene [24], phenyl [25], furan
[26] and benzothiadiazole [27]. Among them, PTZ, a heterocyclic aro-
matic compound containing three fused rings, is one of the most popular
building block, due to its electronic-rich structure and non-planar but-
terfly configuration that can either efficiently provide electrons or
inhibit the intermolecular aggregation [28,29]. In 2007, Tian et al.
firstly reported the PTZ dye, using PTZ as donor, and PCE of 5.5% was
achieved when employed in a DSSC device [30]. Since then, PTZ moiety
dyes, in which the carbazole moiety was grafted to PTZ π-bridge through
meta and para position across a phenyl group. Due to its better molecular
conjugation and more effective ICT process, the para-dye showed better
solar-energy harvesting ability. And also, the more favorable spacial
conformation of the para-dye leads to higher dye loading. All these
contribute to the higher PCE in the para-dye device [23].
In this paper, three novel PTZ dyes are synthesized (CBPTZ-o,
CBPTZ-m and CBPTZ-p). Cz, PTZ, and cyanoacrylic acid are used as
has been widely used in D-
π
-A type organic dyes [19]. Up to now, DSSC
donor, π-bridge, and anchoring acceptor, respectively. A hexyl chain is
devices based on dye comprising PTZ donor have achieved PCE of
introduced in the N-position of PTZ spacer to suppress dye aggregation
[41,42]. The Cz is meta-linked to the PTZ bridge across a phenylene.
These dyes are part of our investigation on PTZ based dyes; and their
analogues with ortho- or para-linked Cz will be investigated later. Spe-
10.06% [28]; PTZ π-bridge containing dyes showed PCE of 8.18% or
12.10% in DSSC devices without [31] or with co-adsorption of cheno-
deoxycholic acid (CDCA) [32].
Besides the selections of donor, acceptor and
π
-bridge in an organic
cifically, the linking position of the acceptor moiety relative to π-bridge
dye with a D-
π-A structure, the attachment of additional D-,
π
-, and/or A-
is varied as ortho-, meta- and para-for the purpose of studying
structure-property relationship as well as the impact on device
performance.
moieties is another dye-design scheme [33,34]. By such, the effective
conjugation and/or the push-pull tendency can be improved, hence
redshifting absorption spectrum for better sunlight harvesting. Addi-
tionally, changing the position of a building-block in a dye is also proved
We find that changing the linking position of the acceptor results in
different absorption intensities and peaks of the ICT band. Additionally,
2

S. Liu et al.
Dyes and Pigments 180 (2020) 108470
the linking position greatly influences the electronic structure, as well as
the molecular stretching in space of the dyes, making them very
different when applied in a DSSC device. The highest PCE device is
achieved from CBPTZ-p (6.63%) while the lowest from CBPTZ-o
(2.99%) under simulated AM 1.5G irradiation (100 mW cmÀ 2). Corre-
sponding reasons are investigated with highlights for further dye design.
acid (574 mg, 2 mmol), K2CO3 (345 mg, 2.5 mmol), Pd[P(C6H5)3]4 (50
mg, 0.05 mmol), TBAB (96.6 mg, 0.3 mmol) and THF (15 mL) were
placed in a three-necked flask and heated under reflux for 13 h under a
nitrogen atmosphere. When the reaction complete, water was added to
quench the reaction. After extracted with DCM three times, then washed
with brine and dried over anhydrous Na2SO4, the crude product was
purified by column chromatography using PE as the eluent to give
compound 3 as yellow solid, yield 56% (338 mg), 1H NMR δH (CDCl3,
400 MHz), 8.15 (2H, d, J ¼ 7.6 Hz), 7.71 (1H, s), 7.62 (2H, d, J ¼ 7.3
Hz), 7.50 (1H, dt, J ¼ 7.0, 2.0 Hz), 7.47–7.37 (6H, m), 7.29 (2H, t, J ¼
7.0 Hz), 7.24–7.20 (2H, m), 6.88 (1H, d, J ¼ 8.4 Hz), 6.69 (1H, d, J ¼ 9.3
Hz), 3.80 (2H, t, J ¼ 7.1 Hz), 1.49 (2H, p, J ¼ 7.6, 6.8 Hz), 1.31–1.28
(6H, m), 0.87 (3H, d, J ¼ 5.7 Hz). MS Found: [M]þ 601.84, 603.84;
2. Experimental
2.1. Materials
All solvents were purified using standard techniques. Commercially
available reagents were used without further purification. All NMR
solvents were used as received.
C
36H31BrN2S requires [M]þ 602.14, 604.14.
2.3.4. 2-(7-(3-(9H-carbazol-9-yl)phenyl)-10-hexyl-10H-phenothiazin-3-
yl)benzaldehyde, 3-(7-(3-(9H-carbazol-9-yl)phenyl)-10-hexyl-10H-
phenothiazin-3-yl)benzaldehyde and 4-(7-(3-(9H-carbazol-9-yl)phenyl)-
10-hexyl-10H-phenothiazin-3-yl)benzaldehyde (4a, 4b and 4c)
2.2. Methods and instrumentations
Experimental details for 1H and 13C NMR, matrix-assisted laser
desorption/ionization time-of-flight mass spectrometry (MALDI-TOF),
ultraviolet–visible spectra (UV–vis), cyclic voltammetry analysis (CV)
can be found in our previous paper [43]. High resolution mass spec-
trometer (HRMS) data were obtained using a Brooke solanX 70 FT-MS.
The molecular geometry optimization and the calculation of the optical
transition were performed using density functional theory (DFT) and the
time-dependent-DFT via Gaussian 09 program under Becke’s 3-parame-
ters employing the Lee-Yang-Perdew function (B3LYP) with the 6-31G
(d) basis set.
Compound 3 (603 mg, 1 mmol), 2-formylphenyl, 3-formylphenyl or
4-formylphenyl boronic acid (180 mg, 1.2 mmol), K2CO3 (345 mg, 2.5
mmol), Pd[P(C6H5)3]4 (50 mg, 0.05 mmol), TBAB (96.6 mg, 0.3 mmol)
and THF (15 mL) were placed in a three-necked flask and heated to
reflux under a nitrogen atmosphere for 13 h. When the reaction com-
plete, water was added to quench the reaction. After extracted with DCM
three times, then washed with brine and dried over anhydrous Na2SO4,
the crude product was purified by column chromatography using PE/
DCM (2:1) as eluent to give compound 4a, 4b or 4c.
Characterization: Compound 4a, yellow solid, yield 71% (446 mg),
2.3. Synthesis
1H NMR δ (CDCl3, 400 MHz), 10.02 (1H, s),8.16 (2H, d), 8.00 (1H, m, J
H
¼ 7.1 Hz), 7.74 (1H, s), 7.67–7.58 (3H, t), 7.50–7.37 (9H, m), 7.30 (1H,
s), 7.21–7.09 (3H, m), 6.95 (2H, t, J ¼ 8.7 Hz), 3.91 (2H, t, J ¼ 7.1 Hz),
1.49 (2H, dt, J ¼ 13.5, 7.4 Hz), 1.34 (6H, d, J ¼ 3.4 Hz), 0.90 (3H, t, J ¼
4.8 Hz). MS Found: [M]þ 628.49; C41H36N2OS requires [M]þ 628.25.
Characterization: Compound 4b, yellow solid, yield 61% (383 mg),
1H NMR δH (CDCl3, 400 MHz), 10.07 (1H, s), 8.17 (2H, d, J ¼ 7.7 Hz),
8.04 (1H, s), 7.81 (2H, t, J ¼ 7.5 Hz), 7.74 (1H, s), 7.65 (2H, d, J ¼ 5.5
Hz), 7.58 (1H, t, J ¼ 7.7 Hz), 7.53–7.37 (9H, m), 7.33–7.27 (2H, m), 6.94
(2H, d, J ¼ 8.6 Hz), 3.91 (2H, t, J ¼ 7.1 Hz), 1.49 (2H, dt, J ¼ 13.5, 7.4
Hz), 1.34 (6H, d, J ¼ 3.4 Hz), 0.90 (3H, t, J ¼ 4.8 Hz). MS Found: [M]þ
628.43; C41H36N2OS requires [M]þ 628.25.
The synthesis route is shown in Scheme 1 and details for synthesis
and characterization are given as following:
2.3.1. 10-hexyl-10H-phenothiazine (1)
Phenothiazine (200 mg, 1 mmol), bromohexane (216.7 mg, 1.3
mmol), KOH (168.3 mg, 3 mmol), tetrabutylammonium bromide (TBAB,
96.6 mg, 0.3 mmol) and dimethyl sulfoxide (DMSO, 10 mL) were placed
in a three-necked flask and allowed to react at room temperature for 10
h under a nitrogen atmosphere. When the reaction complete, water was
added to stop the reaction. After extracted three times with dichloro-
methane (DCM), then washed with brine and dried over anhydrous
Na2SO4, the crude product was purified by column chromatography
using petroleum ether (PE)/DCM (15:1) as eluent to afford compound 1
[44] as colorless oil, yield 88% (250 mg), 1H NMR δ (CDCl3, 400 MHz),
7.21–7.13 (4H, m), 6.97–6.85 (4H, m), 3.86 (2HH, t, J ¼ 7.1 Hz),
1.87–1.77 (2H, m), 1.51–1.40 (2H, m), 1.37–1.28 (4H, m), 0.95–0.86
(3H, m). MS Found: [M]þ 283.14; C18H21NS requires [M]þ 283.43.
Characterization: Compound 4c, yellow solid, yield 60% (377 mg),
1H NMR δ (CDCl3, 400 MHz), 10.02 (1H, s), 8.16 (2H, d), 8.00 (1H, m,
J ¼ 7.1 HHz), 7.74 (1H, s), 7.67–7.58 (3H, t), 7.50–7.37 (9H, m), 7.30
(1H, s), 7.21–7.09 (3H, m), 6.95 (2H, t, J ¼ 8.7 Hz), 3.91 (2H, t, J ¼ 7.1
Hz), 1.49 (2H, dt, J ¼ 13.5, 7.4 Hz), 1.34 (6H, d, J ¼ 3.4 Hz), 0.90 (3H, t,
J ¼ 4.8 Hz). MS Found: [M]þ 628.32; C41H36N2OS requires [M]þ
628.25.
2.3.2. 3,7-dibromo-10-hexyl-10H-phenothiazine (2)
2.3.5. (E)-3-(2-(7-(3-(9H-carbazol-9-yl)phenyl)-10-hexyl-10H-
phenothiazin-3-yl)phenyl)-2-cyanoacrylic acid, (E)-3-(3-(7-(3-(9H-
carbazol-9-yl)phenyl)-10-hexyl-10H-phenothiazin-3-yl)phenyl)-2-
cyanoacrylic acid and (E)-3-(4-(7-(3-(9H-carbazol-9-yl)phenyl)-10-hexyl-
10H-phenothiazin-3-yl)phenyl)-2-cyanoacrylic acid (CBPTZ-o, CBPTZ-m
and CBPTZ-p)
Compound 1 (283 mg, 1 mmol) and CHCl3 (10 mL) were added to a
three-necked flask, and N-bromosuccinimide (NBS, 531 mg, 3 mmol)
was added to the bottle three times in the first 2 h. The mixture was
stirred under an atmosphere of nitrogen for 6 h in an ice water bath.
When the reaction completed, water was added to stop the reaction.
After extracted three times with DCM, then washed with brine and dried
over anhydrous Na2SO4, the crude product was purified by column
chromatography using PE/DCM (1:1) as eluent to afford compound 2
[45] as yellow liquid, yield 70% (309 mg), 1H NMR δ (CDCl3, 400
MHz), 7.25–7.20 (4H, m), 6.68 (2H, d, J ¼ 8.5 Hz), 3.75H(2H, t, J ¼ 7.1
Hz), 1.73 (2H, p, J ¼ 7.5 Hz), 1.39 (2H, p, J ¼ 6.9 Hz), 1.27 (4H, dt, J ¼
7.3, 3.7 Hz), 0.90–0.82 (3H, m). MS Found: [M]þ 438.62, 440.62,
442.62; C18H19Br2NS requires [M]þ 438.96, 440.96, 442.96.
Compound 4a, 4b or 4c (628 mg, 1 mmol), cyanoacetic acid (170
mg, 2 mmol), NH4Ac (77 mg, 1 mmol) and CH3COOH (10 mL) were
placed in a three-necked flask and heated to reflux under a nitrogen
atmosphere for 24 h. When the reaction complete, water was added to
quench the reaction. After extracted with DCM three times, then washed
with brine and dried over anhydrous Na2SO4, the crude product was
purified by column chromatography using methanol (MeOH)/DCM
(1:30) as the eluent to give CBPTZ-o, CBPTZ-m or CBPTZ-p.
Compound CBPTZ-o, crimson solid, yield 95% (650 mg), 1H NMR δ
2.3.3. 3-(3-(9H-carbazol-9-yl)phenyl)-7-bromo-10-hexyl-10H-
phenothiazine (3)
H
(DMSO‑d6, 400 MHz), 8.21 (2H, d, J ¼ 7.8 Hz), 8.11–7.97 (2H, m), 7.78
(1H, s), 7.74–7.60 (3H, m), 7.57–7.35 (9H, m), 7.30–7.19 (2H, m),
Compound 2 (441 mg, 1 mmol), 3-(9H-Carbazol-9-yl)phenylboronic
3

S. Liu et al.
Dyes and Pigments 180 (2020) 108470
Table 1
Parameters for CBPTZ-o, CBPTZ-m and CBPTZ-p based on DFT calculation.
Dipole moment
HOMO (eV)
LUMO (eV)
CBPTZ-o
CBPTZ-m
CBPTZ-p
51.43�
51.51�
51.36�
65.57�
33.41�
32.06�
35.00�
9.96�
0.30�
37.95�
34.02�
10.04�
8.93
5.09
9.11
À 5.22
À 5.21
À 5.21
À 2.55
À 2.72
À 2.74
a Dihedral angle, b torsion angle.
7.10–6.95 (4H, m), 3.81 (2H, s), 1.64 (2H, p, J ¼ 12.1, 9.3 Hz),
133.99, 133.01, 131.87, 131.18, 130.58, 126.91, 126.75, 125.73,
125.69, 124.58, 124.24, 124.07, 123.25, 120.98, 120.52, 116.93,
110.21, 103.48, 47.15, 40.01, 31.31, 26.63, 26.29, 22.54, 14.30. HRMS
Found: [M]þ 695.2608; C46H37N3O2S requires [M]þ 695.2606.
1.40–1.22 (6H, m), 0.77 (3H, d, J ¼ 6.7 Hz). 13C NMR δ (DMSO‑d6, 101
C
MHz), 163.63, 152.19, 144.77, 144.33, 142.48, 141.38, 140.61, 137.99,
134.01, 133.52, 132.20, 131.19, 130.60, 130.48, 130.05, 128.94,
128.33, 128.11, 126.77, 125.83, 125.75, 125.67, 124.59, 124.18,
123.76, 123.25, 120.98, 120.53, 117.34, 116.67, 115.81, 110.22, 47.16,
40.00, 31.32, 26.65, 26.33, 22.54, 14.30. HRMS Found: [M]þ 695.2606;
2.4. DSSC fabrication and photovoltaic measurements
C
46H37N3O2S requires [M]þ 695.2606.
The DSSC fabrication detail is depicted in electronic supporting in-
formation (ESI), and the current-voltage measurement (J-V), incident
photon-to-current efficiencies (IPCEs) and electrochemical impedance
spectroscopy (EIS) can be found in our previous paper [46–48].
Compound CBPTZ-m, crimson solid, yield 75% (513 mg), 1H NMR
δ (DMSO‑d6, 400 MHz), 8.44 (1H, s), 8.35–8.23 (3H, m), 8.01 (1H, d, J
¼H7.9 Hz), 7.90 (2H, d, J ¼ 9.9 Hz), 7.82 (1H, d, J ¼ 8.2 Hz), 7.74 (1H, t,
J ¼ 7.8 Hz), 7.69–7.51 (6H, m), 7.46 (4H, d, J ¼ 5.5 Hz), 7.31 (2H, ddd,
J ¼ 8.0, 5.8, 2.4 Hz), 7.14 (2H, t, J ¼ 9.4 Hz), 3.95 (2H, t, J ¼ 6.7 Hz),
1.48–1.37 (2H, m), 1.31–1.20 (6H, m), 0.84 (3H, t, J ¼ 6.9 Hz). 13C NMR
δC (DMSO‑d6, 101 MHz), 163.65, 153.96, 148.37, 140.62, 140.51,
139.30, 138.10, 137.72, 137.64, 132.94, 132.90, 132.56, 131.85,
131.49, 130.73, 130.49, 129.66, 129.47, 129.34, 128.70, 126.82,
125.92, 124.60, 123.21, 120.95, 120.65, 117.66, 117.10, 116.75,
114.72, 110.13, 47.28, 31.29, 26.44, 25.92, 22.51, 14.27. HRMS Found:
[M]þ 695.2608; C46H37N3O2S requires [M]þ 695.2606.
3. Result and discussions
3.1. Theoretical calculation
Density functional theory (DFT) calculations using Gaussian 09
software (B3LYP/6-31G*) were conducted in order to probe the opti-
mized spatial configuration of the dye molecules and the features of
frontier molecular orbitals in CBPTZ-o, CBPTZ-m and CBPTZ-p. As
shown in the first cell of Table 1, our three dyes are designed in such that
Compound CBPTZ-p, crimson solid, yield 95% (650 mg), 1H NMR δ
H
(DMSO‑d6, 400 MHz), 8.33 (1H, s), 8.26 (2H, d, J ¼ 7.8 Hz), 8.09 (2H, d,
J ¼ 8.4 Hz), 7.89 (3H, s), 7.81 (1H, d, J ¼ 9.2 Hz), 7.73 (1H, t, J ¼ 8.1
Hz), 7.68–7.54 (5H, m), 7.45 (4H, d, J ¼ 5.2 Hz), 7.34–7.26 (2H, m),
7.11 (2H, dd, J ¼ 8.6, 2.1 Hz), 3.99–3.88 (2H, m), 1.45–1.37 (2H, m),
1.27–1.21 (6H, m), 0.83 (3H, t, J ¼ 6.9 Hz). 13C NMR δC (DMSO‑d6, 101
MHz), 163.88, 153.93, 145.15, 144.29, 143.47, 141.36, 140.61, 138.00,
the central PTZ π-bridge connects with the left hand Cz donor via a meta-
phenylene (Ph1), and with the right hand cyanoacrylic acid acceptor (A)
via another phenylene (Ph2) in ortho, meta or para pattern, forming D-
Ph1-π-Ph2-A framework. The only structural difference among the three
dyes is the linking position of the acceptor, therefore all property dif-
ferences are accounted from it.
Fig. 1. Frontier molecular orbital profiles of CBPTZ-o, CBPTZ-m and CBPTZ-p.
4

S. Liu et al.
Dyes and Pigments 180 (2020) 108470
Table 2
ICT absorption of CBPTZ-o, CBPTZ-m and CBPTZ-p from TD-DFT quantum
calculation.
Dye
λ (nm)
f
Transition assignmenta
CBPTZ-o
554
0.01
HOMO → LUMO (97.4%)
HOMO-1 → LUMO (2.6%)
HOMO → LUMO (100%)
HOMO → LUMO (100%)
CBPTZ-m
CBPTZ-p
626
564
0.002
0.31
a
The transition assignments were computed with the TD-DFT (see ESI).
To enable low-energy charge separation in a D- -A dye, alignment of
π
the π-system is important, which can be quantified by considering the
dihedral/twist angles between these moieties because the relative
alignment of these moieties largely dictates the overall molecular
planarity. As informed from the quantum calculation, within the mo-
lecular framework of D-Ph1-π-Ph2-A, the dihedral angles between Cz
and PTZ are almost the same, i.e., 51.4 for CBPTZ-o, 51.5 for CBPTZ-m
and 51.4� for CBPTZ-p. The identical orientation of the carbazole moi-
ety is understood by that the distance between the Cz and the A-moiety
is too far to be spatially influenced by each other. However, the position
of the A-moiety dramatically influences the spatial stretching of both
π
-bridge and the adjacent Ph2 group. When the A-moiety changes from
ortho to meta to para position, twist angles between the PTZ and Ph2
(65.6, 33.4 and 32.1�), and between Ph2 and A-moiety (35.0, 9.9 and
0.3�) monotonically decreases, mainly due to the monotonically
decreased steric-hindrance for rotation movement. Furthermore, the
influences of the A-position on the space-stretching of the anchoring
group itself are large, e.g., the dihedral angle between cyanoacrylic acid
and PTZ changes from 37.9 to 34.0 and to 10.0�. Again, this monotonic
decrease is caused by steric-hinderance decreasing when the dye A-
moiety changed from ortho to meta to para position. Small dihedral/twist
angles between the π-bridge and acceptor parts will facilitate intra-
molecular charge transfer (ICT) and reduce adsorption hinderance on
the TiO2 photoanode.
Fig. 1 shows the frontier molecular orbitals of the three dyes. The
HOMOs of three dyes distribute mainly on the central phenothiazine
Fig. 2. (a) Cyclic voltammograms of CBPTZ-o, CBPTZ-m and CBPTZ-p
measured in DCM and (b) energy levels diagram for CBPTZ-o, CBPTZ-m and
CBPTZ-p dye related DSSC devices.
π
-bridge with slight extension to the adjacent Ph1 and Ph2. The HOMO
density from the Cz group is negligible in all dyes, consisting with their
large dihedral angles (about 51�) between the Cz and the PTZ
π
-bridge,
which limits electron delocalization. The LUMOs of dyes localize
dominantly over the cyanoacrylic acid moiety, spreading over Ph2 in
ortho and meta dye and to both Ph2 and PTZ in the para dye. The degree
of HOMO/LUMO electron delocalization within the molecules is
consistent with their dihedral/twist angles along the molecular axis. The
overlapping between HOMO and LUMO levels is largest in the para-dye,
which agrees well with the most efficient ICT absorption in the para dye
as discussed later. To further access the nature of the ICT transition
process, time-dependent DFT calculations have been done, neglecting
the influence of solvent (see Table 2). The predicted ICT bands are 554
(0.01), 626 (0.002) and 564 nm (0.31) for the ortho-, meta- and para-
dyes, respectively. ICT absorption bands in CBPTZ-m and CBPTZ-p are
solely from HOMO to LUMO, while those in CBPTZ-o consist small part
(2.6%) from HOMO-1 besides the main contribution from HOMO
(97.4%). As illustrated in the electron density of the HOMO-1 levels, that
for CBPTZ-o contains much contribution from the central PTZ moiety
compared to other two dyes, and this may account for its HOMO-1
contribution to the ICT transition. It is abnormal that the CBPTZ-m
possesses the longest wavelength for ICT transition, though this is
consistent with the experimental results as discussed in photophysical
section.
Table 3
Photophysical and electrochemical properties of CBPTZ-o, CBPTZ-m and
CBPTZ-p.
À
π
b
Dye
λπ
*
max [nm] (
ε
λICT max [nm] (
[molÀ 1 dm3
cmÀ 1])a
ε
Eg
HOMOc
(eV)
LUMOd
(eV)
[molÀ 1 dm3
cmÀ 1])a
(eV)
CBPTZ-
278 (39500), 330
(12700)
434 (1500)
530 (1400)
438 (13000)
2.61
2.20
2.20
À 5.21
À 5.21
À 5.22
À 2.60
À 3.01
À 3.02
o
CBPTZ-
m
284 (48800), 328
(19200)
CBPTZ-
260 (44700), 327
(30400)
p
a
Measured in DCM, data are peak wavelength and its extinction coefficient.
Estimated from the cutoff wavelength of absorption spectra in DCM solution.
Measured in DCM with 0.1 M TBAPF6 as electrolyte.
Calculated by HOMO þ Eg.
b
c
d
CBPTZ-o, CBPTZ-m and CBPTZ-p, respectively. Therefore, the DSSC
devices based on CBPTZ-p can be expected a largest VOC compared to
devices based on CBPTZ-o and CBPTZ-m.
Since the dipole moment of a dye has a great influence on the con-
duction band (CB) position of TiO2 absorbed with the dye, the relatively
large dipole moment tends to cause the upward shift of the CB in TiO2
film [28,49,50], increasing device open-circuit photovoltage (VOC). As
can be seen that the calculated dipole moments are 8.9, 5.1 and 9.1 for
3.2. Electrochemical investigations
To evaluate the possibilities of the electron injection from the dye to
TiO2 photoanode, as well as the dye regeneration via redox IÀ /I3À in
DSSC device based on these dyes, their energy levels were investigated
5

S. Liu et al.
Dyes and Pigments 180 (2020) 108470
(Fc/Fcþ) is 0.46 eV. With band gap values (Eg) of 2.61 eV for CBPTZ-o,
2.20 eV for CBPTZ-m and 2.20 eV for CBPTZ-p derived from the ab-
sorption spectra (see Fig. 3a), the corresponding LUMO values of
CBPTZ-o, CBPTZ-m and CBPTZ-p are calculated as À 2.60, À 3.01 and
À 3.02 eV, respectively. Impressively, these HOMO values are almost the
same as their theoretically calculated values, the LUMO values are in the
same tendency as the calculated ones with not quite large deviations,
verifying the sufficient accuracy of our calculation method. The similar
HOMOs are due to the same moieties that dominate the electron cloud of
this level, while the difference of ortho-, meta- and para-anchoring
acceptor contributes to the differences in their LUMO. Specifically, the
LUMO of the ortho dye stands well above the other two, which is due to
the largest steric hinderance for the delocalization of the LUMO elec-
trons in the ortho dye, making the most unstable of these electrons.
As shown in the energy diagram of HOMO and LUMO levels (Fig. 2b),
the LUMO levels of these dyes are all obviously higher than the CB of the
TiO2 semiconductor, which sufficiently guarantees the necessary ther-
modynamic driving force of electron injections from the dye molecules
to the CB of TiO2 photoanode. However, the much higher LUMO of the
ortho-dye relative to TiO2 is expected to cause larger energy loss in
electron injection process in its DSSC device. The HOMO levels of three
dyes are all much lower than the redox potential of IÀ /I3À redox couple,
which ensures the fast regeneration of the dyes from oxidation states.
3.3. Photophysical study
In a DSSC device, one of the determining factors for DSSC perfor-
mance is the behavior of dye absorption, and the study of property-
structure relationship in this aspect helps to tailor dye molecular
structure for good device performance. Therefore, dye absorptions in
DCM solution as well as those adsorbed on TiO2 films were investigated
(Fig. 3). The corresponding spectroscopic parameters are compiled in
Table 3. In solution, all dyes exhibit three absorption bands, peaked at
278/330/434 nm for CBPTZ-o, 284/328/530 nm for CBPTZ-m, and
260/327/438 nm for CBPTZ-p. The structured absorptions in the short
wavelength region from 250 to 380 nm are ascribed to
while those less intense and broad bands with structureless feature in
380–580 nm are ICT process from donor to acceptor moiety. In the π-π*
π-π* transitions;
band, the similar curve profiles and peak positions reflect their struc-
tural similarity, while the subtle difference there signifies the difference
of ortho-, meta- and para-anchoring acceptor. Furthermore, the band
with higher energy should be dominated by Cz, and that with lower
energy contributes mainly from PTZ moiety because they locate at the
similar position of free Cz and PTZ in solution (dot for Cz and dash line
for PTZ in Fig. 3a). When the anchoring group was grafted onto the dye
Fig. 3. UV–Vis absorption spectra of CBPTZ-o, CBPTZ-m and CBPTZ-p (a)
dissolved in DCM and (b) adsorbed on TiO2 film. The dot and dash lines are the
free carbazole and phenothiazine, respectively.
by CV. The curves are shown in Fig. 2a and the corresponding data are
shown in Table 3.
molecule via a para-mode, both Cz and PTZ dominated
are efficient. The meta-dye exhibits the most intensive
π
-
π* transitions
As shown in Fig. 2a, all dyes undergo a reversible oxidation process,
which is preferred for dye regeneration in a DSSC device. Accordingly,
HOMOs of CBPTZ-o, CBPTZ-m and CBPTZ-p are estimated to be À 5.21,
À 5.21 and À 5.22 eV, respectively, referring to equation -(EOX þ 4.8 -
EþFc/Fc), where the internal reference potential of ferrocene/ferrocenium
π-π* transition
mainly from Cz moiety. The ortho-dye presents the smallest π-π* tran-
sition originating mainly from the Cz group.
In the ICT bands, which govern dye suitability for DSSCs, the most
desired attributes are long and broad wavelengths with strong in-
tensities. As reported that dyes with longer molecular axis [51,52],
smaller twist angle [53], and/or efficient molecular orbital overlap be-
tween the ground state and the excited state [51,54], possess relatively
redshift and efficient ICT transition. Abnormally, the meta-dye presents
scissor-like peaks in the longest wavelength and comparable intensity to
the ortho-dye, contradicting with its moderate molecular axis, and
moderate effectiveness for conjugation as implied in the corresponding
dihedral/twist angles in previous calculation section. This phenomenon
is beyond our knowledge; nevertheless, the trend of wavelength/-
intensity of these ICT absorptions is consistent with the calculated result
(Table 2). Among the three dyes, CBPTZ-p presents the strongest ICT
absorption, which is partially attributed to its best molecular orbital
overlap between ground state and excited state as discussed in the
previous calculation section. Furthermore, the para-dye can produce
resonance structures relating to the N atom at the PTZ moiety, which can
Table 4
The fitting decay time (τ) and corresponding amplitude (A) of CBPTZ-o, CBPTZ-
m and CBPTZ-p from TRPL.
Dye
Film A1 (%) τ1 (ns) A2 (%) τ2 (ns) τava (ns) ηinjb (%)
CBPTZ-o
TiO2 89.51 0.10
Al2O3 92.55 0.35
TiO2 91.92 0.10
Al2O3 80.32 0.48
TiO2 96.45 0.05
Al2O3 78.60 0.49
10.49 2.49
0.35
0.90
0.18
1.12
0.10
1.26
61.11
83.93
92.06
7.45
8.08
7.69
1.14
CBPTZ-m
CBPTZ-p
19.68 3.71
3.55
1.43
21.40 4.09
a
The values of τav were determined with τav ¼ A1 τ1 þ A2 τ2
.
τav=TiO2
Calculated by ηinj. ¼ 1À
τav=Al2O3
b
6

S. Liu et al.
Dyes and Pigments 180 (2020) 108470
in Fig. 4a-c, all dyes present excitation wavelength-dependent PL vary-
ing in both wavelength and intensity. When the excitation energy
changes from high to low, all dyes show decreased short-wavelength
Table 5
Dye loading and photovoltaic parameters, as well as data from impedance
analysis.
π-π* emission and increased long wavelength ICT-emission, demon-
Dye
Dye loading
JSC
VOC
FF
PCE
(%)
Rrec
τe
(ms)
(10À 6 mol
(mA
(V)
(%)
(Ω)
strating vividly the excitation population changes upon different exci-
tation energy. Obviously, this PL feature conforms with the electronic
cmÀ 2
0.75
)
cmÀ 2
)
CBPTZ-
5.94
7.33
0.78
0.85
0.89
64.53
66.28
61.72
2.99
4.13
6.63
22.55
24.70
35.65
19.28
23.37
41.55
complicity in dyes with D-π-A structure. Since the exciton in dye ICT
o
state is the most important species for the harvesting of solar energy in a
DSSC device owning to the more spectra-matched nature with solar
CBPTZ-
m
0.91
1.64
irradiation compared to the π-π* part, to further study the behavior of
CBPTZ-
p
12.07
ICT excitons in a dye, excitation experiments monitored at the maximum
of the ICT emission in each dye were conducted and results are shown in
Fig. 4d. These excitation spectra are different to their ICT absorption
one, an indicative of the other excitation route for ICT band besides the
direct ICT band excitation. As verified by the excitation spectra them-
enhance ICT transition with redshifted feature [51,52], coinciding with
the above result.
When adsorbed onto TiO2 surface, the three dyes present ICT ab-
sorption in the same region (inset of Fig. 3b) with an obvious blue-shift
in absorbance maxima compared to that in CH2Cl2 solution, which can
be attributed to the H-aggregation and the deprotonation of –COOH
anchoring group for the strong interactions between the dyes and the
TiO2 interfaces [55–57]. The para-dye exhibits the strongest ICT ab-
sorption on TiO2 substrate, which accounts from both its strongest
extinction coefficient and the highest dye loading (Table 5), agreeing
well with the DSSC performance as discussed later.
selves, both the
π-π* absorption (less than 380 nm) and ICT absorption
(380–580 nm) contribute to the ICT emission, evidencing the partial
transfer of exciton from higher energies of π-π* band to ICT band with
lower energies. This means that the most important exciton in a DSSC
device, the ICT exciton, can be formed by both direct light absorption
and energy transfer from the π-π* exciton. And in the ortho dye, since the
ICT absorption is weak, the formation of its ICT exciton is predominately
from energy transfer from the * exciton. It should be noticed that the
π-π
downward energy transfers from the π-π* exciton to ICT exciton can
To probe the population of excited-state electrons in a dye molecule,
photoluminescence spectra of the three dyes in dilute DCM solution
were investigated by irradiation with light of different wavelengths,
corresponding to high, moderate and low excitation energies. As shown
increase exciton lifetime and hence benefiting to the afterward exciton
dissociation in its DSSC device.
Fig. 4. (a–c) Fluorescence emission spectra and (d) normalized excitation spectra of CBPTZ-o, CBPTZ-m and CBPTZ-p in DCM.
7

S. Liu et al.
Dyes and Pigments 180 (2020) 108470
Fig. 5. Time-resolved photoluminescence decay curves of CBPTZ-o, CBPTZ-m
and CBPTZ-p on TiO2 and Al2O3 films.
3.4. Transient photoluminescence
In order to study the electron dynamics between the dye and the TiO2
photoanode, the time-resolved photoluminescence (TRPL) spectros-
copies based on dye-adsorbed TiO2 films were investigated. To rule out
the influence of non-injecting quenching, PL decays of the dyes absorbed
on Al2O3 were also measured [58].
The TRPL decay curves of TiO2 and Al2O3 films with the dyes are
shown in Fig. 5. Decay times (τ) and corresponding amplitudes (A) are
summarized in Table 4. The measured TRPL signals were fitted by two
exponentials as follows: IPL ¼ A1 exp(-t/τ1) þ A2 exp(-t/τ2). where Ai
denotes the proportion of each exponential that meets the relationship
of A1 þ A2 ¼ 1, and τ1/τ2 represents the time constant of each expo-
nential component. In the Al2O3 films, due to the mismatch between the
CB bottom of Al2O3 and the dye LUMOs, the excited dyes decay to the
ground state through radiative and radiationless transitions, no electron
injection from dye to Al2O3 occurs. The obvious faster PL decay in
CBPTZ-o anchored Al2O3 film compared to other two implies its higher
non-injecting quenching and lower excited state stability, which may
result from the highest steric hinderance by the ortho-anchoring
acceptor in CBPTZ-o. Replacing the Al2O3 with TiO2 films, a dramatic PL
quenching can be observed due to the electron-injection from the
excited dyes into the CB of TiO2. Since the structural heterogeneity of
the film samples often gives rise to multiexponential PL decay [59], we
calculated the amplitude-weighted average lifetimes (τav) of the PL to
simplify the comparison between the PL decays of various samples, as
summarized in Table 4. Furthermore, by the comparison of PL decays
Fig. 6. (a) J-V curves recorded under standard AM 1.5G conditions (100 mW
cmÀ 2) and (b) IPCE spectra for DSSC devices based on CBPTZ-o, CBPTZ-m and
CBPTZ-p.
highest short-circuit photocurrent density (JSC) of 12.07 mA cmÀ 2 and
V
OC of 0.89 V, the CBPTZ-p based DSSC exhibits the best PCE of 6.63%.
In order to study the reason for the different JSC values of these dyes, the
dye loading of CBPTZ-o, CBPTZ-m and CBPTZ-p have been measured.
Higher dye loading generally contributes to higher JSC value. As shown
in Table 5, the values of the dye loading follow the order CBPTZ-p >
CBPTZ-m > CBPTZ-o, consisting with the JSC values.
It should be noted that although the only difference is the employed
dye, these DSSC devices exhibit dye-dependent VOC, e.g. 0.78 V for
CBPTZ-o, 0.85 V for CBPTZ-m and 0.89 V for CBPTZ-p. The slightly
difference between CBPTZ-m and CBPTZ-p may be caused by their
dipole moment difference (5.09 vs 9.11) as discussed in the theoretical
section. However, the moderate dipole moment of 8.93 and the smallest
between TiO2 and Al2O3 samples, we estimated the electron injection
τav=TiO2
efficiencies using: ηinj ¼ 1À
. The injection efficiency can be
τav=Al2O3
calculated as 61.11% for CBPTZ-o, 83.93% for CBPTZ-m and 92.06%
for CBPTZ-p. The combination of these results confirms that the electron
injection of CBPTZ-p is the most effective, which may expect to have a
beneficial effect on the performance of the device.
V
OC of 0.78 V for CBPTZ-o are a contradictory result. In this aspect, the
highest energy loss for electron injection in the CBPTZ-o device due to
its highest LUMO level may take a role. In addition, due to the largest
hinderance to adsorption on TiO2 in CBPTZ-o, its dye loading is small-
est, which provides the most vacancy sites on TiO2 surface available for
the back electron transfer (electron recombination) from TiO2 to the IÀ /
IÀ3 redox in the electrolyte, hence reducing VOC. These two factors are
consistent with literature report [60]. In the IPCE curves, compared to
the other dyes, dye CBPTZ-p exhibits higher IPCE values in all wave-
length regions, with excellent ability to convert light to photocurrent in
3.5. DSSC applications
Finally, the three dyes were applied in DSSC devices conjunction
with TiO2 photoanode, iodine electrolyte and Pt counter electrode. The
current-voltage (J-V) and IPCE curves are shown in Fig. 6. Table 5 gives
the performance parameters. Among the three devices, owning to the
8

S. Liu et al.
Dyes and Pigments 180 (2020) 108470
(
τe) for DSSC can be calculated from the Bode plots (Fig. 7b) by using the
following equation: τe ¼ 1/(2πfp), where fp is the peak frequency at low
frequency region. Besides the dipole moment, the largest τe (41.55 ms)
in CBPTZ-p device can also explain its highest VOC, because longer
electron lifetime means higher electron density in TiO2 conduction band
and lower charge recombination rate.
4. Conclusions
In summary, we have designed/synthesized three organic dyes with
the only structural difference occurring in the ortho-, meta- and para-
linked anchoring acceptor; and applied the dyes in DSSC devices.
Detailed studies demonstrate the properties differences among these
dyes and the possible reasons are given via combination of experimental
and theoretical methods. Particularly, excitation-energy varied emission
spectra and the excitation spectrum of ICT emission are demonstrated as
an effective way to explore the excited state population of the dyes with
D-π-A structure, which generally possess intrinsic photophysical com-
plicity. The best PCE of 6.63% with a JSC of 12.07 mA cmÀ 2 and VOC of
0.89 V is achieved in DSSC device based on the para-dye CBPTZ-p,
ascribing to its better absorption, higher dye loading, good electron
injection efficiency and lower charge recombination rate.
Declaration of competing interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
CRediT authorship contribution statement
Shuaishuai Liu: Validation, Formal analysis, Investigation, Data
curation, Writing - original draft, Writing - review & editing. Yunfei
Jiao: Formal analysis, Investigation. Yongjie Ding: Investigation. Xin-
gli Fan: Investigation. Juan Song: Methodology. Baoxiu Mi: Concep-
tualization, Resources, Writing - review & editing, Supervision.
Zhiqiang Gao: Supervision.
Acknowledgments
This work was supported by the National Natural Science Foundation
of China (Nos. 61474064); the Priority Academic Program Development
of Jiangsu Higher Education Institutions in China (PAPD: YX03001,
YX03002); Jiangsu National Synergetic Innovation Center for Advanced
Materials (SICAM), China; and the Synergetic Innovation Center for
Organic Electronics and Information Displays, China; Postgraduate
Research Practice Innovation Program of Jiangsu Province
(KYCX18_0845).
Fig. 7. (a) Nyquist plots and (b) Bode plots for DSSC devices based on CBPTZ-
o, CBPTZ-m and CBPTZ-p.
400–550 nm (quantum efficiency 70–87%). Apparently, the IPCE ten-
dency of these dyes conforms well to the order of their JSC values. The fill
factors (FFs) of DSSCs based on dyes of CBPTZ-o, CBPTZ-m and CBPTZ-p
are 64.53, 66.28 and 61.72% respectively. These values are quite
similar; and the minor difference is possibly caused by the electrical
difference originating from different dye aggregation and/or variation
in device fabrication.
Appendix A. Supplementary data
Electrochemical impedance spectrum (EIS) can indicate the charge
transfer/collection processes across various interface existing in a DSSC
device, which dictate the device performance. Therefore, the new-dye
based DSSCs were conducted with this experiment under dark condi-
tions and bias of device VOC. The Nyquist and Bode plots are displayed in
Fig. 7 and the corresponding data are listed in Table 5. Normally, under
dark conditions, the semicircle at low frequency region in the Nyquist
plots corresponds to the charge recombination resistance (Rrec) at the
TiO2/dyes/electrolyte interface in a DSSC, with large Rrec values signi-
fying the enhanced hindrance of electron recombination there and an
improved value of photovoltage. As shown in Fig. 7a, the values of Rrec
follow the order CBPTZ-p (35.65 Ω) > CBPTZ-m (24.70 Ω) > CBPTZ-o
(22.55 Ω). This trend is consistent with the trend of VOC, i.e., CBPTZ-p
(0.89 V) > CBPTZ-m (0.85 V) > CBPTZ-o (0.78 V). The electron lifetime
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.dyepig.2020.108470.
References
€
[1] O’Regan B, Gratzel M. A low-cost, high-efficiency solar cell based on dye-sensitized
colloidal TiO2 films. Nature 1991;353:737–40.
[2] Zhang SF, Yang XD, Numata YH, Han LY. Highly efficient dye-sensitized solar cells:
progress and future challenges. Energy Environ Sci 2013;6:1443–64.
[3] Kenji K, Yohei A, Toru Y, Keiji O, Jun-ichi F, Minoru H. Highly-efficient dye-
sensitized solar cells with collaborative sensitization by silyl-anchor and carboxy-
anchor dyes. Chem Commun 2015;51:15894–7.
[4] Anders H, Gerrit B, Sun LC, Lars K, Henrik P. Dye-sensitized solar cells. Chem Rev
2010;110:6595–663.
[5] Huang ZS, Cai C, Zang XF, Iqbal Z, Zeng HP, Kuang DB, et al. Effect of the linkage
location in double branched organic dyes on the photovoltaic performance of
DSSCs. J Mater Chem A 2015;3:1333–44.
9

S. Liu et al.
Dyes and Pigments 180 (2020) 108470
[6] Yao ZY, Wu H, Li Y, Wang JT, Zhang J, Zhang M, et al. Dithienopicenocarbazole as
the kernel module of low-energy-gap organic dyes for efficient conversion of
sunlight to electricity. Energy Environ Sci 2015;8:3192–7.
donors for liquid and quasi-solid-state dye-sensitized solar cells. J Mater Chem A
2014;2:8988–94.
[34] He JX, Hua JL, Hu GX, Yin XJ, Gong H, Li CX. Organic dyes incorporating a
thiophene or furan moiety for efficient dye-sensitized solar cells. Dyes Pigments
2014;104:75–82.
[7] Yao ZY, Zhang M, Wu H, Yang L, Li RZ, Wang P. Donor/acceptor indenoperylene
dye for highly efficient organic dye-sensitized solar cells. J Am Chem Soc 2015;
137:3799–802.
[35] Ram BA, Sandeep BM, Chang GF, Chen-Hsiung H. Effects of number and position of
meta and para carboxyphenyl groups of zinc porphyrins in dye-sensitized solar
cells: structure-performance relationship. ACS Appl Mater Interfaces 2015;7:
1879–91.
[8] Li CT, Wu FL, Liang CJ, Ho KC, Lin JT. Effective suppression of interfacial charge
recombination by a 12-crown-4 substituent on a double-anchored organic
sensitizer and rotating disk electrochemical evidence. J Mater Chem 2017;5:
7586–94.
[36] Aaron SH, Chandra BKC, Habtom BG, Lindsey RS, Francis D’S. Porphyrin-
sensitized solar cells: effect of carboxyl anchor group orientation on the cell
performance. ACS Appl Mater Interfaces 2013;5:5314–23.
[9] Erik G, Hanna E, Sandra F, Haining T, Gerrit B, Anders H, et al. Convergent/
divergent synthesis of a linker-varied series of dyes for dye-sensitized solar cells
based on the D35 donor. Adv. Energy Mater 2013;3:1647–56.
[37] Jonathan R, Dorothy C, Anders H, Elena G. Tetrachelate porphyrin chromophores
for metal oxide semiconductor sensitization: effect of the spacer length and
anchoring group position. J Am Chem Soc 2007;129:4655–65.
[38] Si LP, He HS. Porphyrin dyes on TiO2 surfaces with different orientations: a
photophysical, photovoltaic, and theoretical investigation. J Phys Chem A 2014;
118(19):3410–8.
€
[10] Stefan H, Magdalena M, Amaresh M, Mateusz W, Joel T, Jacques -EM, et al.
Significant improvement of dye-sensitized solar cell performance by small
structural modification in π-conjugated donor-acceptor dyes. Adv Funct Mater
2012;22:1291–302.
[11] Wu YZ, Zhu WH. Organic sensitizers from D-π-A to D-A-π-A: effect of the internal
electron-withdrawing units on molecular absorption, energy levels and
photovoltaic performances. Chem Soc Rev 2013;42:2039–58.
[39] Lin YD, Ching-Ting C, Lin SY, Chang HH, Liu CY, Tahsin JC. Meta versus para
substituent effect of organic dyes for sensitized solar cells. J Photochem Photobiol
Chem 2011;222:192–202.
[12] Zhang WW, Wu YZ, Zhu HB, Chai QP, Liu JC, Li H, et al. Rational molecular
engineering of indoline-based D-A-
π
-A organic sensitizers for long-wavelength-
[40] Walid S, Zeinab MH, Basant AA, Mohamed ME, Rayhan MA, Nageh KA. Position of
responsive dye-sensitized solar cells. ACS Appl Mater Interfaces 2015;7(48):
26802–10.
the anchoring group determined the sensitization efficiency of metal-free D-π-A
dyes: combined experimental and TD-DFT insights. J Photochem Photobiol Chem
[13] Qian X, Yan RC, Xu C, Shao L, Li HM, Hou LX. New efficient organic dyes
employing indeno[1,2-b]indole as the donor moiety for dye-sensitized solar cells.
J Power Sources 2016;332:103–10.
2018;367:128–36.
[41] Yao ZY, Yang L, Cai YC, Yan CC, Zhang M, Cai N, et al. Rigidifying the π-linker to
enhance light absorption of organic dye-sensitized solar cells and influences on
charge transfer dynamics. J Phys Chem C 2014;118(6):2977–86.
[14] Tan LL, Xie LJ, Shen Y, Liu JM, Xiao LM, Kuang DB, et al. Novel organic dyes
incorporating a carbazole or dendritic 3,6-diiodocarbazole unit for efficient dye-
sensitized solar cells. Dyes Pigments 2014;100:269–77.
[42] Zhang L, Yang XC, Wang WH, Gagik GG, Li JJ, Li XX. 13.6% efficient organic dye-
sensitized solar cells by minimizing energy losses of the excited state. ACS Energy
Lett 2019;4:943–51.
[15] Kimin L, Kihyung S, Youngjin K, Jaejung K. Organic sensitizers possessing
carbazole donor and indeno[1,2-b] thiophene spacer for efficient dye sensitized
solar cells. Dyes Pigments 2015;119:41–8.
[43] Liu SJ, Zhang XL, Ou CJ, Wang SL, Yang XL, Zhou XH, et al. Structure-property
study on two new D-A type materials comprising pyridazine moiety and the OLED
application as host. ACS Appl Mater Interfaces 2017;9(262):42–51.
[44] Hua Y, Chang S, Wang HD, Huang DD, Zhao JZ, Chen T, et al. New phenothiazine-
based dyes for efficient dye-sensitized solar cells: positioning effect of a donor
group on the cell performance. J Power Sources 2013;243:253–9.
[16] Zhang MD, Zhan ZY, Bao ZQ, Ju ZM, Wang XY, Zheng HG, et al. Promising alkoxy-
wrapped porphyrins with novel push–pull moieties for dye-sensitized solar cells.
J Mater Chem A 2014;2:14883–9.
[17] Wang L, Yang XC, Wang XN, Sun LC. Novel organic dyes with anchoring group of
quinoxaline-2, 3-diol and the application in dye-sensitized solar cells. Dyes
Pigments 2015;113:581–7.
[45] Buene AF, Uggerud N, Economopoulos SP, Gautun OR, Hoff BH. Effect of π-linkers
on phenothiazine sensitizers for dye-sensitized solar cells. Dyes Pigments 2018;
151:263–71.
[18] Dai XX, Feng HL, Huang ZS, Wang MJ, Wang LG, Kuang DB, et al. Synthesis of
phenothiazine-based di-anchoring dyes containing fluorene linker and their
photovoltaic performance. Dyes Pigments 2015;114:47–54.
[46] Zhu TJ, Mu DH, Qin HL, Song J, Mi BX, Zhao XY, et al. Benefits of dispersion
solvents with more OH groups in electrospray preparation of TiO2 photoelectrode
for the improvement of DSSC performance. Org Electron 2014;15:969–76.
[47] Liu J, Fu X, Cao DP, Mao L, Wang J, Mu DH, et al. Stacked graphene-TiO2
photoanode via electrospray deposition for highly efficient dye-sensitized solar
cells. Org Electron 2015;23:158–63.
[19] Huang ZS, Herbert M, Cao DR. Phenothiazine-based dyes for efficient dye-
sensitized solar cells. J Mater Chem C 2016;4:2404–26.
[20] Xie YS, Wu WJ, Zhu HB, Liu JC, Zhang WW, Tian H, et al. Unprecedentedly
targeted customization of molecular energy levels with auxiliary-groups in organic
solar cell sensitizers. Chem Sci 2016;7:544–9.
[48] Wang C, Zhang XY, Cao DP, Yin HM, Li XY, Cheng PS, et al. In situ preparation of
hierarchically structured dual-layer TiO2 films by E-spray method for efficient dye-
sensitized solar cells. Org Electron 2017;49:135–41.
[21] Zhu HB, Liu B, Liu JC, Zhang WW, Zhu WH. D-A-π-A featured sensitizers by
modification of auxiliary acceptor for preventing “trade-off” effect. J Mater Chem C
2015;3:6882–90.
[49] Vellimalai P, Munavvar Fairoos Mele K, Ambarish Kumar S, Jayaraj N. π-Extended
[22] Audun Formo B, Eline Ekornhol O, Ane Garborg Z, Anders H, Bard Helge H.
Auxiliary donors for phenothiazine sensitizers for dye-sensitized solar cells-how
important are they really? J Mater Chem A 2019;7:7581–90.
[23] Jiao YF, Mao L, Liu SS, Tan TW, Wang DL, Cao DP, et al. Effects of meta or para
connected organic dyes for dye-sensitized solar cell. Dyes Pigments 2018;158:
165–74.
cis-configured unsymmetrical squaraine dyes for dye-sensitized solar cells:
panchromatic response. ACS Appl Energy Mater 2019;2(12):8464–72.
[50] Manik Chandra S, Vediappan S, Munavvar Fairoos Mele K, Vellimalai P, Jayaraj N.
Orthogonally functionalized donor/acceptor homo- and heterodimeric dyes for
dye-sensitized solar cells: an approach to introduce panchromaticity and control
the charge recombination. ACS Appl Mater Interfaces 2017;9(40):34875–90.
[51] Zhang L, Jacqueline MC, Paul GW, Kian SL, Liu XG. Relating electron donor and
carboxylic acid anchoring substitution effects in azo dyes to dye-sensitized solar
cell performance. ACS Sustainable Chem Eng 2013;1:1440–52.
[24] Gao HH, Qian X, Chang WY, Wang SS, Zhu YZ, Zheng JY. Oligothiophene-linked D-
π
-A type phenothiazine dyes for dye-sensitized solar cells. J Power Sources 2016;
307:866–74.
[25] Ambika P, Tina T, Justin T KR, Miao-Syuan F, Kuo-Chuan H. Fine tuning the
absorption and photovoltaic properties of benzothiadiazole dyes by donor-acceptor
interaction alternation via methyl position. Electrochim Acta 2019;304:1–10.
[26] The effect of furan linkers on the properties of cyclic thiourea functionalized
triphenylamine dye sensitizers. Dyes Pigments 2017;139:772–8.
[27] Ren YM, Liu J, Zheng AB, Dong XD, Wang P. 2H-Dinaphthopentacene: a polycyclic
aromatic hydrocarbon core for metal-free organic sensitizers in efficient dye-
sensitized solar cells. Adv Sci 2017;4:1700099–117.
[52] Liu XG, Jacqueline MC, Paul GW, Tze-Chia L, Jignesh R, Anita Z. Molecular origins
of optoelectronic properties in coumarin dyes: toward designer solar cell and laser
applications. J Phys Chem A 2012;116:727–37.
[53] Yu Kyung E, Sung Ho K, In Taek C, Youngjun Y, Jeongho K, Hwan Kyu K.
Significant light absorption enhancement by a single heterocyclic unit change in
the π-bridge moiety from thieno[3,2-b]benzothiophene to thieno[3,2-b]indole for
high performance dye-sensitized and tandem solar cells. J Mater Chem A 2017;5:
2297–308.
[28] Han ML, Zhu YZ, Liu S, Liu QL, Ye D, Wang B, et al. The improved photovoltaic
performance of phenothiazine-dithienopyrrole based dyes with auxiliary acceptors.
J Power Sources 2018;387:117–25.
[54] John M, Martin JS. Assignment of the optical spectrum of metal porphyrin and
phthalocyanine radical anions. J Porphyr Phthalocyanines 2001;5:67–76.
[55] Lin LY, Min-Hsin Y, Chuan Pei L, Jeffery C, Abhishek B, Vittala R, et al. Insights
into the co-sensitizer adsorption kinetics for complementary organic dye-sensitized
solar cells. J Power Sources 2014;247:906–14.
[29] Wen WJ, Yang JB, Hua JL, Tang J, Zhang L, Long YT, et al. Efficient and stable dye-
sensitized solar cells based on phenothiazine sensitizers with thiophene units.
J Mater Chem 2010;20:1772–9.
€
[56] Yang JB, Paramaguru G, Joel T, Thomas M, Yong Joo K, Yi CY, et al. Influence of
[30] Tian HN, Yang XC, Chen RK, Pan YZ, Li L, Anders H, et al. Phenothiazine
derivatives for efficient organic dye-sensitized solar cells. Chem Commun 2007;43:
3741–3.
the donor size in D-π-A organic dyes for dye-sensitized solar cells. J Am Chem Soc
2014;136(15):5722–30.
[57] Se Hun K, Jun C, Chun S, Jin Woong N, Woosung L, Dong Hoe K, et al. The effect of
the number, position, and shape of methoxy groups in triphenylamine donors on
the performance of dye-sensitized solar cells. Dyes Pigments 2015;113:390–401.
[31] Hu Y, Chang S, Huang DD, Zhou X, Zhu XJ, Zhao JZ, et al. Significant improvement
of dye-sensitized solar cell performance using simple phenothiazine-based dyes.
Chem Mater 2013;25:2146–53.
�
�
[58] Liu YH, Cao YM, Zhang WW, Marko S, Ibrahim D M, Pechy Peter, et al. Electron-
[32] Bhanumathi N, Suman K, Ramachandran E, Sudip M, Kothandaraman R,
Dhamodharan R. Novel ethynyl-pyrene substituted phenothiazine based metal free
organic dyes in DSSC with 12 % conversion efficiency. J Mater Chem A 2017;5:
10289–300.
affinity-triggered variations on the optical and electrical properties of dye
[33] Tan LL, Huang JF, Shen Y, Xiao LM, Jun-Min Liu, Kuang DB. Highly efficient and
stable organic sensitizers with duplex starburst triphenylamine and carbazole
10

S. Liu et al.
Dyes and Pigments 180 (2020) 108470
molecules enabling highly efficient dye-sensitized solar cells. Angew Chem Int Ed
organic sensitizers with over 10.47% efficiency for dye-sensitized solar cells. Adv.
Energy Mater. 2015;5:1500300–8.
2018;57:14125–8.
[59] Yu Kyung E, In Taek C, Sung Ho K, Joori L, Jeongho K, Myung Jong J, et al. Thieno
[60] Yao ZY, Guo YX, Wang LQ, Hao Y, Guo Y, Daniele F, et al. Energy-loss reduction as
a strategy to improve the efficiency of dye-sensitized solar cells. Sol. RRL 2019;3:
1900253–62.
[3,2-b][1]benzothiophene derivative as a new π-bridge unit in D-π-A structural
11