Heteroatom-mediated performance of dye-sensitized solar cells based on T-shaped molecules

Article abstract of DOI:10.1016/j.dyepig.2019.02.055

Two novel “sister” organic dyes featuring a T-shape molecular pattern based on phenothiazine/phenoxazine and triphenylamine framework as electron donor core and cyanoacrylic acid as acceptor and anchoring group were designed, synthesized and thoroughly ch

Full text of DOI:10.1016/j.dyepig.2019.02.055

Accepted Manuscript
Heteroatom-mediated performance of dye-sensitized solar cells based on T-Shaped
molecules
Mariana-Dana Damaceanu, Catalin-Paul Constantin, Andra-Elena Bejan, Mihai
Mihaila, Mihaela Kusko, Cristian Diaconu, Iuliana Mihalache, Razvan Pascu
PII:
S0143-7208(18)32033-3
https://doi.org/10.1016/j.dyepig.2019.02.055
DYPI 7389
DOI:
Reference:
To appear in: Dyes and Pigments
Received Date: 13 September 2018
Revised Date: 26 February 2019
Accepted Date: 28 February 2019
Please cite this article as: Damaceanu M-D, Constantin C-P, Bejan A-E, Mihaila M, Kusko M, Diaconu
C, Mihalache I, Pascu R, Heteroatom-mediated performance of dye-sensitized solar cells based on T-
Shaped molecules, Dyes and Pigments (2019), doi: https://doi.org/10.1016/j.dyepig.2019.02.055.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
Heteroatom-Mediated Performance of Dye-Sensitized Solar Cells
based on T-Shaped Molecules
1
*
1
1
2
Mariana-Dana Damaceanu , Catalin-Paul Constantin , Andra-Elena Bejan , Mihai Mihaila ,
2
3
2
2
Mihaela Kusko , Cristian Diaconu , Iuliana Mihalache and Razvan Pascu
1
“
Petru Poni” Institute of Macromolecular Chemistry, Polycondensation and Thermostable
Polymers Department, Aleea Grigore Ghica Voda 41A, Iasi, 700487, Romania
2
National Institute for Research and Development in Microtechnologies (IMT), 126A, Erou
Iancu Nicolae Street, 077190, Bucharest, Romania
3
Honeywell Romania, Sensors and Wireless Laboratory, 3, George Constantinescu, BOC
Tower, 020339 Bucharest, Romania
*
To whom correspondence should be addressed: e-mail: damaceanu@icmpp.ro
1

ACCEPTED MANUSCRIPT
Abstract
Two novel “sister” organic dyes featuring a T-shape molecular pattern based on
phenothiazine/phenoxazine and triphenylamine framework as electron donor core and
cyanoacrylic acid as acceptor and anchoring group were designed, synthesized and thoroughly
characterized. Both dyes were functionalized at the nitrogen atom of
phenoxazine/phenothiazine with the triphenylamine unit containing two hexyloxy substituents
to obtain a T-shape configuration. The photo-optical and electrochemical data corroborated
with time-dependent density functional theory computations and photovoltaic features were
discussed in correlation with the structural architecture of each dye. Their performances in
dye-sensitized solar cells (DSSCs) were comparatively surveyed with a special concern on the
influence of the heteroatom variation on the most important DSSCs characteristics. In spite of
its more favorable optical performance, the cells realized with phenoxazine-based dye are less
efficient than those fabricated with phenothiazine ones. The cause is the stronger electron
interception by the electrolyte in the case of phenoxazine dye which consistently degrades the
open-circuit voltage, corroborated with a better regeneration of the phenothiazine by the
iodine/iodide redox couple. In fact, to the best of our knowledge, this is one of the first studies
which highlights how a single heteroatom in a T-shape dye structure can affect the dye
interaction with the electrolyte and its regeneration, consequently the photovoltaic
performances and brings out new opportunities for developing novel dyes for efficient
DSSCs.
Keywords: phenoxazine, phenothiazine, T-shape structure, dye-sensitized solar cells,
heteroatom-dependent dark current
2

ACCEPTED MANUSCRIPT
1
. INTRODUCTION
Dye-sensitized solar cells (DSSCs) as the most viable alternatives to crystalline Si-based
photovoltaics for converting clean, unlimited sunlight to electricity have attracted significant
research interest due to their low fabrication cost and relatively high power conversion
1
efficiency (PCE) since the first cell reported by Grätzel in 1991. In almost three decades this
type of solar cell has shown much promise in the solar energy sector, the most glaring benefit
2
,3
being the ability to overcome the so-called Shockley-Queisser limit through applying a
smart design. The advantageous traits are not limited to the potential efficiency, but also
include the facility of large-scale manufacture and low impurity sensitivity, besides the
aesthetically pleasing. Two of the most important physical parameters that contribute to the
overall efficiency of a DSSC are the open circuit voltage (V ) and short circuit current (I ).
oc
sc
In separately p- and n-type DSSCs, since V is the difference between the Fermi-level of the
oc
semiconducting oxide and the redox potential of the electrolyte, its value can be modulated be
4
using various electrolytes and/or different semiconductors. On the other hand, the maximum
current delivered by a DSSC is limited by the transport resistance and various recombination
pathways; all these can be prevented via intelligent dye design approach to address the
intermolecular quenching due to aggregation on the semiconductor surface and photo-current
limitations through device optimization fabrication.
DSSCs are typically composed of a mesoporous semiconductor metal oxide film, a
5
dye, an electrolyte as hole transporter, and a counter electrode. The sensitizer is crucial in the
DSSCs structure, because it absorbs sunlight and injects the electron into the TiO2
semiconductor. Most organic sensitizers have a rod-shape structure, which tend to aggregate
on the semiconductor surface, lessening the photovoltaic effect due to the intermolecular
quenching: the molecules are not functionally attached to the TiO surface and thus behave as
2
6
filters. It is worth knowing that a phenoxazine or phenothiazine-based dye contains electron-
3

ACCEPTED MANUSCRIPT
rich nitrogen and oxygen/sulphur heteroatoms in the heterocyclic structure, with high
electron-donating ability. The nonplanar butterfly conformation of phenoxazine/phenothiazine
with two phenyl rings arranged in a small torsion angle related to N(10) and O/S(9) atoms can
lead to a sufficiently inhibition of molecular aggregation and formation of intermolecular
7
,8
excimers. Numerous attempts have been made to molecularly engineer phenoxazine or
phenothiazine dyes in order to broaden the absorption band and increase the efficiency. When
designing the dye, high molar extinction coefficients are of particular importance, because
they reduce parasitic light absorption. One of the strategies to achieve better performances is
to introduce an additional electron donor group into the core of phenoxazine or phenothiazine
sensitizer to form a D-D-π-A structure. In this way, the π- delocalization can be extended and
9
,10
the molar absorptivity of the materials significantly enhanced.
Arylamine derivatives
(carbazole, triphenylamine, indoline and their derivatives) are well known electron-rich
structures that are widely incorporated in phenoxazine- or phenothiazine-based dyes as
additional electron donor moiety. For example, the insertion of thiophene and triphenylamine
units into some phenothiazine sensitizers extended the π-conjugation leading to a larger molar
absorptivity. Meanwhile, this design makes the whole molecule to expand on larger surface
area compared to a molecule of linear shape, leading to the reduction of the dye loading
1
1
capacity. Besides, the red-shift of their absorption spectra and an enhanced efficiency were
1
2
achieved. This strategy avoids the charge recombination process of injected electrons with
the triiodide in the electrolyte and also minimizes the formation of aggregates between dye
molecules. A thorough literature survey evidenced that the vast majority of developed
phenoxazine/phenothiazine sensitizers having a D-D-π-A topology contains the second donor
group connected in most cases to the (C)7 position of these cores (rod-like dyes), and only
few studies are reported so far in which the second donor is directly linked to the (N)10 atom
1
3
of phenoxazine/phenothiazine (T-shape dyes). The optical features indicated that rod-like
4

ACCEPTED MANUSCRIPT
dyes displayed broader and red-shifted absorption waves compared to the T-shaped dyes due
to longer conjugation. But T-shaped dyes were superior because of higher recombination
resistance, lower charge transport resistance, better electron lifetime, and good electron
injection which led to high efficiency due to the favorable VOC and J values.
SC
Following the issues outlined above, it appeared challenging to us to design and
synthesize two novel DD-π-A structurally related phenoxazine/phenothiazine- and
triphenylamine-based dyes and to survey comparatively their physico-chemical properties and
performances as sensitizers in DSSCs. The conceptual design consists in the direct bonding of
triphenylamine core to the (N)10 position and of the cyanoacrylate moiety to the C(3) position
of phenoxazine/phenothiazine, resulting in a T-shaped structural motif. Two additional
hexyloxy chains were attached in the para position of triphenylamine unit in order to improve
its donating capability, suppress the dyes aggregation and minimize the recombination. It is
expected this molecular architecture to provide a satisfied relay for the electron migration
from donor to acceptor and better photovoltaic performances when compared to the rod-
shaped dyes.
2
. EXPERIMENTAL SECTION
1
4,15
Details about starting materials and 1-(hexyloxy)-4-iodobenzene synthesis
are available in
the Supporting Information (SI).
2
.1. Synthesis
The synthetic pathway to phenoxazine/phenothiazine-triphenylamine based dyes (PTH-TPA
and PHX-TPA) is outlined in Scheme 1 and the details are described as follows. All reactions
were performed under nitrogen atmosphere with the use of standard Schlenk technique.
5

ACCEPTED MANUSCRIPT
Scheme 1. Synthetic route to the T-shaped dyes based on phenothiazine or phenoxazine
N-(4-nitrophenyl)phenoxazine (1a). In a 100 mL three-necked round-bottomed flask
equipped with a magnetic stirring bar, nitrogen inlet and outlet and condenser, 4-fluoro-
nitrobenzene (2.3 g, 16.4 mmol), 10H-phenoxazine (2 g, 10.1 mmol), potassium carbonate
(
2.26 g, 16.4 mmol) and DMSO (30 mL) were placed. The reaction mixture was refluxed at
20°C overnight (21 h) when the solution became dark-red. The reaction evolution was
monitored by thin layer chromatography (TLC) using hexane/ethyl acetate, 30/1, v/v. When
0H-phenoxazine was no longer present, the reaction mixture was allowed to cool to room
1
1
temperature and slowly poured into 150 mL of stirred methanol when an orange-red solid
started to precipitate. It was collected by filtration and washed thoroughly with
methanol/water. The resulting nitro compound 1a was dried under vacuum to obtain 2.15 g of
orange-red powder (70% yield).
1
H-NMR (DMSO, 400 MHz), d (ppm): 8.49–8.47 (d, 2H), 7.77–7.75 (d, 2H), 6.83–6.70 (m,
6
H), 6.05–6.05 (d, 2H).
6

ACCEPTED MANUSCRIPT
1
3
C-NMR (DMSO, 400 MHz), δ (ppm): 146.71, 144.97, 143.52, 132.82, 131.37, 126.51,
1
23.76, 122.38, 115.66, 113.94.
N-(4-nitrophenyl)phenothiazine (1b) was synthesized according to the procedure reported
for 1a to afford 2.3 g of yellow powder (73% yield).
1
H-NMR (DMSO, 400 MHz), d (ppm): 8.14–8.12 (d, 2H), 7.65–7.63 (d, 2H), 7.60–7.58 (d,
2
H), 7.52–7.48 (t, 2H) 7.39–7.35 (t, 2H), 7.07–7.05 (d, 2H).
C-NMR (DMSO, 400 MHz), δ (ppm): 150.87, 140.66, 140.18, 133.35, 129.82, 129.32,
28.37, 127.348, 127.24, 126.04.
1
3
1
N-(4-aminophenyl)phenoxazine (2a). In a 250 mL three-necked round-bottomed flask
equipped with a magnetic stirring bar, nitrogen inlet and outlet and condenser was added the
dinitro-derivative 1a (1.8 g, 6 mmol) and a mixture of ethanol and water (120 mL),
EtOH/H O, 20/1, v/v. To the resulting suspension, iron powder (1 mg, 18 mmol) and CaCl2
2
(0.66 g, 6 mmol) were added and stirred at 78°C for 24 h. The reaction evolution was
monitored by TLC using hexane/ethylacetate, 10/1, v/v. After completion, the reaction
mixture was filtered on celite to remove the iron residues, and to the alcoholic solution ice
flakes were added when a white-grey solid started to precipitate. The solid was collected by
filtration and washed thoroughly with water. The resulting amine 2a was dried under vacuum
to obtain 1.40 g of white-grey powder (86% yield).
1
H-NMR (DMSO, 400 MHz), d (ppm): 6.89–6.96 (d, 2H), 6.78–6.76 (d, 2H), 6.69–6.61 (m,
6
H), 5.93–5.91 (d, 2H), 5.44 (s, 2H).
1
3
C-NMR (DMSO, 400 MHz), δ (ppm): 148.90, 143.16, 134.71, 130.44, 125.44, 123.56,
1
20.85, 115.67, 114.94, 113.06.
1
6
N-(4-aminophenyl)phenothiazine (2b) was synthesized following a reported procedure,
using Pd/C and hydrazine monohydrate in ethanol at reflux. Accordingly, 3.8 g of white-grey
7

ACCEPTED MANUSCRIPT
solid (90% yield) were obtained when starting from 2.3 g (14.5 mmol) of N-(4-
nitrophenyl)phenothiazine.
1
H-NMR (DMSO, 400 MHz), d (ppm): 7.01–6.99 (m, 4H), 6.93–6.88 (t, 2H), 6.82–6.77 (m,
4
H), 6.22–6.20 (d, 2H), 5.43 (s, 2H).
1
3
C-NMR (DMSO, 400 MHz), δ (ppm): 148.81, 144.42, 130.91, 127.58, 127.04, 126.24,
1
22.045, 118.37, 115.43, 115.37.
B
is(4-hexyloxy-phenyl)-(4-phenoxazin-10-yl-phenyl)-amine (3a). In a 100 mL three-
necked round-bottomed flask equipped with a magnetic stirring bar, nitrogen inlet and outlet
and condenser was added potassium hydroxide (1.93 g, 34.56 mmol) and copper (I) iodide
(
0.17 g, 0.86 mmol) to a stirred solution of amine 2a (1.2 g, 4.32 mmol), 1-(hexyloxy)-4-
iodobenzene (3.95 g, 12.96 mmol), and 1,10-phenanthroline (0.16 g, 0.86 mmol) in toluene
50 ml). The reaction mixture was heated under reflux for 24 h at 120°C, when the solution
(
became dark red. After that, the reaction mixture was allowed to cool to room temperature
and filtered to remove the insoluble solids. The resulting mixture was poured into 250 mL
water. The crude product was extracted three times with ethyl acetate, and the combined
organic layer was washed with water and dried over anhydrous sodium sulphate. After
removing the solvent under reduced pressure, the oil residue was purified by column
chromatography using a mixture of hexane and ethyl acetate, 99/1, v/v to obtain 0.76 g of
light-yellow oil (30% yield).
1
H-NMR (DMSO, 400 MHz), d (ppm): 7.14–7.12 (m, 6H), 6.96–6.94 (d, 4H), 6.91–6.88 (d,
2
H), 6.71–6.64 (m, 6H), 5.97–5.95 (d, 2H), 3.97–3.93 (t, 4H), 1.73–169 (m, 4H), 1.40–1.44
(m, 4H), 1.32–1.31 (m, 8H), 0.91–0.87 (t, 6H).
1
3
C-NMR (DMSO, 400 MHz), δ (ppm): 155.83, 143.24, 139.48, 138.02, 134.39, 130.82,
1
2
27.56, 123.78, 121.32, 119.87, 117.37, 115.72, 115.23, 113.26, 67.75, 31.04, 28.75, 25.26,
2.12, 13.96.
8

ACCEPTED MANUSCRIPT
Bis(4-hexyloxy-phenyl)-(4-phenothiazin-10-yl-phenyl)-amine (3b) was synthesized
according to the procedure described for 3a to afford 0.81 g of light-yellow oil (34% yield).
1
H-NMR (DMSO, 400 MHz), d (ppm): 7.18–7.16 (m, 6H), 7.04–6.92 (d, 2H), 6.97–6.90 (m,
8
H), 6.85–6.82 (t, 2H), 6.26–6.24 (d, 2H), 3.97–3.94 (t, 4H), 1.73–169 (m, 4H), 1.44–1.40
(m, 4H), 1.33–1.31 (m, 8H), 0.90–0.87 (t, 6H).
1
3
C-NMR (DMSO, 400 MHz), δ (ppm): 156.98, 145.38, 140.63, 131.98, 128.14, 127.67,
1
2
1
27.08, 123.02, 122.98, 120.73, 120.16, 116.40, 116.12, 68.53, 32.13, 29.65, 26.28, 23.10,
2.12, 14.13.
0-(4-(Bis(4-(hexyloxy)phenyl)amino)phenyl)phenoxazinyl-3-carbaldehyde
(4a).
A
Schlenk tube equipped with a magnetic stirring bar, nitrogen inlet and outlet and condenser,
containing anhydrous DMF (0.47 mL, 5.5 mmol) was cooled in an ice bath (0–5 °C). To this
solution, phosphorous oxychloride (0.65 mL, 5.5 mmol) was added dropwise for 30 min,
when a slightly opaque solid started to appear. 10 mL of 1,2-dichloroethane was added to the
formed Vilsmeier reactive and the stirring was continued for 1 hour at 0–5 °C, when a white
opalescent solution was obtained. The triphenylamine-substituted phenoxazine 3a (0.7 g, 1.1
mmol) in 10 mL of 1,2-dichloroethane was added to the above solution and heated to 90 °C
for 24 h. The occurrence of the newly formed aldehyde was monitored by TLC using
hexane:ethylacetate, 10/1, v/v. Upon completion, the reaction mixture was allowed to cool to
room temperature, poured into ice water and neutralized to pH 6–7 by dropping saturated
aqueous sodium hydroxide solution. The mixture was extracted with 250 mL ethyl acetate.
The organic layer was washed three times with water and dried with anhydrous sodium
sulphate, followed by concentration under reduced pressure. The aldehyde was purified by
column chromatography to obtain 0.55 g of yellow-greenish solid (73% yield).
1
H-NMR (DMSO, 400 MHz), δ (ppm): 9.64 (s, 1H,), 7.29–7.27 (d, 1H), 718–7.15 (m, 6H),
7
.09 (s, 1H), 6.97–6.95 (d, 4H), 6.91–6.89 (d, 2H), 6.75–6.73 (m, 3H), 6.09–6.07 (d, 1H),
9

ACCEPTED MANUSCRIPT
6
.00–5.99 (d, 1H), 3.97–3.94 (t, 4H), 1.73–1.67 (m, 4H), 1.44–1.40 (m, 4H), 1.36–1.31 (m,
8
H), 0.90–0.87 (t, 6H).
1
3
C-NMR (DMSO, 400 MHz), δ (ppm): 189.99, 155.86, 148.92, 143.39, 142.93, 139.89,
39.11, 132.69, 130.27, 129.85, 128.22, 127.65, 127.39, 123.96, 122.67, 119.32, 115.59,
15.31, 113.96, 113.78, 112.74, 67.59, 30.96, 28.65, 25.17, 22.04, 13.88.
0-(4-(Bis(4-(hexyloxy)phenyl)amino)phenyl)phenothiazinyl-3-carbaldehyde (4b) was
1
1
1
synthesized according with the procedure detailed for 4a to obtain 0.53 g of yellow-greenish
solid (72% yield).
1
H-NMR (DMSO, 400 MHz), δ (ppm): 9.7 (s, 1H), 7.49 (s, 1H), 7.47–7.45 (d, 1H), 7.20–7.18
(d, 6H), 7.06–7.04 (d, 1H), 6.98–6.96 (m, 5H), 6.92–6.90 (m, 3H), 6.33–6.31 (d, 1H), 6.24–
6
0
.22 (d, 1H), 3.97–3.94 (t, 4H), 1.73–1.67 (m, 4H), 1.44–1.40 (m, 4H), 1.34–1.30 (m, 8H),
.90–0.86/ (t, 6H).
1
3
C-NMR (DMSO, 400 MHz), δ (ppm): 190.70, 156.44, 149.34, 149.30, 142.94, 139.64,
31.27, 130.97, 130.38, 130.29, 128.27, 128.07, 127.71, 127.20, 127.05, 124.14, 119.51,
19.36, 118.46, 116.91, 116.16, 115.67, 68.16, 31.47, 29.17, 25.68, 22.55, 14.38.
-(10-(4-(Bis-(4-hexyloxy-phenyl)-amino)-phenyl)-10H-phenoxazin-3-yl)-2-cyanoacrylic
1
1
3
acid (PHX-TPA). In a Schlenk tube, 4a (0.5 g, 0.76 mmol) dissolved in 4 mL chloroform and
cyanoacetic acid (0.12 g, 2.25 mmol) were placed. This mixture was stirred for 10 minutes
under nitrogen stream, and then piperidine (0.45 g 5.33 mmol) dissolved in 2 mL chloroform
was added dropwise to the mixture. The reaction mixture was kept under stirring and gentle
reflux (65 °C) for 12 h. The reaction was quenched with 60 mL HCl 2 M, when the colour
turned from orange-red to dark-red. The organic phase was extracted three times with DCM
and washed three times with water as the last washing water to have pH = 5.5. After drying
over sodium sulphate and solvent evaporation, the residue was purified by column
1
0

ACCEPTED MANUSCRIPT
chromatography using successively DCM and DCM/methanol 10/1, v/v, to afford 0.33 g of
dark-red powder (66 % yield).
1
H-NMR (DMSO, 400 MHz), δ (ppm): 13.14 (s, 1H, H–1), 7.89 (s, 1H, H–6), 7.46 (s, 1H, H–
8
6
2
1
), 7.30–7.28 (d, 1H, H–18), 7.19–7.15 (m, 6H, H–21,25), 6.97–6.94 (d, 4H, H–24), 6.91–
.89 (d, 2H, H–20), 6.81–6.79 (m, 1H, H–11), 6.75–6.72 (m, 2H, H–12,13), 6.04–6.02 (m,
H, H–14,17), 3.97–3.93 (t, 4H, H–27), 1.74–1.67 (m, 4H, H–28), 1.44–1.40 (m, 4H, H–29),
.32–1.30 (m, 8H, H–30,31), 0.90–0.87 (t, 6H, H–32).
1
3
C-NMR (DMSO, 400 MHz), δ (ppm): 163.39 (C–2), 155.79 (C–26), 148.90 (C–5), 143.02
(C–23), 139.19 (C–19,22), 138.05 (C–9), 132.64 (C–10), 130.21(C–15,16), 129.37 (C–18),
1
2
27.52 (C–24), 127.40 (C–7), 125.03 (C–12), 123.91 (C–13), 122.54 (C–17), 119.51 (C–
0,21), 117.99 (C–11), 115.58 (C–25), 115.34 (C–4), 114.55 (C–14), 113.85 (C–8), 112.85
(C–3), 67.59 (C–27), 30.92 (C–28), 28.63 (C–29), 25.13 (C–30), 21.99 (C–31), 13.83 (C–32).
Elemental analysis calc. for C H N O (M = 721) (%): С 76.56; Н 6.52; N 5.83. Found, %:
4
6
47
3
5
C 76.50; H 6.58; N 5.75.
-(10-(4-(Bis-(4-hexyloxy-phenyl)-amino)-phenyl)-10H-phenothiazin-3-yl)-2-
3
cyanoacrylic acid (PTH-TPA) was synthesized according to the procedure described for 5a
to obtain 0.39 g of red powder (69 % yield).
1
1

ACCEPTED MANUSCRIPT
1
H-NMR (DMSO, 400 MHz), δ (ppm): 13.15 (s, 1H, H–1), 7.81 (s, 1H, H–6), 7.63 (s, 1H, H–
8
), 7.51–7.49 (d, 1H, H–18), 7.20–7.17 (d, 6H, H–21,25), 7.06–7.04 (d, 1H, H–11), 6.98–6.96
(m, 5H, H–12,24), 6.92–6.86 (m, 3H, H–13,20), 6.25–6.22 (m, 2H, H–14,17), 3.97–3.94 (t,
4
0
H, H–27), 1.74–1.68 (m, 4H, H–28), 1.44-1.40 (m, 4H, H–29), 1.30–1.34 (m, 8H, H–30,31),
.90–0.87 (t, 6H, H–32).
1
3
C-NMR (DMSO, 400 MHz), δ (ppm): 163.38 (C–2), 156.38 (C–26), 149.25 (C–5), 143.16
(C–19,22), 139.72 (C–23), 131.30 (C–15,16), 130.66 (C–7), 130.38 (C–18), 128.20 (C–24),
1
2
28.01 (C–13), 127.68 (C–8), 127.33 (C–11), 127.06 (C–9), 123.80 (C–10), 119.67 (C–
0,21), 119.21 (C–14), 118.40 (C–12), 116.61 (C–17), 116.03 (C–4), 116.15 (C–25), 115.79
(C–3), 68.15 (C–27), 31.47 (C–28), 29.17 (C–29), 25.68 (C–30), 22.55 (C–31), 14.38 (C–32).
Elemental analysis calc. for C H N O S (M = 737) (%): С 74.90; Н 6.38; N 5.70; S 4.34.
4
6
47
3
4
Found, %: C 74.83; H 6.42; N 5.78; S 4.28.
2
.2. Measurements
1
13
The H- and C-NMR spectra were registered with a Bruker Avance III 400
spectrometer provided with a 5 mm multinuclear inverse detection probe, operating at 400.1
1
13
1
13
and 100.6 MHz for H and C nuclei, respectively. H and C chemical shifts are afforded in
δ units (ppm) relative to the residual peak of the solvent. All column chromatography
purifications were performed with the use of silica gel 60 (0.063-0.2 mm) as the stationary
phase in columns of different lengths and diameters. The infrared (FTIR) spectra were
1
2

ACCEPTED MANUSCRIPT
recorded on a FTIR Bruker Vertex 70 Spectrophotometer in ATR mode by using samples in
form of powder. Electronic absorption spectra were obtained on a Specord M42 UV-Visible
spectrophotometer at room temperature using freshly prepared solutions at low concentration
−
5
(
≈ 10 M) in various solvents. The absorption spectra of the dye-loaded TiO films were
2
obtained by diffuse reflectance spectroscopy (DRS) using Eding. Inst. FLS920
spectrofluorometer. Photoluminescence (PL) spectra of dye solutions were acquired with a
Perkin Elmer LS 55 instrument. The PL decay profiles were measured using time-correlated
single-photon counting (TCSPC) and pulsed laser excitation (5 mW at 375nm). Cyclic
voltammetric experiments were carried out at room temperature on a Potentiostat-Galvanostat
(PG581, Uniscan Instruments) with a conventional three-electrode configuration consisting of
ITO-coated glass covered with the dye or platinum as working electrode, a platinum wire
auxiliary electrode, and a calomel (SCE) reference electrode. Tetrabutylammonium
perchlorate (TBAP)/acetonitrile (CH CN) was used as supporting electrolyte for the dye
3
coating samples and TBAP/dichloromethane for solution samples having the concentration
−
3
of 1×10 M. All potentials were reported with respect to SCE electrode at a scan rate of 50
mV/s. The potentials are quoted against the ferrocene internal standard (+0.45 vs SCE). The
current-voltage (I-V) characteristics of the cells were measured with Keithley 4200-SCS
semiconductor characterization system, under dark and illumination conditions, using a LED
solar simulator, class ABA (model LSH 7320, Newport, USA). The measurement of incident
photon to current conversion efficiency - IPCE (λ) - spectra was performed by incrementing
the F920 calibrated monocromator over the 300-900 nm wavelength range, and the short-
2
circuit photocurrent density (mA/cm ) was recorded with PGSTAT302N potentiostat. The
electrochemical impedance spectroscopy (EIS) measurements of the cells were recorded with
2
Autolab/PGSTAT302N electrochemical workstation, under 100 mW/cm light illumination,
in the frequency range from 50 mHz to 100 kHz, with logarithmic point spacing, at open
1
3

ACCEPTED MANUSCRIPT
circuit potential (OCP) bias, using an alternating low signal with 10 mV amplitude. The
impedance spectra were analyzed on the basis of the electrical equivalent circuit
Rs(Q1R1)(Q2R2)(Q3R3), using the ZSimpWin software, which takes into consideration the
phenomena encountered at the Pt counterelectrode, sensitized TiO photoanode and within the
2
electrolyte, respectively. In order to compensate the non-ideal behavior, the constant phase
elements (Q) replace the simple capacitances in the model. Accordingly, it allows the
extraction of series resistance (R ), charge transfer resistance at Pt/electrolyte interface (R ,
s
1
R ) and at TiO /dye/electrolyte interface (R , R ), or chemical capacitance of the TiO film
ce
2
2
r
2
(C2).
2
.3. Fabrication of DSSCs
A 10µm thick mesoporous TiO layer, with a 4µm scattering layer deposited on it (Solaronix,
2
Switzerland), both treated in a TiCl atmosphere, was functionalized using a 0.3mM solution
4
of the dye in dichlormethane (DCM), for an immersion time of 4 or 6 hours. Thereafter, the
film, rinsed with ethanol and soaked in nitrogen flux, was encapsulated using a platinum
counter electrode and a 60µn thick Surlyn sealant (Solaronix). The cells have been
impregnated using two different electrolytes (HI-30 and AN-50, Solaronix). Both are based
¯
¯
¯
on I /I redox couple, but with different I concentrations, 30 mM (HI-30) and 50 mM (AN-
3
3
5
0), with acetonitrile as solvent and ionic liquid, lithium salt and pyridine derivative as
additives. The Surlyn sealant has been heated for two different times (2 and 3 minutes). After
the TiO film impregnation, the hole in the counter electrode was sealed by heating a Surlyn
2
2
sheet covered by a glass cover. The device area was 0.36 cm .
1
4

ACCEPTED MANUSCRIPT
2
.4. Quantum chemical simulations
In order to avoid geometry optimization difficulties due to the floppiness of the long alkyl
chains, slightly modified structural motifs of PTH-TPA and PHX-TPA were involved in
simulations by replacing the hexyl unit with the less voluminous methyl group (PTH-TPAm
and PHX-TPAm, SI, figure S1). The ground-state geometries of PTH-TPAm and PHX-TPAm
were fully optimized without any symmetry constraints using the density functional theory
1
7,18
(DFT) level with the B3LYP exchange-correlation functional
and 6-311++G** basis set
1
9-21
22
on all atoms
. This approach for geometry optimization has been noted to be robust.
The UV-Vis absorption spectra of both dyes were simulated by computing the excited
states and oscillator strengths for the lowest 30 singlet transitions using the time-dependent
(TD) DFT with the B3LYP functional and the 6-311++G** basis set using the B3LYP/6-
311++G** optimized geometry in the ground state. The simulated UV-Vis spectrum of PHX-
TPAm in vacuo is in good qualitative/semi-quantitative agreement with the experimental one,
while for PTH-TPAm the peak around 400nm is missing. UV-Vis absorption spectra of PHX-
2
3
TPAm and PTH-TPAm were also simulated using CAM-B3LYP , which is a range-
separated hybrid functional and it has been found to successfully describe charge-transfer
24,25
excited states.
However, for both molecules the whole simulated UV-Vis spectrum using
CAM-B3LYP is blue-shifted by 0.5-0.6 eV. This behaviour has been previously reported for
2
6
phenothiazine-based molecules.
All computations DFT and TD-DFT were carried out with the Gaussian 09 program in
2
7
a computer workstation; molecular structures and HOMO-LUMO orbitals figures, along
2
8
with UV-Vis spectra were attained with GaussView Version 5.
1
5

ACCEPTED MANUSCRIPT
3
. RESULTS AND DISCUSSIONS
3
.1. Molecular design and synthesis
Novel DD–π–A type organic dyes (PTH-TPA and PHX-TPA) with a T shape configuration
have been developed by inserting a subordinate donor unit in the conventional design of
metal-free organic sensitizers with the aim to facilitate the electron migration, to improve
photostability and to inhibit the dye aggregation. The developed sensitizer combines as
structural features the triphenylamine containing two hexyloxy substituents and
phenothiazine/phenoxazine as donor framework (DD), the vinylene bridge as π-linker (π), and
the cyanoacrylic acid as both anchoring group and electron acceptor (A). The long hexyl
groups on the triphenylamine core were used aiming at increasing the push strength,
1
4
suppressing the aggregation and minimizing the recombination. Having in mind the
proposed design a scenario of photovoltaic mechanism can be rationalized as follows. When
the dyes harvest sunlight, the electrons move from the triphenylamine core under the pushing
force of hexyloxy units to the phenothiazine/phenoxazine unit, being further driven through
the π-conjugated vinyl bridge to the cyanoacrylic acid acceptor by means of dye
photoexcitation. The generated photo-excited electrons are then injected into the conduction
band of TiO via the anchoring carboxyl group. Next, the dye captures an electron from the
2
redox couple and comes back to the initial state (dye regeneration), leaving a positive charge
(a hole) in the electrolyte. Since the mechanism of charge separation responsible for the
generation of the photovoltaic effect is accomplished, the structural motif of our metal-free
organic sensitizers suggests that they are suitable for use in DSSCs.
The developed structural pattern of PTH-TPA and PHX-TPA sensitizers was fully
1
13
1
proved by spectral methods, as H-, C-NMR and FTIR. In the H-NMR spectra (Figure 1)
the most shielded signal at 7.82 ppm for PTH-TPA and 7.88 ppm for PHX-TPA was
attributed to the vinyl proton H-6 of cyanoacrylic acid.
1
6

ACCEPTED MANUSCRIPT
1
Figure 1. H-NMR spectra of PTH-TPA and PHX-TPA sensitizers
PTH-TPA
PHX-TPA
1
7

ACCEPTED MANUSCRIPT
The protons belonging to the phenothiazine/phenoxazine core gave splitted patterns that are
partially overlapped with the signals corresponding to the aromatic protons of the
triphenylamine unit. The protons belonging to phenothiazine/phenoxazine in the vicinal
positions of vinylene linkage were evidence by a singlet (H-8) at 7.63 ppm for PTH-TPA and
7
.46 ppm for PHX-TPA and a doublet (H-18) at 7.51-7.49 for PTH-TPA and 7.30-7.28 ppm
for PHX-TPA. The less shifted signal in the aromatic region, at 6.26-6.23 and 6.04-6.02 ppm
for PTH-TPA and PHX-TPA, respectively, was assigned to the protons H-14 and H-17 of the
phenothiazine/phenoxazine rings. The protons belonging to the phenylene units of
triphenylamine fall at 7.20–7.17 ppm (d, H-21, H-25), 6.98–6.95 ppm (m, H-24, overlapped
signal) and 6.92–6.90 ppm (d, H-20, overlapped signal) for PTH-TPA and 7.19–7.15 ppm (m,
H-21, H-25), 6.97–6.94 ppm (d, H-24) and 6.91–6.89 ppm (d, H-20) for PHX-TPA. The
protons belonging to the hexyloxy groups were evidenced for both dyes at similar shielding.
The triplet at 3.97–3.94 ppm was associated with the CH protons linked to O of hexyloxy
2
group (H-17). The other four methylene groups were grouped in three multiplets at 1.74–1.68
ppm (H-28), 1.44–1.40 ppm (H-29) and 1.32–1.31 ppm (H-30, H-31). Finally, the protons of
1
3
the end-methyl group (H-32) appeared at 0.90–0.87 ppm as a triplet. The C-NMR spectra
evidenced the presence of carboxylic group in PTH-TPA and PHX-TPA dyes by the most
shielded signal at 163.38-163-39 ppm, while the nitrile group appeared at 116.03 and 115.34
ppm for PTH-TPA and PHX-TPA, respectively (Figure S2, SI).
FTIR spectra of PTH-TPA and PHX-TPA sensitizers showed characteristic absorption
-1
bands of the newly formed vinylene linkage at 826-818 cm , whereas the FTIR bands due to
-1
the aldehyde (approx. 1700 cm ) of the starting reagents completely disappeared, undeniable
proving the formation of cyanoacrylic acid core under the Knoevenagel condensation
conditions. The carboxylic functional group was found by the absorption bands at 3485-3465
-1
-1
cm (OH stretching, wide band) and 1718-1724 cm (C=O stretching), while CN groups was
1
8

ACCEPTED MANUSCRIPT
-1
evidenced by the absorption band at 2218-2221 cm . The IR bands at approx. 3061-
−
1
3
041 cm originate from the C–H stretching in phenylene units, while the system band at
-1
approx. 2954-2857 cm corresponds to the aliphatic C-H stretching of the hexyloxy groups.
The -C=C- stretching in phenylene and phenothiazine/phenoxazine rings falls at 1585-1601
-1
-1
and 1504 cm , while C–O–C stretching in hexyloxy units can be identified at 1241-1244 cm
-1
as strong band. The absorption band at 801 cm in FTIR spectrum of PTH-TPA was
associated with stretching vibrations of C-S bond in phenothiazine core, whereas the medium-
-1
strong band at 1021 cm found in FTIR spectrum of PHX-TPA was attributed to the
C–O–C= vibration or heteroatom ring deformation of the phenoxazine ring. The FTIR
spectra of PTH-TPA and PHX-TPA are provided in Figure S3, SI.
=
3
.2. Electrochemistry and electronic structure
Cyclic voltammetry (CV) was employed to survey the oxidation and reduction potentials of
PTH-TPA and PHX-TPA sensitizers, and to further asses the HOMO and LUMO energy
levels for an accurate evaluation of the electron transfer from the excited dye molecules into
the conductive band of TiO and the dyes regeneration. The redox waves of the PTH-TPA and
2
PHX-TPA in DCM/TBAP electrolyte system at a scan rate of 50 mV/s are shown in Figure 2.
A small oxidation wave with an onset around + 0.67 V vs SCE, followed by other four peaks,
could be observed in the CV curve of PTH-TPA upon positive potential sweep. This
irreversible oxidation is assumed to occur at the free electron pair of the S atom of the
phenothiazine core, as this oxidation is seen only for PTH-TPA. This dye undergoes further in
the first scan a quasi-reversible oxidation process at the N atom of the phenothiazine at the
same potential at which the oxidation takes place at the N atom of the phenoxazine in PHX-
TPA. It was followed by the quasireversible oxidation of the triphenylamine unit. This
oxidation wave shows a slight shift correlated with the electronic effect of the substituent on
1
9

ACCEPTED MANUSCRIPT
the TPA moiety. It appears that phenothiazine has a better electron donating effect than
phenoxazine since it leads to a lower oxidation potential of the TPA core in PTH-TPA (1.10V
vs SCE) compared to PHX-TPA (1.13V vs SCE).
Figure 2. Redox waves of PTH-TPA and PHX-TPA in DCM/TBAP:
a) anodic waves; b) cathodic waves
a)
b)
PHX-TPA
FTH-TPA
0
.0
-
5
PHX-TPA
PTH-TPA
red
4
2
.0x10
.0x10
ox
-
5
-
-
6.0x10
1.2x10
-5
ox
-4
0
.0
red
0
.0
0.5
1.0
1.5
2.0
-2.0
-1.5
-1.0
-0.5
0.0
E (V)
E (V)
By further scanning the potential in the anodic direction both dyes registered two
irreversible oxidation processes which are most likely due the dimerization of the formed
cationic radicals. These peaks disappeared on repeated cycling, being an indicative that the
formed dimers are soluble oligomeric products. Instead, the oxidation processes at the lone
electron pair of the N atoms of PHX-TPA, either from phenothiazine/phenoxazine or
triphenylamine are maintained over at least 10 scans between 0 and 1.5 V, being
accomplished by a small increase of the peaks intensity during the measurement, most
probably due to the deposition of a material on the electrode (Figure S4b, SI). Different
behavior was observed for PTH-TPA for which these oxidation processes appear completely
irreversible, being accompanied by a visible decrease of current intensity (Fig. S3a, SI). We
may assume that released electrons are well-captured by the electron-withdrawing
2
0

ACCEPTED MANUSCRIPT
cyanoacrylic acid, and in the absence of an external electron donor the reduction of radical
cations to neutral states became difficult.
Upon scanning in the negative direction to -2 V vs SCE, both dyes exhibited reduction
waves in accord with the electron deficient nature of the electron acceptor units, indicating
electron transport characteristics. The phenothiazine-based dye displays a large reduction
wave centered at -1.39 V vs SCE, being attributed to the formation of radical anions of the
electron-withdrawing cyanoacrylic acid unit (Figure 2b). A similar trend was noticed for
PHX-TPA, but at slightly more positive potential value (centred at -1.17 V vs SCE). A further
oxidation to neutral species of the formed radical anions was observed for both molecules at -
1
.27 V vs SCE.
Evaporation of a drop of dye solutions in DCM (0.5 % concentration) onto ITO
electrodes resulted in thin dye layers with characteristic surface waves quite different from
those recorded for solutions (Figure 3).
Figure 3. CV diagrams of a) PHX-TPA and b) PTH-TPA films during the positive potential
sweep
a)
b)
-4
2
.0x10
scan 1
scan 2
scan 3
-4
8.0x10
scan 1
scan 2
scan 3
-4
1
.0x10
ox
ox
-4
4.0x10
0
.0
0
.0
red
1.5
red
0
.0
0.5
1.0
1.5
2.0
0
.0
0.5
1.0
2.0
E (V)
E (V)
There are three oxidation peaks in the anodic region of PHX-TPA in the first oxidation
cycle, at 0.89 V, 1.25 and 1.81 V vs SCE due to the formation of radical cations and dications
2
1

ACCEPTED MANUSCRIPT
of triphenylamine and/or phenoxazine. In the reverse scan, a single reduction peak centered at
0
.62 V was recorded. In the second oxidation cycle, the three oxidation processes were
preserved, but shifted to 0.94, 1.19 and 1.54 V vs. SCE, being accompanied by drastically
intensity decrease. We may assume that a migration of radical cations occurs up to full
2
9
rd
stabilization by resonant reactions since the dye is fully conjugated. Starting with the 3
scan, only two overlapped oxidation peaks remained prevalent. The oxidation processes in the
following cycles could not be detected due to the total detachment of the dye film from the
ITO electrode. Unlike PHX-TPA film, during first potential sweeping of PTH-TPA in the
anodic region, two reduction bands with maximum at 0.95 and 0.77 V vs SCE appear after the
dye oxidation at 0.92 V, 1.15 and 1.39 V vs SCE (Figure 3b). Moreover, the detachment of
the film from the electrode did not occurred, and only instability of the oxidation processes
peaked at 1.15 and 1.39 V vs SCE was observed.
The peak potential due to the reduction of cyanoacrilic acid units to the correspondic
radical anions (-1.58 V vs SCE for PHX-TPA and -1.54 eV vs SCE for PTH-TPA) registered
an important shift to more negative values in films compared to solutions, proving that the
process of electron acceptance is much difficult in condensed phase. No subsequent oxidation
peak was detected upon reverse scanning (Figure S5, SI). A very good electrochemical
stability of both dye films in the cathodic area has been noted.
The energy of the highest occupied molecular orbital (HOMO) and lowest unoccupied
molecular orbital (LUMO) of PHX-TPA and PTH-TPA sensitizers were calculated from the
ox
red
oxidation (E onset) and reduction (E onset) onset potentials vs SCE. Different LUMO and
HOMO energy values were found for PHX-TPA (-3.17 eV and -5.07 eV) compared to PTH-
TPA (-3.2 eV and -5.18 eV), most probably due an increased electron delocalization between
donor and acceptor as a result of a better planarity of the former in condensed phase. As a
2
2

ACCEPTED MANUSCRIPT
consequence, a lower energy bandgap (1.9 eV) was estimated for PHX-TPA compared to
PTH-TPA (1.98 eV).
By comparing the edge of the conduction band of TiO and the potential of the redox
2
¯
¯
couple (I /I ) with HOMO and LUMO energy levels of sensitizers used in DSSCs, the energy
3
3
0-32
barriers can be established.
The band structure of PHX-TPA and PTH-TPA is presented
¯
¯
in Figure 4. It suggests that both dyes have more negative HOMO energy than the I /I redox
3
couple, leading to dyes regeneration with the selected electrolyte. The more positive LUMO
energy relative to the conduction band of TiO ensures an energetically favorable downhill
2
electron injection.
Figure 4. Energy levels diagram of PHX-TPA and PTH-TPA sensitizers
3
.3. Photo-optical features
The photophysical properties of PHX-TPA and PTH-TPA were investigated by UV-Vis and
photoluminescence (PL) spectroscopy, both in solution and solid state. UV-Vis absorption
spectra of the two compounds measured in solvents of different polarities are shown in Figure
S6, SI, and the pertaining data are listed in Table S1, SI. Figure 5 displays the absorption
spectra of PHX-TPA and PTH-TPA dye in dichloromethane (DCM) and acetonitrile (ACN)
2
3

ACCEPTED MANUSCRIPT
solutions in comparison with computed spectra of PHX-TPAm and PTH-TPAm with time-
dependent DFT using the B3LYP functional: excitation energies with their oscillator strengths
(
atomic units), and UV-Vis spectrum simulated by Gaussian broadening of excitation energies
with a broadening parameter of 0.33eV).
(
In DCM solutions two distinctive absorption bands were observed for PHX-TPA and
three distinctive bands for PTH-TPA. The first band in the region of 280−330 nm is
originating from the phenothiazine/phenoxazine and triphenylamine localized π−π*
transitions that partially overlap with the absorption band associated with the localized π−π*
transitions in the conjugated skeleton formed by donor-acceptor framework. Due to the T-
shape configuration, this band registered a bathochromic shift compared to the analogous
rigid-rod dyes, where the extension of conjugation takes place through the
phenothiazine/phenoxazine ring via the C3 atom. The electronic coupling between
phenothiazine/phenoxazine and the acceptor unit is poor in the T-shaped dyes due to the
1
3
perpendicular orientation of the phenothiazine/phenoxazine and conjugation pathway.
Figure 5. Experimental UV-Vis spectra of PHX-TPA (a) and PTH-TPA (b) in DCM and ACN
solutions versus simulated UV-Vis spectra of PHX-TPAm and PTH-TPAm
(a)
(b)
2
4

ACCEPTED MANUSCRIPT
The band appearing at the longer wavelength region, 492 nm for PHX-TPA and 477 nm for
PTH-TPA is mainly ascribed to the charge transfer from the donor comprising
phenoxazine/phenothiazine - triphenylamine framework to the cyanoacryclic acid acceptor.
The origin of this electronic absorption was confirmed by TD-DFT which proved that it
originates intrinsically from the intramolecular charge-transfer (ICT) transition from the
highest occupied molecular orbital (HOMO) to the lowest unoccupied molecular orbital
(
LUMO). Further discussion on this topic is carried out in Section 3.4. Interestingly, PHX-
TPA and PTH-TPA display longer wavelength CT transition compared to the analogous
1
3
fluorene-containing T-shaped dyes. This clearly indicates that our T–shape DD-π-A system
generates improved donor − acceptor interactions and consequently the light harvesting
properties.
Moreover, the CT absorption of dye possessing phenothiazine linker appeared 15 nm
blue-shifted compared to phenoxazine-based dye attributable to the more twisted
3
3
conformation of phenothiazine with respect to phenoxazine,
which hampers the
donor−acceptor interaction. The most intriguing feature is the long wavelength band
extending beyond 400 nm in the spectra of PTH-TPA. It may belong to a CT state or to
localized π−π* transitions in the conjugated skeleton. The simulated UV-Vis spectrum for
PTH-TPA predicted no excitations in this region. Therefore, it appeared reasonable to assume
that this band is very likely due to the presence of the protonated form of PTH-TPA, as
2
6
previously reported. To prove this, the UV-Vis spectrum of PTH-TPAm protonated at the N
atom in the thiazine cycle (H+ PTH-TPAm) was simulated (Figure S7, SI). The obtained
2
6
spectrum qualitatively correlates with the experimental spectrum reported in ref. , in which
the low energy transition around 500 nm almost disappeared. The molecular orbitals of
H+PTH-TPAm show a disruption of the conjugation in the thiazine ring due to the presence
of the proton, and the orbitals are spatially separated on the two halves of the phenothiazine
2
5

ACCEPTED MANUSCRIPT
fragment (Figure 6). It is apparent that LUMO becomes less and less delocalized as we go
+
from PHX-TPAm to PTH-TPAm, and to H ·PTH-TPAm.
Figure 6. Molecular orbitals of PHX-TPAm, PTH-TPAm, and protonated PTH-TPAm
+
(
H ·PTH-TPAm) computed with time-DFT using the B3LYP functional and 6-311++G**
basis set (energies are in eV).
2
6

ACCEPTED MANUSCRIPT
To gain further insight into the photo-physical processes of these donor-acceptor dyes,
their absorption behaviour in different solvents was investigated (Figure S6, SI). Both dyes
showed a negative solvatochromic behaviour (blue shift of the absorption maximum with
solvent polarity increase), indicating a better stabilization by solvation of dye molecules in
ground state with respect to the excited state with increasing solvent polarity. It is well-known
that a larger absorption range is favorable to a good performance of DSSCs, as more photons
can be harvested. According to the absorption maxima and molar extinction coefficients in
different solvents, it is clear that DCM is the most suitable solvent for TiO photoanode
2
functionalization due to a better ability of both dyes to harvest light in this medium, which
would lead to a higher photocurrent.
Due to the donor-acceptor arrangement, the photophysical features of PHX-TPA and
PTH-TPA are predictably solvent dependent. Thus, the PL spectra were recorded in different
polarity media by exciting at the ICT absorbance maxima. Figure 7 displays normalized
fluorescence spectra of PHX-TPA and PTH-TPA dyes in the chosen solvents.
Figure 7. Normalized PL spectra of dyes in various solvents
PHX-TPA
PTH-TPA
1
0
0
0
0
0
.0
.8
.6
.4
.2
.0
1.0
0.8
0.6
0.4
0.2
0.0
ACN
DMF
EtOH
THF
DCM
CHCl3
ACN
DMF
EtOH
THF
DCM
CHCl3
4
00
500
600
700
800
400
500
600
700
800
Wavelength (nm)
Wavelength (nm)
By exciting at maximum ICT absorbance wavelength a single PL emission was registered,
except PTH-TPA in THF solution that shows two PL maxima. It can be fully attributed to the
2
7

ACCEPTED MANUSCRIPT
1
spin-singlet charge-transfer excited state ( CT) → S transition from the intramolecular
0
charge transfer (ICT) excited state. The splitting of the ICT band in THF solution in two
peaks are probably due to the twisted conformations adopted by the PTH-TPA in this solvent
that impart the conjugated core in two individual fluorophores. A new CT band appears,
suggesting that the charge transfer process is considerable forbidden in this solvent. Such
features indicate that the fluorescence behavior of the dyes is strongly depending on the
relationship between molecular conformation and excited states. Maximum emission
wavelength values of PHX-TPA and PTH-TPA in selected solvents and solid state are
collected in Table S1, SI. The ICT band showed a negative solvatoflurochromism for both
dyes, except ACN, when a strong bathochromic shift of the ICT band with increasing solvent
polarities occurred. A large solvatochromic shift with polarity decreasing is generally
determined by a material that holds a strong ICT character. Here, the strongest electron
density accumulation on the acceptor unit was achieved in PHX-TPA along with the largest
solvatochromic shift of 100 nm.
Fluorescence lifetime measurements were recorded for both dyes in DCM using a
pulsed picosecond 375 nm diode laser. The fluorescence decay curves were convoluted with
the instrument response and fitted to a bi-exponential function and the resulting decay time
constants are presented in the Figure 8. The average lifetime of PHX-TPA (3.6 ns) is
significantly longer than the PTH-TPA lifetime (2.6 ns) which indicates that the time spent by
PHX-TPA in the excited states is longer than in the case of PTH-TPA, therefore the
probability of electron injection into TiO appears to be higher for PHX-TPA. As a
2
consequence, one would expect better performance for solar cells realized with PHX-TPA.
2
8