Organic dyes end-capped with perfluorophenyl anchors: Synthesis, electrochemical properties and assessment of sensitization capacity of titania photoanodes

Article abstract of DOI:10.1016/j.dyepig.2017.09.007

In the present work, organic sensitizers are synthesized and attached on TiO₂ photoanodes via Ti-O-C bonds. All sensitizers, designed for this purpose are symmetrical and have two perfluorophenyl end groups, which can lead to stable non-hydroly

Full text of DOI:10.1016/j.dyepig.2017.09.007

Accepted Manuscript
Organic dyes end-capped with perfluorophenyl anchors: Synthesis, electrochemical
properties and assessment of sensitization capacity of titania photoanodes
Panagiotis Giannopoulos, Dimitrios Raptis, Krystallia Theodosiou, Aikaterini K.
Andreopoulou, Charalampos Anastasopoulos, Alexandros Dokouzis, George
Leftheriotis, Panagiotis Lianos, Joannis K. Kallitsis
PII:
S0143-7208(17)31390-6
10.1016/j.dyepig.2017.09.007
DYPI 6231
DOI:
Reference:
To appear in: Dyes and Pigments
Received Date: 20 June 2017
Revised Date: 28 July 2017
Accepted Date: 6 September 2017
Please cite this article as: Giannopoulos P, Raptis D, Theodosiou K, Andreopoulou AK, Anastasopoulos
C, Dokouzis A, Leftheriotis G, Lianos P, Kallitsis JK, Organic dyes end-capped with perfluorophenyl
anchors: Synthesis, electrochemical properties and assessment of sensitization capacity of titania
photoanodes, Dyes and Pigments (2017), doi: 10.1016/j.dyepig.2017.09.007.
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
Graphical Abstract
ORGANIC SENSITIZERS
di thien-R
di EDOT-R
TiO2 photoanodes
Two organic dye sensitizer families attached onto TiO2 photoanodes via non-hydrolysable C-
O-Ti bonds are “fighting” over DSSCs and PCEs. Superiority of EDOT-bearing dyes is
demonstrated.

ACCEPTED MANUSCRIPT
Organic Dyes End-Capped with Perfluorophenyl Anchors:
Synthesis, Electrochemical Properties and Assessment of
Sensitization Capacity of Titania Photoanodes
1
2
3
4
5
6
Panagiotis Giannopoulos,a Dimitrios Raptis,b,c Krystallia Theodosiou,d
Aikaterini K. Andreopoulou,a,b,* Charalampos Anastasopoulos,a Alexandros
Dokouzis,d George Leftheriotis,d Panagiotis Lianos,b,c Joannis K. Kallitsis,a,b
7
8
9
10
11
12
13
14
15
16
17
18
19
20
a
Department of Chemistry, University of Patras, University Campus, Rio-Patras, GR26504,
Greece
b
Foundation for Research and Technology Hellas / Institute of Chemical Engineering Sciences
(FORTH/ICE-HT), Platani Str., Patras, GR26504, Greece
c Department of Chemical Engineering, University of Patras, 26504 Patras, Greece
d Department of Physics, University of Patras, 26504 Patras, Greece.
Corresponding Author:
E-mail: andreopo@upatras.gr
21
22
Department of Chemistry, University of Patras, University Campus, Rio-Patras, GR26504,
Greece
23
[1]

ACCEPTED MANUSCRIPT
1
Abstract
2
3
In the present work, organic sensitizers are synthesized and attached on TiO2 photoanodes via Ti-
O-C bonds. All sensitizers, designed for this purpose are symmetrical and have two
perfluorophenyl end groups, which can lead to stable non-hydrolysable bonds on the TiO2
surface. Broad absorption in the UV-Vis region and low band gap energy levels have been
achieved through the alternation of carbazole- and benzothiadiazole- based units. All dyes and
their corresponding photoanodes, were characterized via various complementary techniques,
including Cyclic Voltammetry measurements and sensitization capacity in Dye-sensitized solar
cell tests.
4
5
6
7
8
9
10
11
12
13
14
15
Keywords: organic dyes; perfluorophenyl anchors; benzothiadiazole-alt-carbazole; 3,4-
ethylenedioxythiophene; DSSCs; Photoelectrochromic devices.
[2]

ACCEPTED MANUSCRIPT
1
1. Introduction
2
3
Since their discovery, Dye-sensitized solar cells (DSSCs) have inaugurated a very interesting
field of research in the general frame of mesoscopic solar cells.[1,2] A great number of different
approaches have been employed for their study, but still, the full replacement of standard silicon-
based photovoltaics, remains a tough task, even after decades of intensive research and significant
progress in the field. Nevertheless, DSSCs show significant advantages, such as low cost and
environmentally friendly fabrication process, even if their efficiency is still much lower,
compared to inorganic solar cells.[3,4,5,6] This has led to increasing efforts for market
penetration of DSSCs, over the last years. A typical DSSC consists of two conducting substrates
that act as working and counter electrode, separated by an electrolyte that contains a redox
couple. Dye molecules, which are anchored to a semiconductor (usually TiO2 or ZnO) deposited
on the anode, absorb photons, leading to the excitation of electrons from a lower to a higher
energy state of the dye (sensitizer).[7,8] Thus, electrons from the excited state of the dye are
injected into the conduction band of the semiconductor. The dye should have suitable anchoring
groups to facilitate the formation of stable bonds with the semiconductor. The most commonly
used anchor is the carboxylic group, since it can form ester type bonds with the hydroxyl groups
on the TiO2 surface.[9,10] Other groups, such as sulfonic or phosphonic, have also been proven
functional,[11,12,13] but have a significant drawback compared to the carboxylic unit: their use is
limited in pure organic solvents, since they are not stable in aqueous environments. For this
reason, our group has developed an alternative anchoring methodology, which is based on the
perfluorophenyl unit. Through this route, the fluorine placed on the para- position can react under
mild alkaline conditions with the hydroxyl groups of the TiO2 surface, forming stable, non-
hydrolysable Ti-O-C bonds.[14,15]
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
The injected electrons are collected at the transparent conducting oxide (TCO), move through an
external circuit towards the counter electrode and are exchanged with a redox couple in solution.
[3]

ACCEPTED MANUSCRIPT
1
2
This electron transfer is facilitated in the presence of an appropriate electrocatalyst. The redox
shuttle finally compensates the holes left in the dye sensitizer and thus the cell functions in a
-
3
cyclic manner.[16,17,18] Liquid electrolytes, based in the I-/I3 couple, have proven to be the most
4
suitable, between several candidates, even though other couples such as Co3+/Co2+ have showed
higher redox potential which leads to enhanced power conversion efficiencies, but still suffer
from slow mass transport and slow regeneration of Co2+ at the cathode.[19,20,21,22,23] On the
other hand, liquid electrolytes present important disadvantages, such as solvent evaporation and
device short lifetime, which has led to the search for viscous variants, such as ionic liquids (ILs)
or polymer electrolytes (PEs).[24,25,26,27,28]
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Regarding the DSSC operation, two major electrochemical limitations must be taken care of. At
first, the lowest unoccupied molecular orbital (LUMO) of the sensitizer, has to be higher than the
conduction band of the semiconductor, in order to encourage spontaneous electron injection.
Secondly, the dye molecules may be regenerated, when their highest occupied molecular orbital
(HOMO) is lower than the redox level of the electrolyte. It is clear, that the energy levels of the
dye are very important, and the synthesis of the suitable dye is relevant with the oxide/electrolyte
used.[29,30]
An additional important feature for the efficiency of dyes employed in DSSCs is the anchoring
moiety or moieties used for their connection with the TiO2 NPs. Two (or more) anchoring groups
are mainly used in Ruthenium (or other metal) based sensitizers,[3,31] while for most organic
sensitizers the synthetic routes typically employed lead to only one anchoring group (D-
π-A
structure).[32,33,34] However, some reports suggest that only one anchor may cause drawbacks
in the sensitizer’s adsorption geometry on TiO2 nanoparticles and the overall charge
injection.[35,36,37] In some polymeric dyes on the other hand, the α,ω-end modification of the
polymeric chain enables the formation of sensitizers with two anchoring groups, a synthetic
strategy also used by us in a previous study.[15] More particularly, those polymeric entities based
on the combination of benzothiadiazole and N-alkyl-2,7-carbazole have attracted our attention
[4]

ACCEPTED MANUSCRIPT
1
2
because of the high conjugation level, low bandgap and the good charge carrier transport and
mechanical properties they provide, which lead to high power conversion efficiencies in
OPVs.[38]
3
4
In the present work, we have synthesized novel small organic dyes of defined chemical structures
and length based on thiophene derivatives, carbazole electron donating groups and benzothiazole
electron accepting groups. Two end-perfluorophenyl units have been incorporated as anchoring
groups for the subsequent sensitization of TiO2 for DSSCs. The synthesis of the herein reported
dyes was accomplished via two different synthetic paths. Furthermore, their optical and
electrochemical characterization is presented. Finally, initial DSSC tests of the four novel
chromophores have been performed, comparing the effect of the different structural motives of
the new chromophores and of the side solubilizing chains employed.
5
6
7
8
9
10
11
12
13
14 2. Experimental Section
15
16
17
18
19
20
21
2.1 Materials
4,7-Di-2-thienyl-2,1,3-benzothiadiazole (di-thien-BTZ),[39] 4,7-di-(5-bromothiophene-2-yl-
2,1,3-benzothiadiazole (di-Br-di-thien-BTZ),[40] 2,7-dibromo-N-heptadecan-9-yl-carbazole (di-
Br-HD-carbazole),[38 2,7-dibromo-N-2-ethylhexyl-carbazole (di-Br-EH-carbazole),[41] 2,7-
di(4,4,5,5-tetramethyl-1,3,2-dioxaborolan)- -2-ethylhexyl-carbazole (di-boronic ester-EH-
carbazole),[38 2,7-di(4,4,5,5-tetramethyl-1,3,2-dioxaborolan)- -heptadecan-9-yl-carbazole (di-
boronic ester-HD-carbazole)[38 and the catalyst palladium (II)tetrakis triphenyl phosphine
]
Ν
]
Ν
22
23
24
25
]
[Pd(PPh3)4][42] were prepared according to literature.
All solvents and reagents were purchased from Sigma Aldrich, Alfa Aesar or Acros Organics
and were used without further purification unless otherwise stated. All reactions were carried out
[5]

ACCEPTED MANUSCRIPT
1
2
under argon atmosphere. Tetrahydrofuran (THF) was dried and distilled just before use over
sodium wire in the presence of benzophenone. Commercial nanocrystalline titania Degussa P25
(specific surface area 50 m2/g) was used in all cell constructions and Millipore water was used in
all experiments. SnO2:F transparent conductive electrodes (FTO, Resistance 8 Ω/square) were
3
4
5
purchased from Pilkington. ITO transparent conductive electrodes (Rsheet<15 Ω/square) were
6
purchased from Aldrich. TiO2 nanoparticles prior to their reactions with the perfluorophenyl
o
7
functionalized moieties were calcined at 400 C for 3-4 hr to remove all hydrogen-bonded water
8
molecules.[43]
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
2.2 Characterization methods
1H, and 19F NMR spectra were recorded on Bruker Advance DPX 400.13, and 376.5 MHz
spectrometers, respectively, with CDCl3 as solvent containing TMS as internal standard.
Thermogravimetric analysis (TGA) was carried out on 8 mg samples contained in alumina
o
crucibles in a Labsys TM TG apparatus of Setaram under nitrogen and at a heating rate of 10 C
min-1.
IR spectra were recorded on a Perkin-Elmer 16PC FTIR spectrometer and on an ALPHA-P
diamond ATR spectrometer of Bruker Optics GmbH.
UV-Vis spectra in solution were recorded using a Hitachi U-1800 spectrophotometer.
Photoluminescence (PL) spectra in solution were recorded using a Perkin–Elmer LS45
luminescence spectrometer, after excitation at the UV-Vis absorption maxima.
UV–Vis absorption spectra and diffuse reflectance spectra (DRS) of sensitized photoanodes
were recorded with a Shimadzu UV-2600 spectrophotometer equipped with integration sphere.
The electrochemical behaviour of the fabricated materials was investigated using cyclic
voltammetry (CV thereafter). CV experiments were carried out in a three electrode cell.[44,45]
Two kinds of working electrodes have been used. An ITO/Glass with the synthesized sensitizers
[6]

ACCEPTED MANUSCRIPT
1
2
to be measured drop-casted on the ITO conductive side and a dye- sensitized TiO2/FTO/Glass. An
Ag/AgCl electrode served as reference and a platinum wire was used as the counter electrode.
Thin films of the fabricated materials were drop casted on ITO coated glass slides, (Rsheet<15
3
4
Ω/square), preheated at 80 °C for 20 min, from precursor solutions of chloroform. The resulting
5
films were further annealed at 80 °C for 15 min. Dye sensitized TiO2/FTO/Glass was synthesized
as described below. An Autolab PGSTAT 302N electrochemical analyzer connected to a personal
computer running the NOVA 1.8 software was used for data collection and analysis. All
experiments were carried out at a scan rate of 0.1 V/s. Tetrabutylammonium hexafluorophoshate
(TBAPF6) 0.1 M in Acetonitrile (CH3CN) was used as supporting electrolyte. Before carrying out
the measurements the cell was purged with pure argon for 20 min to remove diluted gasses. The
reference electrode potential was calibrated against Ferrocene/Ferrocenium (Fc/Fc+) after each
voltammetry run.
6
7
8
9
10
11
12
13
14
15
The corresponding HOMO, LUMO energy levels and the energy gap (Eg) were calculated with
use of the redox potentials by the following empirical equations:[44]
EHOMO = e Eox − EFerrocene − 5.2 eV
(1)
[
]
]
(
)
ons
1/2
ELUMO = e Ered − E Ferrocene − 5.2 eV
(2)
(3)
[
(
)
ons
1/2
ox
ons
red
ons
Eg = e
(E
− E
)
16
17
18
19
20
21
22
Furthermore, the value of -5.2 eV in equations (1), (2) emerged from calibration of the Ag/AgCl
electrode versus a normal hydrogen electrode ( NHE) and considering that the NHE redox
potential corresponds to -4.6 eV on the zero vacuum level scale.[46,47]
[7]

ACCEPTED MANUSCRIPT
1
2
2.3 Synthetic procedures
3
2.3.1 Synthesis of the organic dye precursors
4
5
Synthesis of 4,7-di-2-(3,4-ethylenedioxythiophene)-2,1,3-benzothiadiazole (di-EDOT-BTZ)
4,7-Di-bromo-2,1,3-benzothiadiazole 7.00g (23.81 mmol, 1eq) were dissolved in 110 mL
dimethylacetamide (DMAc), followed by the addition of 10 eq of 3,4-ethylenedioxythiophene
(EDOT). Afterwards, CH3COOK (7.01 g, 71.43 mmol, 3 eq) and Pd(OAc)2 (53.45 mg, 0.24
6
7
8
o
9
mmol, 0.01 eq) were added to the solution, which was stirred at 155 C under Ar for 18 h. The
10
11
12
13
14
15
16
17
18
19
20
21
22
mixture was precipitated in 500 mL of a brine solution and extracted using ethyl acetate (EtOAc).
The organic layer was dried using MgSO4, filtered and the solvent was removed under reduced
pressure. The resulted material was finally dried in a vacuum oven at 45 oC for 18 h. Yield 90%.
1H NMR (CDCl3,
δ ppm): 4.30-4.33 (m, 2H), 4.39-4.42 (m, 2H), 6.57 (s, 2H), 8.39 (s, 2H)
Synthesis of 4,7-di-(5-bromo-3,4-ethylenedioxythiophene-2-yl)-2,1,3-benzothiadiazole (di-Br-di-
EDOT-BTZ)
Di-EDOT-BTZ 6.00g (14.41 mmol, 1 eq) were dissolved in 250 mL of dimethylformamide
(DMF) and afterwards N-Bromosuccinimide (NBS, 5.77 g, 32.41 mmol, 2.25 eq) was added to
that solution in 3 portions over 60 minutes. The solution was then stirred at RT for 18 h, followed
by the addition of an aqueous Na2SO3 saturated solution. The desired product was taken after
filtration and recrystallization using DMF/MeOH: 1/10 v/v. Yield 52%:
1H NMR (CDCl3,
δ ppm): 4.39-4.41 (m, 4H), 8.37 (s, 2H)
[8]

ACCEPTED MANUSCRIPT
1
2
2.3.2 Synthesis of the small organic dye sensitizers
3
Synthesis of bis-{4,7-[2-pentafluorophenyl-7-thiophene-N-2-ethylhexyl-carbazole]} (dithien-EH
sensitizer)
4
5
(i) via Direct coupling: Di-thien-BTZ (70.00 mg, 0.235 mmol, 1 eq) and di-Br-EH-carbazole
(822.00 mg, 1.88 mmol, 8 eq) were dissolved in 10 mL of DMAc, in a round bottom flask
equipped with a stirring bar and a reflux condenser, under inert atmosphere, followed by the
addition of CH3COOK (78 mg, 0.7 mmol, 3 eq) and Pd(OAc)2 (1.00 mg, 3.50 umol, 0.015 eq).
The mixture was refluxed for 24hr, cooled at RT and a solution of perfluorophenyl boronic acid
(104.60 mg, 0.49 mmol, 2.1 eq) in 0.6 mL of DMAc was then added for another 24hr under
reflux. After cooling at RT, the mixture was poured into 50 mL of MeOH, stirred for 2hr, filtered
and consecutively washed with boiling petroleum ether, MeOH, H2O and again with MeOH,
before drying under vacuum at 45oC for 18 h. Yield 58%.
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
(ii) via Suzuki coupling: Di-Br-di-thien-BTZ (50.00 mg, 0.11 mmol, 1 eq) and di-boronic
ester-EH-carbazole (462.00 mg, 0.09 mmol, 8 eq) were dissolved in 10 mL of degassed toluene,
in a round bottom flask equipped with a stir bar and a reflux condenser, under inert atmosphere,
followed by the addition of 0.4 mL Na2CO3 (aq. solution, 0.87 mmol, 8 eq) and 7.6 mg Pd(PPh3)4
(6.00 umol, 0.06 eq). The mixture was refluxed for 48h, cooled at RT and
bromopentafluorobenzene (76 uL, 0.61 mmol, 5.6 eq) was then added and stirred for another 24 h
under reflux. The mixture was cooled at RT, filtered and extracted using CHCl3 and H2O. The
organic layer was dried over MgSO4, filtered and the solvent was removed in a rotary evaporator.
The final product thus obtained, was dried under vacuum at 45 oC for 18 h. Yield 82%.
1H NMR (CDCl3,
4.20 (m, 4H), 7.35 (d, 2H), 7.51 (d, 4H), 7.69 (s, 4H), 7.94 (s, 2H), 8.08 (d, 4H), 8.21 (d, 2H). 19F
δ ppm): 0.83-1.03 (two t, 12H), 1.04-1.5 (three broad, 16H), 2.12 (m, 2H),
NMR (CDCl3,
162.52 (m, 4F).
δ ppm): -142.82 – -143.02 (two m, 4F), -156.09 – -156.24 (m, 2F), -162.16 – -
[9]

ACCEPTED MANUSCRIPT
1
2
Analogous conditions were used for the synthesis of the remaining sensitizers using both the
Direct coupling and the Suzuki coupling conditions in each case (Table 1)
3
•
•
•
bis-{4,7-[2-pentafluorophenyl-7-thiophene-N- heptadecan-9-yl-carbazole]}
4
(dithien-HD sensitizer): Yield 30% via Direct arylation, 65% via Suzuki coupling.
5
1H NMR (CDCl3,
δ ppm):0.76-0.87 (t, 12H), 0.98 (m, 4H), 1.04-1.32 (broad, 44H), 1.97
6
(m, 4H), 2.34 (m, 4H), 4.70 (m, 2H), 7.53-7.60 (two broad, 2H), 7.66 (broad, 4H), 7.88
(broad, 2H), 7.96 (s, 2H), 8.02 (broad, 2H), 8.12 (m, 4H), 8.21 (d, 2H). 19F NMR (CDCl3,
7
8
δ ppm): -142.86 – -143.11 (two m, 4F), -156.21 – -156.29 (m, 2F), -162.39 – -162.50 (m,
9
4F).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
bis-{4,7-[2-pentafluorophenyl-7-(3,4-ethylenedioxythiophene)-N-2-ethylhexyl-
carbazole]} (diEDOT-EH sensitizer): Yield 80% via Direct arylation, 70% via Suzuki
1
coupling. H NMR (CDCl3,
δ ppm): 0.86-0.94 (two t, 12H), 1.25-1.40 (m, 16H), 2.00-
2.02 (quint, 2H), 4.01-4.12 (m, 4H), 4.41-4.55 (two m, 8H), 7.32 (d, 2H), 7.51 (s, 2H),
7.74 (d, 2H), 7.93 (d, 2H), 7.95 (s, 2H), 8.04 (d, 2H), 8.45 (s, 2H). 19F NMR (CDCl3,
δ
ppm): -142.85 – -143.09 (two m, 4F), -156.22 – -156.29 (m, 2F), -162.33 – -162.59 (m,
4F).
bis-{4,7-[2-pentafluorophenyl-7-(3,4-ethylenedioxythiophene)-N-heptadecan-9-yl-
carbazole]} (diEDOT-HD sensitizer): Yield 67% via Direct arylation, 50% via Suzuki
coupling. 1H NMR (CDCl3,
δ ppm): 0.82 (t, 12H), 0.98 (m, 4H), 1.03-1.34 (b, 44H), 1.94
(m, 4H), 2.33 (m, 4H), 4.49 (d, 4H), 4.52 (d, 4H), 4.70 (quint, 2H), 7.67, (broad, 4H),
7.89 (s, 2H), 8.03 (s, 2H), 8.11-8.13 (dd, 4H), 8.5 (s, 2H). 19F NMR (CDCl3,
δ
ppm): -
142.87 – -143.10 (two m, 4F), -156.23 – -156.30 (m, 2F), -162.37 – -162.45 (m, 4F).
[10]

ACCEPTED MANUSCRIPT
1
2
2.4 Dye Sensitization of TiO2
3
2.4.1 Preparation of dye-sensitized powder TiO2 NPs
4
The general method for sensitizing powder TiO2 NPs or of photoanodes of deposited TiO2
NPs on onto glass substrates, with organic molecules bearing perfluorophenyl units has been
described also previously.[14] In particular for the powder TiO2 NPs these were thermally treated
prior to their sensitization at 400 oC for 3-4 h.
5
6
7
8
Briefly, 1eq (0.07 mmol) of sensitizer, 4 eq (0.28 mmol) of K2CO3, and 50 mg of powder
TiO2 NPs were dispersed in acetone and heated at reflux for 48 h. The as synthesized hybrids
were filtrated and thoroughly washed with acetone, THF and water to remove unreacted
molecules and salts, until a colourless filtrate was obtained. The dye-sensitized TiO2 NPs were
dried in a vacuum oven overnight at 40 oC.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
2.4.2 Preparation of sensitizer/TiO2 physical blends
TiO2 NPs in their powder form and the organic sensitizer were dispersed in 20 mL of acetone.
The composition of each blend was calculated from the sensitization percentage obtained from
TGA data of the dye-sensitized TiO2 nanoparticles. The mixture was then sonicated for 30min,
stirred at 60 oC for 48 h and finally purified, using conditions similar to the as described above for
the dye-sensitized TiO2 NPs.
2.4.3 Deposition of TiO2 on FTO
Nanoparticulate titania films (np-TiO2) were deposited on FTO electrodes. An FTO glass
electrode was cut in the appropriate dimensions and was carefully cleaned first with soap and then
by sonication in ethanol and acetone. A thin layer of compact titania was first sprayed over a FTO
electrode using 0.2 mol L−1 diisopropoxytitanium bis(acetylacetonate) solution in ethanol and
then it was heated at 500 °C for 1 h. On this thin bottom layer, a titania paste made of P25
[11]

ACCEPTED MANUSCRIPT
1
2
nanoparticles was applied by screen printing. 90 mesh screen was used. The film was heated
again up to 500 °C for 20 min. Then, the film was treated with a 0.04 mol L−1 aqueous solution of
TiCl4 for 30 min at 70 °C, followed by heating up to 500 °C for 20 min. The geometrical area of
the film was 0.16 cm2 (0.4 cm × 0.4 cm).
3
4
5
6
2.4.4 Preparation of dye-sensitized TiO2 photoanodes
7
A TiO2 photoanode was placed in a solution of sensitizer (0.07 mmol, 1 eq) and K2CO3 (0.28
mmol, 4 eq) in acetone and heated at reflux. After 48h the mixture was cooled at RT and the
photoanode was taken out, washed using subsequently acetone, THF, water and acetone, before
drying under vacuum overnight.
8
9
10
11
12
13
14 3. Results and Discussion
15
16
17
18
19
20
21
22
23
24
25
The main scope of this work was the design and synthesis of novel sensitizers for DSSCs with
enhanced absorption in the visible area of the solar spectrum, based on benzothiadiazole and
carbazole units connected in an alternating and regular manner as in the PCDTBT polymer
electron donor case. Such low band gap organic semiconductors developed for organic
photovoltaics (OPVs) generally possess good charge transport characteristics and have shown
improved power conversion efficiencies (PCE). For such low band gap semiconductors several
monomers have been developed and combined in order to lead to the final polymer electron
donors.[48,49,50,51] Among these monomers, carbazoles are often preferred due to their
chemical and environmental stability, along with good charge transport properties. Solubility
issues often encountered due to the rigidity of carbazoles are easily overcome using bulky side
[12]

ACCEPTED MANUSCRIPT
1
2
aliphatic chains.[52,53] Especially secondary alkyl chains attached on the nitrogen atom of
carbazole have led to better solubility and efficiency of the polymer electron donors thereof in
OPVs.[54]
3
4
Another important feature of the herein reported sensitizers is the incorporation of end-
perfluorophenyl functionalities. In general the perfluorophenyl unit has proven an excellent
choice for the development of functionalized semiconductors as precursors of hybrid materials
with carbon nanostructures (CNTs, fullerenes etc.)[55,56,57] or with inorganic nanoparticles (e.g.
TiO2 NPs),[14,15] as electron withdrawing groups for the energy levels modulation of the
semiconductors,[58,59] or for the insertion of additional functionalities via nucleophilic aromatic
substitution.[60]
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
3.1 Synthesis of the sensitizers
For the preparation of the desired small organic dyes bearing carbazole and benzothiadiazole
units, the first step was the di-substitution of benzothiadiazole by either thienyl or 3,4-
ethylenedioxythienyl rings. This was accomplished in a single step via Direct arylation
conditions,[61,62] as can be seen in Scheme 1. The Direct arylation proceeds using palladium
acetate [Pd(OAc)2] as catalyst, potassium acetate (CH3COOK) as base in dimethylacetamide
(DMAc). Thus, the 4,7-di-2-thienyl-2,1,3-benzothiadiazole (di-thien-BTZ)[39] and the 4,7-di-2-
(3,4-ethylenedioxythiophene)-2,1,3-benzothiadiazole
(di-EDOT-BTZ)
were
prepared,
respectively. The direct synthesis of these two central moieties of the final sensitizers, was
efficient and of high yields, namely 50% and 90% respectively.
[13]

ACCEPTED MANUSCRIPT
1
2
Scheme 1. Synthesis of di-thien-BTZ and, di-EDOT-BTZ via Direct arylation, and of their
3
4
respective di-bromo derivatives.
5
In another step, was performed the synthesis of the two different 2,7-dibromo-N-alkyl-
carbazoles and in particular 2,7-dibromo-N-2-ethylhexyl-carbazole (di-Br-EH-carbazole) and
2,7-dibromo-N-heptadecan-9-yl-carbazole (di-Br-HD-carbazole). Di-Br-EH-carbazole was
prepared in a one-step reaction between the commercially available 2,7-dibromo-9H-carbazole
6
7
8
9
and 2-ethylhexyl bromide, using acetone as solvent.[41] On the other hand, di-Br-HD-carbazole
10
11
12
13
14
15
16
17
was prepared in a sequence of four reactions,[38] since this synthetic route leads to the desirable
product in high purity and better yield, in comparison to other reported procedures.[63,64]
After the successful synthesis of the di-Br-alkyl-carbazoles and of the central di-substituted-
BTZ, the final step was the connection of the two different species and the insertion of the
perfluorophenyl anchoring end-groups. For this scope, two coupling methods were employed,
namely the Direct arylation[61] (Scheme 2) and the Suzuki coupling (Scheme 3).[65]
Through the Direct arylation route (Scheme 2) and with a monomers’ ratio of 8/1 di-Br-alkyl-
carbazoles to di-substituted-BTZ, the reactive intermediates with two end bromine groups were
[14]

ACCEPTED MANUSCRIPT
1
2
3
prepared, which subsequently reacted with perfluorophenyl boronic acid giving the desired
sensitizers (Table 1).
4
5
Scheme 2. Synthesis of EH and HD decorated dithien- and diEDOT- sensitizers via Direct
6
7
arylation through a one-pot, two-step reaction.
8
In order to synthesize the desired sensitizers via Suzuki coupling (Scheme 3) the
dibrominated benzothiadiazole derivatives were required, in particular the 4,7-di-(5-
bromothiophene-2-yl)-2,1,3-benzothiadiazole (di-Br-di-thien-BTZ)[40] and the 4,7-di-(5-
9
10
11
12
13
14
bromo-3,4-ethylenedioxythiophene-2-yl)-2,1,3-benzothiadiazole
(di-Br-di-EDOT-BTZ)
(Scheme 1). Moreover, the di-Br-alkyl-carbazoles were transformed to their respective boronic
esters (Scheme S1).[38] Typical Suzuki coupling conditions using Pd(PPh3)4 as catalyst, aqueous
Na2CO3 as base in toluene were employed.[65] The monomers’ ratio was 8/1 di-boronic ester-
[15]

ACCEPTED MANUSCRIPT
1
2
3
alkyl-carbazole to di-brominated-BTZ in order to provide the reactive intermediate with two end
boronic ester units that further on reacted with perfluorophenyl bromide in a second step.
4
5
Scheme 3. Synthesis of dithien-alkyl and diEDOT-alkyl sensitizers via Suzuki coupling through a
6
7
one-pot, two-step reaction.
8
Comparison of the two different synthetic strategies showed that the Suzuki coupling reaction
was more efficient for the preparation of the thienyl bearing sensitizers, namely the dithien-EH
and the dithien-HD, both in terms of purity and reactions’ yields as can be seen in Table 1. On
the other hand, Direct arylation was more efficient, for the preparation of the EDOT-sensitizers,
namely the diEDOT-EH and diEDOT-HD. This fact is mostly attributed to the availability of all
positions of the thienyl ring as reactive sites for the Direct arylation reaction in contrast to the
EDOT bearing molecules. The incorporation of the ethylenedioxythienyl group limits undesirable
byproducts produced during Direct arylation, since there is only one available hydrogenated
9
10
11
12
13
14
15
[16]

ACCEPTED MANUSCRIPT
1
2
position that can react with the respective bromide. Thus, the Direct arylation route provided
increased efficiencies for the preparation of the EDOT bearing sensitizers. On the other hand, the
Suzuki coupling method is to be chosen for the preparation of the dithien-EH and the dithien-
HD sensitizers.
3
4
5
The chemical structures and the abbreviations of all synthesized sensitizers are shown in
Scheme 4. In all cases the NMR characterization of the precursors and of the final sensitizers
confirmed the materials proposed chemical structures. Additionally, 19F NMR spectroscopy gave
proof of the attachment of the end-perfluorophenyl units on the final sensitizers. Analytic data of
the 19F NMR examination of the prepared sensitizers are provided in the Experimental Section
while a representative spectrum is shown in Figure S1 for the diEDOT-HD dye as prepared via
the Direct arylation procedure.
6
7
8
9
10
11
[17]

ACCEPTED MANUSCRIPT
1
2
Scheme 4. Chemical structures of the four sensitizers
3
4
5
6
7
8
9
3.2 Sensitization of powder TiO2 NPs
Based on our previous experience, the use of perfluorophenyl units as anchoring groups of
polymer sensitizers provides an effective methodology for the sensitization of TiO2 NPs and
photoanodes for DSSCs.[14,15] Therefore, also in this work, the perfluorophenyl end-
functionalities of the small organic sensitizers described above, reacted under mild alkaline
conditions, namely potassium carbonate as base in acetone, with the hydroxyl groups of TiO2
[18]

ACCEPTED MANUSCRIPT
1
2
NPs. Thus, non-hydrolysable bonds are formed “anchoring” the organic sensitizers onto the TiO2
surface (Scheme 5). The organic-inorganic hybrid materials created where thoroughly washed
with acetone, THF, water and again acetone in order to remove any unreacted species and all
inorganic salts.
3
4
5
The modification percentages of TiO2 NPs with the four small organic dyes are given in
Table 1 as calculated from thermogravimetric analysis (TGA). Modification percentages up to
22%wt were recorded, but no particular trend was observed regarding the origin of the sensitizers,
meaning via Direct arylation or Suzuki coupling, and their efficacy in TiO2 modification. This
fact also points out to the successful introduction of the perfluorophenyl functionalities regardless
of the synthetic method employed. Moreover, and due to the smaller size of the herein employed
sensitizers in comparison to previously reported polymeric analogues materials[14,15] higher
modification percentages were accomplished, 18-22%wt over the previous 10-13%wt
percentages. Figure 1 shows the TGA analyses of the dithien-alkyl sensitizers and of the
respective dithien-alkyl/TiO2 NP hybrids. The TGA analyses of the diEDOT-alkyl sensitizers and
of the diEDOT-alkyl/TiO2 hybrids are provided in Figure S3 of the Sup. Info. section. The
organic molecular sensitizers degrade around 400oC, whereas pure TiO2 NPs show negligible
weight loss up to 800oC. On the other hand, the hybrid materials in all cases show higher stability
compared to the organic sensitizers, both in terms of onset temperature of degradation and of
weight loss at 800oC, proving the successful binding of the sensitizer on the TiO2 surface.
Further confirmation of the formation of the sensitized TiO2 NPs was provided via ATR
spectroscopy as shown in Figure 2 for the dithien-alkyl materials. The ATR spectra of the
diEDOT-alkyl sensitizers and of the respective hybrids with TiO2 NPs are given in Figure S2.
The absorption bands of pure TiO2 NPs are located in-between 1750-1250cm-1.[66] After the
TiO2 NPs sensitization, new peaks owing to the organic molecules are observed with more
pronounced those at 1223cm-1, 1155cm-1, 2935cm-1 and 2860cm-1. More importantly, the
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
[19]

ACCEPTED MANUSCRIPT
1
2
absorption bands between 1300cm-1 and 1100cm-1, prove the formation of Ti-O-C bonds, in
combination with the presence of C-F streching vibrations.[67,68]
3
The optical properties of the organic sensitizers and of the hybrid organic-inorganic materials
were examined by UV-Vis and PL spectroscopies. The absorption spectra of the organic
sensitizers dithien-alkyl and diEDOT-alkyl are given in the Sup. Info. section, Figures S4 and S5,
respectively, in comparison to the commercial polymeric semiconductors used as benchmark
materials in photovoltaic applications, namely the regioregular-poly(3-hexyl-thiophene) rr-P3HT
4
5
6
7
8
and
the
poly[N-9′-heptadecanyl-2,7-carbazole-alt-5,5-(4′,7′-di-2-thienyl-2′,1′,3′-
9
benzothiadiazole)] PCDTBT. The main absorption bands of the dithien-alkyl molecules are found
at 360nm and 510nm, whereas for the diEDOT-alkyl at 390nm and 570nm. Compared to the
polymeric electron donor rr-P3HT it is obvious that both the dithien-alkyl and the diEDOT-alkyl
sensitizers surpass its absorption to higher wavelengths. More importantly, the absorption spectra
of the small organic sensitizers resemble that of the low-band gap PCDTBT. Especially, in the
case of the EDOT bearing sensitizers, diEDOT-EH and diEDOT-HD, their absorption is identical,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
if not better, than of the high molecular weight polymeric PCDTBT. Τhese excellent absorption
characteristics of the herein presented small-organic semiconductors in the visible region of the
light spectrum, shows their potentiality as sensitizers in DSSCs.
The photoluminescence characterization of the organic and of the hybrid materials is in
Figures 3 and S6 for the dithien-alkyl and the diEDOT-alkyl materials, respectively. Moreover,
physical blends of TiO2 and the sensitizers were prepared in order to distinguish differences of
direct attachment of the sensitizers onto the TiO2 NPs surface versus simple mixtures of the two
components. The composition of each blend was based on the TiO2 modification percentage as
calculated from TGA analyses. As can be seen in Figures 3 and S6, the photoluminescence
behavior of the sensitized nanoparticles strongly differs from the pure organic sensitizers and the
physical blends, after excitation at the absorption maximum of sensitizers. In particular, both the
sensitizer and the physical blend show a broad emission bands around 600nm and 650nm for the
[20]

ACCEPTED MANUSCRIPT
1
2
3
4
5
dithien-alkyl and the diEDOT-alkyl materials, respectively. On the other hand, the sensitized NPs
demonstrate photoluminescence quenching, which is attributed to the covalent attachment of the
sensitizers onto the TiO2 surface. This direct attachment allows electronic interactions between
TiO2 and the sensitizer in contrast to the physical blends. This behavior provides additional proof
of the success of the sensitization route employed herein.
6
7
Scheme 5. General synthetic strategy for the preparation of dithien-alkyl/TiO2 and diEDOT-
8
9
alkyl/TiO2 hybrids.
[21]

ACCEPTED MANUSCRIPT
1
2
Table 1. Reaction yields of the synthesized sensitizers (Scheme 4) and modification percentages
of TiO2 nanoparticles after their sensitization.
Direct arylation
Suzuki coupling
TiO2 NPs
a/a
Sensitizer
Yield (%)
Yield (%)
modification (%)
58
30
80
67
82
65
70
50
20.5
22
1
2
3
4
dithien-EH
dithien-HD
diEDOT-EH
diEDOT-HD
22
18
3
4
100
90
80
70
60
50
40
30
20
TiO2
dithien-EH sensitizer
dithien-HD sensitizer
dithien-EH/TiO2
dithien-HD/TiO2
100
200
300
400
500
600
700
800
Temperature (oC)
5
6
Figure 1. TGA analyses of thienyl bearing sensitizers dithien-alkyl, of TiO2 nanoparticles, and of
7
8
9
hybrid materials dithien-alkyl/TiO2
[22]

ACCEPTED MANUSCRIPT
TiO2
dithien-EH/TiO2 hybrid
dithien-HD/TiO2 hybrid
dithien-EH sensitizer
dithien-HD sensitizer
3500
3000
2500
2000
1500
cm-1
1000
Wavenumbers
(
)
1
2
Figure 2. ATR spectra of thienyl bearing sensitizers dithien-alkyl, of TiO2 nanoparticles, and of
3
4
hybrid materials dithien-alkyl/TiO2.
TiO2
dithien-EH sensitizer
dithien-HD sensitizer
dithien-EH/TiO2 hybrid
dithien-HD/TiO2 hybrid
dithien-EH
dithien-HD
/TiO2 blend
/TiO2 blend
550
600
650
700
750
800
Wavelength (nm)
5
6
Figure 3. PL spectra of TiO2 nanoparticles, dithien-alkyl sensitizers, hybrid materials dithien-
alkyl/TiO2 and of the respective dithien-alkyl sensitizer/TiO2 physical blends in CHCl3 solutions,
after excitation at 510nm.
7
8
[23]

ACCEPTED MANUSCRIPT
1
2
3.3 Sensitization of Photoanodes
For the preparation of the sensitized TiO2 photoanode electrodes, the deposited
nanoparticulate titania films on FTO electrodes were immersed into acetone solutions of the
small-organic perfluorophenyl functionalized dyes in the presence of K2CO3. After 48 h at reflux,
the sensitized photoanodes were removed from the reaction medium and washed thoroughly with
acetone, THF, water and finally acetone. Figure 4 shows representative examples of TiO2
sensitized photoanodes using the 4 types of small organic sensitizers presented in this study.
Uniformly colored photoanodes were obtained that preserved their color even after their extended
washing for the removal of un-attached organic species and of inorganic salts. Notably, the
diEDOT-alkyl sensitizers afforded more intense and darker colored photoanodes in consistency
with their deeper absorption in the red region of the visible spectrum. This phenomenon is also
evident in the diffuse reflectance absorption spectra (DRS) of the sensitized photoanodes as
shown in Figure 5.
3
4
5
6
7
8
9
10
11
12
13
14
15
Figure 4. Images of the photoanodes sensitized with a) dithien-EH, b) diEDOT-EH, c) dithien-
16
HD and d) diEDOT-HD sensitizer.
[24]

ACCEPTED MANUSCRIPT
0,5
0,4
0,3
0,2
0,1
0,0
(b)
(d)
(a)
(c)
400
500
600
700
800
Wavelength (nm)
1
2
Figure 5. Reflection-absorbance spectra of np-TiO2 on FTO photoanodes sensitized with: (a)
3
4
dithien-EH,( b) diEDOT-EH, (c) dithien-HD and (d) diEDOT-HD.
5
6
3.4 Electrochemical characterization
7
Cyclic voltammetry was used for the estimation of HOMO and LUMO energy levels of the
investigated dyes. Knowledge of these values is important for the efficient design and operation
of a DSSC. In order to assess the repeatability of the method, the CV measurements for each
compound were repeated for different samples. Negligible deviations were recorded in some
cases in the HOMO, LUMO and Eg levels of up to 0.1eV, which can be considered as the
experimental error of the procedure.
8
9
10
11
12
13
14
15
16
Table 2 contains the cyclic voltammetry results of the investigated materials. HOMO, LUMO
and Eg levels were calculated according to the empirical equations (1), (2) and (3) given in the
Experimental Section. Reduction and oxidation onset potentials were deduced from reduction and
oxidation peaks respectively, shown in Figure 6.
[25]

ACCEPTED MANUSCRIPT
1
2
Table 2. Reduction and oxidation onset potentials, HOMO, LUMO energy levels and Eg of the
sensitized np-TiO2 on FTO and of films of the organic sensitizers deposited on ITO glass.
Eonsox (V)
Eonsred (V)
Eg
Sample
Homo (eV) Lumo (eV)
vs Ag/AgCl vs Ag/AgCl (eV)
1.36
1.66
1.29
1.56
0.73
0.87
0.94
0.83
-0.79
-0.46
-1.00
-0.69
-1.17
-0.95
-0.89
-0.92
2.2
2.1
2.3
2.3
1.9
1.8
1.8
1.8
-6.1
-6.2
-6.0
-6.2
-5.4
-5.5
-5.6
-5.5
-3.9
-4.1
-3.7
-3.9
-3.5
-3.7
-3.8
-3.7
dithien-EH sensitizer on ΙΤΟ
dithien-EH/TiO2/FTO
dithien-HD sensitizer on ITO
dithien-HD/TiO2/FTO
diEDOT-EH sensitizer on ITO
diEDOT -EH/TiO2/FTO
diEDOT-HD sensitizer on ITO
diEDOT-HD/TiO2/FTO
3
[26]

ACCEPTED MANUSCRIPT
1
2
Figure 6. Cyclic voltammograms of (a) dithien-EH/ITO (b) dithien-EH/TiO2/FTO (c) dithien-
HD/ITO (d) dithien-HD/TiO2/FTO (e) diEDOT-EH/ITO (f) diEDOT-EH/TiO2/FTO, inset cyclic
voltammogram of TiO2/FTO (g) diEDOT-HD/ITO (h) diEDOT-HD/TiO2/FTO, inset cyclic
voltammogram of TiO2/FTO. Scan rate 0.1V/s.
3
4
5
[27]

ACCEPTED MANUSCRIPT
1
2
Figure 7. Schematic energy levels diagram for the dye sensitized TiO2/FTO/Glass samples.
3
4
Effect of substrate.
5
In order to investigate the influence of substrates on the electrochemical characteristics of the
compounds, two different materials were used: (i) “standard” ITO commonly used as substrate in
CV experiments of thin films and (ii) np-TiO2/FTO as used in DSSCs simulating thus the
behavior of the compounds directly within the DSSC devices. In the latter case, the dyes were
anchored on the TiO2 surface as in the preparation of the TiO2 photoanodes.
6
7
8
9
10
11
12
13
14
15
16
At first the plain substrates were measured. ITO did not exhibit any significant features within
the range (-2.5V, +2.5V). On the other hand, TiO2/FTO exhibited a reduction curve as shown in
the insets of Figures 6f and 6h, which yielded the TiO2 conduction band level at -4.3eV, as is
schematically shown in Figure 7. The optical band gap of TiO2/FTO was also measured and it
was combined with the above electrochemical result to give the valence band level of -7.6eV.
As can be seen in Table 2, no significant changes were detected in the energy gap of the
compounds on different substrates. The small differences observed are within the experimental
[28]