Christian B Nielsen1, Andrew J P White1, Iain McCulloch1,2. 1. †Department of Chemistry and Centre for Plastic Electronics, Imperial College London, London SW7 2AZ, United Kingdom. 2. ‡Physical Sciences and Engineering Division, King Abdullah University of Science and Technology (KAUST), Thuwal 23955-6900, Saudi Arabia.
Abstract
The 4,7-dithieno-2,1,3-benzothiadiazole (DTBT) moiety and its fluorinated counterpart are important π-conjugated building blocks in the field of organic electronics. Here we present a combined experimental and theoretical investigation into fundamental properties relating to these two molecular entities and discuss the potential impact on extended π-conjugated materials and their electronic properties. While the fluorinated derivative, in the solid state, packs with a cofacial overlap smaller than that of DTBT, we report experimental evidence of stronger optical absorption as well as stronger intra- and intermolecular contacts upon fluorination.
The 4,7-dithieno-2,1,3-benzothiadiazole (DTBT) moiety and its fluorinated counterpart are important π-conjugated building blocks in the field of organic electronics. Here we present a combined experimental and theoretical investigation into fundamental properties relating to these two molecular entities and discuss the potential impact on extended π-conjugated materials and their electronic properties. While the fluorinated derivative, in the solid state, packs with a cofacial overlap smaller than that of DTBT, we report experimental evidence of stronger optical absorption as well as stronger intra- and intermolecular contacts upon fluorination.
The highly electron-deficient
2,1,3-benzothiadiazole unit is one
of the most popular building blocks in organic electronics. Especially
when designing molecularly hybridized push–pull type materials,
where the alternating arrangement of electron-rich and electron-deficient
units along the π-conjugated backbone effectively controls the
frontier molecular orbitals, 2,1,3-benzothiadiazole is often the electron-deficient
unit of choice.[1−4] To further fine-tune the frontier molecular orbitals as well as
other important materials parameters such as solubility and crystallinity,
chemical modifications of the 2,1,3-benzothiadiazole unit have attracted
much interest.[5−8] In particular, fluorination has been shown to be an effective way
to lower the highest occupied molecular orbital (HOMO) and thus obtain
a better performing material for organic photovoltaics because of
an improved open-circuit voltage.[9−11]While the effect
of fluorination on the frontier energy levels
is well-understood,[9,11] experimental details about the
underlying reasons for observed differences relating to solubility,
crystallinity, charge carrier mobility, and bulk heterojunction blend
morphologies with fullerene acceptors are lacking. Here, we compare
the two well-known chromophores 4,7-dithieno-2,1,3-benzothiadiazole
(DTBT) and its fluorinated derivative, 5,6-difluoro-4,7-dithieno-2,1,3-benzothiadiazole
(DTF2BT), depicted in Figure 1. A detailed
study of these two simple model compounds, which are frequently occurring
building blocks in numerous high-performing organic electronic materials,
highlights some important changes in physical properties upon fluorination.
Figure 1
Molecular structures of DTBT (R = H) and DTF2BT (R = F)
with different
coplanar geometries defined by dihedral (torsional) angles θA and θB.
Results
and Discussion
DTBT and DTF2BT were synthesized according
to literature procedures,
and single crystals were obtained from slow evaporation of hexane
and toluene solutions, respectively. As shown in Figure 1, three coplanar conformations exist for the two chromophores.
In the obtained crystals, both DTBT and DTF2BT show a strong preference
for the trans–cis conformer as illustrated in Figure 2.
Figure 2
Crystal structures of DTBT and DTF2BT with major occupancy orientations
displayed (see the Supporting Information for the minor occupancy orientations).
Molecular structures of DTBT (R = H) and DTF2BT (R = F)
with different
coplanar geometries defined by dihedral (torsional) angles θA and θB.Crystal structures of DTBT and DTF2BT with major occupancy orientations
displayed (see the Supporting Information for the minor occupancy orientations).Both crystal structures are disordered though. In the DTBT
crystal,
approximately 71% of the molecules are in the trans–cis conformation,
while the remaining molecules adopt the cis–cis conformation.
The DTF2BT crystal, meanwhile, shows minor occupancies of both the
trans–trans (∼10%) and cis–cis (∼27%)
conformations. For trans–cis DTBT, dihedral angles θA and θB are 2.03° and 4.26°, respectively,
while a slightly higher degree of coplanarity is observed for trans–cis
DTF2BT with dihedral angles of 0.52° and 3.18°, respectively,
as illustrated in Figure 2. The increased coplanarity
of DTF2BT is reflected in short intramolecular contacts; in particular,
the S–F and S–N distances (2.71 and 2.84 Å, respectively)
are significantly shorter than the sum of the van der Waals radii.
For comparison, the S–H and S–N intramolecular distances
are 2.64 and 2.88 Å, respectively, in DTBT. Although the thiophenes
are obviously disordered in both crystals, the stronger relative representation
of the trans conformation in the DTF2BT crystal as well as the higher
degree of coplanarity and the short intramolecular S–F contacts
are all indications of a planarizing effect from the fluorination
of benzothiadiazole.In contrast to the single-crystal structures,
quantum mechanical
calculations using Gaussian at the B3LYP/6-31G* level of theory predict
the trans–trans conformation to be most stable for both DTBT
and DTF2BT, with the trans–cis conformer being destabilized
by 2.8 and 2.0 kJ mol–1, respectively (Table S1
of the Supporting Information). While the
torsional degree of freedom has previously been studied theoretically
for the two model compounds,[11] we used
variable-temperature 1H NMR to investigate this aspect
experimentally. Although the spectral line broadening upon cooling
(Figure 3) appears to be slightly more pronounced
for DTF2BT than for DTBT, we were not able to reach the coalescence
point in deuterated chloroform at 213 K or in deuterated tetrahydrofuran
at 183 K, where kT is on the order of 1.5 kJ mol–1 (Figures S1–S3 of the Supporting Information). This indicates that the activation
barrier for rotation (θA and θB)
in solution is significantly smaller than what is predicted theoretically
in vacuum.[11,12]
Figure 3
Aromatic region of 1H NMR spectra
of DTBT (left) and
DTF2BT (right) recorded in CDCl3 at 293, 263, and 233 K.
Aromatic region of 1H NMR spectra
of DTBT (left) and
DTF2BT (right) recorded in CDCl3 at 293, 263, and 233 K.UV–vis spectroscopy was
used to investigate the optical
properties of DTBT and DTF2BT in solution as illustrated in Figure 4 and summarized in Table 1. Both model compounds show two absorption bands around 300 and 400–450
nm. While the high-energy absorption band coincides for the two compounds
and is unaffected by changes in solvent polarity, the low-energy absorption
feature is red-shifted approximately 20 nm for DTBT compared to that
for DTF2BT. Moreover, DTF2BT shows a stronger solvatochromic effect
with a 10 nm blue-shift when going from n-hexane
to acetonitrile. We also note that the molar extinction coefficient
[at both absorption features (Table 1)] is
consistently higher for DTF2BT than for DTBT, which of course is of
paramount importance when considering these materials for organic
photovoltaic applications.[8]
Figure 4
Normalized UV–vis
spectra of DTBT (solid lines) and DTF2BT
(dashed lines) in hexane (black), dichloromethane (blue), and acetonitrile
(red) solutions.
Table 1
Experimental and Theoretical Optical
Properties of DTBT and DTF2BT in Solution
solvent
λmax (nm), experimental
ε (×103 M–1 cm–1), experimental
λmax (nm), theoreticala
oscillator strength,
theoreticala
DTBT
n-hexane
306, 446
28.5, 15.5
325, 531
0.663, 0.364
dichloromethane
309, 445
28.9, 15.2
325, 531
0.638, 0.371
acetonitrile
306, 441
28.8, 14.9
325, 529
0.639, 0.357
DTF2BT
n-hexane
304, 428
33.4,
16.7
323, 518
0.771, 0.367
dichloromethane
306, 424
31.0, 16.5
323, 513
0.753, 0.386
acetonitrile
302, 418
29.2, 15.8
322, 510
0.741, 0.375
Determined using Gaussian at the
B3LYP/6-31G* level of theory.
Quantum
mechanical calculations were again used to support the
experimental details, and as depicted in Table 1, there is a fairly good agreement between the experimental and theoretical
optical properties. The optical transition around 300 nm is predominantly
from the HOMO to LUMO+1 transition and has π–π*
character, while the lower-energy feature is dominated by the HOMO
to LUMO transition with strong intramolecular charge transfer (ICT)
character. This also explains why the solvatochromism is observed
for only the low-energy absorption band.Normalized UV–vis
spectra of DTBT (solid lines) and DTF2BT
(dashed lines) in hexane (black), dichloromethane (blue), and acetonitrile
(red) solutions.Turning our attention
to the crystal packing of DTBT and DTF2BT
(Figure 5 and Table 2), we note that DTBT is an orthorhombic crystal, while DTF2BT is
distorted to a monoclinic crystal system with a β angle of 104.82°.
Viewing the two crystals along their a-axes reveals
that both pack in a herringbone type arrangement as often seen for
small π-conjugated molecules.[13] As
illustrated in Figure 5C, the interplanar distances
(d2) between adjacent molecules are nearly
identical for the two structures with values of 3.41–3.44 Å
for DTBT and 3.44–3.46 Å for DTF2BT. The slipping distances
(d1) are slightly larger for DTF2BT (0.94–0.95
Å) than for DTBT (0.75–0.79 Å), which could potentially
affect the charge transport in DTF2BT-based materials adversely.[14,15]
Figure 5
Crystal packing of DTBT (left) and DTF2BT (right)
viewed along
the a-axis (A) and b-axis (B) and
showing the slipping distance (d1) and
the interplanar distance (d2) (C).
Table 2
Crystal Systems, Space Groups, and
Lattice Parameters for DTBT and DTF2BT Crystals
DTBT
DTF2BT
crystal
system
orthorhombic
monoclinic
space group
Pbca
P21/c
a, b, c (Å)
12.73, 9.88, 20.05
15.73, 4.81,
17.75
α, β, γ (deg)
90, 90, 90
90, 104.82, 90
Determined using Gaussian at the
B3LYP/6-31G* level of theory.Crystal packing of DTBT (left) and DTF2BT (right)
viewed along
the a-axis (A) and b-axis (B) and
showing the slipping distance (d1) and
the interplanar distance (d2) (C).Calculated permanent dipole moments
for the three different coplanar
conformations for both molecules are depicted in Figure 6. The major occupancy trans–cis conformations have
dipole moments of 1.18 D (DTBT) and 0.70 D (DTF2BT). Importantly,
for DTBT, the dipole moment is directed away from the electron-deficient
thiadiazole ring, while the dipole moment of DTF2BT, because of its
highly electron-withdrawing fluorine substituents, is directed toward
the thiadiazole ring as illustrated in Figure 6. For DTBT, the other conformation present in the crystal (cis–cis)
has an even stronger dipole moment of 2.21 D in the same direction,
while the two minor conformations of DTF2BT have oppositely directed
dipole moments.[16]
Figure 6
Calculated
dipole moments for the three coplanar conformations
of DTBT and DTF2BT (A), the alignment of dipole moments in the π-stacking
direction (B), and intermolecular H–F interactions in the DTF2BT
crystal (C).
We believe that
the large dipole moment of DTBT and the resulting
strong dipole–dipole interactions are responsible for the antiparallel
alignment of adjacent DTBT molecules depicted in Figure 6B. DTF2BT, on the other hand, has a much smaller permanent
dipole moment, and adjacent molecules are consequently observed to
align in a parallel fashion along the π-stacking direction (Figure 6B). The change in direction of the dipole moment
for DTF2BT with different conformations could potentially account
for some stabilizing dipole–dipole interactions in the crystal,
but we find it unlikely to be the major driving force for this parallel
packing. Instead, we note that adjacent π-stacks of DTF2BT have
particularly close H–F contacts (2.51 Å) as illustrated
in Figure 6C. Each DTF2BT molecule partakes
in two intermolecular H–F interactions, while no similar intermolecular
interactions could be observed for DTBT. Although we believe that
these observations are the major factors governing the molecular packing
motifs of DTBT and DTF2BT, it is worth noting that there are other
potential intermolecular interactions that could affect the molecular
packing.[17]Taking into account the
different conformations and their different
permanent dipole moments, we think it is worth noting that the polarity
of the solvent used during solution processing of DTBT- and DTF2BT-containing
materials is likely to affect the distribution of conformations and
thus also the solid state packing.[16] In
that context, it is also worth noting that the two crystal structures
are obtained from solvents with slightly different polarities.Calculated
dipole moments for the three coplanar conformations
of DTBT and DTF2BT (A), the alignment of dipole moments in the π-stacking
direction (B), and intermolecular H–F interactions in the DTF2BT
crystal (C).The thermal properties
of DTBT and DTF2BT were investigated with
differential scanning calorimetry (DSC). In both cases, only one phase
transition from solid to isotropic melt was observed in the temperature
range of 0–300 °C as illustrated in Figure 7. DTBT has a melting point of 119.9 °C with an enthalpy
of melting of 23.0 kJ/mol, while DTF2BT melts at 213.6 °C with
an enthalpy of melting of 25.8 kJ/mol. As the two chromophores have
similar molecular symmetries and comparable conformational disorders
in terms of cis/trans isomerism, the much higher melting point of
DTF2BT compared to that of DTBT is most likely to stem from stronger
intermolecular forces in the case of DTF2BT. This is in good agreement
with the sharper crystallization peak for DTF2BT upon cooling as well
as the crystal packing discussed above. Although DTBT shows intermolecular
dipole–dipole interactions, these are likely weak because of
the slip-stacked nature of the packing with each 2,1,3-benzothiadiazole
unit interacting with neighboring thiophene units rather than neighboring
2,1,3-benzothiadiazole units, which is evident from Figure 5C. DTF2BT, on the other hand, shows strong intermolecular
H–F interactions as well as intramolecular S–F and S–N
interactions that help to minimize conformational disorder; interactions
that are likely to be factors greatly contributing to the stronger
crystal lattice of DTF2BT.
Figure 7
DSC traces of DTBT (---) and DTF2BT (—)
recorded at 10 °C
min–1 under nitrogen.
DSC traces of DTBT (---) and DTF2BT (—)
recorded at 10 °C
min–1 under nitrogen.
Conclusions
In summary, to unambiguously compare the important
chromophores
DTBT and DTF2BT, a task that is often obscured by other parameters
such as molecular weight differences and solubility issues upon comparison
of the corresponding polymers, their crystal structures were determined
and thoroughly analyzed in this work. These data were supported by 1H NMR, UV–vis, and DSC data as well as quantum mechanical
calculations. While DTBT packs with a slightly better cofacial overlap,
DTF2BT shows more prominent intra- and intermolecular interactions,
which can be particularly important for long-range charge transport
in organic electronic materials. 1H NMR studies furthermore
indicated that the flanking thiophene groups can rotate freely even
at low temperatures, which is in agreement with the conformational
disorder observed for both structures in the solid state. UV–vis
spectroscopy showed that DTF2BT is slightly blue-shifted and has a
molar absorptivity higher than that of DTBT, which is of obvious importance
for photovoltaic applications. DTBT has a permanent dipole moment
significantly larger than that of DTF2BT, which affects the solid
state packing and is also thought to affect the distribution of conformational
orientations in solutions of varying polarity.
Authors: Sankar Subramanian; Sung Kyu Park; Sean R Parkin; Vitaly Podzorov; Thomas N Jackson; John E Anthony Journal: J Am Chem Soc Date: 2008-02-09 Impact factor: 15.419
Authors: Xinran Zhang; Hugo Bronstein; Auke J Kronemeijer; Jeremy Smith; Youngju Kim; R Joseph Kline; Lee J Richter; Thomas D Anthopoulos; Henning Sirringhaus; Kigook Song; Martin Heeney; Weimin Zhang; Iain McCulloch; Dean M DeLongchamp Journal: Nat Commun Date: 2013 Impact factor: 14.919
Authors: Hoi Nok Tsao; Don M Cho; Insun Park; Michael Ryan Hansen; Alexey Mavrinskiy; Do Y Yoon; Robert Graf; Wojciech Pisula; Hans Wolfgang Spiess; Klaus Müllen Journal: J Am Chem Soc Date: 2011-02-03 Impact factor: 15.419
Authors: Nicholas E Jackson; Brett M Savoie; Kevin L Kohlstedt; Monica Olvera de la Cruz; George C Schatz; Lin X Chen; Mark A Ratner Journal: J Am Chem Soc Date: 2013-07-09 Impact factor: 15.419
Authors: Bob C Schroeder; Raja Shahid Ashraf; Stuart Thomas; Andrew J P White; Laure Biniek; Christian B Nielsen; Weimin Zhang; Zhenggang Huang; Pabitra Shakya Tuladhar; Scott E Watkins; Thomas D Anthopoulos; James R Durrant; Iain McCulloch Journal: Chem Commun (Camb) Date: 2012-06-29 Impact factor: 6.222
Authors: O Kwon; V Coropceanu; N E Gruhn; J C Durivage; J G Laquindanum; H E Katz; J Cornil; J L Bredas Journal: J Chem Phys Date: 2004-05-01 Impact factor: 3.488
Authors: Christian B Nielsen; Raja Shahid Ashraf; Neil D Treat; Bob C Schroeder; Jenny E Donaghey; Andrew J P White; Natalie Stingelin; Iain McCulloch Journal: Adv Mater Date: 2014-12-15 Impact factor: 30.849
Authors: Benedito A L Raul; Yuriy N Luponosov; Wenyan Yang; Nikolay M Surin; Olivier Douhéret; Jie Min; Thomas L C Jansen; Sergei A Ponomarenko; Maxim S Pshenichnikov Journal: Sci Rep Date: 2020-12-03 Impact factor: 4.379