Journal of Luminescence 205 (2019) 7–13

Contents lists available at ScienceDirect

Journal of Luminescence
journal homepage: www.elsevier.com/locate/jlumin

The effects of dye aggregation on the performance of organic dyes in dye- T

sensitized solar cells: From static model to molecular dynamics simulation
⁎ ⁎
Weiyi Zhang, Yue Zhang, Huishuang Su, Xinrui Zhu, Li Wang , Jinglai Zhang
College of Chemistry and Chemical Engineering, Henan University, Kaifeng, Henan 475004, PR China

A R T I C LE I N FO A B S T R A C T

Keywords: The properties of three D1-π-D2-A dyes with dithiafulvenyl as donor (D1), phenothiazine as ancillary donor
Dye-sensitized solar cells (D2), cyanoacrylic acid as acceptor along with different π groups, diphenyl for 1, dithienyl for 2, and thio-
Dye-TiO2 adsorbed system pheneyl-benzothiadiazole for 3, are studied. Besides the isolated dyes, the monomeric and dimeric adsorptions
Aggregation are also studied by combination of first principle and molecular dynamic simulations. There is no distinct dif-
Electronic coupling
ference among three dyes on the basis of only isolated dyes. When the dye-TiO2 adsorption is considered, 3 is
Molecular dynamics
excluded because of its inferior items. However, it is still a difficult task to differentiate 1 and 2 even if the
aggregation effect is included. The fluctuation of 1 is larger than that of 2 during the dynamic simulation
resulting in the larger electronic coupling. The aggregation, especially for the aggregation during the dynamic
simulation, is necessary to evaluate the performance of organic dye. Additionally, a new D-π-A dye 4 with the
identical donor, π-bridge and acceptor is designed to compare with 2. 2 is totally superior to 4 indicating that
suitable groups and appropriate configuration are both important to enhance the performance of organic dye.

1. Introduction along with cyanoacrylic acid as acceptor linked through different π-

As the important means for conversion solar radiation to electricity
directly, dye-sensitized solar cells (DSSCs) have attracted extensive
interests due to easy fabrication, high efficiency, and low costs [1].
Further improvement of the efficiency is the prerequisite to accomplish
the large scale applications. Although optimization of each component
would improve the overall performance, the sensitizers play the most
essential role in increasing the power conversion efficiency [2].
Metal-free organic dyes have been the powerful alternatives for
polypyridyl Ru(II) complex and zinc porphyrin dyes because of their
cheap raw materials, facile synthesis, and easily tunable structures [3].
D-π-A (donor-π-acceptor) is the most common configuration for organic
dyes, which is favorable for intramolecular charge transfer leading to a
high efficiency.
As a smaller version of the fulvene family, dithiafulvenyl (DTF) is
considered as a good choice to be terminal electron-donating group in
D-π-A system for DSSC applications [4]. As a strong electron donor, it
facilitates the electron injection from excited state dye molecules to the
conduction band of semiconductor. Its non-planar molecular config-
uration is also helpful to suppress the aggregation and the recombina-
tion of injected electrons with the redox couple [4]. In 2012, DTF is
firstly taken as donor to synthesize organic dyes using D-π-A approach
bridges [5]. After that, the design of D-π-A sensitizers for DSSCs is ac-
celerated by employment of DTF as donor.
Phenothiazine (PTZ) derivatives are also promising donors in or-
ganic sensitizers in DSSCs owing to their excellent electron donating
abilities. Their butterfly configurations would impede molecular ag-
gregation resulting in better electron separation. Lots of D-π-A organic
dyes have been developed with PTZ derivatives as donor [6].
Since both DTF and PTZ are potential candidates for high-efficiency
dyes, Jia et al. have constructed new organic dyes, WD15 (See Fig. S1),
by introduction of DTF and PTZ units as building blocks simultaneously
achieving the best efficiency of 6.63% [6]. Besides the chemical prop-
erties of components, the configuration is also the other critical item to
affect the performance. On the basis of WD15, different π groups are
inserted between DTF and PTZ to develop new organic dyes with D1-π-
D2-A configuration [7].
A great variety of components as well as configurations results in
substantial organic dyes. To develop organic dyes with desired prop-
erties, it is necessary to build an efficient structure-property relation-
ship. The performance of DTF-C sensitizers (See Fig. S1) is quantita-
tively evaluated [5]. Moreover, theoretical studies have been included
in most of experimental investigations to uncover the electron dis-
tribution and variation of energy level of frontier molecular orbital

⁎
Corresponding authors.
E-mail addresses: chemwangl@henu.edu.cn (L. Wang), zhangjinglai@henu.edu.cn (J. Zhang).

https://doi.org/10.1016/j.jlumin.2018.08.084
Received 15 May 2018; Received in revised form 3 August 2018; Accepted 30 August 2018
Available online 31 August 2018
0022-2313/ © 2018 Elsevier B.V. All rights reserved.
W. Zhang et al.
Scheme 1. The sketch map structures of all dyes.
(FMO). However, the influence of aggregation is not included to con-
sider the performance of organic dyes, which is an important item
especially for PTZ and DTF units.
On the basis of 1, other two organic dyes, 2, and 3, are theoretically
designed with dithienyl and thiopheneyl-benzothiadiazole as π groups
(See Scheme 1) to explore the influence of π group. Besides the isolated
dyes and dyes adsorbed on TiO2 surface, the influence of dye ag-
gregation on the performance is also investigated. Furthermore, the
dynamic property is also calculated with the goal to build more precise
structure-property relationship.
2. Computational details
All ground state geometries of the isolated dyes were optimized by
Becke's three-parameter nonlocal-exchange functional with the non-
local correlation of Lee-Yang-Parr method (B3LYP) [8,9] with the
6–31 G(d,p) basis sets [10]. The positive vibrational frequencies testi-
fied that the optimized structures were local minima on their respective
potential energy surfaces. Based on above results, the time-dependent
density functional theory (TD-DFT) was utilized to simulate the ab-
sorption spectra. A series of different functionals, B3LYP, CAM-B3LYP
[11], LC-BLYP [12,13], M06-2 ×[14], and PBE0 [15] were performed
for dye 1 with the polarized continuum model (PCM) [16,17] in tet-
rahydrofuran (THF) medium to determine the more suitable functional
(See Fig. S2). As compared with the experimental result, both the
8
Journal of Luminescence 205 (2019) 7–13
absorption shapes and the absorption bands simulated by CAM-B3LYP
method are more accurate and reasonable. Therefore, the TD-CAM-
B3LYP functional was employed to study the absorption spectra for
other two dyes. All aforementioned calculations were carried out by the
Gaussian 09 software [18].
All periodic DFT calculations were carried out by the plane-wave
code VASP (Vienna ab initio simulation package) [19–21] with the
generalized gradient approximation (GGA) using Perdew-Burke-Ern-
zerhof (PBE) exchange-correlation function, supplemented with Grim-
me's D2 dispersion correction [22,23]. The energy cutoff was set to be
400 eV and the optimization would stop when the force on each atom
was smaller than 0.1 eV/Å. Both monomeric adsorption form and π-π
dimeric adsorption form were considered. To gain insight into the in-
terfacial properties between dye and TiO2, the 4 × 6 × 6 and 6 × 8 × 6
TiO2 anatase (101) supercells for monomer-adsorption and π-π dimeric-
adsorption, respectively, were prepared as the adsorption surface. A
vacuum buffer space of 21, 23, and 44 Å was added in x-, y-, and z-
direction for monomer-adsorption, and 31, 31, and 40 Å was added in
x-, y-, and z-direction for π-π dimeric-adsorption, respectively. Fur-
thermore, the densities of states (DOS) and projected density of states
(PDOS) were calculated at the same level to deeply analyze the ad-
sorption properties.
For the dyes adsorbed on the TiO2 surface, we carried out molecular
dynamics simulations for 10 ps with an integration time step of 1 fs at
T = 298 K in the constant volume constant internal energy (NVE en-
semble). The dynamic simulation was implemented in the DFTB+ 1.3
program package [24]. The SK-parameters of mio-1–1 [25–27] for C, N,
O, H, and S and tiorg-0–1 [28] for Ti were employed.
3. Results and discussion
3.1. Isolated dyes
Three investigated dyes have the similar structural parameters since
the only difference for them is the π-bridge, which is tabulated in Table
S1. The corresponding energy levels of FMOs are shown in Fig. 1. In the
one hand, the lowest unoccupied molecular orbital (LUMO) energy of
the dye is higher than the energy level of TiO2 conduction band
(-4.0 eV) [29], which would provide enough force for electron injec-
tion. In the other hand, the highest occupied molecular orbital (HOMO)
energy of the dye is more stable than the Nernst potential of iodine/
triiodide redox electrolytes (-4.8 eV) [30], which ensures the oxidized
sensitizers to accept the electrons from the electrolytes to accomplish
the whole cell circuit. When the diphenyl group is replaced by dithienyl
or thiopheneyl-benzothiadiazole, the LUMO energy level is stabilized
and the HOMO energy level is inverse. As a result, the HOMO-LUMO
energy gap is narrowed, which is helpful to enhance the light har-
vesting.
Fig. 1. Energy diagram of HOMO and LUMO for all dyes calculated at the CAM-
B3LYP/6–31 G(d,p) level of theory, TiO2, and the electrolyte.
W. Zhang et al.
Fig. 2. The UV/Vis absorption spectra of all dyes calculated in tetrahydrofuran
solvent at the CAM-B3LYP/6–31 G(d,p) level of theory.
On the basis of the optimized geometries, the simulated absorption
spectra are presented in Fig. 2. The maximum absorption wavelength
(λmax) of 1 (389 nm) is in good agreement with the experimental result
(382 nm). The λmax(s) of 2 and 3 are both red-shifted as compared with
that of 1 suggesting the wider light harvesting region. Although the
λmax of 3 is larger than that of 2, the absorption strength of former is
smaller. According to the following equation [31]:
LHE = 1 − 10 −f (1)
the light harvesting efficiency (LHE) is proportional to the oscillator
strength of the dye at the λmax. The weaker absorption strength in-
dicates the less power conversion efficiency (PCE).
3.2. Overall efficiency
The overall conversion efficiency (η) is related with Jsc, Voc, and FF:
Voc Jsc
η = FF
Is (2)
in which Jsc is short-circuit current density, Voc is open-circuit voltage,
and FF is fill factor. Normally, η is determined by both Jsc and Voc, since
FF is an incidental factor. It is impossible to calculate the accurate va-
lues of Jsc and Voc, therefore, some related items are developed to
qualitatively determine the PCE including ΔGinject, λtotal, and LHE. Their
detailed definitions please refer to references 31–36. The corresponding
values are listed in Table 1. The 2 has the most negative ΔGinject in-
dicating the more sufficient force to promote electron injection
from excited dyes to the conduction band (CB) of TiO2. At the
same time, it also has the larger λtotal suggesting the smaller charge
transfer rate constant (k) on the basis of the equation of
k = V 2 (π /ℏ2kB Tλtotal )1/2 exp(−λtotal /4kB T ) [37]. The contradictory re-
sults would also be suitable for other two dyes 1 and 3. Moreover, 3 has
the smallest Voc, which is also unfavorable for the improvement of PCE.
In general, it is difficult to make a clear sequence for the PCE of three
dyes since no one has all superiority. Even if a relative sequence is
made, the difference between various items is too small to distinguish
them. Some of them are even in an acceptable computational error. To
Table 1
Estimated electrochemical parameters for the dyes.
Dye dye
EOX (eV) dye *
EOX (eV) ΔGinject(eV)
1 6.34 3.15 − 0.85
2 5.99 3.07 − 0.93
3 6.07 3.54 − 0.46
4 6.31 3.79 − 0.21
Journal of Luminescence 205 (2019) 7–13
build a reliable structure-property relationship, the dye-TiO2 adsorbed
system should be studied rather than only the isolated dyes.
3.3. Adsorbed dyes onto the TiO2
When the dye is adsorbed on the TiO2 surface, the energy gap would
be varied. Correspondingly, some related properties would also be
changed. It is necessary to study the dye-TiO2 adsorbed system, which
is more close to the real environment than the isolated dye. The bi-
dentate bridging adsorption model via the carboxyl group is employed
along with one proton transferred to nearby surface oxygen. The TiO2
(101) surface is employed since it is the most stable and popular surface
[38]. The optimized adsorption structures simulated by the first prin-
ciple are plotted in Fig. 3. The adsorption has negligible effect on the
structures of dye. The adsorption energy (Eads ) of 3 is more positive
than that of 1 and 2 (See Table 2) indicating the weaker binding, which
is unfavorable for the electron injection. The Eads is defined by the
following relationship: as Eads = Edye + TiO2 − (Edye + ETiO2) , where Edye ,
ETiO2 , and Edye + TiO2 refer to the total energies of the isolated dye, TiO2,
and dye-TiO2 complex, respectively.
After adsorption, the energy gap is greatly varied as compared with
the isolated dyes. On the basis of the optimized dye-TiO2 adsorbed
system, DOS and PDOS for the pure and adsorbed dyes are shown in
Fig. 4 along with those of TiO2 surface. After adsorption, the LUMO of
adsorbed dye moves up to more positive values. In contrast, the CB of
TiO2 shifts towards the inverse direction. As a result, the energy gap
between the LUMO of adsorbed dye and the CB of TiO2 is enlarged as
compared with that of isolated dye. However, 3-TiO2 adsorption system
has the smallest energy gap between the LUMO of adsorbed dye and the
CB of TiO2 with the value of 0.99 eV, which is not beneficial for the
electron injection. To further confirm the inferior performance of
electron from the excited dye to the CB of TiO2, the electron injection
time (in fs) is calculated [39–41]. 3 has the longer electron injection
time of 1.66 fs than other two dyes with the values of 1.23 fs for 1 and
1.21 fs for 2 indicating the worst electron injection efficiency.
For isolated dyes, 3 still has some advantageous to improve the PCE
including the maximum absorption wavelength and the smallest λtotal.
When the dye-TiO2 adsorption system is considered, 3 is firstly ex-
cluded due to all inferior items. There is an interesting thing that the
performance of 1-TiO2 is better or comparable as compared to that of 2-
TiO2 for all employed factors. However, the difference between them is
too small to differentiate them. Therefore, other factors should be fur-
ther considered.
3.4. Comparison with D-π-A configuration dye
Although 1 and 2 have the different π groups, they have the same
performance as far. If the same π group is employed in different con-
figurations, the performance would be kept or varied? A new D-π-A dye
4 is designed with the same donor, π group, and acceptor to construct 2.
The corresponding results are also shown in Table 1 and Table 2. For
the isolated 4, it has the more positive ΔGinject, the larger λtotal, and the
smaller LHE than 2 indicating the less PCE. Even if the adsorbed system
is considered, 4 still has the more positive adsorbed energy, the smaller
energy gap between the LUMO of dye and the CB of TiO2, and the
longer electron injection time than 2. The 4-TiO2 system has the most
LHE Voc(eV) λh(eV) λe(eV) λtotal(eV)
0.86 2.72 0.32 0.31 0.63
0.99 2.69 0.39 0.30 0.68
0.92 2.37 0.33 0.24 0.57
0.97 2.23 0.33 0.42 0.75
9
W. Zhang et al. Journal of Luminescence 205 (2019) 7–13

Table 2
The calculated adsorption energy (Eads), H-L energy gaps before (Eg1) and after
adsorption (Eg2) on TiO2 surface, energy difference between the CB of TiO2 and
the LUMO of dye before (ΔE1) and after adsorption (ΔE2) on TiO2 surface and
electron injection time τinj for monomer.
Dye Eads (eV) Eg1 (eV) Eg2 (eV) ΔE1 (eV) ΔE2 (eV) τinj (fs)

Monomer-1 − 4.46 0.55 0.70 0.12 1.24 1.23

Monomer-2 − 3.80 0.16 0.60 0.32 1.18 1.21
Monomer-3 − 2.49 0.41 0.43 0.20 0.99 1.66
Monomer-4 − 1.96 0.51 0.56 0.14 1.09 1.50

Fig. 3. The graph is the adsorption configuration for dye-TiO2 complexes of

Monomer-1, Monomer-2, and Monomer-3, respectively.

positive adsorbed energy, which is even worse than 3-TiO2 system.

However, aforementioned other two items of dye 4 are still better than
those of dye 3. It is testified that the suitable group would play more Fig. 4. Calculated total density of states (DOS) and projected density of states
important role in enhancing PCE especially when it is combined with (PDOS) for clean TiO2 surface and interfaces of TiO2 adsorbed with 1-monomer
appropriate configuration. (a), 2-monomde (b), and 3-monomer (c).

10
W. Zhang et al. Journal of Luminescence 205 (2019) 7–13

Table 3
The calculated adsorption energy (Eads), H-L energy gap before (Eg1) and after
adsorption (Eg2) on TiO2 surface, energy difference between the CB of TiO2 and
the LUMO of dye before (ΔE1) and after adsorption (ΔE2) on TiO2 surface, and
the electron injection time τinj for π-π dimeric architecture.
Dye Eads (eV) Eg1 (eV) Eg2 (eV) ΔE1 (eV) ΔE2 (eV) τinj (fs)

D-1 − 37.77 0.40 0.58 0.21 1.09 1.29

D-2 − 36.09 0.08 0.44 0.37 1.04 0.68

3.5. Dyes aggregation

Besides to improve the electron injection and transfer, the reduction

of deactivation process is the other pathway to enhance the overall
performance indirectly. Aggregation is an inevitable deactivation
pathway, which should be carefully considered. Although some re-
searches have reported it previously, it is rarely to be included in the
study to design new dyes.
After the π-π dimeric architecture is adsorbed on the TiO2 (101)
surface, the adsorption energy for 2 with the value of −36.09 eV is still
more positive than that for 1 (See Table 3), which is consistent with the
monomer-adsorption. Moreover, the differences between CB of TiO2
and LUMO of adsorption dye are 1.09 eV for D-1 (π-π dimeric archi-
tecture of 1) and 1.04 eV for D-2 (π-π dimeric architecture of 2). There
is no great difference between monomeric and dimeric adsorption for
not only the absolute value but also the difference between 1 and 2.
However, the injection time of D-2 (0.68 fs) is less than of D-1 (1.29 fs).
In general, there is not big difference between monomeric and dimeric
adsorption.
There is electronic coupling between stacking monomers, which is
detrimental for the electron injection into the TiO2 resulting in the less
Fig. 5. Evolution of the total electronic energy of the systems as a function of
PCE. The electronic coupling is calculated by the direct method, which
the simulation time.
has been widely employed in previous literature [42,43]. The direct
method is defined via the relationship:

0, sitea
Vab = ϕLUMO 0 0, site b 0, sitea 0, site b
/ HOMO F ϕLUMO / HOMO = ϕLUMO / HOMO h core ϕLUMO / HOMO

+∑l (occ ) ⎛⎜ ϕLUMO

0, sitea 0 0, site b − 0, sitea 0, site b 0 0 ⎞
/ HOMO ϕl ϕLUMO / HOMO ϕLUMO / HOMO ϕLUMO / HOMO ϕl ϕl ⎟
⎝ ⎠
(3)

where Vab is the charge transfer integral for the electron/hole and
0, sitea 0, site b
ϕLUMO / HOMO and ϕLUMO / HOMO represent the LUMOs/HOMOs of two ad-
jacent molecules a and b when no intermolecular interaction is pre-
sented. F° is the Fock operator and its density matrix is constructed from
orbitals of two adjacent molecules [44]. The calculated Vab for 2 is
larger than that for 1. All aforementioned items are based on the static
model, while the dye would be distorted. Therefore, it is necessary to
consider them by dynamic model.
As shown in Fig. 5, there is small fluctuation in the total simulated
10 ps except for in the initial 0.1 ps. Initially, the adsorbed dyes are all
perpendicular to the TiO2 (101) surface. Then, there is noticeable bend
Fig. 6. Intermolecular electronic couplings of 10 selected snapshots in dy-
between dye and TiO2 (101) surface and between individual component
namics. D-1 and D-2.
of dye during the dynamics simulations. Three snapshots at 0, 5, and
10 ps are extracted to illustrate the variation of structures (inset of
Fig. 5). The electronic coupling is greatly affected by the distortion of factors. However, it is still hard to differentiate 1 and 2 because of the
dye. Eleven electronic couplings are calculated with an interval of one similar performance. This problem is still kept even if the dimeric ad-
ps, which are plotted in Fig. 6. In the initial or final status, the elec- sorption model is employed. It is resolved when the dynamic properties
tronic coupling of 2 is larger than that of 1. However, the extent of is considered. The adsorption model and dynamic simulation are ne-
fluctuation for 2 is smaller than that of 1. The average electronic cou- cessary to build reliable structure-property relationship for dyes applied
pling of 2 is smaller than that of 1, which agrees well with the less in DSSCs.
electron injection time of 2, especially for the dimeric adsorption.
Three models are considered in this work to evaluate the perfor- 4. Conclusion
mance of dye. For the isolated dye, it is impossible to screen out the dye
with the better PCE since they have the contradictory items in quali- On the basis of reported 1, other two D1-π-D2-A dyes, 2 and 3, are
tative comparison. Next, when the monomeric adsorption is studied, 3 theoretical designed with different π groups. Their performance is
is excluded because of the inferior factors in almost every determined qualitatively compared by three different models to screen out the more

11
W. Zhang et al.
efficient dye. When only isolated dye is studied, no one would stand out
since no one has all superiority. After adsorption, 3 has the less per-
formance than other two dyes with the inferior items. However, 1 and 2
have the similar adsorption energy, energy gap between the CB of TiO2
and the LUMO of dye, and injection time. The corresponding results are
kept even when the dimeric adsorption is considered. However, the
distortion of dye 1 is greater than that of dye 2 resulting in the larger
average electronic coupling during the simulated time. The aggregation
plays a more important role in distinguishing the performance of dif-
ferent dyes. Moreover, the dynamic model should be considered rather
than only the static model. Another D-π-A dye, 4, is designed with the
same donor, π group, and acceptor with 2. However, its performance is
much less than 2. The suitable combination of configuration and π
group would reach the final goal.
Acknowledgements
We thank the National Supercomputing Center in Shenzhen
(Shenzhen Cloud Computing Center) for providing computational re-
sources and software. This work was supported by the National Natural
Science Foundation of China (21476061, 21503069, 21676071),
Program for He’nan Innovative Research Team in University
(15IRTSTHN005).
Appendix A. Supporting information
Supplementary data associated with this article can be found in the
online version at https://doi.org/10.1016/j.jlumin.2018.08.084.
References
[1] J. Gong, K. Sumathy, Q. Qiao, Z. Zhou, Review on dye-sensitized solar cells (DSSCs):
advanced techniques and research trends, Renew. Sust. Energ. Rev. 68 (2017)
234–246.
[2] Y.K. Eom, J.Y. Hong, J. Kim, H.K. Kim, Triphenylamine-based organic sensitizers
with π-spacer structural engineering for dye-sensitized solar cells: synthesis, theo-
retical calculations, molecular spectroscopy and structure-property-performance
relationships, Dyes Pigments 136 (2017) 496–504.
[3] Y. Wu, M. Marszalek, S.M. Zakeeruddin, Q. Zhang, H. Tian, M. Grätzel, W. Zhu,
High-conversion-efficiency organic dye-sensitized solar cells: molecular engineering
on D-A-π-A featured organic indoline dyes, Energy Environ. Sci. 5 (2012)
8261–8272.
[4] N. Duvva, U. Chilakamarthi, L. Giribabu, Recent developments in tetrathiafulvalene
and dithiafulvalene based metal-free organic sensitizers for dye-sensitized solar
cells: a mini-review, Sustain. Energy Fuels 1 (2017) 678–688.
[5] K. Guo, K. Yan, X. Lu, Y. Qiu, Z. Liu, J. Sun, F. Yan, W. Guo, S. Yang, Dithiafulvenyl
unit as a new donor for high-efficiency dye-sensitized solar cells: synthesis and
demonstration of a family of metal-free organic sensitizers, Org. Lett. 14 (2012)
2214–2217.
[6] Z. Wan, C. Jia, Y. Wang, J. Luo, X. Yao, Significant improvement of phenothiazine
organic dye-sensitized solar cell performance using dithiafulvenyl unit as additional
donor, Org. Electron. 27 (2015) 107–113.
[7] X. Zhang, F. Gou, D. Zhao, J. Shi, H. Gao, Z. Zhu, H. Jing, π-Spacer effect in di-
thiafulvenyl-π-phenothiazine dyes for dye-sensitized solar cells, J. Power Sources
324 (2016) 484–491.
[8] A.D. Becke, Density-functional thermochemistry. III. The role of exact exchange, J.
Chem. Phys. 98 (1993) 5648–5652.
[9] C. Lee, W. Yang, R.G. Parr, Development of the colle-salvetti correlation-energy
formula into a functional of the electron density, Phys. Rev. B 37 (1988) 785–789.
[10] P.M.W. Gill, B.G. Johnson, J.A. Pople, M.J. Frisch, The performance of the Becke-
Lee-Yang-Parr (B-LYP) density functional theory with various basis sets, Chem.
Phys. Lett. 197 (1992) 499–505.
[11] T. Yanai, D.P. Tew, N.C. Handy, A new hybrid exchange-correlation functional
using the Coulomb-Attenuating method (CAM-B3LYP), Chem. Phys. Lett. 393
(2004) 51–57.
[12] H. Iikura, T. Tsuneda, T. Yanai, K. Hirao, A long-range correction scheme for
generalized-gradient-approximation exchange functionals, J. Chem. Phys. 115
(2001) 3540–3544.
[13] Y. Tawada, T. Tsuneda, S. Yanagisawa, A long-range-corrected time-dependent
density functional theory, J. Chem. Phys. 120 (2004) 8425–8433.
[14] Y. Zhao, D.G. Truhlar, The M06 suite of density functionals for main group ther-
mochemistry, thermochemical kinetics, noncovalent interactions, excited states,
and transition elements: two new functionals and systematic testing of four M06-
12
Journal of Luminescence 205 (2019) 7–13
class functionals and 12 other functionals, Theor. Chem. Acc. 120 (2008) 215–241.
[15] J.P. Perdew, K. Burke, M. Ernzerhof, Generalized gradient approximation made
simple, Phys. Rev. Lett. 77 (1996) 3865–3868.
[16] M. Cossi, V. Barone, B. Mennucci, J. Tomasi, Ab initio study of ionic solutions by a
polarizable continuum dielectric model, Chem. Phys. Lett. 286 (1998) 253–260.
[17] E. Cancès, B. Mennucci, J. Tomasi, A. New, integral equation formalism for the
polarizable continuum model: theoretical background and applications to isotropic
and anisotropic dielectrics, J. Chem. Phys. 107 (1997) 3032–3041.
[18] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman,
G. Scalmani, V. Barone, B. Mennucci, G.A. Petersson, H. Nakatsuji, M. Caricato,
X. Li, H.P. Hratchian, A.F. Izmaylov, J. Bloino, G. Zheng, J.L. Sonnenberg, M. Hada,
M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda,
O. Kitao, H. Nakai, T. Vreven, J.A. Montgomery Jr., J.E. Peralta, F. Ogliaro,
M. Bearpark, J.J. Heyd, E. Brothers, K.N. Kudin, V.N. Staroverov, T. Keith,
R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J.C. Burant, S.S. Iyengar,
J. Tomasi, M. Cossi, N. Rega, J.M. Millam, M. Klene, J.E. Knox, J.B. Cross,
V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R.E. Stratmann, O. Yazyev,
A.J. Austin, R. Cammi, C. Pomelli, J.W. Ochterski, R.L. Martin, K. Morokuma,
V.G. Zakrzewski, G.A. Voth, P. Salvador, J.J. Dannenberg, S. Dapprich,
A.D. Daniels, O. Farkas, J.B. Foresman, J.V. Ortiz, J. Cioslowski, D.J. Fox, Gaussian
09, Revision C.01, Gaussian, Inc., Wallingford CT, 2010.
[19] G. Kresse, J. Furthmüller, Efficient iterative schemes for ab initio total-energy cal-
culations using a plane-wave basis set, Phys. Rev. B 54 (1996) 11169–11186.
[20] G. Kresse, J. Furthmüller, Efficiency of ab-initio total energy calculations for metals
and semiconductors using a plane-wave basis set, Comp. Mater. Sci. 6 (1996)
15–50.
[21] G. Kresse, J. Hafner, Ab initio molecular dynamics for liquid metals, Phys. Rev. B 47
(1993) 558–561.
[22] R. Elmér, M. Berg, L. Carlén, B. Jakobsson, B. Norén, A. Oskarsson, G. Ericsson,
J. Julien, T.F. Thorsteinsen, M. Guttormsen, G. Løvhøiden, V. Bellini, E. Grosse,
C. Müntz, P. Senger, L. Westerberg, K+ emission in symmetric heavy ion reactions
at subthreshold energies, Phys. Rev. Lett. 77 (1996) 4884–4886.
[23] T. Bučko, J. Hafner, S. Lebègue, J.G. Ángyán, Improved description of the structure
of molecular and layered crystals: ab initio DFT calculations with van der waals
corrections, J. Phys. Chem. A 114 (2010) 11814–11824.
[24] B. Aradi, B. Hourahine, T. Frauenheim, DFTB+, a sparse matrix-based im-
plementation of the DFTB method, J. Phys. Chem. A 111 (2007) 5678–5684.
[25] M. Elstner, D. Porezag, G. Jungnickel, J. Elsner, M. Haugk, T. Frauenheim, S. Suhai,
G. Seifert, Self-consistent-charge density-functional tight-binding method for si-
mulations of complex materials properties, Phys. Rev. B 58 (1998) 7260–7268.
[26] T.A. Niehaus, M. Elstner, T. Frauenheim, S. Suhai, Application of an approximate
density-functional method to sulfur containing compounds, J. Mol. Struct.
(Theochem) 541 (2001) 185–194.
[27] M. Gaus, Q. Cui, M. Elstner, DFTB3: extension of the self-consistent-charge density-
functional tight-binding method (SCC-DFTB), J. Chem. Theory Comput. 7 (2011)
931–948.
[28] G. Dolgonos, B. Aradi, N.H. Moreira, T. Frauenheim, An improved self-consistent-
charge density-functional tight-binding (SCC-DFTB) set of parameters for simula-
tion of bulk and molecular systems involving titanium, J. Chem. Theory Comput. 6
(2010) 266–278.
[29] R. Yan, X. Qian, Y. Jiang, Y. He, Y. Hang, L. Hou, Ethynylene-linked planar rigid
organic dyes based on indeno[1,2-b]indole for efficient dye-sensitized solar cells,
Dyes Pigments 141 (2017) 93–102.
[30] R. Maragani, R. Misra, M.S. Roy, M.K. Singh, G.D. Sharma, D-π-A)2-π-D-A type
ferrocenyl bisthiazole linked triphenylamine based molecular systems for DSSC:
synthesis, experimental and theoretical performance studies, Phys. Chem. Chem.
Phys. 19 (2017) 8925–8933.
[31] H.S. Nalwa, Handbook of Advanced Electronic and Photonic Materials and Devices:
Semiconductors 1 Academic Press, 2001.
[32] R. Katoh, A. Furube, T. Yoshihara, K. Hara, G. Fujihashi, S. Takano, S. Murata,
H. Arakawa, M. Tachiya, Efficiencies of electron injection from excited N3 dye into
nanocrystalline semiconductor (ZrO2, TiO2, ZnO, Nb2O5, SnO2, In2O3) films, J.
Phys. Chem. B 108 (2004) 4818–4822.
[33] A. Datta, S. Mohakud, S.K. Pati, Electron and hole mobilities in polymorphs of
benzene and naphthalene: role of intermolecular interactions, J. Chem. Phys. 126
(2007) 144710–144717.
[34] A. Datta, S. Mohakud, S.K. Pati, Comparing the electron and hole mobilities in the α
and β phases of perylene: role of π-stacking, J. Mater. Chem. 17 (2007) 1933–1938.
[35] V. Mohan, A. Datta, Structures and electronic properties of Si-substituted benzenes
and their transition-metal complexes, J. Phys. Chem. Lett. 1 (2010) 136–140.
[36] S. Mohakud, S.K. Pati, Large carrier mobilities in octathio[8]circulene crystals: a
theoretical study, J. Mater. Chem. 19 (2009) 4356–4361.
[37] R.A. Marcus, Electron transfer reactions in chemistry. Theory and experiment, Rev.
Mod. Phys. 65 (1993) 599–610.
[38] M. Pastore, F. De Angelis, Computational modelling of TiO2 surfaces sensitized by
organic dyes with different anchoring groups: adsorption modes, electronic struc-
ture and implication for electron injection/recombination, Phys. Chem. Chem.
Phys. 14 (2012) 920–928.
[39] P. Persson, M.J. Lundqvist, R. Ernstorfer, W.A. Goddard III, F. Willing, Quantum
chemical calculations of the influence of anchor-cum-spacer groups on femtosecond
electron transfer times in dye-sensitized semiconductor nanocrystals, J. Chem.
Theory Comput. 2 (2006) 441–451.
[40] F. Labat, I. Ciofini, H.P. Hratchian, M. Frisch, K. Raghavachari, C. Adamo, First
W. Zhang et al.
principles modeling of eosin-loaded ZnO films: a step toward the understanding of
dye-sensitized solar cell performances, J. Am. Chem. Soc. 131 (2009)
14290–14298.
[41] F. Labat, T. Le Bahers, I. Ciofini, C. Adamo, First-principles modeling of dye-sen-
sitized solar cells: challenges and perspectives, Accounts Chem. Res. 45 (2012)
1268–1277.
[42] X. Yang, Q. Li, Z. Shuai, Theoretical modelling of carrier transports in molecular
semiconductors: molecular design of triphenylamine dimer systems,
13
Journal of Luminescence 205 (2019) 7–13
Nanotechnology 18 (2007) 424029.
[43] S. Yin, Y. Yi, Q. Li, G. Yu, Y. Liu, Z. Shuai, Balanced carrier transports of electrons
and holes in Silole-based compoundss-a theoretical study, J. Phys. Chem. A 110
(2006) 7138–7143.
[44] G. Nan, L. Wang, X. Yang, Z. Shuai, Y. Zhao, Charge transfer rates in organic
semiconductors beyond first-order perturbation: from weak to strong coupling re-
gimes, J. Chem. Phys. 130 (2009) 024704.