Accepted Manuscript

Research paper

An approximate procedure for profiling dye molecules with potentials as sen-

Kayode Sanusi, Nafisat O. Fatomi, Adegoke O. Borisade, Yusuf Yilmaz, Ümit

PII: S0009-2614(19)30207-6
DOI: https://doi.org/10.1016/j.cplett.2019.03.028
Reference: CPLETT 36320

To appear in: Chemical Physics Letters

Received Date: 24 November 2018

Please cite this article as: K. Sanusi, N.O. Fatomi, A.O. Borisade, Y. Yilmaz, U. Ceylan, A. Fashina, An approximate
procedure for profiling dye molecules with potentials as sensitizers in solar cell application: A DFT/TD-DFT
approach, Chemical Physics Letters (2019), doi: https://doi.org/10.1016/j.cplett.2019.03.028

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.
An approximate procedure for profiling dye molecules with potentials as
sensitizers in solar cell application: A DFT/TD-DFT approach
Kayode Sanusi1,*, Nafisat O. Fatomi1, Adegoke O. Borisade2, Yusuf Yilmaz3, Ümit Ceylan4 and
Adedayo Fashina5

1
Department of Chemistry, Obafemi Awolowo University, Ile-Ife, Nigeria
2
Centre for Energy, Research and Development, Obafemi Awolowo University, Ile-Ife, Nigeria
3
NT Vocational School, Gaziantep University, 27310 Gaziantep, Turkey
4
Department of Medical Services and Techniques, Vocational High School Health Services,
Giresun University, 28100 Giresun, Turkey
5
School of Pharmacy and Biomolecular Sciences, Liverpool John Moores University, James
Parsons Building Byrom Street, Liverpool, L3 3AF UK

*Corresponding Author: [email protected]

Tel: +234-803-756-8619, +234-705-560-2893

Abstract

A DFT/TD-DFT method was employed to study the photovoltaic properties of some metal-free
and metal containing organic compounds (A-G) for dye sensitized solar cells (DSSCs)
application. Fluorescence emission ( ) and charge collection ( ) efficiencies were obtained by
deductive logic, and were used to determine the dyes’ incident photon conversion efficiency
(IPCE). These parameters were used to evaluate the dye’s suitability for photovoltaic technology.
The new approach could be a simple method of predicting photovoltaic activities while avoiding
the laborious experimental trial and error procedures, though it cannot be used to predict the
actual efficiency of a DSSC.

Keywords: Diffusion coefficient, photoluminescence, emission spectra, photosensitizers,

1
1. Introduction

Dye sensitized solar cells (DSSCs) are hinged on the principle of charge separation at an
interface of two materials with different electron transfer mechanisms. The device creates a
platform where the task of light absorption and charge carrier transport can be separated, unlike
the classical semiconductor which performs both functions [1,2]. Consequently, DSSCs provide
a more technically and economically viable option compared to the present-day p-n junction
photovoltaic devices. In a typical DSSC (Fig. S1, supporting document), light is absorbed by a
sensitizer (dye) grafted to the surface of TiO2 semiconductor thin film. A charge separation then
occurs at the interface between the sensitizer and the TiO 2 via a photo-induced electron transfer
from the dye to the conduction band (CB) of the semiconductor. The electron-hole pair generated
afterwards is transported to the external circuit through the charge collectors. The original state
of the dye molecule is regenerated by a redox couple system (usually an organic substance, such
as the iodide/triiodide couple), which itself is regenerated at the counter electrode by electrons
passing through the load [3].

Investigations have shown that the efficiency of the photo-induced electron injection into the
TiO2 conduction band depends strongly on the photophysicochemical properties of the dye
sensitizers [1,3]. A highly efficient photo-induced electron injection process is critical to
producing a high-performance DSSC [4,5]. This implies that the dye component of the device
usually plays a vital role in the injection process, thus, the sensitizer is undoubtedly one of the
major components of a DSSC.

Accurate theoretical prediction of the photophysicochemical properties, such as, the dye lowest
unoccupied molecular orbital (LUMO) positions relative to the TiO2 conduction band edge ( ),
pi-electron diffusion coefficients ( ), fluorescence emission factor ( ), and ground state molar
absorptivity coefficients ( ) in the visible or near infrared (near IR) region could provide useful
information for a modulated molecular architectural design of efficient dye systems. These
parameters in addition to the incident photon conversion efficiency (IPCE), charge collection
efficiency ( ), and light harvesting efficiency (LHE) would be useful in establishing the
structural requirements for prototype dye molecules in DSSC fabrication [4,6 – 13].

Notably, the DSSC is a good example of a multicomponents system (i.e. it contains the dye, the
conducting glass substrates, metal-oxide semiconductor, Pt or carbon photo-anode and an
electrolyte redox couple), where the function of the overall device is better than the sum of the
properties of each component as a consequence of the many complex interactions occurring
between them [13-20]. Although the properties predicted for an isolated component often show a
reasonable correlation with the overall performance of the cell [4,6-12], our intentions for
developing this model were not to predict the actual efficiency of a DSSC device. The aim was
to provide a guide to designing sensitive dyes at ambient temperature (298.15 K) for
employment in DSSC application.
2
Seven potential dye molecules were investigated (Fig. 1); namely, homoleptic Ru(II) complex of
4’-(trans-2-methyl-2-butenoic acid)-terpyridine (A) [21], heteroleptic Ru(II) complex of 5-(trans-
2-methyl-2-butenoic acid)-1,10-phenanthroline (B) [22], Y123 dye (C) [5], heteroleptic Ru(II)
complex of 5-(9-anthracenyl-10-(trans-2-methyl-2-butenoic acid)-1,10-phenanthroline (D) [21],
YD-2 (E) [5], YD-0 (F) and fullerene-A composite (G). It should be noted that molecules A – E
have been previously synthesized with Y123 and YD-2 photovoltaic activities already tested
[5,21,22]. Molecules F and G were derived by modifying E and A respectively.

2. Methodology

2.1. Computational details

All calculations were performed in gas phase on the Gaussian 03W program suite using the
hybrid Becke three-parameter and Lee-Yang-Parr (B3LYP) exchange and correlation functionals
respectively [23-25]. Basis function suitability for each molecule was determined by performing
geometry optimization calculations using all the basis sets available in the program suite. While
some of the basis functions did not converge and were dropped, the total energy values of those
that converged were compared, and the basis function with the lowest total energy value for each
molecule was chosen (Table S0, supporting document). The assumption is that molecules will
adopt the geometrical configuration for which their total energy is minimum. Therefore, the
method (i.e. for each molecule) that gives the minimum total energy will be appropriate in
describing the molecular ground and excited states properties. The adopted benchmarking
method was necessary to ensure that the properties of the molecules were not over- or under-
estimated resulting from the use of inappropriate basis sets. Vibrational frequency calculations
were performed with the basis set determined via optimization for molecules A – E , but
LANL2MB were used for F and G (Table S0). The signs of the frequency data were used to
classify the optimized structures as minima and maxima stationary states on the potential energy
surface (PES), or as completely unrealistic. Time-dependent self-consistent-field density
functional theory (TD-SCF DFT) method was used to determine the electronic properties of the
studied molecules.

2.2. Methods of calculation

The incident photon conversion efficiency (IPCE), charge collection efficiency ( ), electron
injection efficiency ( ) and light harvesting efficiency (LHE), have been shown to be related
by Eqs. 1 and 2 [6,12]:

(1)

3
 (2)

where f is the oscillator strength that corresponds to the maximum absorption wavelength ( )
in the visible or near IR region.

The free energy of electron injection ( ) is obtained from Eq. 3 [7],

(3)

where and are the dye’s excited state oxidation potential and the ground state reduction
potential of the semiconductor (TiO2) respectively. The employed was given in the literature
as -4.21 eV [26], while the was calculated using Eq. 4 [7].

(4)

The dye’s ground state oxidation potential is given by [10], while

The release of energy by the excited dye molecule (dye*) into the CB of the TiO2 has been
assumed to be predominantly a diffusion process. The diffusion coefficient Dπ (of the pi system)
can thus be calculated using the Stokes’ equation, Eq. 5 [27]:

(5)

where is the Boltzmann constant in J.K-1, is the minimum temperature in Kelvin (298.15 K
has been adopted as the minimum), is the dye molecular radius and is the viscosity of the
medium. Since the investigations were carried out in gas phase, the viscosity of an inert gas, He
at 300 K (20.0 × 10-6 Pa.s) which is expected to have little or no interaction with the dye was
employed [28].

Molecular radii ( ) are assumed to be equal to the dyes’ respective Onsager cavity radii, ,
which is obtained from the molecular volume, as given by Suppan’s equation [29], Eq. 6:

(6)

where is the molecular weight of the dyes, is the density of the He gas (9.00 ×10-2 kg/m3 at
STP), and is the Avogadro’s number.

4
Here, we assumed the electron injection efficiency ( ) of the dye will be equal to its
fluorescence emission factor, , which may be obtained by using Eq. 7:

(7)

where and are the integrated emission and absorption coefficients respectively,
representing the areas under the emission and absorption curves. The emission curve may be
obtained by extrapolation from the absorption spectrum, using the reverse factor ( ) described by
Eq. 8:

(8)

The reverse factor or fluorescence probability factor ( ) is a function of four components. and
are the numbers of lower and upper states molecules respectively; is the dipole moment
exchange from the ground to the excited states (unitless), and is the energy difference (in eV)
between the two states. may be defined as the fraction of the absorbed energy that is released
to the ground state via fluorescence.

Assuming that the transitions only occur between the singlet states, such that the selection rule,
is obeyed, and that bodies which absorb light must also emit a fraction of it. Then, may
be explicitly written as Eq. 9:

(9)

where and are the reverse factor and the energy difference between a lower and a
upper level respectively, and is the radiative lifetime by assuming a pulse width of 1 ns for the
incident photon. and represent the lower and the upper level respectively. Thus, the emission
coefficient ( ) can be expressed as a product of the reverse factor and the absorption
coefficient per transition, Eq. 10:

(10)

where and are the emission and absorption coefficients for each transition
from a higher to a lower state, and a lower to a higher state respectively. The emission spectrum
may then be obtained by plotting the emission coefficient values and the corresponding
wavelength q, , in nm.

5
 (11)
where corresponds to the same set of wavelengths initially absorbed.

The charge collection efficiency, , can be expressed as the ratio of the diffusion coefficient to
the square of the potential difference between the LUMO of the dye and the CB edge of the TiO 2
semiconductor, Eq. 12:

(12)

3. Results and Discussion

3.1. Optimized geometry and electronic properties

The accuracy of a predicted molecular property depends on the correctness of the optimized
structure. The optimization procedure employed here involved the use of several different basis
functions to determine the conformation of each molecule that will have the lowest energy –
conventionally known as the optimized structure. The optimized structures of the studied
molecules are given in Fig. S2 (supporting document). Validation of the optimized structures as
stable or transition state structures were done by frequency calculation. The frequency
calculation method assumes that a transition state molecule should possess at least one imaginary
bond, and so would have a negative vibrational frequency. Whereas, a stable molecule should
have no imaginary bonds, therefore, would have no negative frequency. For the seven molecules
under investigation (Fig. 1 and Fig. S2), the calculated vibrational frequencies are all positive,
indicating that they are all stable.

Figure S3 in the supporting document shows the calculated IR data of all the studied molecules.
To show that the theoretically modified structures (A, B and E transformed to G, D and F
respectively) have different vibrational frequencies from the original structures, the calculated IR
data of the respective molecules and their derivatives were compared (Fig. S4). Interestingly, the
functional group changes that occur when the fullerene became tethered to molecule A were not
clearly observed (Fig. S4(i)). This observation confirmed our previous finding with single-walled
carbon nanotube when it was covalently linked with phthalocyanine [30]. This suggests that the
functional group changes that occur in carbon nanomaterials covalently linked with organic
molecules are not readily detectable by IR spectroscopy. The IR absorption spectrum of the

6
fullerene-A composite (G) is significantly similar to that of molecule A (Fig. S4(i)), making it
difficult to ascertain if there was any covalent linkages between the fullerene and molecule A. In
order to provide an evidence for the linkage, so that the possibility that the fullerene has been
detached from the composite during the geometry optimization process can be excluded, a
Raman calculation using DFT method was carried out on molecule A and the composite G (Fig.
S5). Differences in the spectrum of A which are due to the changes in the functional group as a
result of the covalent linkage with fullerene moiety have been indicated on the spectra in Fig. S5.
The other organic dyes after derivatizations showed clear functional group changes in the IR
spectra which have been clearly indicated by the boxes in Figs. S4(ii) and S4(iii)).

Ground state electronic absorption spectra obtained for the molecules by time-dependent self-
consistent field method (TD-SCF) are given in Fig. S6. The absorption wavelengths have been
presented against both the absorptivity coefficient and the oscillator strength (f) values to be able
to rationalize the transition strengths. The true picture of the transition probability per molecule
could be seen from the oscillator strengths, as against the epsilon (‘molar absorptivity’) which
depends on the molecule’s molecular weight. Since the f values give the extent of the transition
strength and probability, electronic transitions from the ground states to the first excited singlet
states in molecules C, E, F and A are expected to be strong and most probable than those of
molecules B, D and G with C expected to have the strongest and most probable transitions (Fig.
S6). Table 0 shows the summary of the oscillator strengths (f) in the visible or near infrared
region for each investigated molecule, and the same trend, wherein f for A< F<E<C was
observed. This observation suggests that molecule C with an f value of 0.57 is expected to show
the strongest transition intensity in the spectral range 400 – 800 nm. Therefore, for all intents and
purposes, G, a dye-fullerene dyad, may be considered to be poorly absorbing in the specified
spectral range (Fig. S6), and thus will be unsuitable for applications requiring absorption of
photons in this range. The molecules’ absorption wavelengths and molar absorptivity ( ) within
the spectral range (400 – 800 nm) have been given in Table S0. These two values for each
molecule in this range also reflect on the energy of the absorbed incident photon and the intensity
of the absorption, except that the takes the molecular weight into consideration unlike the
oscillator strength which does not. This may explain why the trend observed for f differs largely
from what was obtained for (F<A<E<C).

Interesting electronic behaviors in the molecules were observed by structurally modifying some
of the molecules to get new ones. By inserting a Ru metal to replace the Zn in E, and removing
the decorating tert-butyl and hexyl substituents on the meso phenyl and diphenyl amine groups
respectively, we formed a new molecule F (Fig. 1). A smaller band gap relative to E was
obtained for F (Table S0), which brings its LUMO closer to the CB edge of the TiO2 (given by
values in Table 0), and showed a red-shifted absorption spectrum (Fig. S7(i)); although with
lower extinction coefficients, especially in the spectral range of interest. However, with the
replacement of the enoic acid group in B now attached directly to the anthracenyl group in D, we

7
obtained a wider band gap for D (Table S0), which takes its LUMO farther from the CB of the
TiO2, and showed a blue-shifted absorption spectrum with lower extinction coefficients relative
to B (Fig. S7(ii), Table S0). Similarly, we observed a significant shift in the absorption
wavelength of A into the red region after it was covalently linked to fullerene to form G (Fig.
S7(iii)), but with no much difference in their values (Table 0), because the gap narrowing was
achieved by HOMO destabilization. The intense conjugation that the fullerene introduced upon
linkage to A could account for the significant narrowing of the composite band gap (Table S0).
The molecular orbitals (MOs) that mostly contribute to the transitions responsible for the
observed absorptions at the are shown in Fig. S8. Generally, it can be concluded that the
differences in the absorption spectra of the molecules and their respective derivatives can be
linked to the differences in the MOs.

The emission of photons as fluorescence by molecules from the first excited singlet state (S1) to
the ground singlet state (S0) depends on some factors such as the nature of molecules, solvent
media, the incident photon energy and its pulse width. These factors in turn determine how the
molecular electronic excited state dynamics play out. However, since the study was carried out in
silico and without incorporating any solvent models, effect of solvent molecules and incident
photon energy, together with the pulse widths have been excluded in the interpretation of the
results. Therefore, the calculated excited state properties have been assumed to depend on the
geometrical configurations of the studied molecules only.

Notably, the molecule’s excited state energy could be deactivated by various competitive excited
state processes which include internal conversion, vibrational relaxation, intersystem crossing
and fluorescence emission [31]. It is possible for a highly absorbing molecule to fluoresce
greatly, if the excited state conditions favor deactivation by fluorescence. In addition, a poorly
absorbing molecule could have poor emission provided that the little energy absorbed is largely
consumed through other processes other than fluorescence. Furthermore, molecules which show
strong absorptions could have poor emission if the deactivation to the S0 state is by other
processes different from fluorescence. And finally, molecules with relatively poor absorption
could still fluoresce considerably, if the excited state conditions favor deactivation by
fluorescence. Subsequently, it can be seen from this work that some of the molecules showing
intense absorption in the specified spectral range have poor fluorescence emission (Table 0). The
reason for this has been attributed to the overriding effects of other processes in deactivating the
molecule’s excited state. However, some with relatively low absorptions but with considerably
large fluorescence may be said to have emitted a large fraction of their absorbed energy (Table
0).

The TD-DFT calculated absorption and the deduced emission spectra using (equations 8 – 11)
for molecules A, C, D and E are presented in Fig. 2 as representatives. The deduced emission
spectra showed the expected emission at longer wavelengths relative to the absorptions (Fig. 2);
and other typical characteristics of an emission curve [27,31]. For instance, the emission

8
spectrum generated for E is a mirror image of its absorption spectrum (Fig. 2). Both the emission
and absorption wavelengths for each molecule have been given in Table S0, and the difference
(shift) between them can easily be estimated. The fluorescence lifetimes as given by equation 9
are presented in Table 0. The trend observed in the emission factors, , revealed that the
fluorescence emissions are higher in molecules B, F and G compared to the others; which
include the reference cyanidin molecule (Fig. S9), a known photosensitizer in dye sensitized
solar cells’ application [32-34].

It should be of interest that both the absorbed and emitted energy by the fullerene-A composite
(G) did not fall within the target spectral range, 400 – 800 nm (Fig. S10). As a result, the
composite may likely not be relevant to DSSC application. An explanation to this observation
may be given in terms of the location of the HOMO on the components of the composite (G).
Normally, a donor-acceptor system like this would have the acceptor moiety designated as
LUMO and is expected to be the fullerene; whereas, the donor which is expected to be the dye
would be designated as the HOMO. However, in the MO configurations shown in Fig. S8G, the
HOMO is situated on the fullerene thereby giving the fullerene the status of the donor, while the
dye assumes the status of the acceptor. This result is in contrasts with the fundamentally
established description of dye-fullerene interactions, which stipulates that the dye will always be
the donor while the fullerene will always be the acceptor [35]. The low obtained for A, C, D
and E could be attributed to the charges in A, and, the steric effect in C, D and E. The higher
of molecule F compared to E, in spite of the similarity in their structures may be due to the
increase in the degree of structural hindrance in E relative to F.

3.2. Dye electron injection into TiO2 and its default state regeneration by the
iodide/triiodide redox couple system

Suitable dye sensitizers for DSSCs must absorb most of the radiation of the solar spectrum in the
visible and near-IR regions, and be able to produce a large photocurrent response [36,37]. In
addition to this, the HOMO must be situated below the iodide/triiodide redox couple of the
electrolyte, while the LUMO should be above the CB edge of the TiO2 semiconductor. In this
work, we employed the experimentally determined TiO2 CB edge reported by Xu and Schoonen
[26]. It is on record that the energy difference between the LUMO of the dye and the CB edge of
the TiO2 semiconductor ( ) determines the rate of electron injection and consequently the
photocurrent density in DSSCs [35]. Therefore, may represent, together with its importance in
the estimation of values, a predictive tool for new dye sensitizers [36,37]. For a negative
value, the dye excited state would lie below the TiO2 conduction band edge, and would imply an
unfavorable electron injection (+ve ∆Ginj) for the molecule being considered.

The estimated values are presented in Table 0 together with the pictorial representations in
Fig. 10(i & ii). These results clearly revealed that the LUMO of molecules C, D, E and F lie
above the CB edge of the semiconductor TiO2, thus suggesting that the dyes would allow for a

9
rapid transfer of electrons into the energy state of the TiO2 [19,36,37]. A positive value was
also obtained for the reference cyanidin molecule (Table 0). However, it has been noted that the
ideal value for an efficient charge injection should be ~ 0.40 eV above the TiO2 conduction
band edge [36,38]; therefore, one may consider molecule F, on account of the value, as the
only favored sensitizer even above the cyanidin (Table 0). Another important parameter is the
gap between the iodide/triiodide redox couple energy level and the HOMO level of the dye,
which gives information about the ease of electron replacement by the redox couple to the dye
after it (the dye) has released electrons to the semiconductor. For a ready transfer, the dye’s
HOMO should be located below the redox couple potential as shown for molecule A, B, F and G
in Fig. 3(i). Notably from Fig. 3(i & ii), only F out of the molecules C, D, E and F that gave
positive values, has its HOMO below the redox couple potential. These observations make F
the only molecule with potentials for both electron accepting from the redox couple system and
electron releasing to the TiO2 semiconductor. It is also worthy of note that the reference cyanidin
molecule with the calculated HOMO value of -4.00 eV (Table 0) showed poor electron accepting
tendency relative to the position of the redox couple in the photoelectrochemical energy levels.
Since the two conditions as provided by the value, and the dye-HOMO/redox couple potential
difference, favored molecule F, it is therefore considered the best electron donor molecule to the
semiconductor, as well as the best electron acceptor from the redox couple system.

3.3. Photo-voltaic properties

The estimated molecular radii ( ) and the diffusion coefficients are given in Table 0. The
results were found to fall within the range of the literature values [39], suggesting that all the
assumptions made, as well as the method of calculations, are reasonably adequate. In addition,
was found to be inversely proportional to the size of the dye molecule, revealing its expected
dependence on the dye molecular radii [39]. The results of the photovoltaic parameters are
presented in Table 0. The LHE were found to be considerably high in molecules A, C, E and F,
indicating that they would absorb more strongly within the specified spectral range of interest
when compared to B, D and G. The LHE obtained for the reference cyanidin is comparable to
that of molecule A which like the cyanidin possesses a net positive ionic charge. A negative
value indicates that the CB band edge of TiO2 lies below the dye LUMO, favoring the
electron injection process as suggested by the values. One would realize from this
interpretation that the inferences obtained from the data of and are the same, showing
that there is a strong correlation between the two parameters, and is such that for a negative
value, a positive value would be expected and vice-versa. Negative and positive
values were found for molecules C, D, E, F and cyanidin.

The charge collection efficiency ( ) may be defined as the probability of electron availability at
the TiO2-dye interface. We have defined it as the ratio of the diffusion coefficient ( ) to the
square of the potential difference between the LUMO of the dye and the CB edge of the TiO2

10
semiconductor (Eq. 12). The obtained for the molecules showed that F has the highest
value with the overall results following the trend E < C < G < A < D < B < F. This suggests that
electrons will mostly be available in the TiO2-F interface than the other TiO2-dye interfaces.
However, the incident photon conversion efficiency (IPCE) as determined using equation 1, was
observed to follow the trend A < E < G < D < C < B < F, showing A and F as the sensitizers with
the lowest and highest ability to convert incident photon to charges respectively. In these
arrangements according to the molecules’ and IPCE values, we found the reference cyanidin
taking up the third and the fourth positions respectively. The relatively high and IPCE values
obtained for F when compared to the other dyes may be attributed to the favorably aligned
positions of its LUMO and HOMO with respect to the TiO 2 CB edge, and the redox couple
energy level respectively. However, for molecule B, and IPCE values were high due to the
high fluorescence factor which compensate for its unfavorable LUMO position relative to the
TiO2 CB edge, and the favorable position of its HOMO relative to the redox couple energy level.
It should be noted that the IPCE value calculated for F ( 4.82 × 10-10) is higher than B (8.01 × 10-11)
and the next closest molecule C (3.75×10-13) by one and three orders of magnitude respectively
(Table 0). The unfavorable HOMO position of the cyanidin relative to the iodide/triiodide redox
couple level, may account for its much lower IPCE (2.46 × 10-13) value when compared to those
of molecules F, B and C. Thus, it may be concluded that molecules F and B possess the required
photophysicochemical and photovoltaic properties needed for solar cells applications; which may
be based on the account of their structural simplicity (lower steric effects) relative to the other
molecules, narrower HOMO-LUMO gaps, and the absence of molecular charges, while still not
ruling out the importance of the central Ru metal in enhancing photoluminiscence. Additionally,
the observed results of cyanidin are shown to possess considerably high IPCE and values as
previously reported before [34]. This method also predicted that cyanidin would be a good
photosensitizer as previously reported in the literature. Therefore, the new approach may be used
to predict the photovoltaic activities of dyes, without going through the laborious experimental
trial and error methods.

4. Conclusion

A density functional theoretical (DFT) method was employed to study the photo-voltaic and
photophysicochemical properties of a highly conjugated metal-free and some metal containing
organic compounds (A-G) for application in dye sensitized solar cells (DSSCs). Time-dependent
density functional theory (TD-DFT) methods have been used to generate the electronic
absorption spectra of the molecules. Lowest unoccupied molecular orbital (LUMO) relative to
TiO2 conduction band ( ) positions, pi-electron diffusion coefficients ( ) and molar
absorptivity coefficients ( ) at maximum absorption wavelengths of the molecules in the visible
or near infrared (near IR) region have been determined. The fluorescence emission ( ) and
charge collection ( ) efficiencies were obtained theoretically for the first time to the best of our
knowledge, using the calculated emission spectra data, and values respectively. Consequently,

11
the determined and values have enabled the calculation of the dyes’ incident photon
conversion efficiency (IPCE) which hitherto have not been obtained theoretically. The calculated
properties have been employed in predicting the dyes suitability for application in photo-voltaic
technology. The overall results showed that molecules F, B and C have attractive properties that
qualify them for employment in solar cells fabrication. The new approach is a simple method of
predicting photovoltaic activities of dyes without going through the laborious experimental trial
and error methods. It should however be noted that, while this simple method may be used as a
guide for designing sensitive dyes for DSSC application, it cannot be used to predict the actual
efficiency of a DSSC device.

Acknowledgements

This work was supported by the Federal Government of Nigeria Tertiary Education Trust Fund
(FGN-TETFund) and Obafemi Awolowo University, Ile-Ife, Nigeria.

12
References

[1.] B. O’Regan, M. Grätzel (1991). A low-cost, high-efficiency solar cell based on dye-
sensitized colloidal TiO2 films, Nature, 353, 737.
[2.] M. Grätzel (2001). Photoelectrochemical cell, Nature, 414, 338.
[3.] M. Grätzel (2003). Dye-sensitized solar cells. J. Photochem. Photobiol. C: Photochem. Rev.,
4, 145.
[4.] U. Mehmood, I.A. Hussein, M. Daud, S. Ahmed, K. Harrabi (2015). Theoretical study of
benzene/thiophene based photosensitizers for dye sensitized solar cells (DSSCs), Dyes and
Pigments, 118, 152.
[5.] A.O. Adeloye, P.A. Ajibade (2014). Towards the development of functionalized
polypyridine ligands for Ru(II) complexes as photosensitizers in dye-sensitized solar cells
(DSSCs), Molecules, 19, 12421.
[6.] J. Wang, H. Li, N.-N. Ma, L.-K. Yan, Z.-M. Su (2013). Theoretical studies on organoimido-
substituted hexamolybdates dyes for dye-sensitized solar cells (DSSCs), Dyes and Pigments, 99,
440.
[7.] J. Zhang, Y.-H. Kan, H.-B. Li, Y. Geng, Y. Wu, Z.-M. Su (2012). How to design proper π-
spacer order of the D-π-A dyes for DSSCs? A density functional response, Dyes and Pigments
95, 313.
[8.] W. Fan, D. Tan, W.-Q. Deng (2012). Acene-modified triphenylamine dyes for dye-sensitized
solar cells: a computational study, Chemphyschem., 13, 2051.
[9.] D. Rocca, R. Gebauer, F. De Angelis, M.K. Nazeeruddin, S. Baroni (2009). Time-dependent
density functional theory study of squaraine dye-sensitized solar cells, Chem. Phys. Lett., 475,
49.
[10.] W. Sang-aroon, S. Saekow, V. Amornkitbamrung (2012). Density functional theory study
on the electronic structure of Monascus dyes as photosensitizer for dye-sensitized solar cells J.
Photochem. Photobiol. A Chem., 236, 35.
[11.] X. Zarate, E. Schott, T. Gomez, R. Arratia-Pérez (2013). Theoretical study of sensitizer
candidates for dye-sensitized solar cells: Peripheral substituted dizinc pyrazinoporphyrazine–
phthalocyanine complexes, J. Phys. Chem. A., 117, 430.
[12.] W. Fan, W. Deng (2013). Incorporation of thiadiazole derivatives as π‐ spacer to construct
efficient metal‐ free organic dye sensitizers for dye-sensitized solar cells: A theoretical study
Commun. Comput. Chem., 1, 152.
[13.] M. Grätzel (2005). Solar energy conversion by dye-sensitized photovoltaic cells, Inorg.
Chem., 44, 6841.
[14.] S. Ardo, G.J. Meyer (2009). Photodriven heterogeneous charge transfer with transition-
metal compounds anchored to TiO2 semiconductor surfaces, Chem. Soc. Rev., 38, 115.

13
[15.] L.M. Peter (2007). Characterization and modeling of dye-sensitized solar cells, J. Phys.
Chem. C, 111, 6601.
[16.] L.M. Peter (2007). Dye-sensitized nanocrystalline solar cells, Phys. Chem. Chem. Phys. 9,
2630.
[17.] J. Bisquert, D. Cahen, G. Hodes, S. Ruhle, A. Zaban (2004). Physical chemical principles
of photovoltaic conversion with nanoparticulate, mesoporous dye-sensitized solar cells, J. Phys.
Chem. B, 108, 8106.
[18.] B.C. O’Regan, J. Durrant (2009). Kinetic and energetic paradigms for dye-sensitized solar
cells: Moving from the ideal to the real, Acc. Chem. Res., 42, 1799.
[19.] A. Hagfeldt, G. Boschloo, L. Sun, L. Kloo, H. Pettersson (2010). Dye-sensitized solar cells
Chem. Rev., 110, 6595.
[20.] A. Hagfeldt, G. Boschloo, H. Lindstrom, E. Figgemeier, A. Holmberg, V. Aranyos, E.
Magnusson, L. Malmqvist (2004). A system approach to molecular solar cells, Coord. Chem.
Rev., 248, 1501.
[21.] A.O. Adeloye, T.O. Olomola, A.I. Akinbulu, P.A. Ajibade (2012). A high molar extinction
coefficient bisterpyridyl homoleptic Ru(II) complex with trans-2-methyl-2-butenoic acid
functionality: Potential dye for dye-sensitized solar cells Int. J. Mol. Sci., 13, 3511.
[22.] A.O. Adeloye, P.A. Ajibade (2014). A high molar extinction coefficient Ru (II) complex
functionalized with cis-Dithiocyanato-bis-(9-anthracenyl-10-(2-methyl-2-butenoic acid)-1, 10-
phenanthroline): potential sensitizer for stable dye-sensitized solar cells, J. Spectrosc., 2014, 1.
[23.] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman, J.A.
Montgomery Jr., T. Vreven, K.N. Kudin, J.C. Burant, J.M. Millam, S.S. Iyengar, J. Tomasi, V.
Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G.A. Petersson, H. Nakatsuji, M. Hada,
M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H.
Nakai, M. Klene, X. Li, J.E. Knox, H.P. Hratchian, 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, P.Y. Ayala, K. Morokuma, G.A. Voth, P. Salvador, J.J. Dannenberg, V.G.
Zakrzewski, S. Dapprich, A.D. Daniels, M.C. Strain, O. Farkas, D.K. Malick, A.D. Rabuck, K.
Raghavachari, J.B. Foresman, J.V. Ortiz, Q. Cui, A.G. Baboul, S. Clifford, J. Cioslowski, B.B.
Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R.L. Martin, D.J. Fox, T. Keith, M.A.
Al-Laham, C.Y. Peng, A. Nanayakkara, M. Challacombe, P.M.W. Gill, B. Johnson, W. Chen,
M.W. Wong, C. Gonzalez, J.A. Pople, In Gaussian 03, Revision E.01, Gaussian, Inc.,
Wallingford CT, 2004.
[24.] (a) A.D. Becke (1993). Density‐ functional thermochemistry. III. The role of exact
exchange, J. Chem. Phys., 98, 5648; (b) A.D. Becke (1992). Density‐ functional
thermochemistry. I. The effect of the exchange‐ only gradient correction, J. Chem. Phys., 96,
2155; (c) A.D. Becke (1992). Density‐ functional thermochemistry. II. The effect of the
Perdew–Wang generalized‐ gradient correlation correction, J. Chem. Phys., 97, 9173.

14
[25.] C. Lee, W. Yang, R.G. Parr (1988). Development of the Colle-Salvetti correlation-energy
formula into a functional of the electron density, Phys. Rev. B, 37, 785.
[26.] Y. Xu, M.A.A. Schoonen (2000). The absolute energy positions of conduction and valence
bands of selected semiconducting minerals, Am. Miner., 85, 543.
[27.] J.R. Lakowikz. Principles of Fluorescence Spectroscopy, 2nd edn, p. 239. Kluwer
Academic/Plenum Publishers, New York (1999).

[28.] J. Kestin, K. Knierim, E.A. Mason, B. Najafi, S.T. Ro, M. Waldman, (1984). Equilibrium
and transport properties of the noble gases and their mixtures at low density, J. Phys. Chem. Ref.
Data, 13, 229.
[29.] (a) P. Suppan (1983). Excited-state dipole moments from absorption/fluorescence
solvatochromic ratios Chem. Phys. Lett., 94, 272. (b) S. B. Bulgarevich, D. Ya. Movshovich
(1992). The Onsager approximation and nonspecific intermolecular interactions in polar media,
Mendeleev Commun., 2, 16.
[30.] K. Sanusi, E.K. Amuhaya, T. Nyokong (2014). Enhanced optical limiting behavior of an
indium phthalocyanine−single-walled carbon nanotube composite: An investigation of the
effects of solvents, J. Phys. Chem. C, 118, 7057.
[31.] U. Noomnarm, R.M. Clegg (2009). Fluorescence lifetimes: Fundamentals and
interpretations, Photosynth Res., 101, 181.
[32.] N-A. Sanchez-Bojorge, L-M. Rodriguez-Valdez, D. Glossman-Mitnik, N. Flores-Holguin
(2015). Theoretical calculation of the maximum absorption wavelength for cyanidin molecules
with several methodologies, Comput. Theor. Chem., 1067, 129.
[33.] S. Meng, J. Ren, E. Kaxiras (2008). Natural dyes adsorbed on TiO2 nanowire for
photovoltaic applications: Enhanced light absorption and ultrafast electron injection, Nano Lett.,
8, 3266.
[34.] A. Lim, N.T.R.N. Kumara, A.L. Tan, A.H. Mirza, R.L.N. Chandrakanthi, M.I. Petra, L.C.
Ming, G.K.R. Senadeera, P. Ekanayake (2015). Potential natural sensitizers extracted from the
skin of Canarium odontophyllum fruits for dye-sensitized solar cells, Spectrochim. Acta Part A:
Mol. Biomol. Spectrosc. 138, 596.
[35.] M. Olguin, L. Basurto, R.R. Zope, T. Baruah, (2014).The effect of structural changes on
charge transfer states in a light-harvesting carotenoid-diaryl-porphyrin-C60 molecular triad, J.
Chem. Phys., 140, 204309.
[36.] M.P. Balanay, D.H. Kim (2008). DFT/TD-DFT molecular design of porphyrin analogues
for use indye-sensitized solar cells, Phys. Chem. Chem. Phys., 10, 5121.
[37.] F. De Angelis, S. Fantacci, A. Selloni (2008). Alignment of the dye's molecular levels with
the TiO2 band edges in dye-sensitized solar cells: A DFT–TDDFT study, Nanotech., 19, 424002.
[38.] R. Mosurkal, J.-A. He, K. Yang, L. A. Samuelson, J. Kumar (2004). Organic
photosensitizers with catechol groups for dye-sensitized photovoltaics, J. Photochem. Photobiol.
A, 168, 191.

15
[39.] A. Ogunsipe (2018). Solvent effects on the spectral properties of Rhodamine 6G:
Estimation of ground and excited state dipole moments, J. Solution Chem., 47, 203.

16
Figure 1: Structural features of the investigated molecules.

17
Figure 2: TD-DFT generated absorption and deduced emission spectra of the studied molecules; with molecules A,
C, D and E shown as representatives

Figure 3: Plot showing (i) the positions of molecular energy levels with respect to the TiO2 conduction band edge
and iodide-triiodide redox couple positions (ii) the molecules whose first excited states (LUMO) are higher than the
conduction band edge of TiO2 semiconductor.

18
Table 0: Photophysicochemical and photo-voltaic properties of the dyes

Dye δp
(nm) ×10-10 *
f LHE ∆Ginj IPCE (eV)
m2s-1 (ns)
A 15.0 7.28 0.19 0.35 4.17 6.36×10 -5
5.84 × 10 -11
1.29×10 -15 10.5 -3.53
-9 -11
B 15.0 7.26 0.04 0.09 0.79 0.86 1.05 × 10 8.01×10 48.0 -0.83
C 17.6 6.20 0.57 0.73 -3.28 1.00×10-2 5.14 × 10-11 3.75×10-13 21.0 3.47
D 17.1 6.40 0.04 0.09 -2.81 3.00×10-2 6.61 × 10-11 1.82×10-13 17.2 3.11
E 17.6 6.21 0.44 0.64 -3.38 4.00×10-3 4.56 × 10-11 1.17×10-13 16.5 3.69
F 15.7 6.97 0.29 0.49 -1.19 0.30 3.29 × 10-9 4.82×10-10 30.3 0.46
18.7 5.84 6.00 16.5
G 0.01 3.78 0.20 5.17 × 10-11 1.35×10-13 -3.36
×10-3
Cya 10.8 10.1 0.19 0.36 -3.34 8.5×10-3 8. 04 × 10-11 2.46×10-13 25.9 3.54
*
Only the oscillator strengths for the maximum absorption at wavelengths in the visible or near infrared region are
given. Cya = cyanidin.

19
20
Dye sensitized solar cells
Photo-induced electron transfer
Incident photon conversion efficiency
Time-dependent self-consistent-field density functional theory (TD-SCF DFT) method

21