Highly efficient donor-acceptor hydrazone dyes-inorganic Si/TiO2 hybrid solar cells

Article abstract of DOI:10.1016/j.saa.2015.02.108

We have synthesized the two donor-bridge-acceptor organic dyes (hydrazone dye 1 (HD1) and hydrazone dye 2 (HD2)) with the aim to enhance intra-molecular charge transfer then characterized by FTIR and NMR. The ground state geometries have been optimized at

Full text of DOI:10.1016/j.saa.2015.02.108

Accepted Manuscript
Highly efficient donor-acceptor hydrazone dyes-inorganic Si/TiO2 hybrid solar
cells
Abdullah G. Al-Sehemi, Ahmad Irfan, Mohrah Abdullah M. Al-Melfi
PII:
S1386-1425(15)00282-6
http://dx.doi.org/10.1016/j.saa.2015.02.108
SAA 13410
DOI:
Reference:
Spectrochimica Acta Part A: Molecular and Biomo-
lecular Spectroscopy
To appear in:
Received Date:
Revised Date:
Accepted Date:
27 December 2014
16 February 2015
22 February 2015
Please cite this article as: A.G. Al-Sehemi, A. Irfan, M.A.M. Al-Melfi, Highly efficient donor-acceptor hydrazone
dyes-inorganic Si/TiO2 hybrid solar cells, Spectrochimica Acta Part A: Molecular and Biomolecular
Spectroscopy (2015), doi: http://dx.doi.org/10.1016/j.saa.2015.02.108
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.

Highly efficient donor-acceptor hydrazone dyes-inorganic Si/TiO2 hybrid solar cells
Abdullah G. Al-Sehemia,b,c*, Ahmad Irfana∗, Mohrah Abdullah M. Al-Melfia
a Department of Chemistry, Faculty of Science, King Khalid University, Abha 61413,
P.O. Box 9004, Saudi Arabia
b
Unit of Science and technology, Faculty of Science, King Khalid University, Abha
61413, P.O. Box 9004, Saudi Arabia
c Research Center for Advanced Materials Science, King Khalid University, P.O. Box
9004, Abha 61413, Saudi Arabia
Abstract
We have synthesized the two donor-bridge-acceptor organic dyes (hydrazone dye 1 (HD1)
and hydrazone dye 2 (HD2)) with the aim to enhance intra-molecular charge transfer then
characterized by FTIR and NMR. The ground state geometries have been optimized at
three different levels of theories, i.e., B3LYP/6-31G*, B3LYP/6-31G** and Hartee-Fock
HF/6-31G**. The absorption spectra and oscillator strengths in different solvents have
been computed and compared with the experimental data. The vibrational spectral
assignments have been performed on the recorded FTIR spectra based on the theoretical
predicted wavenumbers at three different levels of theories. The effect of different
solvents (CHCl3, CH3CN and C2H5OH) has been studied on the absorption wavelengths.
Furthermore, we have computed the ionization potentials, electron affinities and
∗ Corresponding author: Ahmad Irfan
E-mail: irfaahmad@gmail.com
Tel.:0096672418632
Fax:0096672418426
A.G. Al-Sehemi
E-mail: agmasq@gmail.com
1

reorganization energies of studied compounds and shed light on the charge transport
properties. The hetero-junction solar cell devices were fabricated by organic-inorganic
hetero-junction (Si/TiO2/dye) then the efficiency has been measured by applying the
incident power 30, 50 and 70 mW/cm2. The maximum efficiency 3.12% has been
observed for HD1.
Keywords: Organic-inorganic hybrid solar cells; Hydrazones; Fabrication; Efficiency:
Density functional theory
1. Introduction
The record efficiency of conventional solar cells (inorganic materials) has been reached
up to 25% [1]. Nowadays, researchers are focusing to design materials and techniques to
fabricate devices which have low processing cost, easy to fabricate and environmental
friendly. The organic solar cells have attained efficiency up to 10% [2]. The major issue
for organic solar cell materials is not only the low efficiency until now but also stability
issue [3]. In recent years, dye-sensitized solar cells [4-8] and hybrid solar cells using
nanoparticles like TiOx [9], ZnO [10], CdSe [11], CdS [12], PbS [13], CuInS2 [14] gained
interest. The Si based hybrid solar cells have been studied intensively [15-17]. The
organic-inorganic hybrid solar cells are good and efficient as both of the materials take
part to enhance the efficiency [18, 19] and inorganic part which also act as acceptor is
environmentally stable and excitons absorbed here [3].
With the aim to improve the efficiency of device, we have synthesized donor-
bridge-acceptor efficient organic dyes, i.e., hydrazone dye 1 (HD1) and hydrazone dye 2
(HD2), see Fig. 1. We discussed the exciton binding energy and dissociation of excitons
on the basis of energies of frontier molecular orbitals as well as favorable band alignment
2

has been discussed thoroughly. By applying the DFT approach, we shed light on the
absorption and IR spectra then compared with the available experimental data. The
charge transfer behavior was studied on the basis of ionization potential, electron affinity
and reorganization energy. The effect of three different level of theories, i.e., B3LYP/6-
31G*, B3LYP/6-31G** and HF/6-31G** has been discussed. The structures-properties
relationship has been discussed. To improve efficiency we have fabricated the
dye/TiO2/Si hybrid solar cells. The paper has been divided as; first section deals with the
computational and experimental methodology. In section 3.1, we discussed the charge
transfer properties and chemical descriptors. The main focus of this study is detailed
investigation about the fabrication techniques and efficiency measurements.
Insert Fig. 1
2. Methodology
2.1. General experimental methods
Infra-red (IR) spectra of crystalline compounds were determined using a Thermo
1
scientific smart omni-transmission. H and 13C nuclear magnetic resonance (NMR)
TM
spectra were recorded on a Bruker at 500 MHz Ultra Shield
at room temperature in
deutrated dimethyl sulfoxide (DMSO-d6) (CH3SOCH3 have two signals in 1H-NMR at δ
2.52 singlet, 3.38 singlet). UV-Vis spectra were recorded with UV-1800 Shimadzu.
Melting points (mps) were determined with a Stuart SMP11 without correction.
2.1.1. Physical data
For nuclear magnetic resonance (NMR) spectra, chemical shifts and expressed in ppm on
the δ- scale relative to the internal standard (TMS). The following abbreviations are used
s- singlet, t- triplet, q- quartet, d- doublet, m- multiplet; J is the coupling constant (Hz).
3

2.1.2. General procedure for hydrazones
We have synthesized the hydrazone based sensitizers by the same method as in previous
studies [20, 21]. The hydrazone derivatives would be prepared through direct
condensation between the corresponding aromatic aldehydes and phenyl hydrazine.
Equimolar quantities of phenylhydrazine and the aldehydes would be boiled in ethanol
for an hour. The precipitated hydrazones would be filtered, washed and dried. The pure
hydrazones would be obtained after recrystallization from ethanol.
2.1.3. Preparation of 1-phenyl-2-(3,4,5-trimethoxybenzylidene)hydrazine (1)
The 3,4,5-trimethoxybenzaldehyde (9.8g, 0.05 mol) was added to phenylhydrazine (4.9
ml, 0.05 mol) and added 30 ml of ethanol absolute as solvent. The flask was related with
condenser under reflux system and solution boiled at 170-180 ˚C for 1h then transfer
solution to beaker and stilled in room temperature or cooled, the precipitated hydrazones
would be collected and filtered, washed with ethanol then dried. The product
crystallization from ethanol gave yellow crystals of title compound (1-phenyl-2-(3,4,5-
trimethoxybenzylidene)hydrazine) m.p 125 ˚C. δ H (DMSO, 500 MHz) 6.74-7.24 (7H,
m, Ar-H), 7.80 (1H, s, CH═N), 10.32 (1H, s, NH); δ C (DMSO, 500 MHz) 55.83,56.04
and 60.06 (3xOCH3), 102.82-145.27(8C-Ar), 153.13 (CH═N).
2.1.4. Preparation of 2-ethoxy-4-((2-phenylhydrazono)methyl)phenol (2)
The 3-ethoxy-4-hydroxybenzaldehyde (8.3g, 0.05mol) was added to phenylhydrazine
(4.9 ml, 0.05mol) and added 30 ml of ethanol absolute as solvent. The flask was related
with condenser under reflux system and solution boiled at 170-180 ˚C for 1h then transfer
solution to beaker and stilled in room temperature or cooled, the precipitated hydrazones
would be collected and filtered, washed with ethanol then dried. The product
4

crystallization from ethanol gave crystals of title compound (2-ethoxy-4-((2-
phenylhydrazono)methyl)phenol) δ H (DMSO, 500 MHz) 1.37 (3H, t, J 6.5-7, CH3),
4.09(2H, q, J 7, CH2), 6.69-7.24 (8H, m, Ar-H), 7.76 (1H, s, CH═N), 9.16(1H, s, OH),
10.07 (1H, s, NH); δ C (DMSO, 500 MHz) 14.74 (CH3), 63.82(CH2), 109.87, 111.72,
115.54, 118.14, 119.67, 127.35, 129.0, 137.27, 145.62 and 147.26 (8C-Ar), 147.39
(CH═N).
2.1.5. General procedure for dyes
The new chromospheres were prepared by direct tricyanovinylation of hydrazones. A
solution of the requisite hydrazone 1 (0.01 mol) and tetracyanoethylene (TCNE) in DMF
o
(20 mL) was stirred at 60-90 C for 8-12h or boiled at high temperature up to 400˚C for
8h. The solvent was removed and the residual solid was collected and recrystallized from
toluene/ petroleum ether mixture.
2.1.6. Preparation of 2-{4-[2-(3,4,5-trimethoxybenzylidene)hydrazine]phenyl}ethylene-
1,1,2-tricarbonitrile (HD1)
The 1-phenyl-2-(3,4,5-trimethoxybenzylidene)hydrazine (1) (2.86g, 0.01 mol) was
dissolved in 20 ml DMF as solvent then added 1.28g of TCNE direct gave dark color
solution. The solution was boiled at 278-285˚C under reflux for 8h. The solvent was
removed and the solid residual was collected and recrystallized from toluene/ petroleum
ether mixture to get violet crystals of (HD1) (3.46g 89.4%) m.p 248˚C. δ H (DMSO,
500 MHz) 3.72 (3H, s, OCH3), 3.87 (6H, s, 2xOCH3), 7.1, 8.01 and 8.02 (6H, m, Ar-H),
8.06 (1H, s, CH═N), 11.85 (1H, s, NH exchange with D2O); δ C (DMSO,500 MHz)
55.99 and 60.13 (3xOCH3), 79.12 and 144.47 (C═C), 114.01, 114.22 and 114.83
5

(3xC≡N), 104.27, 113.09, 119.36, 129.73, 132.76, 137.12, 139.20 and 151.19 (8C-Ar),
153.21(CH═N).
2.1.7. Preparation of 2-{4-[2-(3-ethoxy-4-hydroxybenzylidene}hydrazino]phenyl)-
ethylene-1,1,2-tricarbonitrile (HD2)
The 2-ethoxy-4-((2-phenylhydrazono)methyl)phenol (2) (2.56g, 0.01 mol) was dissolved
in 20 ml DMF as solvent then added 1.28g of TCNE direct gave dark color solution.
Then the solution was boiled at 210˚C under reflux for 8h. The solvent was removed and
the solid residual was collected and recrystallized from toluene/ petroleum ether mixture
to get violet crystals of (HD2) (3.02g 84.6%) m.p 225-227˚C. δ H (DMSO) 1.38 (3H, t,
J 7, OCH2CH3), 4.1 (2H, q, J 6.9-7, OCH2CH3), 6.85-8.01 (7H, m, Ar-H), 8.04 (1H, s,
CH═N), 9.54 (1H, s, OH exchange with D2O), 11.77(1H, s, NH exchange with D2O); δ
C (DMSO) 14.68 (CH3CH2O), 63.92 (CH3CH2O), 77.59 and 145.65 (C═C), 114.25,
114.44 and 114.89 (3xC≡N), 110.61, 115.60, 119.07, 121.94,125.57, 132.85, 136.59,
147.23 and 149.35 (9C-Ar), 151.27(CH═N).
2.2. Computational details
The density functional theory (DFT) is good approach to reproduce the
experimental data as well as to predict the properties of interests [5, 6, 22-24]. The
B3LYP has been proved an efficient approach to reproduce the experimental data for the
small molecules [25]. Moreover, by using the large fraction of HF exchange charge
transfer states have been discussed previously [26]. The B3LYP functional has been
applied to compute and reproduce the absorption wavelengths of different organic dyes
(indigo, azobenzene, phenylamine, hydrazone, and anthraquinone) and 0.12 eV average
deviation was observed for hydrazone dyes [27]. Recently, we showed that the B3LYP is
6

good choice to reproduce the excitation energies for hydrazone based dyes [28-30]. We
found that B3LYP (Polarizable continuum model (PCM), methanol) is better, accurate
and more reasonable choice than BHandHLYP, LC-BLYP and CAM-B3LYP (PCM,
methanol) to reproduce the experimental data. In the present study we have computed the
excitation energies by using TD-B3LYP functional. The effect of solvents (CHCl3,
CH3CN and C2H5OH) has been studied on the absorption spectra.
Geometry optimization for all dyes have been performed using DFT at the
B3LYP/6-31G*, B3LYP/6-31G** and HF/6-31G** levels of theories. The vibrational
modes were examined by using the ChemCraft program. The absorption spectra and
energy gap for dyes were calculated in different solvents by using polarizable continuum
model (PCM) [31] at TD-B3LYP/6-31G* level of theory [32]. The ionization potentials,
electron affinities and reorganization energy of all dyes have been performed by using the
B3LYP/6-31G* level of theory. All of the calculations were performed by using
Gaussian-09 program package [33].
3. Results and discussion
3.1. Electronic properties and absorption spectra
The computed HOMOs energies (EHOMO), LUMOs energies (ELUMO) and HOMO-
LUMO energy gaps (Egap) have been illustrated in Fig. 2. In solvents the EHOMO level
elevated as compared to gas phase. In CHCl3, CH3CN and C2H5OH the EHOMO of HD1
are -0.12, -0.17 and -0.16 eV high, respectively as compared to the EHOMO in gas phase.
The EHOMO of HD2 in CHCl3, CH3CN and C2H5OH are -0.14, -0.18 and -0.18 eV higher
than the EHOMO without solvent, respectively. The solvents have no significant effect to
elevate or diminish the ELUMO. The root cause of reduced Egap in solvents is the elevation
7

of EHOMO. We observed that the effect of CH3CN and C2H5OH solvents on the EHOMO and
ELUMO is similar to lift up or reduce resulting corresponding Egap. The larger Egap in CHCl3
revealing that all the studied dyes would be red shifted in CH3CN and C2H5OH.
Insert Fig. 2
The HOMO and LUMO energies of Si are –5.43 and -3.92 eV [34, 35]. The
HOMO and LUMO energies of TiO2 are –7.40 and -4.20 eV [36]. We have observed the
Nested band alignment in the donor and acceptor frontier molecular orbitals by
considering the Si as acceptor only. The successful operation of a photovoltaic device
requires a staggered band alignment heterojunction which allocate electrons to transport
to the cathode and holes to the anode. By considering the average values both for Si and
TiO2, the valance band energy has been found -6.41 eV while the conduction band energy
-4.06 eV (Si/TiO2). It is expected that Si/TiO2 acceptor changed the alignment to
staggered band alignment heterojunction, see Fig. 2.
In hybrid solar cells, excitons formed in the donor material are dissociated at the
donor–acceptor (D–A) interface. The force required to overcome the exciton binding
energy is provided by the energy level offset of the lowest unoccupied molecular orbital
(LUMO) of the donor and the conduction band edge of the acceptor materials. We found
energy level offset 0.74 and 0.83 eV for HD1 and HD2, respectively to overcome the
exciton binding energy. In HD1 less force is required to overcome the exciton binding
energy as compared to other dye. For dissociation of excitons formed in the acceptor
material, the energy offset of the highest occupied molecular orbital (HOMO) of the
donor and the valence band edge of the acceptor materials is required. We found energy
8

level offset 0.57 and 0.68 eV for HD1 and HD2, respectively to dissociate of excitons. In
HD1 less force would be required for dissociation of excitons. It is expected that HD1
would be more efficient because of less force would be required to overcome the exciton
binding energy as well as dissociation of excitons.
Insert Table 1
In HD1 the maximum absorption wavelength has been observed at 529 nm in
chloroform and 532 nm in acetonitrile, no significant effect has been observed in
absorption spectra toward red shift by changing chloroform to CH3CN. In HD2
substitution of m-OC2H5, p-OH showed absorption band at 538 nm in chloroform, 543
nm in CH3CN and 547 nm in ethanol.
The dyes were measured in various solvents having different polarity, see Table 1.
The trend of absorption spectra toward red shift of HD1 and HD2 in different solvents is
CHCl3 < CH3CN < CH3CH2OH. The maximum absorption spectra computed at TD-
B3LYP/6-31G* level of theory in chloroform is 568 and 587 nm for HD1 and HD2,
respectively. The maximum absorption wavelengths in acetonitrile are 570 and 593 nm
for HD1 and HD2, respectively. In ethanol, we observed the maximum absorption
wavelength at 570 and 593 nm for HD1 and HD2, respectively which are in reasonable
agreement with the experimental evidence.
3.2. FTIR spectra
Vibrational spectral assignments were recorded based on the theoretical predicted
wavenumbers by DFT/B3LYP/6-31G*, B3LYP/6-31G** and HF/6-31G** level of
9

theories. ChemCraft graphical interface was used to assign the calculated harmonic
wavenumbers using scaled displacement vectors to identify the motion of modes. The
computed wavenumbers by DFT are in good agreement with the experimental values, see
Fig. 3. The IR spectra calculated were made for a free molecule in vacuum while
experiments were performed for solid sample. For this reason we used scale factors in
theoretical method. In this study, we have used the scaling factor 0.937 for investigated
dyes except HD2 (0.930) for DFT calculations and 0.866 and 0.845 scaling factor for
HD1 and HD2, respectively for HF calculations.
The FTIR spectra exhibited three important absorption bands; the first band has
been observed at 3271 and 3250 cm-1 for the (νN-H) mode for HD1 and HD2, respectively.
In the present case DFT calculations gave 3382 and 3349 cm-1 at B3LYP/6-31G* and
3399 and 3367 cm-1 at B3LYP/6-31G** while the HF calculations gave 3319 and 3239
cm-1 for HD1 and HD2, respectively. The NH stretching wavenumber is red shifted in IR
from the computed wavenumber, which indicates the weakening of the N―H bond
resulting in hydrogen bonding. The strong and sharp second band appears in the region of
2219-2215 which was attributed to the cyano group. The B3LYP/6-31G* and B3LYP/6-
31G** calculations gave these modes at 2200 and 2183 cm-1 while HF calculations gave
at 2265 and 2210 cm-1 for HD1 and HD2, respectively. The third absorption band in the
region of 1610 and 1608 cm-1 ascribed for the C═N in HD1 and HD2, respectively. The
DFT calculations gave at 1574 and 1564 cm-1 at B3LYP/6-31G* level of theory and 1570
and 1561 cm-1 at B3LYP/6-31G** level of theory while the HF calculations gave at 1656
and 1617 cm-1 for HD1 and HD2, respectively. The absorption of OH group appear in
HD2 in the range 3143 ̵
3443 cm-1 (FTIR), from 3139 to 3577 cm-1 for DFT and from
10

3261 to 3564 cm-1 for HF. The N-H stretching vibration showed only one band at 3335
cm−1 [37]. The sharp band has been observed at 3265 cm-1 in the FTIR spectrum for the
title compounds. The DFT calculations gave 3361 and 3379 cm-1 for B3LYP6-31G* and
B3LYP6-31G**, respectively. The HF calculations gave band at 3462 cm-1. Effect of
hydrogen bonding on an O–H stretching vibration appeared in the region (3500−2500)
cm−1 [37]. The ʋO─H mode is interference with N─H stretching mode and seen broad band
in the region about 3143 cm-1. In DFT calculations gave 3139 and 3152 cm-1 for
B3LYP/6-31G* and B3LYP/6-31G** level of theories, respectively. The HF calculations
gave at 3261 cm-1.
For the existence of benzene rings in a structure, the C─H and C═C─C vibrations
is varying and depending on the number and type of substitutions. In case of disubstituted
on benzene ring, the ʋC─H symmetric and asymmetric showed in region (3080-3030) cm-1
and DFT calculations gave the νC─H modes in the range 3060-2894 cm-1. In the present
case, the DFT calculations gave the νC─H modes in the range 3055-2982 cm-1 at
B3LYP/6-31G* and in the range 3049-2979 cm-1 at B3LYP/6-31G** level of theories. In
HF calculations showed bands in the range 2909-2837 cm-1. For C═C─C stretching in
aromatic rings are different in case 1,2-disubstituted from 1,4-disubstituted on benzene
ring, the C═C stretching vibration showed in the region (1577 4) cm-1 and (1579 6) cm-
1 . In the present case, the ʋC═C mode of aromatic rings occurs at 1656 cm-1 in IR spectra.
The ʋC═C mode appeared at 1576 and 1573 at B3LYP/6-31G* and B3LYP/6-31G** level
of theory, respectively. The HF calculations gave 1620 cm-1 for 1,2-disubstituted ring but
for 1,4-disubstituted the ʋC═C mode of aromatic rings has been obsered at 1560, 1558 and
1531 cm-1 for B3LYP/6-31G* , B3LYP/6-31G** and HF level of theories, respectively.
11

The ʋC═N mode has been reported in the range (1605-1665) cm-1 by A. M. Asiri et
al [38]. In present case the strong band has been observed at 1601 cm-1 in the FTIR
spectra (C═N stretching mode). The DFT calculations gave same mode at 1570 and 1566
cm-1 at B3LYP/6-31G* and B3LYP/6-31G** level of theories, respectively. In HF
calculations this mode has been observed at 1549 cm-1. The C≡N stretching vibration has
been reported in the range 2235-2215 cm-1. The sharp and strong band for ʋC≡N mode
appeared 2219 cm-1 in FTIR spectra. The DFT calculations gave ʋC≡N mode in the range
2178-2201 cm-1 at B3LYP/6-31G* and B3LYP/6-31G** while HF calculations gave in
the range 2217-2224 cm-1. The C─O stretching mode reported in literature in the range
1000-1300 cm-1. In title compounds, the ʋC─O mode observed at 1294 cm-1 in FTIR while
calculations showed at 1250 and 1206 cm-1 for DFT and HF, respectively. The C=C in
aromatic rings show at 1598, 1588, 1494, and 1456 cm-1 [39]. The computed modes at
different level of theories have been tabulated in Tables S1-S4 (Detail can be found in
supporting information).
Insert Fig. 3
3.3. Chemical descriptors
The vertical ionization potentials (IPv), adiabatic ionization potentials (IPa),
vertical electron affinities (EAv), adiabatic electron affinities (EAa), hole reorganization
energies λ(h) and electron reorganization energies λ(e) have been tabulated in Table 2.
Previously, it was concluded that higher EAv would be favorable for the generation of
free [40]. The better hole transfer materials have smaller value of ionization potentials
[41] while better electron transfer materials have higher electron affinities. The high EA
12

of HD1 is revealing that it would be more appropriate to generate free electrons and holes.
The calculated hole reorganization energy of HD2 is smaller than the electron
reorganization energies revealing that this dye might be better hole transport material.
Insert Table 2
The electronegativity (χ), hardness (η), electrophilicity (ω), softness (S) and
electrophilicity index (ωi) at the B3LYP/6-31G*, B3LYP/6-31G**, and HF/6-31G**
level of theories have been presented in Tables S5. It has been observed that the trend of
the electronegativity (χ), hardness (η), electrophilicity (ω), softness (S) and
electrophilicity index (ωi) at all the level of theories is similar. The χ and ω are larger for
HD1. The significant effect to reduce the ωi has been observed in HD1 while reverting
effect has been observed in HD2 in which strong activating group ethoxy has been
substituent.
3.4. Fabrication
The Si wafer was immersed for ten seconds in HF solution (water/HF 49%) for
etching purpose. Then it was rinsed in a mixture of ethanol, acetone and de-ionized water
with molar ratios 1:1:3, respectively to eliminate the fluoride ions. The substrates were
rinsed with dry nitrogen gas. Recently we showed that the nanoporous TiO2 prepared by
sol-gel method from the mixture of H2O2 and HNO3 has better porosity about 32% and
showed only pure anatase phase with energy gap of 3.2 eV. The the sol-gel solution was
used to deposit a titanium oxide films on glass and silicon substrates by spin-coating
method. All films were prepared at spin-coated at rate 2000 rpm (revolution per minute)
13

for 1 min. On pre-cleaned substrates (glass and SiO2 etched- silicon), then dried in oven
(60 oC for 10 min) and repeated pervious steps for 3 layers of coating. For building solar
cell device; about 20 nm nanoporous TiO2 film has been deposited on SiO2 etched-Si
substrate. Then about 400 nm of dye was deposited on the top of TiO2 film. Gold is used
as a top metal electrode by use of a suitable mask to control the shape of the electrode.
The dye thin films were prepared by conventional thermal evaporation technique
at a pressure of about 10-6 mbar. The powder was loaded into a molybdenum cell with
nozzle of 2 mm in diameter on the top. The flat glass; glass coated-TiO2; and silicon
substrates were located above 20 cm from the dye. The film thicknesses were controlled
by using a quartz crystal thickness monitor and subsequently confirmed by ellipsometry
measurements. The films were deposited at room temperature and the rate of deposition
was 0.2nm/s. The low deposition rate was adjusted for better film quality and crystallinity
[42]. The dye layer was prepared by thermal evaporation on the etched surface of Si
substrate. Before depositing of the dye, a thin layer of nanoporous TiO2 (20 nm) was
deposited by sol-gel method.
3.5. Efficiency
Fig. 4 showed the current-voltage characteristics curves with open-circuit voltage
and short-circuit current under different illumination power, i.e., 30, 50 and 70 mW/Cm2.
The photovoltaic properties of heterojunction organic-inorganic hybrid solar cells
(Si/TiO2/dye) based on HD1 and HD2 have been shown in Table 3. The power
conversion efficiency of HD1 has been observed 2.87 with Jsc 2.41 mA/cm2, Voc of 570
mV and FF of 63% when the Pin was 30 mW/cm2. By changing the Pin to 50 mW/cm2 the
HD1 gave Jsc 3.95 mA/cm2, Voc of 615 mV and FF of 64.20% with efficiency 3.12%. For
14

HD1, 2.71% efficiency has been examined when Pin was 70 mW/cm2 which boosts up the
Jsc 4.92 mA/cm2, Voc of 636 mV and FF of 61%. The power conversion efficiency 1.70%
has been measured for HD2 with Pin 30 mW/cm2 and Jsc 1.51 mA/cm2, Voc of 502 mV and
FF of 67.21%. The HD2 gave Jsc 2.66 mA/cm2, Voc of 533 mV and FF of 64.14% with
improved efficiency 1.82% when Pin was 50 mW/cm2. The HD2 gave Jsc 4.93 mA/cm2,
Voc of 592 mV and FF of 46.26% with Pin 70 mW/Cm2 which improved the efficiency to
1.93%. It can be noticed that by increasing the Pin from 30 mW/cm2 to 50 mW/Cm2 Jsc,
Voc and FF improved resulting enhanced the efficiency, see Table 3.
Insert Fig. 4 and Table 3
Conclusions
The following conclusions have been drawn by our dual approach investigations:
1- Elevation of EHOMOs reduces the Egap in solvents while ELUMOs have no significant
effect to elevate or diminish it. The average values of EHOMOs and ELUMOs for
Si/TiO2 and donor dye showed the staggered band alignment.
2- The energy level offset values revealed that HD1 would be more efficient because
less force would be required to overcome the exciton binding energy and
dissociation of excitons.
3- The computed absorption wavelengths and vibrational spectral modes are in good
agreement with the experimental data.
4- The high EA of HD1 is revealing that it would be more appropriate to generate
free electrons and holes.
5- The direct correlation between the adiabatic electron affinity/ionization potential
and the Voc has been observed.
15

6- The HD1 showed good efficiency which reached to the 3.12%. To further
increase the efficiency different metal oxides are under synthetic process in this
regards which would be investigated in near future.
Acknowledgement
Mohrah Abdullah M. Al-Melfi thanks King Abdul Aziz city of Science and technology
(KACST) for their financial support by the grant No. GSP-28-33.
Electronic supplementary material
The online version of this article (doi:-------) contains supplementary material.
References
[1] M.A. Green, K. Emery, Y. Hishikawa, W. Warta, E.D. Dunlop, Progress in
Photovoltaics: Research and Applications, 20 (2012) 12-20.
[2] M.A. Green, Progress in Photovoltaics: Research and Applications, 9 (2001) 123-135.
[3] S. Ren, L.-Y. Chang, S.-K. Lim, J. Zhao, M. Smith, N. Zhao, V. Bulović, M.
Bawendi, S. Gradečak, Nano Lett., 11 (2011) 3998-4002.
[4] A. Irfan, A.G. Al-Sehemi, S. Muhammad, Synth. Met., 190 (2014) 27-33.
[5] A. Irfan, H. Aftab, A.G. Al-Sehemi, J. Saudi. Chem. Soc., 18 (2014) 914-919.
[6] A. Irfan, R. Jin, A.G. Al-Sehemi, A.M. Asiri, Spectrochimica Acta Part A: Molecular
and Biomolecular Spectroscopy, 110 (2013) 60-66.
[7] A. Irfan, N. Hina, A. Al-Sehemi, A. Asiri, J. Mol. Model., 18 (2012) 4199-4207.
[8] D. Gebeyehu, C.J. Brabec, N.S. Sariciftci, D. Vangeneugden, R. Kiebooms, D.
Vanderzande, F. Kienberger, H. Schindler, Synth. Met., 125 (2001) 279-287.
[9] P.A. van Hal, M.M. Wienk, J.M. Kroon, W.J.H. Verhees, L.H. Slooff, W.J.H. van
Gennip, P. Jonkheijm, R.A.J. Janssen, Adv. Mater., 15 (2003) 118-121.
[10] W.J.E. Beek, M.M. Wienk, R.A.J. Janssen, Adv. Funct. Mater., 16 (2006) 1112-
1116.
[11] A.P. Alivisatos, Science, 271 (1996) 933-937.
[12] N.C. Greenham, X. Peng, A.P. Alivisatos, Phys. Rev. B, 54 (1996) 17628-17637.
[13] S.A. McDonald, G. Konstantatos, S. Zhang, P.W. Cyr, E.J.D. Klem, L. Levina, E.H.
Sargent, Nat Mater, 4 (2005) 138-142.
[14] E. Arici, N.S. Sariciftci, D. Meissner, Adv. Funct. Mater., 13 (2003) 165-171.
[15] A.G. Al-Sehemi, A. Irfan, M.A.M. Al-Melfi, A.A. Al-Ghamdi, E. Shalaan, J.
Photochem. Photobiol., A: Chem., 292 (2014) 1-9.
[16] M. Wright, A. Uddin, Sol. Energy Mater. Sol. Cells, 107 (2012) 87-111.
[17] S.-H. Tsai, H.-C. Chang, H.-H. Wang, S.-Y. Chen, C.-A. Lin, S.-A. Chen, Y.-L.
Chueh, J.-H. He, ACS Nano, 5 (2011) 9501-9510.
[18] T. Xu, Q. Qiao, Energy Environ. Sci., 4 (2011) 2700-2720.
[19] Y. Zhou, M. Eck, M. Kruger, Energy Environ. Sci., 3 (2010) 1851-1864.
16

[20] A.G. Al-Sehemi, A. Irfan, A.M. Asiri, Y.A. Ammar, Spectrochimica Acta Part A:
Molecular and Biomolecular Spectroscopy, 91 (2012) 239-243.
[21] A.G. Al-Sehemi, A. Irfan, A.M. Asiri, Y.A. Ammar, J. Mol. Struct., 1019 (2012)
130-134.
[22] A. Irfan, A.G. Al-Sehemi, S. Muhammad, Journal of Quantum Chemistry, 2014
(2014) 6.
[23] A. Irfan, A.G. Al-Sehemi, J. Saudi. Chem. Soc., 18 (2014) 574-580.
[24] A. Irfan, A.G. Al-Sehemi, M.S. Al-Assiri, J. Mol. Graphics Modell., 44 (2013) 168-
176.
[25] B.M. Wong, J.G. Cordaro, J. Chem. Phys., 129 (2008) -.
[26] R.J. Magyar, S. Tretiak, Journal of Chemical Theory and Computation, 3 (2007)
976-987.
[27] D. Guillaumont, S. Nakamura, Dyes Pigm., 46 (2000) 85-92.
[28] A. Al-Sehemi, M. Al-Melfi, A. Irfan, Struct. Chem., 24 (2013) 499-506.
[29] A. Al-Sehemi, A. Irfan, A. Asiri, Theor. Chem. Acc., 131 (2012) 1199-1208.
[30] A.G. Al-Sehemi, A. Irfan, A.M. Asiri, Y.A. Ammar, J. Mol. Struct., (2012).
[31] C. Amovilli, V. Barone, R. Cammi, E. Cancès, M. Cossi, B. Mennucci, C.S. Pomelli,
J. Tomasi, Recent Advances in the Description of Solvent Effects with the Polarizable
Continuum Model, in: L. Per-Olov (Ed.) Adv. Quantum Chem., Academic Press, 1998,
pp. 227-261.
[32] C. Zhang, W. Liang, H. Chen, Y. Chen, Z. Wei, Y. Wu, J. Mol. Struct. (TheoChem),
862 (2008) 98-104.
[33] M. J. Frisch et al., Inc., Wallingford, CT, (2009).
[34] C.-Y. Liu, Z.C. Holman, U.R. Kortshagen, Nano Lett., 9 (2008) 449-452.
[35] C.-Y. Liu, Z.C. Holman, U.R. Kortshagen, Adv. Funct. Mater., 20 (2010) 2157-2164.
[36] C.Y. Kuo, W.C. Tang, C. Gau, T.F. Guo, D.Z. Jeng, Appl. Phys. Lett., 93 (2008)
033307-033303.
[37] B.H. Stuart, Infrared Spectroscopy: Fundamentals and Applications, Wiley, 2004.
[38] A.M. Asiri, A.A. Bahajaj, I.M.I. Ismail, N.A. Fatani, Dyes Pigm., 71 (2006) 103-108.
[39] R. Aich, F. Tran-Van, F. Goubard, L. Beouch, A. Michaleviciute, J.V. Grazulevicius,
B. Ratier, C. Chevrot, Thin Solid Films, 516 (2008) 7260-7265.
[40] L. Sun, F.-Q. Bai, Z.-X. Zhao, B.-Z. Yang, H.-X. Zhang, J. Polym. Sci., Part B:
Polym. Phys., 48 (2010) 2099-2107.
[41] R.K. Chan, S.C. Liao, Can. J. Chem., 48 (1970) 299-305.
[42] L.B. Freund, Suresh, S., , Thin film materials: Stress, Defect formation and surface
evolution, in, Cambridge University Press, Cambridge University Press, 2009.
17

Figure Captions
Fig. 1. The synthesized organic dyes HD1 and HD2.
Fig. 2. The HOMOs energies, LUMOs energies and energy gap of Dyes in different
solvents (top) and gas phase (bottom) (in eV).
Fig. 3. The FTIR spectra of the studied compounds.
Fig. 4. The current-voltage characteristics curves with open-circuit voltage and short-
circuit current under different illumination power, HD1 (left) and HD2 (right).
18

Table 1
Calculated (calc) and experimental (exp) absorption spectra (nm) and oscillator strength
(ƒ) in different solvents.
calc CHCl3
Exp
CH3CN
exp
EtOH
exp
λ
calc
calc
calc
λ
(ƒ)
λ
(ƒ)
λ
(ƒ)
λ
(ƒ)
(ƒ)
λ
(ƒ)
HD1
HD2
532 1.01 570 0.71 540 1.06 570 0.71
330 0.22 419 0.17 332 0.2 419 0.17
529 0.47 568 0.71
333 0.12 418 0.18
288 0.02
285 0.02
288 0.03
543 0.56 593 0.58 547 0.52 593 0.58
338 0.09 388 0.05 342 0.11 455 0.32
538 0.44 587 0.6
340 0.05 383 0.04
288 0.01
288 0.01
293 0.01 388 0.05
19

Table 2
The vertical and adiabatic ionization potentials (IPv/a), vertical and adiabatic electronic
affinities (EAv/a), hole and electron reorganization energies (λh/e) of hydrazone dyes (in
eV) at the B3LYP/6-31G* level.
Dyes
HD1
HD2
IPv
IPa
EAv
2.14
2.08
EAa
1.87
1.77
λh
λe
6.70
6.78
7.09
7.01
0.778 0.523
0.460 0.593
20

Table 3
The organic-inorganic hybrid solar cell device performance parameters.
Complexes Incident power Voc
Jsc
FF
Efficiency
(%)
(mW/cm2)
(V)
(mA/cm2)
2.41
HD1
HD2
30
50
70
30
50
70
0.57
0.62
0.64
0.50
0.53
0.59
0.63 2.87
0.64 3.12
0.61 2.71
0.67 1.70
0.65 1.82
0.46 1.93
3.95
4.92
1.51
2.66
4.93
21

HD1
NC
C
N
HO
CN
N
(dye4)
H CH CO
CH
NH
HD2
3
2
Fig. 1.
22

HD2
HD1
HD1 HD2
-3.92
-4.20
-5.43
Si
-7.40
TiO2
Fig. 2.
23

90
80
70
60
50
40
30
20
10
100
80
60
40
20
0
HD2
HD1
4000
3500
3000
2500
2000
1500
1000
500
4000
3500
3000
2500
2000
1500
1000
500
Wavenumber cm-1
Wavenumber cm-1
Fig. 3.
24

HD1
HD2
Fig. 4.
25

Research highlights
ꢀ D-bridge-A hydrazone dyes were synthesized and characterized by UV-Vis, FTIR,
and NMR techniques.
ꢀ The effect of different solvents has been studied on the absorption wavelengths.
ꢀ Computed absorption and IR spectra are in good agreement with the experimental
data.
ꢀ Direct correlation between IPa/EAa and Voc has been observed.
ꢀ Hetero-junction solar cell devices were fabricated and 3.12% efficiency has been
observed for HD1.
26