Journal of ELECTRONIC MATERIALS

https://doi.org/10.1007/s11664-018-6406-3
Ó 2018 The Minerals, Metals & Materials Society

Efficiency Enhancement of CH3NH3SnI3 Solar Cells by Device

Modeling

FAISAL BAIG,1,2 YOUSAF HAMEED KHATTAK ,1,2,4 BERNABÉ MARÍ,1

SAIRA BEG,3 ABRAR AHMED,2 and KHURRAM KHAN2

1.—School of Design Engineering, Departament de Fı́sica Aplicada, Universitat Politécnica de

Valencia, Camı́ de Vera, Valencia, Spain. 2.—Electrical Engineering Department, Federal Urdu
University of Arts, Science and Technology, Islamabad, Pakistan. 3.—COMSATS Institute of
Information Technology, Islamabad, Pakistan. 4.—e-mail: [email protected]

Lead halide perovskite solar cells (LHPSC) have great potential, with con-
version efficiency exceeding 20%. However, their toxic nature and difficult
fabrication prevent their consideration for commercial applications. To ad-
dress this, numerical analysis was performed to propose a structure for lead-
free perovskite solar cells with MASnI3 as absorber layer. Device modeling for
Cd1xZnxS as electron transport layer (ETL) and methylamine tin halide as
hole transport layer (HTL) was carried out using a solar cell capacitance
simulator. The simulation results revealed the dependence of the open-circuit
voltage (VOC), short-circuit current (JSC), fill factor (FF), and power conversion
efficiency on the HTL valence-band offset, absorber layer thickness, absorber
layer doping concentration, ETL band offset, minority-carrier diffusion length,
and defects at the HTL–absorber and absorber–ETL interfaces. An optimum
absorber layer thickness was confirmed, being well consistent with the range
for practical absorber layer designs. Moreover, conversion efficiency of 18.71%
was found for absorber thickness of 500 nm and doping concentration of
1 9 1016 cm3. These results will provide important guidelines for design of
low-cost perovskite solar cells.

Key words: Cd1xZnxS, SCAPS, MASnBr3, MASnI3, perovskite, solar cell

INTRODUCTION 1.3 eV, being considered a possible replacement for

use in LHPSCs. Due to its narrower bandgap, it can
Lead halide perovskite solar cells (LHPSCs) have
cover a wider range of the visible light spectrum
attracted wide attention in recent years because of
than LHPSCs. Tin (Sn) perovskite-based planar-
their good absorption properties, optimal bandgap,
structure solar cells have been developed with
good carrier diffusion length, and low-cost process-
heterostructure architecture and optical bandgap
ing.1–4 The power conversion efficiency (PCE) of
ranging from 1.3 eV to 1.4 eV, reaching efficiency of
perovskite solar cells has been enhanced from 3.8%
3%.11–13 The major limitation when using Sn-based
to 22% over the last few years.5–10 Despite this rapid
perovskites is oxidation of Sn from Sn2+ to Sn4+ in
PCE enhancement and simplification of the fabri-
air, which limits device performance. With exten-
cation process of LHPSCs, their stability and the
sive research, development of encapsulation pro-
toxic nature of lead have hindered commercializa-
cesses can increase the stability of Sn-based
tion of perovskite solar cells. Theoretical and exper-
perovskite materials. Addition of tin fluoride
imental studies have shown that
(SnF2) in the fabrication process reduces the chance
CH3NH3SnI3(MASnI3) has an optimal bandgap of
of oxidation of Sn2+ to Sn4+.14 Moreover, as in
LHPSCs, the bandgap can also be tuned between
1.3 eV and 2.15 eV by substituting iodine (I) with
(Received April 4, 2018; accepted May 23, 2018) bromide (Br), and addition of Br increases the
bandgap. Such tunability of the bandgap provides
 Baig, Khattak, Marı́, Beg, Ahmed, and Khan

an opportunity to use Sn-based perovskites for solar gallium (di)selenide (CIGS) and CdTe solar cells and
cell applications as a possible replacement for has been widely used for their optimal design.20
LHPSCs.15 However, recently, SCAPS has also emerged as an
Despite the rapid improvement in Sn-based per- alternative numerical tool for analysis of kesterite
ovskite materials, PCE values achieved for Sn- and perovskite solar cells.21–25
based perovskite solar cells remain very low. This Our approach for improvement of MASnI3 solar
is mainly due to a lack of understanding of device cell performance is based on the following approach:
properties and the effect of the band structure on
device performance. Indeed, there are no reports on  Validation of simulated results for Spiro-MeO-
simulations for Sn-based perovskite solar cells with TAD as hole transport layer (HTL)
Cd(Zn,S) as electron transport layer (ETL) and  Analysis of the effect of MASnBr3 as HTL to
MASnBr3 as hole transport layer (HTL). We report replace Spiro-MeOTAD in SCAPS simulation
herein the theoretically proven and experimentally  Analysis of the effect of the HTL valence-band
expected effects of the conduction-band offset (CBO) offset on solar cell performance
of the ETL/MASnI3 layer on the performance of  Optimization of absorber layer thickness and
perovskite solar cells. To analyze this effect, it is doping concentration
necessary to control the bandgap of the CdS ETL  Replacement of TiO2 with CdS ETL layer
layer, which can be achieved by adding a controlled  Analysis of the conduction-band offset to im-
amount of zinc (Zn) to the ETL chemical bath prove device efficiency
solution containing (Cd) and sulfur (S) ions, as  Analysis of the effect of the absorber layer
discussed in literature.16–19 Cd1xZnxS thin film was minority-carrier diffusion length on the thick-
deposited by chemical bath deposition (CBD) tech- ness of the absorber layer
nique. The parameters of the bath used for fabrica-  Comparison of results
tion of Cd1xZnxS film on transparent conducting
oxide (TCO) were similar to those used elsewhere.17 The results will provide useful guidelines for design
Various zinc/cadmium ratios (x = 0%, 3%, 5%, and of high-performance MASnI3-based solar cells.
10%) were used in this experiment. After film
fabrication on TCO substrate, optical analysis was SIMULATION PROCEDURES
performed, revealing a shift in the bandgap of the
Cd1xZnxS layer with the Zn concentration. Optical A planar heterostructure solar cell based on
analysis of Cd1xZnxS layers with different Zn MASnI3 absorber material is considered, with the
concentrations is shown in Figs. 1 and 2. following layer configuration: glass substrate/trans-
In this work, device simulations were carried out parent conducting oxide/CdxZn1xS (ETL)/MASnI3
using solar cell capacitance simulator (SCAPS) (absorber)/MASnBr3 (HTL)/back contact, as shown
software developed by the University of Gen. in Fig. 3.
SCAPS is a one-dimensional software used to Simulation parameters for different layers were
calculate the band diagram, I–V characteristic, carefully taken from experimental and theoretical
quantum efficiency/spectral response (QE), and work. The primary parameters of the proposed
alternating-current (AC) quantities. This program structure used in the SCAPS simulation are sum-
was mainly developed for analysis of copper indium marized in Table I.11,12,15,26 Defects in the absorber

Fig. 1. Transmittance of Cd1xZnxS. Fig. 2. Plot of (Ahv)2 versus photon energy (eV).
Efficiency Enhancement of CH3NH3SnI3 Solar Cells by Device Modeling

layer were considered to be neutral Gaussian, with Fig. 6 that, for a bandgap of 2.2 eV, the PCE tends
characteristic energy of 0.1 eV and energy above the to increase up to electron affinity of 3.3 eV, but
valence band of 0.6 eV. Defects at the ETL–absorber decreases thereafter. This occurs because, with
layer and absorber–HTL interfaces were set to be increase of the EA, the valence-band position
neutral, with single defects having energy of 0.6 eV changes for HTL, resulting in an increase in the
above the valence band with total concentration of VBO (vHTL  vAbsorber ). Moreover, with such increase
interface defects of 1.0 9 1017 cm2. The character- in the VBO, the band structure of the valence band
istics of the defects in the absorber layer and at the at the interface changes from a cliff to spike, as
interface region are presented in Tables II and III. shown in Fig. 5. However, with further increase in
the VBO, the spike height increases, thereby
RESULTS AND DISCUSSION increasing the resistance for flow of holes towards
the back contact and causing a decrease in the PCE
Effect of MASnBr3 HTL on Solar Cell
of the solar cell.
Performance
The J–V characteristics obtained for the solar Effect of Absorber Thickness and Doping
cells with two different HTLs are shown in Fig. 4. Concentration on Solar Cell Performance
According to Fig. 4, on changing the HTL layer from
The absorber thickness has a direct effect on the
Spiro-MeOTAD to MASnBr3, there is an increase in
performance of the solar cell and was thus changed
the short-circuit current (JSC) and open-circuit
from 0.1 lm to 1 lm. The PCE, JSC, and VOC tended
voltage (VOC). This mainly occurs because of the
to increase with increasing thickness, but became
valence-band offset of the HTL layer with the
constant above thickness of 0.5 lm. Meanwhile, the
absorber layer. With Spiro-MeOTAD as HTL layer,
FF tended to increase with thickness at the start,
the band structure exhibits a cliff-like structure
but with further increase in thickness, started to
that reduces the activation energy of the solar cell,
decrease, mainly due to the increase in the series
while for MASnBr3, the band structure exhibits a
resistance with increasing thickness. Figure 7
small spike, due to which the activation energy
shows the results for the effect of the absorber layer
becomes equal to the bandgap energy, thus increas-
thickness on the performance of the solar cell.
ing the open-circuit voltage of the solar cell.27 The
MASnI3 is unstable in air because of the oxidiza-
band structure for the HTL interface with absorber
tion process from Sn2+ to Sn4+, which acts as a p-
layer is shown in Fig. 5 for Spiro-MeOTAD (a) and
type dopant for the perovskite absorber layer. This
MASnBr3 (b).
self-doping process can greatly affect the perfor-
Figure 6 shows the effect of the band offset on the
mance of solar cells based on an MASnI3 absorber
solar cell performance. It can be observed from
layer. Therefore, to investigate the effect of the
acceptor concentration on the solar cell perfor-
mance, the doping level was varied from

Table II. Interface defects (neutral)

Defect Layer CdZnS/ MASnI3/
Property MASnI3 MASnBr3

Capture cross-section 1.0 9 1015 cm2 1.0 9 1015 cm2

area of electron/hole
Density of defects 1.0 9 1017 cm2 1.0 9 1017 cm2
Fig. 3. Solar cell structure.

Table I. Parameters for SCAPS11,12,15,26

SCAPS Parameter FTO (SnO2) ETL (CdZnS) Absorber (MASnI3) HTL (MASnBr3)

Thickness (nm) 500 0.1–1000 0.1–1000 200

CB effective DOS (cm3) 1019 2.8 9 1018 1 9 1018 1 9 1018
VB effective DOS (cm3) 1019 3.9 9 1018 1 9 1018 1 9 1018
Doping density (cm3) n: 2 9 1019 n: 1 9 1017 p: 1 9 1014–1 9 1017 p: 1 9 1018
Permittivity 9 10 8.2 8.2
Electron affinity (eV) 4 4.4–3.6 4.17 3.39
Bandgap (eV) 3.50 2.42–3.2 1.3 2.15
Electron/hole mobility (cm2/V s) 100/25 100/25 1.6/1.6 1.6/1.6
 Baig, Khattak, Marı́, Beg, Ahmed, and Khan

Table III. Defect layer properties (neutral)

Defect Layer
Property MASnI3 MASnBr3

Capture cross-section 1.0 9 1015 cm2 1.0 9 1015 cm2

area of electron/hole
Density of defects 1.0 9 1015 cm3 1.0 9 1015 cm3

Fig. 6. Effect of electron affinity on solar cell performance.

Fig. 4. Effect of HTL on solar cell performance.

Fig. 7. Effect of absorber layer thickness on solar cell performance.

Fig. 5. Band interface for HTL with absorber layer.

after reaching an optimal acceptor concentration,

1 9 1013 cm3 to 1 9 1017 cm3. The results for the the PCE starts to decrease whereas the FF tends to
effect of the acceptor concentration on JSC and VOC increase. This increase in the FF depends upon JSC
are shown in Fig. 8. JSC reaches a maximum value and VOC, as depicted in Eq. 1,
at acceptor concentration of 1 9 1016 cm3, whereas
VOC starts to decrease at acceptor concentration of VMax IMAX
FF ¼ : ð1Þ
1 9 1016 cm3. This behavior can be well explained VOC ISC
based on the quantum efficiency (QE) results in
Fig. 9: with increasing acceptor concentration, the
QE increases to a value of 1 9 1016 cm3, indicating Effects of ETL Bandgap and Electron Affinity
that the rate of carrier photogeneration increases on Solar Cell Performance
for given incident photon number. Hence, an opti-
mal acceptor carrier concentration is necessary for On replacing the TiO2 layer with CdS, the
improvement in JSC and the photoabsorption effi- conversion efficiency of the solar cell is reduced,
ciency. The acceptor concentration also affects the because of the large conduction-band offset (CBO)
PCE and FF of the solar cell: with increase in the between the ETL and absorber layer, which gives
carrier concentration, the PCE and FF increase, but rise to interface recombination. Thus, the CBO can
Efficiency Enhancement of CH3NH3SnI3 Solar Cells by Device Modeling

the bandgap, the electron affinity of Cd1xZnxS will

change, greatly affecting the band structure
between the ETL and absorber layer. The effect of
the CBO on the solar cell performance and band
structure is shown in Fig. 10b–d. The cliff-like band
structure (Fig. 10c) does not impede flow of elec-
trons towards the front contact, but the activation
energy for carrier recombination (Ea) becomes lower
than the absorber bandgap Eg, where Ea is repre-
sented by Eg  |CBO|. The main recombination
process is interface recombination when Ea is less
than the bandgap of the absorber layer.28–30 Thus,
the CBO directly correlates with the PCE of the
solar cell, as Ea directly correlates with VOC, and the
negative CBO reduces VOC. When the CBO is
positive, the PCE is excellent, but further increase
in the CBO reduces the PCE of the solar cell. When
the CBO is positive, a spike is formed at the buffer–
Fig. 8. Effect of acceptor carrier concentration on solar cell perfor-
absorber interface (Fig. 10d), and increase in the
mance. spike impedes the flow of photogenerated electrons
towards the front electrode, thus affecting the PCE
of the solar cell, as shown in Fig. 10b.

Effect of Absorber Layer Defects on Solar Cell

Performance
By setting the band offset of the HTL and ETL,
one can improve the PCE of the solar cell by
suppressing carrier recombination at the interface,
but the PCE of the solar cell still cannot reach the
maximum theoretical conversion efficiency. The
main reason is the defect density in the absorber
layer of the solar cell, as the density of such defects
in the absorber layer directly affects the lifetime
and diffusion length of photogenerated carriers.
This can be well expressed using Eqs. 2 and 3. The
relation between the diffusion length and the car-
rier mobility and lifetime is expressed in Eq. 2.
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
lðe;hÞ kT
LD ¼ slifetime ; ð2Þ
q
Fig. 9. QE as function of acceptor concentration versus wavelength.

where LD is the diffusion length, lðe;hÞ are the

greatly affect solar cell performance, based on the electron and hole mobility, and slifetime is the
plot in Fig. 10a of the J–V characteristic for two minority-carrier lifetime. Meanwhile, slifetime
different ETL layers with electron affinity of 4.4 eV depends upon the defect trap density and capture
for CdS and 4.27 eV for TiO2.23 Therefore, by cross-section area for electrons and holes. The
adjusting the CBO of the ETL/absorber, one can relation between slifetime and the bulk defect density
improve the solar cell efficiency, and the reason for is expressed in Eq. 3.
the selection of CdS is because of its bandgap 1
tunability. The bandgap of CdS can be tuned by slifetime ¼ ; ð3Þ
Nt dvth
adding Zn into the CdS bath solution, as discussed
above. As discussed in ‘‘Introduction’’ section, the where d represents the capture cross-section area
bandgap of Cd1xZnxS can be varied from 2.42 eV to for electrons and holes, mth represents the thermal
3.2 eV by changing the Zn concentration in the bath velocity of carriers (107 cm/s), and Nt is the density
solution. Therefore, the bandgap of the Cd1xZnxS of traps. Therefore, based on Eqs. 2 and 3, with
layer was changed in SCAPS, assuming that only increase in the trap density, the carrier diffusion
the conduction-band minimum changes with the length will reduce, thus reducing the chance of
bandgap of Cd1xZnxS whereas the valence-band carriers reaching the contact. The diffusion length
maximum remains constant. Thus, with change in can greatly affect the solar cell performance, and
 Baig, Khattak, Marı́, Beg, Ahmed, and Khan

Fig. 10. (a) Effect of ETLs on solar cell J–V characteristic curve. (b) Effect of CBO of Cd1xZnxS on solar cell performance. (c) Band structure for
vETL>vAbsorber . (d) Band structure for vETL<vAbsorber .

limiting the of thickness of absorber layer. This

effect of the diffusion length with change in the
thickness of absorber material versus PCE is shown
in Fig. 11, according to which the value of the
diffusion length limits the optimal absorber layer
thickness for which one obtains the maximum
conversion efficiency. The effect of the diffusion
length on the optimal thickness is clear for diffusion
length of 0.5 lm and 0.1 lm: for 0.5 lm, the optimal
thickness of the absorber layer for which the
maximum PCE is achieved is 0.6 lm, whereas for
0.1 lm, the optimal absorber layer thickness is
0.4 lm. Based on the results shown in Fig. 11, it can
be concluded that the optimal thickness of the
absorber layer will be limited by the minority
carrier-diffusion length and density of defects in
the absorber layer.

Fig. 11. PCE versus absorber thickness for different diffusion

COMPARISON OF RESULTS lengths.

The results obtained in each analysis step herein

are plotted together in Fig. 12, revealing that the
final optimized cell with Cd1xZnxS as ETL and solar cells with PCE of 18.71%. Details of the solar
MASnBr3 as HTL represents a good alternative for cell performance parameters for each analysis
design of high-efficiency MASnI3 absorber layer herein are presented in Table IV.
Efficiency Enhancement of CH3NH3SnI3 Solar Cells by Device Modeling

Table IV. Comparison of results

Analysis Step VOC (V) JSC (mA/cm22) FF (%) PCE (%)

With Spiro-MeOTAD as HTL 0.55 25.82 63.31 8.97

With MASnBr3 as HTL 0.67 26.54 65.02 11.53
Optimization of absorber thickness 0.77 30.74 67.31 15.96
Optimization of acceptor doping concentration 0.78 30.75 67.06 16.13
ETL CBO adjustment 0.87 31.42 68.88 18.71

REFERENCES
1. S. De Wolf, J. Holovsky, S.J. Moon, P. Löper, B. Niesen, M.
Ledinsky, F.J. Haug, J.H. Yum, and C. Ballif, J. Phys.
Chem. Lett. (2014). https://doi.org/10.1021/jz500279b.
2. C. Quarti, E. Mosconi, J.M. Ball, V. D’Innocenzo, C. Tao, S.
Pathak, H.J. Snaith, A. Petrozza, and F. De Angelis, En-
ergy Environ. Sci. (2016). https://doi.org/10.1039/
C5EE02925B.
3. Q. Dong, Y. Fang, Y. Shao, P. Mulligan, J. Qiu, L. Cao, and
J. Huang, Sci. Express (2015). https://doi.org/10.1126/
science.aaa5760.
4. W. Liao, D. Zhao, Y. Yu, N. Shrestha, K. Ghimire, C.R.
Grice, C. Wang, Y. Xiao, A.J. Cimaroli, R.J. Ellingson, N.J.
Podraza, K. Zhu, R.G. Xiong, and Y. Yan, J. Am. Chem.
Soc. (2016). https://doi.org/10.1021/jacs.6b08337.
5. A. Kojima, K. Teshima, Y. Shirai, and T. Miyasaka, J. Am.
Chem. Soc. (2009). https://doi.org/10.1021/ja809598r.
6. J.-H. Im, C.-R. Lee, J.-W. Lee, S.-W. Park, and N.-G. Park,
Nanoscale (2011). https://doi.org/10.1039/c1nr10867k.
7. H.S. Kim, C.R. Lee, J.H. Im, K.B. Lee, T. Moehl, A. Mar-
chioro, S.J. Moon, R. Humphry-Baker, J.H. Yum, J.E.
Moser, M. Grätzel, and N.G. Park, Sci. Rep. (2012). https://
Fig. 12. Comparison of results. doi.org/10.1038/srep00591.
8. J. Burschka, N. Pellet, S.J. Moon, R. Humphry-Baker, P.
Gao, M.K. Nazeeruddin, and M. Grätzel, Nature (2013).
https://doi.org/10.1038/nature12340.
CONCLUSIONS 9. M. Liu, M.B. Johnston, and H.J. Snaith, Nature (2013).
Lead-free perovskite solar cells were simulated https://doi.org/10.1038/nature12509.
10. M.A. Green, K. Emery, Y. Hishikawa, W. Warta, E.D.
using SCAPS software. A numerical model was Dunlop, D.H. Levi, and A.W.Y. Ho-Baillie, Prog. Photovolt.
analyzed for a lead-free MASnBr3 HTL layer with Res. Appl. (2016). https://doi.org/10.1002/pip.2855.
Cd1xZnxS as ETL for an MASnI3 absorber layer. 11. Y. Yu, D. Zhao, C.R. Grice, W. Meng, C. Wang, W. Liao,
The effect of the HTL layer band offset was ana- A.J. Cimaroli, H. Zhang, K. Zhu, and Y. Yan, RSC Adv.
(2016). https://doi.org/10.1039/C6RA19476A.
lyzed; based on the results, it can concluded that the 12. T. Fujihara, S. Terakawa, T. Matsushima, C. Qin, M. Ya-
HTL band offset can greatly affect the solar cell hiro, and C. Adachi, J. Mater. Chem. C (2017). https://doi.
performance. An optimal thickness (0.5 lm) and org/10.1039/C6TC05069G.
optimal doping concentration (1 9 1016 cm3) of the 13. T.M. Koh, T. Krishnamoorthy, N. Yantara, C. Shi, W.L.
absorber layer were identified, exceeding which will Leong, P.P. Boix, A.C. Grimsdale, S.G. Mhaisalkar, and N.
Mathews, J. Mater. Chem. A (2015). https://doi.org/10.
lead to degradation of solar cell performance. The 1039/C5TA00190K.
effect of two different ETLs on the solar cell 14. Z. Zhao, F. Gu, Y. Li, W. Sun, S. Ye, H. Rao, Z. Liu, Z. Bian,
performance was analyzed, revealing that adjusting and C. Huang, Adv. Sci. (2017). https://doi.org/10.1002/
the band offset of the Cd1xZnxS layer can lead to advs.201700204.
15. F. Hao, C.C. Stoumpos, D.H. Cao, R.P.H. Chang, and M.G.
PCE of 18.71%. The effect of the defect density in Kanatzidis, Nat. Photonics (2014). https://doi.org/10.1038/
the absorber layer was analyzed, revealing that nphoton.2014.82.
increase of the defect density will limit the thick- 16. J.S.J. Song, S.S. Li, S. Yoon, W.K. Kim, J.K.J. Kim, J.
ness of the absorber layer. The results presented Chen, V. Craciun, T.J. Anderson, O.D. Crisalle, and F.R.F.
herein will provide a baseline for selection of Ren, Growth and characterization of CdZnS thin film buf-
fer layers by chemical bath deposition, in Conference Re-
alternative ETL and HTL layers and design of cord of the Thirty-First IEEE Photovoltaics Specialists,
efficient solar cells. Conference 2005 (2005). https://doi.org/10.1109/pvsc.2005.
1488166.
ACKNOWLEDGEMENTS 17. S. Ullah, H. Ullah, M. Mollar, and B. Mari, Fabrication of
Cd1xZnxS buffer layer deposited by chemical bath depo-
This work was supported by Ministerio de Econ- sition for photovoltaic applications, in 2016 International
omı́a y Competitividad (ENE2016-77798-C4-2R) Renewable Sustainable Energy Conference. IEEE (2016).
and Generalitat Valenciana (Prometeus 2014/044). https://doi.org/10.1109/irsec.2016.7983932.
 Baig, Khattak, Marı́, Beg, Ahmed, and Khan

18. L.L. Baranowski, S. Christensen, A.W. Welch, S. Lany, 24. Y.H. Khattak, F. Baig, T. Hanae, S. Ullah, B. Marı́, S. Beg,
M. Young, E.S. Toberer, and A. Zakutayev, Mater. and H. Ullah, Curr. Appl. Phys. (2018). https://doi.org/10.
Chem. Front. (2017). https://doi.org/10.1039/C6QM002 1016/j.cap.2018.03.013.
91A. 25. Y.H. Khattak, F. Baig, S. Ullah, B. Marı́, S. Beg, and H.
19. K.T.R. Reddy and P.J. Reddy, J. Phys. D Appl. Phys. Ullah, Optik (Stuttg) (2018). https://doi.org/10.1016/j.ijleo.
(1992). https://doi.org/10.1088/0022-3727/25/9/011. 2018.03.055.
20. M. Burgelman, P. Nollet, and S. Degrave, Thin Solid Films 26. H.-J. Du, W.-C. Wang, and J.-Z. Zhu, Chin. Phys. B (2016).
(2000). https://doi.org/10.1016/S0040-6090(99)00825-1. https://doi.org/10.1088/1674-1056/25/10/108802.
21. K.R. Adhikari, S. Gurung, B.K. Bhattarai, and B.M. Sou- 27. T. Minemoto and M. Murata, Sol. Energy Mater. Sol. Cells
case, Phys. Status Solidi Curr. Top. Solid State Phys. (2015). https://doi.org/10.1016/j.solmat.2014.10.036.
(2016). https://doi.org/10.1002/pssc.201510078. 28. U. Rau and H.W. Schock, Appl. Phys. A Mater. Sci. Process.
22. T. Minemoto and M. Murata, J. Appl. Phys. (2014). https:// (1999). https://doi.org/10.1007/s003390050984.
doi.org/10.1063/1.4891982. 29. M. Turcu and U. Rau, J. Phys. Chem. Solids (2003). https://
23. Y.H. Khattak, F. Baig, S. Ullah, B. Marı́, S. Beg, and H. doi.org/10.1016/S0022-3697(03)00137-9.
Ullah, J. Renew. Sustain. Energy (2018). https://doi.org/10. 30. K. Tanaka, T. Minemoto, and H. Takakura, Sol. Energy
1063/1.5023478. (2009). https://doi.org/10.1016/j.solener.2008.09.003.