Journal of Computational Electronics

https://doi.org/10.1007/s10825-019-01433-0

Performance enhancement of CIGS‑based solar cells by incorporating

an ultrathin ­BaSi2 BSF layer
Sayed Rezwanul Islam Biplab1 · Md. Hasan Ali1 · Md. Mahabub Alam Moon1 · Md. Firoz Pervez2 ·
Md. Ferdous Rahman1,3 · Jaker Hossain3

© Springer Science+Business Media, LLC, part of Springer Nature 2019

Abstract
Conventional copper indium gallium diselenide (CIGS)-based solar cells offer higher efficiency than other second-generation
technologies such as hydrogenated amorphous silicon (a-Si:H)- or cadmium telluride (CdTe)-based solar cells, but higher
manufacturing cost due to the use of the rare metals indium and gallium. The purpose of the work presented herein is to
improve the efficiency of such devices by using cheaper materials. Accordingly, a back-surface field layer made of low-cost
and widely available barium silicide (­ BaSi2) with a thickness of 0.3 µm is introduced for the first time into the basic CIGS
solar cell structure consisting of Al/ZnO/CdS/CIGS/Mo, resulting in the alternative structure of Al/FTO/CdS/CIGS/BaSi2/
Mo, with fluorine-doped tin oxide (FTO) as the window layer. One-dimensional simulations of the solar cell capacitance are
employed to study the photovoltaic parameters such as the power conversion efficiency, short-circuit current density, open-
circuit voltage, fill factor, and quantum efficiency of the devices. The thickness of the CIGS absorber layer is varied from 0.1
to 3 µm to optimize the device. Besides, the effects of the acceptor ion and bulk defect densities in the CIGS absorber layer,
cell resistances, and operating temperature on the overall performance are also investigated. The proposed structure offers
an efficiency of 26.24% with a thin CIGS layer of only 0.8 µm. In addition to reduced CIGS thickness and cost, the presented
approach results in CIGS solar cells with enhanced performance compared with previously reported conventional designs.

Keywords Back-surface field · BaSi2 · CIGS · Photovoltaic · SCAPS-1D · Solar cell

1 Introduction efficiency of CIGS solar cells is sufficiently high, while

their outdoor performance is admirable [1]. Moreover, this
Solar cells are required for efficient, cost-effective, and chalcogenide material represents a perfect candidate mate-
highly proficient conversion of sunlight to electrical energy rial for the p-type absorber layer in CIGS-based solar cells
at low cost. Current energy needs worldwide indicate the [2]. Although CIGS suffers from high manufacturing cost,
requirement for cost-effective, efficient power conversion it provides better performance than other thin-film solar cell
using copper indium gallium diselenide [Cu(In,Ga)Se2, technologies such as CdTe or a-Si:H [3]. This high cost is
CIGS]-based thin-film solar cells (TFSCs). The conversion due to the use of the rare metals gallium (Ga) and indium
(In). Due to their scarcity, the CIGS layer should be as thin
as possible, to minimize the cost of the absorber material [4].
* Md. Mahabub Alam Moon CIGS comprises four elements, namely copper (Cu),
mmamoon9357@gmail.com indium (In), gallium (Ga), and selenium (Se). A bandgap of
1
Department of Electrical and Electronic Engineering, 1.1 eV [5] and an absorption coefficient of ­105 cm−1 [6] have
Begum Rokeya University, Rangpur, Rangpur 5400, been reported for CIGS. The bandgap (Eg) of Cu(In1−xGax)
Bangladesh Se2 depends on the Ga concentration (x) according to Eq. (1)
2
Institute of Electronics, Atomic Energy Research [7],
Establishment, Bangladesh Atomic Energy Commission,
Dhaka 1207, Bangladesh Eg = 1.06 + 0.39238x + 0.24762x2 . (1)
3
Solar Energy Laboratory, Department of Electrical
and Electronic Engineering, University of Rajshahi,
Rajshahi 6205, Bangladesh

13
Vol.:(0123456789)
 Journal of Computational Electronics

Since the value of Eg depends on the proportion of Ga in In the work described herein, ­BaSi2 material is exploited
the CIGS layer, it can be easily tailored from 1.02 to 1.69 eV for the first time as a BSF layer in a CIGS solar cell. The
by varying the Ga/In ratio in the CIGS material [6, 8]. influence of including such a BSF is explored based on a
The experimental efficiency of CIGS thin-film solar cell comparison with similar materials. The results of this work
(TFSCs) has been reported to be 19.2% [9], 19.9% [10], and reveal that B­ aSi2 is the best material for the BSF layer of
23.4% [11], while based on numerical analysis, the theo- CIGS solar cells when compared with other BSF-aided
retical efficiency has been reported to be 21.3% [5], 22.02% structures [5, 18–20]. In addition, fluorine-doped tin oxide
[12], 22.1% [13], and 22.67% [14]. The optimum value for (FTO) is employed for the window layer of the CIGS solar
the thickness of the CIGS absorber layer has been found cell.
to be 1 µm [5], 1.4 µm [12], 2 µm [14], and 2.5 µm [13].
With a 1-µm-thick CIGS absorber layer and an additional
1-µm-thick Si absorber layer, an efficiency of 21.3% was 2 Device and simulation
confirmed [5]. In addition, an efficiency of 24.45% has also
been achieved by using a 1-µm-thick CIGS absorber layer Figure 1a shows the conventional structure (Al/ZnO/CdS/
with an additional 0.3-µm SnS BSF layer [15]. An efficiency CIGS/Mo) of the CIGS solar cell with a p-type CIGS
of 24.55% was obtained by using a CIGS absorber layer with absorber having Eg of 1.1 eV [5] and n-type CdS buffer layer
a thickness of 2.2 µm and bandgap of 1.9 eV [2]. Besides, a having Eg of 2.4 eV [26], creating a junction between them,
conversion efficiency of 26% was reported for a 3-µm-thick while the bandgap of the ZnO window layer is 3.3 eV [27].
CIGS absorber layer with a 0.03-µm ZnS buffer layer [16]. The proposed TFSC CIGS configuration (Al/SnO2/CdS/
When using a 4-µm CIGS absorber layer with Eg of 1.15 eV, CIGS/BaSi2/Mo) is depicted in Fig. 1b, including a thin,
an efficiency of 26.15% was found [17]. The efficiency of 0.3-µm ­BaSi2 BSF layer to significantly reduce the mate-
fabricated CIGS solar cells was increased from 9% to 14% rial cost without compromising the performance. The ­BaSi2
by the use of a molybdenum selenide (­ MoSe2) BSF layer BSF layer with Eg of 1.3 eV [22] is inserted between the
[18]. The theoretical efficiency of CIGS solar cells was CIGS layer and the molybdenum (Mo) back contact layer.
enhanced from 19.48% to 24% [19] and from 17.99% to Furthermore, the traditional ZnO window layer is changed
25.29% [20] by using a tin selenide (­ SnSe2) and tin sulfide to a FTO ­(SnO2:F) layer having Eg of 3.6 eV [28]. This
(SnS) BSF layer, respectively. This review of the literature new solar cell design comprises a 0.8-µm-thick p-type
indicates that the CIGS absorber layer is usually 1–4 µm (NA = 1 × 1018 cm−3) CIGS absorber layer, a 0.3-µm-thick
thick [2, 5, 12–17, 20]. It is obvious that the cost of the p+-type (NA = 5 × 1018 cm−3) ­BaSi2 BSF layer on the back
solar cell will increase as the thickness of the CIGS absorber Mo layer, a 0.05-µm-thick n-type CdS buffer layer, and a
layer is increased. In this regard, as a potential candidate 0.05-µm-thick ­SnO2 window layer. Aluminum (Al) is used
material for the back-surface field (BSF) layer, the emerging as the front grid contact, having a metal surface work func-
semiconducting barium silicide ­(BaSi2) could be utilized in tion (WF) of 4.06 eV [3]. The molybdenum (Mo) back con-
CIGS-based solar cells to reduce the thickness of the CIGS tact layer has a WF of 4.95 eV [14].
absorber layer.
BaSi2 comprises barium (Ba) and silicon (Si). Barium
has an atomic number of 56 and is found in period 6 of the
Periodic Table, having soft elemental form with silvery-gray
color [21]. The unit cell of the semiconductor depends on
silicon (Si), which has an atomic number of 14 and is found
in period 3, having four outer shell electrons. Orthorhom-
bic ­BaSi2 has been applied in TFSC applications because
of its wide bandgap from 1.1 to 1.35 eV and excellent sta-
bility at standard pressure and temperature [22]. Ba and
Si are Earth-abundant materials, so the incorporation of a
thin film of B­ aSi2 could greatly reduce the fabrication cost
of CIGS solar cells. The absorption coefficient of B ­ aSi2 is
about 30 [23] to 40 [24] times greater than that of single-
crystalline silicon (c-Si). Based on both experimental and
theoretical approaches [21–24], ­BaSi2 is considered to be
a perfect absorber material for solar cells. The potential of
­BaSi2 material in TFSCs has been explored [25] based on Fig. 1  The CIGS-based solar cell structures: a the conventional cell,
its fundamental properties. b the proposed cell

13
Journal of Computational Electronics

For the design and understanding of chalcogenide-

based solar cells, numerical simulations have been of great
importance through the years. In this work, the solar cell
capacitance simulator-one dimensional (SCAPS-1D) [28]
simulation program is employed. SCAPS is based on three
important equations of semiconductor physics: Poisson’s
equation, and the continuity equations for free electrons
and free holes. In SCAPS, almost all the parameters can be
graded, including the bandgap, electron affinity, mobility,
doping, thermal speed, effective density of states, etc. Along
with the standard test condition (STC) for the illumination
spectrum (global air mass, AM 1.5G), a variety of other illu-
mination spectra, including AM0, AM1.5D, monochromatic,
white, etc., are available.
Table 1 presents the values for the parameters of the lay-
ers used in the solar cells that were input into the simulation.
The amount of current passing through the heterojunc-
Fig. 2  The energy band diagram of the proposed CIGS solar cell with
tion is determined by the band alignment. The energy band a ­BaSi2 BSF layer
diagram for the proposed cell is illustrated in Fig. 2 based on
the energy band panel data obtained after the SCAPS simu-
lation, illustrating the bandgap and layer thickness of each
material. From Fig. 2, one can easily identify the band bend-
ing that occurs between the ­BaSi2 and CIGS junction due to
the different doping concentration levels used in this study.

Table 1  The materials properties used in the simulation

Parameter (unit) n+-Type window (ZnO) n+-Type window n-Type buffer (CdS) p-Type absorber p+-Type BSF ­(BaSi2)
­(SnO2) (CIGS)

Thickness (µm) 0.05 0.05 0.05 0.8* 0.3*

Bandgap (eV) 3.3 [2, 5, 28] 3.6 [28] 2.4 [29] 1.1 [5, 7] 1.3 [22]
Electron affinity (eV) 4.6 [5, 28] 4 [28] 4.4 [5, 7] 4.2 [7] 3.3 [22]
Dielectric permittivity 9 [2, 5, 28] 9 [28] 10 [29] 13.6 [5, 7] 11.17 [30]
(relative)
CB effective DOS 2.2 × 1018 [28] 2.2 × 1018 [28] 2.2 × 1018 [29] 2.2 × 1018 [5, 23] 2.6 × 1019 [22]
­(cm−3)
VB effective DOS 1.8 × 1019 [28] 1.8 × 1019 [28] 1.8 × 1019 [7] 1.8 × 1019 [7] 2 × 1019 [22]
­(cm−3)
Electron thermal veloc- 1 × 107 1 × 107 1 × 107 1 × 107 1 × 107
ity (cm/s)
Hole thermal velocity 1 × 107 1 × 107 1 × 107 1 × 107 1 × 107
(cm/s)
Electron mobility 100 [2, 28] 100 [28] 100 [5, 7] 100 [2, 5] 820 [22]
­(cm2/V s)
Hole mobility 25 [2, 28] 25 [28] 25 [5, 7] 25 [2, 7] 100 [22]
­(cm2/V s)
Donor density, ND 1 × 1019 1 × 1019 1 × 1018 0 0
­(cm−3)
Acceptor density, NA 0 0 0 1 × 1018* 5 × 1018
­(cm−3)
Defect type SA SA SA SD SD
Defect density ­(cm−3) 1 × 1015 1 × 1015 1 × 1015 1 × 1014* 1 × 1015

CB conduction band, VB valence band, DOS density of states, SA single acceptor, SD single donor
*A variable field

13
 Journal of Computational Electronics

This diagram confirms the possibility that the proposed solar

cell will be able to achieve the desired performance.

3 Results and discussion

To simulate the optoelectronic characteristics of the CIGS

solar cell, SCAPS software is applied to reveal the effect of
decreasing the thickness of the absorber layer (CIGS). The
numerical results also illustrate the effect of the introduc-
tion of the new ­BaSi2 BSF layer on the device performance.
The impact of varying the thickness of the CIGS layer, the
defects, and the temperature on the photovoltaic parameters
was studied for both layouts, under irradiation by the global
air mass (AM) 1.5G solar spectrum at temperature of 300 K
and incident solar power of 1000 W/m2. First, the series and
shunt resistances of the solar cell are considered to be zero
and infinitely large, respectively, which reflects the case of
an ideal solar cell.

3.1 The effect of the thickness of the CIGS absorber

layer

3.1.1 Case I: conventional design (without ­BaSi2)

The conventional layout of a CIGS solar cell (Fig. 1a) is first

examined analytically. The findings are evaluated by com- Fig. 3  The effect of varying the CIGS thickness with and without the
parison with the results for theoretically modeled [5, 31] and ­BaSi2 BSF layer
experimentally fabricated [9, 10, 32–34] CIGS cells. The
impact of the thickness of the -type CIGS absorber on the
photovoltaic (PV) performance of the cell is shown in Fig. 3,
where the thickness of the absorber layer is varied from 0.1
to 3 µm. In this case, the parameter values for the different
layers as specified in Table 1 are applied at a temperature of
300 K under 1-sun illumination.
A huge number of photons incident from the solar illu-
mination are absorbed when the thickness of the CIGS
absorber layer is increased, thereby increasing the effi-
ciency (η). Indeed, the efficiency is 7.09% and 25.1% for the
absorber with thickness of 0.1 and 3 µm, respectively. For
the 3-µm-thick CIGS absorber, the fill factor (FF) is 85.1%,
the open-circuit voltage (VOC) is 0.649 V, and the short-
circuit current density (JSC) is 40.4 mA/cm2, in compari-
son with analogous reference values of 85% for the FF [16,
35], 0.64–0.69 V for VOC [9, 10, 31, 32], and ~ 40 mA/cm2
for JSC [34]. Figure 3 presents the increase in the efficiency
with increasing thickness. Also, VOC and JSC varied in pro-
Fig. 4  The variation of Q with wavelength (λ) for the CIGS solar cell
portion to the thickness of the absorber layer. The quantum
efficiency (QE) is the ratio of the output charge carrier to
the absorbing photons. The variation of the QE of the basic CIGS cell with respect to the thickness of the CIGS layer is
shown in Fig. 4.

13
Journal of Computational Electronics

3.1.2 Case II: the CIGS solar cell with a ­BaSi2 BSF layer to Eq. (2) [41]. However, the FF still retains a good value
for solar cell operation.
A high impurity doping concentration at the back side of
VMPP × JMPP
the solar cell can prevent the strong recombination of the FF = , (2)
minority charge carriers (electrons) at the metallic back VOC × JSC
contact layer. This can be accomplished by introducing a where VMPP and JMPP denote the voltage and current, respec-
back-surface field (BSF) layer with a doping concentration tively, at the maximum power point. The quantum efficiency
that is greater than that in the active absorber layer. In the (QE) can be calculated as the ratio of the current flowing to
present study, an ultrathin ­BaSi2 layer is incorporated to the external circuit to the number of incident charge carriers.
act as such a BSF layer (with doping of 5 × 1018 cm−3), If an electron–hole pair is generated by each photon with
adjacent to the CIGS absorber layer (with doping of a particular wavelength, the external quantum efficiency
1 × 1018 cm−3). The interface between the p+-type ­BaSi2 (EQE) is equal to 1. The impact of the different thicknesses
and the p-type CIGS behaves like a p–n junction, forming in the new layout on the QE is shown in Fig. 4. The QE
an electric field that acts as a barrier to the flow of minor- of the CIGS/BaSi2 solar cell first increases with increas-
ity carriers to the rear surface. By reflecting the minor- ing CIGS layer thickness, based on the increased capture
ity carriers, the B ­ aSi 2 BSF layer will thus increase the of photons by thicker absorber layers. At certain points in
short-circuit current while decreasing the dark current. the wavelength range, all the curves start to drop towards
The incorporation of the BSF layer thereby minimizes the zero QE, because each material can only absorb photons in
surface recombination rate and thus enhances the perfor- a particular wavelength range of the visible light spectrum.
mance of the solar cell. Indium and gallium are high-cost materials in CIGS-
In this study, the parameter values indicated in Table 1 are based solar cells [4]. Hence, producing absorbers with
applied at a temperature of 300 K under 1-sun illumination. greater thickness of CIGS is not recommended. Similarly,
First, the thickness of the B­ aSi2 layer is varied from 0.1 to minimizing the thickness of the CIGS material can decrease
1 µm. This initial assessment reveals that the efficiency of the amounts of Ga and In required, which will ultimately
the cell becomes stable in the range from 0.3 to 1 µm. Hence, lower the manufacturing cost of such devices. B ­ aSi2 is
a 0.3-µm-thin ­BaSi2 BSF layer is considered optimal and cheaper and more widely available than Ga and In, and
applied in the remaining investigation. The traditional CIGS moreover these results indicate higher efficiency when using
solar cell with a 0.8-µm absorber layer shows an efficiency of the new cost-effective solar cell layout compared with the
19.72%, but after addition of the 0.3-µm thin p+-type ­BaSi2 previous design. The results of this investigation could thus
layer, the modified CIGS solar cell exhibits an efficiency of guide manufacturing companies towards more profitable
26.24% for the same thickness of 0.8 µm for the CIGS layer, production of CIGS solar cells.
revealing a great enhancement in efficiency for the same
absorber layer thickness. This result is in good agreement
with experimental works [18, 36, 37] as well as theoreti- 3.2 Impact of the CIGS acceptor density
cal research in literature [5, 19, 20, 38–40]. The key reason on the performance of the solar cell
for this is that both ­BaSi2 and CIGS act as absorbers (with
a total thickness of 1.1 µm) in the solar cell, whereby the The acceptor density (NA) in the CIGS layer is varied from
­BaSi2 layer helps to increase the absorption of photon energy 1 × 1012 to 1 × 1020 cm−3 to investigate its effect on the per-
by the cell. A large number of photons can thus be captured formance of the CIGS solar cell. In this case, the acceptor
and more electron–hole pairs generated accordingly. Thus, ­ aSi2 layer is 5 × 1018 cm−3, while the donor
density for the B
when moving from the conventional cell design to that pro- density in the CdS layer and S ­ nO2 layer is kept constant
posed herein, JSC increases from 34.1 to 40.559 mA/cm2 at 1 × 1018 cm−3 and 1 × 1019 cm−3, respectively. Figure 5
while VOC also increases from 0.691 to 0.8429 V. Very simi- shows the effect of varying the acceptor density in the CIGS
lar behavior of the JSC and VOC parameters was observed in layer. The results reveal that changing the NA value in the
Refs. [2, 5, 6]. Similarly, on incorporation of the BSF layer, CIGS material from 1 × 1012 to 1 × 1020 cm−3 increases VOC
JSC increases from 29.25 to 33.47 mA/cm2 [19] and VOC from 0.704 to 0.963 V, JSC from 40.18 to 40.5 mA/cm2, the
from 0.62 to 0.79 V [20]. FF from 39.79% to 83.9%, and the efficiency from 16.7% to
All of these effects enhance the efficiency of the solar 32.7%. Varying the acceptor density in the CIGS absorber
cell, greatly improving the PV performance of the traditional layer changes all the PV parameters in an analogous way
CIGS-based cell. However, the fill factor decreases from to that reported in Refs. [5, 16]. A very similar phenomena
83.8% to 76.763%. This occurs because both VOC and JSC has been reported when changing the NA value in the CdTe
increase, whereas VMPP and JMPP do not increase accordingly absorber layer with using a vanadium pentoxide ­(V2O5) BSF
but rather decrease. Thus, the FF clearly reduces according in CdTe solar cells [38].

13
 Journal of Computational Electronics

Fig. 6  The effect of varying the defect density in the CIGS layer on
Fig. 5  The impact of the acceptor density in the CIGS layer on the the characteristics of the proposed solar cell
characteristics of the solar cell

can reach the junction and generate current is increased. In

Considering the cost and growth rate of device fabrica- other words, the higher the defect density, the lower the rate
tion, the optimal value of the acceptor impurity density in of creation of electron–hole pairs. As described in Refs. [5,
the CIGS material for the proposed design is found to be 6, 13, 15], defects can easily deteriorate the photovoltaic
1 × 1018 cm−3. performance, thereby diminishing the efficiency of the solar
cell. In this regard, to improve the output of the proposed
3.3 The impact of the bulk defect density cell, the defect density must be limited to as low as possible
in the CIGS absorber layer on the proposed and should not exceed the limit of 1 × 1014 cm−3.
solar cell
3.4 The effect of the resistances on the PV
To study the effect of the bulk defect density in the CIGS performance of the proposed solar cell
layer on the photovoltaic performance of the solar cell, the
single-donor-type bulk defect density is varied from 1 × 1012 The effect of the series (Rs) and shunt resistances (Rsh) on the
to 5 × 1017 cm−3, while keeping the defect densities of all photovoltaic device is studied using SCAPS-1D software.
the other layers constant at the values specified in Table 1. The Rs of the solar cell results from the bulk resistance, the
Figure 6 shows the impact of the different defect densities on front and back metallic contact resistance, and the circuit
the performance parameters of the solar cell. The values of terminal resistance [5]. The key effect of the series resistance
VOC and JSC decrease as the defect density is increased; JSC is to diminish the fill factor [42], meaning that a FF of 100%
drops from 40.87 to 28.5 mA/cm2, while VOC falls from 0.97 cannot be achieved. Lower series and higher shunt resist-
to 0.516 V. The conversion efficiency thus falls from 34.27% ances must be achieved to increase the efficiency. The values
(for a defect density of 1 × 1012 cm−3) to 9.11% (for a defect of these resistances also affect VOC and JSC. Using SCAPS-
density of 5 × 1017 cm−3). As the defect density is increased, 1D, the value of Rs is varied to observe the impact on the
the probability of carrier recombination before the carriers performance of both the traditional and proposed solar cells.

13
Journal of Computational Electronics

Fig. 8  The impact of varying Rsh on the characteristics of the pro-

Fig. 7  The impact of varying Rs on the characteristics of the proposed
posed cell
cell

The value of Rs is varied from 0 (ideal case) to 6 Ω cm2, (efficiency), and the FF are temperature-dependent param-
while the shunt resistance (Rsh) is fixed at 105 Ω cm2. Fig- eters. Note that VOC varies inversely proportionally with the
ure 7 illustrates the effect of changing Rs on the proposed temperature, because the saturation current decreases rapidly
cell including the ­BaSi2 BSF layer, indicating that the effi- at increasing temperature, due to the increase in the reverse
ciency of the solar cell is significantly degraded by increas- saturation current. VOC is a saturation current-dependent
ing Rs, as also observed in literature [5, 16]. These results parameter. The short-circuit current density (JSC) remains
reveal a decrease in the efficiency from 26.24% to 18.59% almost constant with increasing temperature. The proposed
when changing Rs over the mentioned range. solar cell layout also shows a good response to the tempera-
Fixing the series resistance, Rs at 0.5 Ω cm2, the value of ture variation. When the temperature is low, the efficiency
Rsh is now varied from 10 to 1 × 107 Ω cm2. Figure 8 illus- is high enough. At a temperature of 273 K, the efficiency is
trates the results when changing the Rsh value of the cell as 25.81%, but it reduces to 17.9% at 473 K. At much higher
mentioned, showing that the efficiency of the proposed solar temperature, the efficiency was also degraded in Refs. [2,
cell increases with increasing Rsh. The efficiency is found to 5, 21].
increase from 25.08% to 25.56% on increasing the Rsh value The lower efficiency observed at higher temperatures
of the proposed solar cell including the ­BaSi2 BSF layer. is due to the temperature dependence of parameters such
as the carrier concentrations, electron and hole mobilities,
3.5 The impact of the cell operating temperature and energy bandgaps of the materials [32]. The Eg values
become unbalanced as T is increased, and the resulting elec-
The performance of the CIGS solar cells with and with- tron–hole recombination in the unstable bandgaps dimin-
out the ­BaSi2 BSF layer is examined while varying the cell ishes the conversion efficiency [14]. For both cases, viz.
temperature (T) from 273 to 473 K, with all other param- with and without the BSF, the short-circuit current density
eters kept constant as specified in Table 1. Figure 9 illus- remains almost constant. With the B ­ aSi2 BSF layer, JSC is
trates the effect of such variation of the temperature on JSC, always ~ 6.6 mA/cm2 higher than in the conventional struc-
VOC, the FF, and the efficiency for both cases, viz. with ture across the entire temperature variation. It is obvious
and without the B ­ aSi2 BSF layer. The characteristics of the from Fig. 9 that both the cell efficiency and open-circuit
solar cell depend strongly on temperature, since VOC, JSC, 𝜂 voltage decrease linearly for the case of the basic CIGS cell.

13
 Journal of Computational Electronics

Fig. 10  The J–V characteristic of the CIGS-based solar cells

Table 3 presents a comparison of the PV parameters

obtained by some theoretical and experimental analyses of
CIGS solar cells, in comparison with the results of the cur-
rent work. It is evident from Table 3 that, with only a thin,
800-nm CIGS absorber layer, the TFSC cell proposed herein
could be commercially more attractive compared with the
other CIGS-based structures mentioned in Table 3. Also, the
structure provides a route to decrease the absorber material
cost of CIGS-based solar cells. The proposed solar cell is
thus superior to the other CIGS solar cell structures [2, 5,
Fig. 9  The temperature dependence of the PV parameters of the cell
9–20].

However, on inclusion of the ­BaSi2 BSF layer, these two 3.7 An evaluation of the potential of the ­BaSi2 BSF
parameters reduce in a less linear fashion, clearly confirm- in CIGS solar cells
ing the overall improvement achieved by the addition of the
ultrathin ­BaSi2 BSF layer to the conventional CIGS solar Materials such as SnS, S ­ nSe2, Si, M
­ oSe2, etc. have been
cell. employed as a BSF layer in CIGS solar cells and analyzed
either theoretically or experimentally. The beneficial effect
of using a ­BaSi2 BSF in CIGS solar cells and a comparative
3.6 The J–V characteristic of the CIGS‑based solar summary versus similar materials are presented in Table 4.
cells Among the mentioned BSF layers, ­BaSi2 exhibits a rela-
tively higher absorption coefficient. Considering its ready
Figure 10 displays the current–voltage (J–V) characteris- availability in the Earth’s crust, ­BaSi2 enables lower-cost
tic curves of the CIGS solar cells. Table 2 summarizes the fabrication compared with SnS, ­SnSe2, or M ­ oSe2. Moreo-
photovoltaic parameters of the proposed solar cell structure ver, its bandgap can be adjusted to the nearly optimal value
with a 0.8-µm-thick CIGS absorber layer, which is seen to of 1.4 eV by introducing strontium (Sr) impurities [21]. It
be superior to the performance of the conventional CIGS also possesses a long minority-carrier diffusion length and
solar cell with either a 0.8-µm- or 3-µm-thick absorber, as lifetime [21, 25]. The simulation results clearly verify the
observed in the table. However, the FF is reduced by about potential of such a ­BaSi2 BSF layer to enhance the PV per-
8% in the proposed cell (with a 0.8-µm-thick CIGS absorber) formance of CIGS solar cells. Furthermore, based on the
compared with the conventional cell (with a 3-µm-thick results presented throughout the current paper and consid-
CIGS absorber). The reason behind this decrement is ering the advantages summarized in Table 4, ­BaSi2 can be
explained in Sect. 3.1 (case II), although the value is still considered to be the best material for the BSF layer of CIGS-
a good result. based TFSCs.

13
Journal of Computational Electronics

Table 2  A summary of the simulation results

Cell type CIGS thick- BaSi2 thick- JSC (mA/cm2) JMPP (mA/cm2) VOC (V) VMPP (V) FF (%) η (%)
ness (µm) ness (µm)

Conventional 3.0 – 40.40 38.700 0.731 0.649 85.1 25.10

Conventional 0.8 – 34.10 32.296 0.691 0.610 83.8 19.71
Proposed 0.8 0.3 40.56 37.060 0.843 0.708 76.8 26.24

Table 3  The photovoltaic No. Type of research CIGS layer VOC (V) JSC (mA/cm2) FF (%) η (%) Ref.
performance of the proposed thickness (nm)
cell in comparison with other
reported CIGS solar cells 1 Experimental 2000 0.671 34.90 77.60 18.10 [32]
2 Experimental 1000 0.689 35.71 78.12 19.20 [9]
3 Experimental 2200 0.690 35.50 81.20 19.90 [10]
4 Experimental – 0.741 37.80 80.60 22.60 [33]
5 Experimental – 0.734 39.58 80.40 23.85 [34]
6 Theoretical 3000 0.640 34.60 79.50 17.70 [31]
7 Theoretical 1000 0.743 34.47 83.09 21.30 [5]
8 Theoretical 1000 0.780 38.66 80.00 24.45 [15]
9 Theoretical 2200 1.020 27.66 86.88 24.55 [2]
10 Theoretical 3000 0.824 36.90 85.50 26.00 [16]
11 Theoretical 4000 0.800 38.20 75.00 26.15 [17]
12 Theoretical 2500 0.798 38.66 85.15 26.30 [35]
13 Theoretical 800 0.843 40.56 76.80 26.24 [a]

[a] The cell proposed herein

Table 4  The influence of the No. Type of research Absorber BSF η without η with BSF (%) Enhancement Ref.
back-surface field (BSF) layer BSF (%) in η (%)
in comparison with similar
research 1 Experimental Si ZnS 6.40 11.02 72.19 [36]
2 Experimental Si Al 12.96 13.75 6.09 [37]
3 Experimental CIGS MoSe2 9 14 55.55 [18]
4 Theoretical CdTe V2O5 19.58 23.50 20.00 [38]
5 Theoretical CZTS CZTS 12.05 14.11 17.09 [39]
6 Theoretical ZnTe Sb2Te3 7.14 18.33 156.72 [40]
7 Theoretical CZTSSe SnS 12.30 17.25 40.24 [43]
8 Theoretical CIGS Si 16.39 21.30 29.96 [5]
9 Theoretical CIGS SnSe2 19.48 24 23.20 [19]
10 Theoretical CIGS SnS 17.99 25.29 40.57 [20]
11 Theoretical CIGS BaSi2 19.71 26.24 33.13 [a]

[a] The cell proposed herein

4 Conclusions investigated. A thin, 0.3-µm ­BaSi2 layer is inserted into the

basic CIGS solar cell structure of Al/ZnO/CdS/CIGS/Mo
SCAPS-1D simulation is applied to study the behavior with a ZnO window, resulting in an alternative structure
of thin-film solar cells based on CIGS absorber material. of Al/FTO/CdS/CIGS/BaSi2/Mo with a FTO window. The
The effects of the thickness, doping concentration, bulk photovoltaic parameters of the modeled cells are evaluated
defect density, resistances, and temperature on the opto- numerically and compared with those reported in other
electronic output parameters of the cells are thoroughly theoretical and experimental studies. The addition of the
back-surface field (BSF) layer made from B ­ aSi 2 results

13
 Journal of Computational Electronics

in a significant enhancement of the current, voltage, and 8. Ando, Y., Ishizuka, S., Wang, S., Chen, J., Islam, M.M., Shibata,
efficiency while also notably reducing the thickness of H., Akimoto, K., Sakurai, T.: Relationship between bandgap grad-
ing and carrier recombination for Cu(In,Ga)Se2-based solar cells.
the absorber material. B ­ aSi 2 can thus be said to repre- Jpn. J. Appl. Phys. (2018). https:​ //doi.org/10.7567/JJAP.57.08RC0​
sent a favorable material for the BSF in CIGS-based solar 8
cells. The optimized thickness of the FTO window, CdS 9. Ramanathan, K., Contreras, M.A., Perkins, C.L., Asher, S.,
buffer, CIGS absorber, and ­BaSi2 BSF layers is found to Hasoon, F.S., Keane, J., Young, D., Romero, M., Metzger, W.,
Noufi, R., Ward, J., Duda, A.: Properties of 19.2% efficiency ZnO/
be 0.05 µm, 0.05 µm, 0.8 µm, and 0.3 µm, respectively. CdS/CuInGaSe2 thin-film solar cells. Prog. Photovolt. Res. Appl.
The acceptor impurity concentration in the CIGS layer (2003). https​://doi.org/10.1002/pip.494
is chosen to be 1 × 10 18 cm −3. For a certain number of 10. Repins, I., Contreras, M.A., Egaas, B., DeHart, C., Scharf, J.,
defects, this structure presents an efficiency (η) of 26.24% Perkins, C.L., To, B., Noufi, R.: 19.9%-efficient ZnO/CdS/CuIn-
GaSe2 solar cell with 81.2% fill factor. Prog. Photovolt. Res. Appl.
along with a VOC of 0.843 V, JSC of 40.56 mA/cm2, and FF (2008). https​://doi.org/10.1002/pip.822
of 76.76%. The results of this work provide design guide- 11. Green, M.A., Hishikawa, Y., Dunlop, E.D., Levi, D.H., Hohl-
lines for the introduction of an ultrathin B­ aSi2 BSF layer Ebinger, J., Yoshita, M., Ho-Baillie, A.W.Y.: Solar cell efficiency
into a conventional CIGS solar cell to enhance its overall tables (version 53). Prog. Photovolt. Res. Appl. (2019). https​://
doi.org/10.1002/pip.3102
performance while significantly reducing the cost of the 12. Daoudia, A.K., Hassouani, Y.E., Benami, A.: Investigation of the
absorber material. effect of thickness, band gap and temperature on the efficiency
of CIGS solar cells through SCAPS-1D. Int. J. Eng. Tech. Res.
Acknowledgements The authors would like to thank Marc Burgelman (IJETR) 6, 71–75 (2016)
and his team at the Department of Electronics and Information Systems 13. Mostefaoui, M., Mazari, H., Khelifi, S., Bouraiou, A., Dabou, R.:
(ELIS), University of Gent, Belgium for providing the SCAPS software Simulation of high efficiency CIGS solar cells with SCAPS-1D
package, version 3.3.07. software. Energy Proc. (2015). https​://doi.org/10.1016/j.egypr​
o.2015.07.809
Funding This research did not receive any specific grants from funding 14. Robin, M.S.R., Mansoor, M., Rasmi, M., Sarkar, M.S.Z., Rabbi,
agencies in the public, commercial, or not-for-profit sectors. A.S.M., Mamun, A.: Numerical modeling and analysis of ultra
thin film Cu(In,Ga)Se2 solar cell using SCAPS-1D. In: Interna-
tional Conference on Electrical Engineering and Information and
Compliance with ethical standards Communication Technology (iCEEiCT) (IEEE), 22–24th Sept.
2016, MIST, Dhaka, Bangladesh. https​://doi.org/10.1109/CEEIC​
Conflict of interest The authors declare that they have no conflicts of T.2016.78731​69
interest. 15. Benabbas, S., Rouabah, Z., Heriche, H., Chelali, N.: A numerical
study of high efficiency ultra-thin CdS/CIGS solar cells. Afr. J.
Sci. Technol. Innov. Dev. (2016). https​://doi.org/10.1080/20421​
338.2015.11189​29
16. Sylla, A., Touré, S., Vilcot, J.-P.: Numerical modeling and simu-
References lation of CIGS-based solar cells with ZnS buffer layer. Open J.
Model. Simul. (2017). https​://doi.org/10.4236/ojmsi​.2017.54016​
1. Alhammadi, S., Park, H., Kim, W.K.: Optimization of intrinsic 17. AlZoubi, T., Moustafa, M.: Numerical optimization of absorber
ZnO thickness in Cu(In, Ga)Se2-based thin film solar cells. Mate- and CdS buffer layers in CIGS solar cells using SCAPS. Int.
rials (2019). https​://doi.org/10.3390/ma120​91365​ J. Smart Grid Clean Energy (2019). https​://doi.org/10.12720​/
2. Bouich, A., Hartiti, B., Ullah, S., Ullah, H., Touhami, M.E., San- sgce.8.3.291-298
tos, D.M.F., Mari, B.: Experimental, theoretical, and numerical 18. Kohara, N., Nishiwaki, S., Hashimoto, Y., Negami, T., Wada, T.:
simulation of the performance of ­CuInxGa(1–x)Se2-based solar Electrical properties of the Cu(In,Ga)Se2/MoSe2/Mo structure.
cells. Optik (2019). https​://doi.org/10.1016/j.ijleo​.2019.02.067 Sol. Energy Mater. Sol. Cells (2001). https​://doi.org/10.1016/
3. Moon, M.M.A., Rahman, M.F., Hossain, J., Ismail, A.B.M.: S0927​-0248(00)00283​-X
Comparative study of the second generation a-Si:H, CdTe, and 19. Ahamed, E.M.K.I., Bhowmik, S., Matin, M.A., Amin, N.: Highly
CIGS thin-film solar cells. Adv. Mater. Res. (2019). https​://doi. efficient ultra thin Cu(In,Ga)Se2 solar cell with tin Selenide BSF.
org/10.4028/www.scien​tific​.net/AMR.1154.102 In: 2017 International Conference on Electrical, Computer and
4. Candelise, C., Spiers, J.F., Gross, R.J.K.: Materials availability Communication Engineering, February 16–18, 2017, Cox’s Bazar,
for thin film (TF) PV technologies development: a real concern? Bangladesh. IEEE. https:​ //doi.org/10.1109/ECACE.​ 2017.791294​ 2
Renew. Sustain. Energy Rev. (2011). https​://doi.org/10.1016/j. 20. Benabbas, S., Heriche, H., Rouabah, Z., Chelali, N.: Enhancing
rser.2011.06.012 the efficiency of CIGS thin film solar cells by inserting novel
5. Heriche, H., Rouabah, Z., Bouarissa, N.: New ultra thin CIGS back surface field (SnS) layer. In: 2014 North African Work-
structure solar cells using SCAPS simulation program. Int. shop on Dielectric Materials for Photovoltaic Systems, October
J. Hydrogen Energy (2017). https ​ : //doi.org/10.1016/j.ijhyd​ 26–27, 2014, Algeria. IEEE. https​://doi.org/10.1109/NAWDM​
ene.2017.02.099 PV.2014.69976​11
6. Guirdjebaye, N., Ouédraogo, S., Ngoupo, A.T., Tcheum, G.L.M., 21. Moon, M.M.A., Ali, M.H., Rahman, M.F., Kuddus, A., Hossain,
Ndjaka, J.M.B.: Junction configurations and their impacts on J., Ismail, A.B.M.: Investigation of thin-film p-BaSi2/n-CdS het-
Cu(In,Ga)Se2 based solar cells performances. Opto-Electron. Rev. erostructure towards semiconducting silicide based high efficiency
(2019). https​://doi.org/10.1016/j.opelr​e.2019.02.001 solar cell. Phys. Scr. (2019). https​://doi.org/10.1088/1402-4896/
7. Oyedele, S.O., Soucase, B.M., Aka, B.: Numerical simulation and ab49e​8
performance optimization of Cu(In,Ga)Se2 solar cells. IOSR J. 22. Doorene, S.V.: Barium disilicide: Development of a novel, low
Appl. Phys. (2016). https​://doi.org/10.9790/4861-08040​40111​ cost and earth abundant absorber material for thin film solar cell

13
Journal of Computational Electronics

applications. MS Thesis, Sustainable Energy Technology, Delft 34. Green, M.A., Dunlop, E.D., Levi, D.H., Hohl-Ebinger, J., Yosh-
University of Technology, Delft, Netherlands, June 6, 2017. https​ ita, M., Ho-Baillie, A.W.Y.: Solar cell efficiency tables (version
://repos​itory​.tudel​ft.nl/islan​dora/objec​t/uuid%3A177​b86f2​-cb70- 54). Prog. Photovolt. Res. Appl. (2019). https​://doi.org/10.1002/
4e74-b02b-b0ee4​21c7e​36 (2017). Accessed 25 July 2019 pip.3171
23. Khan, M.A., Suemasu, T.: Donor and acceptor levels in impurity- 35. Fridolin, T.N., Maurel, D.K.G., Ejuh, G.W., Bénédicte, T.T.,
doped semiconducting ­BaSi2 thin films for solar-cell application. Marie, N.J.: Highlighting some layers properties in performances
Phys. Status Solidi A (2017). https​://doi.org/10.1002/pssa.20170​ optimization of CIGSe based solar cells: case of Cu(In, Ga)Se–
0019 ZnS. J. King Saud Univ. Sci (2018). https​://doi.org/10.1016/j.
24. Deng, T., Sato, T., Xu, Z., Takabe, R., Yachi, S., Yamashita, Y., jksus​.2018.03.026
Toko, K., Suemasu, T.: p-BaSi2/n-Si heterojunction solar cells on 36. Yang, X., Chen, B., Chen, J., Zhang, Y., Liu, W., Sun, Y.: ZnS
Si(001) with conversion efficiency approaching 10%: comparison thin film functionalized as back surface field in Si solar cells.
with Si(111). Appl. Phys. Express (2018). https:​ //doi.org/10.7567/ Mater. Sci. Semicond. Process. (2018). https​://doi.org/10.1016/j.
apex.11.06230​1 mssp.2017.08.011
25. Suemasu, T., Usami, N.: Exploring the potential of semiconduct- 37. Kaminski, A., Vandelle, B., Fave, A., Boyeaux, J.P., Nam, L.Q.,
ing ­BaSi2 for thin-film solar cell applications. J. Phys. D Appl. Monna, R., Sarti, D., Laugier, A.: Aluminium BSF in silicon
Phys. (2017). https​://doi.org/10.1088/1361-6463/50/2/02300​1 solar cells. Sol. Energ. Mater. Sol. Cells. (2002). https​://doi.
26. Kodalle, T., Choubrac, L., Arzel, L., Schlatmann, R., Barreau, org/10.1016/S0927​-0248(01)00185​-4
N., Kaufmann, C.A.: Effects of KF and RbF post deposition treat- 38. Kuddus, A., Rahman, M.F., Ahmmed, S., Hossain, J., Ismail,
ments on the growth of the CdS buffer layer on CIGS thin films— A.B.M.: Role of facile synthesized ­V2O5 as hole transport layer
a comparative study. Sol. Energy Mater. Sol. Cells (2019). https:​ // for CdS/CdTe heterojunction solar cell: validation of simulation
doi.org/10.1016/j.solma​t.2019.10999​7 using experimental data. Superlattices Microstruct. (2019). https​
27. Frisk, C., Platzer-Björkman, C., Olsson, J., Szaniawski, P., Wätjen, ://doi.org/10.1016/j.spmi.2019.10616​8
J.T., Fjällström, V., Salomé, P., Edoff, M.: Optimizing Ga-profiles 39. Benzetta, A.H., Abderrezek, M., Djeghlal, M.E.: Contribution to
for highly efficient Cu(In, Ga)Se2 thin film solar cells in simple improve the performances of ­Cu2ZnSnS4 thin-film solar cell via
and complex defect models. J. Phys. D Appl. Phys. (2014). https​ a back-surface field layer. Optik (2019). https​://doi.org/10.1016/j.
://doi.org/10.1088/0022-3727/47/48/48510​4 ijleo​.2018.12.048
28. Burgelman, M., Decock, K., Niemegeers, A.,Verschraegen, J., 40. Sohid, S.B., Kabalan, A.: Numerical analysis of ZnTe based solar
Degrave, S.: SCAPS Manual (version: 3.3.07). Department of cell with ­Sb2Te3 back surface field layer using SCAPS-1D. In:
Electronics and Information Systems, University of Gent, Bel- 2018 IEEE 7th World Conference on Photovoltaic Energy Con-
gium. http://scaps​.elis.ugent​.be (2018). Accessed 5 January 2019 version, June 10–15, 2018, USA. IEEE. https​://doi.org/10.1109/
29. Dabbabi, S., Nasr, T.B., Kamoun, T.: CIGS solar cells for space pvsc.2018.85478​00
applications: numerical simulation of the effect of traps created 41. Paul, D.I.: Experimental characterisation of photovoltaic modules
by high-energy electron and proton irradiation on the perfor- with cells connected in different configurations to address nonu-
mance of solar cells. JOM (2018). https​://doi.org/10.1007/s1183​ niform illumination effect. J. Renew. Energy (2019). https​://doi.
7-018-2748-9 org/10.1155/2019/51682​59
30. Huang, J., Lee, K., Tseng, Y.: Analysis of the high conversion 42. Series Resistance. https​://www.pvedu​catio​n.org/pvcdr​om/solar​
efficiencies ß-FeSi2 and ­BaSi2 n-i-p thin film solar cells. J. Nano- -cell-opera​tion/serie​s-resis​tance​ (2019). Accessed 15 July 2019
mater. (2014). https​://doi.org/10.1155/2014/23829​1 43. Omrani, M.K., Minbashi, M., Memarian, N., Kim, D.H.: Improve
31. Gloeckler, M.: Device physics of Cu(In, Ga)Se2 thin-film solar the performance of CZTSSe solar cells by applying a SnS BSF
cells. Ph.D. dissertation, Colorado State University, Fort Collins layer. Solid-State Electron. (2018). https​://doi.org/10.1016/j.
(2005) sse.2017.12.004
32. Nakada, T., Mizutani, M.: 18% efficiency Cd-free Cu(In,Ga)Se2
thin film solar cells fabricated using chemical bath deposition Publisher’s Note Springer Nature remains neutral with regard to
(CBD)-ZnS buffer layers. Jpn. J. Appl. Phys. (2002). https​://doi. jurisdictional claims in published maps and institutional affiliations.
org/10.1143/JJAP.41.L165
33. Jackson, P., Wuerz, R., Hariskos, D., Lotter, E., Witte, W., Pow-
alla, M.: Effects of heavy alkali elements in Cu(In, Ga)Se2 solar
cells with efficiencies up to 22.6%. Phys. Status Solidi Rapid Res.
Lett. (2016). https​://doi.org/10.1002/pssr.20160​0199

13