Optimization of the Output Parameters in
Kesterite-based Solar Cells by AMPS-1D
Joël Hervé Nkuissi Tchognia*, Youssef Arba, Khalid
Dakhsi, Bouchaib Hartiti, Abderraouf Ridah
Department of Physics, Faculty of Science and Technique
Hassan II University of Casablanca, Mohammedia, Morocco
*
hervetchognia@gmail.com
Philippe Thevenin
Department of Physics, Supelec
University of Lorraine,
Metz, France
Abstract— the presence of defects, grain boundaries and other
impurities in Cu2ZnSnS4 (CZTS) thin film material is
detrimental on output performances of fabricated solar cells.
Moreover the formation of non stable and high resistance
contacts to CZTS contribute also to lower the values of short
circuit density ( ), open circuit voltage ( ), fill factor ( )
and therefore the efficiency of thin film solar cells. This study
evaluates the baseline ZnO/CdS/CZTS cell structure using
AMPS-1D program and quantifies reasons for the efficiency
limits predicted in such a device. The effect of back contact
material, acceptor concentration as well as the absorber
thickness on CZTS cell performance is also studied. The solar cell
with a back contact work function of 5.6 eV, an acceptor
concentration of 4x1015cm-3 and an absorber layer’s thickness of
2250 nm showed the best performance with a conversion
efficiency of 15.68%. This presented work is essential because it
can contribute to fully understand the performance of CZTS
solar cell and to fabricate cells with a precise control of the
process parameters with a reduced surface as well as bulk
recombination and secondary phases.
Keywords: CZTS, solar cells, numerical simulation, thickness, back
metal contact, acceptor concentration
I. INTRODUCTION
Kesterite Cu2ZnSnS4 (CZTS) is a quaternary
semiconducting material which has been used recently as an
absorber in thin film solar cells. CZTS offers favorable optical
and electronic properties that are advantageous for thin film
photovoltaic applications. The kesterite compound CZTS is
composed of abundant, non-toxic elements and has been
studied by several groups over recent years as a promising
alternative to the binary CdTe and quaternary CuInGa (S, Se)2
(CIGS) absorber layers in thin film solar cells [1, 2]. Although
the CIGS cell conversion efficiency has recently reached up to
20% [3], the scarcity of indium and gallium has pushed
978-1-4673-7894-9/15/$31.00 ©2015 IEEE
Joël Hervé Nkuissi Tchognia, Jean-Marie Ndjaka
Department of Physics, Faculty of Science
University of Yaoundé I, P.O Box 812, Yaoundé, Cameroon
researchers working in this field to give a broad interest to the
CZTS material due to its excellent material properties such as
a direct and tunable band gap energy in the range of 1.4-1.6
eV and large absorption coefficient over 104 cm-1 [4, 5].
However, because of the quaternary phase, structural, optical
and electrical properties can be influenced by the deviation of
the stoichiometry composition which will result in formation
of native defects [6]. The current record for CZTS
photovoltaic efficiency is 12.6% in 2014 which was reported
by Wang et al. [7]. They have partially selenized the CZTS
using hydrazine pure-solution targeting a Cu-poor and Zn-rich
condition. However, focusing on pure sulfur based kesterite
material allows retention of a non toxic absorber. The study of
the phase diagram of this system has shown that impurities
including ternary and quaternary compounds are easier to
form than CZTS because single phase CZTS material can be
formed only in a very small region [8]. The defects as well as
the smaller grain size of CZTS lead to a shorter diffusion
length of carriers [9]. The efficiency in CZTS film based solar
cells is limited by low open circuit voltage ( ) due to
recombination losses, high series resistance ( ), and double
diode behavior in the current–voltage characteristic
particularly in thicker films [1]. Modeling current-voltage
characteristic is widely used in the analysis of single junction
solar cells to predict their performances [9]. Since
heterostructures like CZTS solar cells are complex in nature,
the need of numerical simulation is relevant to predict the
effect of various model parameters on the output performance
and to test the variability of the proposed structure. The aim of
this work is first to model baseline CZTS solar cell which
means that no kind of defects having discrete or Gaussian
distribution in materials and at interfaces are considered or
assumed in order to evaluate the theoretical efficiency
attainable by such a device, then to study the effect of back
metal contact material work function, acceptor concentration,
absorber thickness on the overall performances of the
traditional cell. We have used the AMPS-1D program in our
present study and details of structure and simulation of the
device are stated in following sections.
II. DEVICE STRUCTURE AND SIMULATION
In this study, numerical modeling of CZTS thin film solar cell
has been carried out by AMPS-1D program. The computer
simulation tool AMPS-1D (Analysis of Microelectronic and
Photonic Structures) [10] is employed by specifying the
semiconductor parameters in each defined layer of the cell
structure as input parameters to the simulations. Developed by
the Stephen Fonash’s group at the Pennsylvania State
University in USA, the uniqueness of this software lies in its
ability to examine a variety of device structures such as: [11]
homo-junction and hetero-junction p-n and p-i-n solar cells
and detectors, multi-junction solar cell structures, novel
photovoltaic and optoelectronic structures, Schottky barrier
devices with optional back layers [11,12]. It is the only
software that incorporates all recombination and defect charge
distribution mechanisms compared to other simulation
programs like SCAPS-1D and PISCES [13]. It evaluates the
recombination profile, band diagram under steady-state and
carrier transport in one dimension based on the Poisson’s
equation, the continuity equation for free electrons and the
continuity equation for free holes given respectively by [11]:
(1)
1 (2)
1 (3)
Where the electrostatic potential , the free electron n
and free hole p, trapped electron and trapped hole as
well as ionized donor-like doping and ionized acceptor-
like concentrations are all given in terms of the different
band gap energy levels by the Boltzmann and Fermi-Dirac
distributions, and are a function of the coordinate position " ".
and are the electron and hole current densities
respectively. The term is the optical carrier generation rate
due to illumination and is the total recombination rate
accounting for direct (band to band) and indirect (Shockley-
Read-Hall) recombination traffics.
Figure 1. Typical CZTS solar cell structure
The CZTS solar cell structure used in this simulation
consists of the following layers: ZnO window layer, CdS
buffer, CZTS absorber and Mo back contact metal on glass
substrate. Fig.1 illustrates the vertical cross section of the
typical cell. The device is considered to be illuminated from
the front ZnO to the end of the device with the AM1.5
spectrum. Outputs such as current-voltage characteristics
under dark and under illumination, the quantum efficiency, the
reflectivity and internal quantities such as the electric field, the
free and trapped carrier concentrations, the electron and hole
currents, the recombination and generation rates can be
extracted from the solutions provided by an AMPS simulation.
The basic input parameters used in this simulation were
selected based on literature, theory, in some cases reasonably
estimated and are specific to the device and materials
properties. The front contact is considered as a standard ohmic
contact and the back contact is concerned for optimizing the
cell performance.
III. RESULTS AND DISCUSSION
A. Basic device simulation
The AMPS-1D program has been used to explore the
output characteristics of traditional Mo/CZTS/CdS/ZnO solar
cells by varying the acceptor concentration of the absorber as
well as its thickness. The effect of electron work function of
the back metal contact in the range of 5-6 eV is also
investigated. In the baseline case, platinum was used as back
contact instead of molybdenum because of its high electron
work function of 5.7 eV. The choice of platinum in this work
is also done assuming its ideal ohmic contact to the CZTS
based on theoretical work function value. The use of this metal
in such as a role would be cost prohibitive; however the aim is
to evaluate the maximum theoretical efficiency attainable with
CZTS-based solar cells.
The resulting J-V characteristics and quantum
efficiency (QE) outputs are shown in Fig.2. The simulation
results predicted a maximal theoretical efficiency of 19.04%
attainable by CZTS-based solar cells. The drop at ~0.820
wavelength corresponds to the CZTS band gap. We can see
with the QE plot that more than 85% of the photons of the
visible range of the solar radiation are absorbed by the (Cm-3)
material. Effective density of states Nv 1.8x1019 1.8x1019 1.8x1018
(Cm-3)
Electron affinity CHI (eV) 4.40 3.80 4.10
25
Gaussian defect NG (Cm-3) 1x1017 1x1018 2.61x1015

20 Fig. 3 shows the simulated J-V characteristics of the

2
CZTS thin film solar cell obtained with the data of TABLE I.
Jsc = 21.687 mA/Cm
15
The results of our simulation were in good agreement with
Jsc (mA/Cm2)

Voc = 1.023 V
FF = 0.859 reported experimental results of the first Mo/CZTS/CdS/ZnO
10
 19.046 % solar cell. We can see in TABLE 2 a comparison with the
available data in literature [1, 14-17].
5
The open circuit voltage (Voc), quantum efficiency and
series resistances (Rs) are the three fundamental factors which
0
0,0 0,2 0,4 0,6 0,8 1,0 limit the performance of CZTS solar cell [9]. The short current
Voc (V) density (Jsc) of the solar cell is affected by the electron
transport properties at the ZnO/CdS interface and also by
shading losses caused by the front metal contact on the top of
0,9
ZnO layer. Normally, 11% area of PV cell front surface is
occupied by metallic contacts for charge carriers’ collection
[9]; this can limit the Jsc and can induce series resistive
Quantum Efficiency (%)

0,8

components degrading the carrier collection efficiency and

0,7 therefore the performance. This aspect must be well controlled
in order to improve the performance of solar cells.
0,6

B.  Effect of back metal work function

0,5

To evaluate the effect of the electron work function of

0,4
0,4 0,5 0,6 0,7 0,8 0,9 the back metallic contact, its value was varied from 5 to 6 eV
m) by step of 0.1 eV in this simulation. There is an incredible
Figure 2. J-V and QE plots of the baseline CZTS solar cell
improvement in solar cell parameters when increasing the
electron work function of the back metal contact as revealed
The basic input parameters used in this simulation for from Fig.4.
the traditional case where molybdenum plays the role of back
contact material are shown in TABLE I. The front surface 14
reflectivity and the back surface reflectivity are set to 0.1 and
0.8 respectively in order to reflect experimental quantum 12

efficiency data. The default illumination temperature is set to

10
300 K and the default illumination spectrum is set to the
Jcc (mA/Cm2)

global standard AM 1.5. 8

Jcc = 13.001 mA/Cm
2

TABLE I. Main material properties used in the simulation

Vco = 0.566 V
6
Material window buffer absorber FF = 60.8
ZnO CdS CZTS 4  4.478 %
Energy band gap Eg (eV) 3.30 2.40 1.50
Thickness (nm) 200 60 2000 2
Donor concentration ND (cm-3) 1x1018 1x1018 0
Acceptor concentration NA 0 0 4x1016 0
(cm-3) 0,0 0,2 0,4 0,6

Electron Mobility MUN 100 100 100 Vco (V)

(Cm2/Vs)
Hole Mobility MUP (Cm2/Vs) 25 25 25
Relative permittivity EPS 9 10 10 Figure 3. Simulated J-V characteristics of the traditional Mo/CZTS/CdS/ZnO
Effective density of states Nc 2.2x1019 2.2x1018 2.2x1018 solar cell
 length is a constraint. In thin film solar cells particularly, a
1,1 large part of the entire device’s thickness is occupied by the
Jcc
1,0
Vco absorber layer which plays the role of the active layer and in
0,9 FF
Effi
which the majority of photons having an energy greater than
0,8
or equal to the band gap energy of the material is absorbed. It
Normalized value

0,7

0,6 is therefore important and essential to control the absorber

0,5 layer thickness for achieving the consistent output
0,4 performance of CZTS solar cells. To reach this goal, the
0,3
dependency of CZTS cell performance on the absorber layer
0,2
thickness is studied by varying the thickness from 800 to 4000
0,1

0,0
nm. Fig.5 shows the effect of CZTS thin film thickness on the
5,0 5,1 5,2 5,3 5,4 5,5 5,6 5,7 5,8 5,9 6,0
cell performance parameters such as Jsc, Voc, FF and
Electron work function of back metal contact (eV)
efficiency.
Figure 4. Simulated normalized output parameters of the CZTS solar cell as a Remarkable improvements were observed in both
function of back metal work function values of Jsc and Voc and hence in the efficiency when the
TABLE 1 Output characteristics of various reported CZTS solar cell and their thickness increases to higher values. Jsc is improved for CZTS
comparison with our simulation thickness up to 2250 nm. From Fig.5, we can observe
individual effect of thickness on the cell performance where
Cell structure Jsc Voc FF η Ref Fabricati
(mA/Cm (V) (%) (%) on the Jsc shows substantial rise compared to Voc which is
2
) technique mostly expected. When the absorber layer is thicker, more
Mo/CZTS/CdS/Z 13 0.54 59.8 4.1 [18] Co- photons with longer wavelength are absorbed and thus more
nO 1 evaporatio
electron-hole pairs are generated. Hence more free electrons
n
Mo/CZTS/CdS/Z 13 0.56 60.8 4.48 This modeled will be collected at the front and this will increase the Jsc and
nO 6 wor therefore the efficiency. Due to the increase of Jsc, the
k efficiency curve increases following the same evolution as the
BM/CZTS/CdS/Z 19.99 0.95 0.82 15.6 This optimized Jsc curve and saturates above 2250 nm. However, in order to
nO 6 1 8 wor
k
get a higher efficiency as well as cost effectiveness, an
Single junction 29 1.23 90 32.2 [19] Schockley optimum thickness is desirable for having a tradeoff between
CZTS solar cell -Queisser cell efficiency and cost for mass production. Our simulation
(ideal) limit indicates that a CZTS thickness up to 2250 nm is sufficient for
BM= back metal with work function of 5.6 eV photons absorption of the AM1.5G radiation due to the high
absorption coefficient and direct band gap of CZTS material.
One can see in Fig.4 that the efficiency increases
considerably as the back metal work function increases from
5.0 to 5.6 eV which is due to the increase in Jsc, Voc and FF
values. As the back metal work function increases, the 1,1
Jcc
schottky barrier created at the back increases and prevents 1,0
Vco
0,9 FF
strong recombination of charge carriers. This effect can 0,8
EFFI

improve the value of Voc and therefore the efficiency. No

Normalized value

0,7

change in values of Jsc, Voc and FF is observed for back 0,6

metal work function value between 5.6 and 6 eV and this 0,5

0,4
cause the efficiency to remain constant in this range. We can 0,3

thus say that a back metal work function of 5.6 eV is the 0,2

optimum value for achieving CZTS solar cell with good 0,1

0,0
performance as shown our simulation results. 1000 1500 2000 2500 3000 3500 4000

Thickness of the CZTS thin film (nm)

C. Effect of the CZTS absorber layer’s thickness Figure 5.Calculated normalized output parameters of the CZTS solar cell as a
function of the thickness
Thickness is an important parameter for semiconductor
devices especially in thin film where the carrier diffusion
D.  Effect of the acceptor concentration in the CZTS
absorber
Acceptor defects, characterized with their
concentration NA can be considered as another parameter
which limits the solar cell performance. We have varied in our
simulation NA values in the range 1013-1018 cm-3 in order to
evaluate its effect on solar cell performance. Fig.6 shows the
simulation results of the effect of NA on output parameters Jsc,
Voc, FF and efficiency. In practice, the value of NA is one of
the important parameters for p-n junction devices and the
simulated results must be more relevant with a limited
quantity of acceptor concentration for efficient CZTS solar
cell. As carrier concentration increases, the semiconductor
material becomes degenerate and this can be detrimental for
the output performance as shown on Fig.6. In Fig.6, we can
see that the Jsc decreases with the carrier concentration while
the Voc increases with it.
The variations in the values of Jsc and Voc caused by
NA can be explained by the following simple PN junction
model [20].
(4)
This phenomenon reduces the possibility of free
electrons to be collected at the front after been generated by
incident light and this can result to lower the Jsc. A higher
content of NA in the material will also cause a drop in the
value of FF since long wavelength photons are absorbed
deeply in the material and the electrons created there are more
dependent on diffusion effect to be collected effectively.
E. Optimized CZTS thin film solar cell
J-V characteristics of optimized CZTS solar cell are
shown in Fig.7. A better performance showed an efficiency of
15.68% with Jsc=19.99 mA/Cm², Voc=0.95V and FF=82.1%.
A thickness of 2250 nm, acceptor concentration of 4x1015cm-3
and back metal work function value of 5.6 eV were found to
be optimum parameters values for best output performance of
the CZTS thin film solar cell. The quantum efficiency plot
shows that there is an increase in photons absorption in the
visible range for the optimized structure compared to the
typical CZTS-based solar cell structure.
20
18

16
2
14 Jsc = 19.994 mA/Cm
Voc = 0.956 V
Jcc (mA/cm )
2

12

10 FF = 0.821
ln 1 8
 15.680%
(5)
6

2
These equations show that the saturation current I0 will 0
0,0 0,2 0,4 0,6 0,8 1,0
decrease if NA increases and Voc will therefore increase if I0 Vco (V)
decreases. This is probably due to the recombination process
which is enhanced by increasing NA.
Typical structure
0,9 Optimized structure
1,1
Quantum efficiency (%)

1,0
0,8
0,9 Jsc
0,8 Voc
FF 0,7
Normalized value

0,7 EFFI
0,6 0,6

0,5

0,4 0,5

0,3

0,2 0,4

0,1
0,4 0,5 0,6 0,7 0,8 0,9
0,0
wavelength (m)
1E14 1E15 1E16 1E17 1E18

-3
Acceptor concentration (Cm ) Figure 7. J-V characteristics and quantum efficiency of optimized CZTS solar
cell
Figure 6.Calculated normalized output parameters of the CZTS solar cell as a
function of acceptor concentration
 IV. CONCLUSION
In this study, baseline CZTS/CdS/ZnO solar cell has
been numerically evaluated for its conversion efficiency.
Platinum has been used in a first time as back metal contact
due to its high work function value and also due to its good
ohmic contact to the CZTS absorber. The use of this metal
was just to evaluate the maximum theoretical efficiency
attainable with such a device. This solar cell showed a
conversion efficiency of 19.046% with Jsc=21.687 mA/cm²,
Voc= 1.023V and FF=85.9%. In a second time typical
molybdenum was used as back metal contact and an efficiency
of 4.47% was obtained from numerical simulations close to
the first reported experimental efficiency of 4.1%. The back
metal work function, the acceptor concentration and the
absorber thickness were adjusted and optimized in order to
improve the output performance of this solar cell. The
optimum values of these parameters were 5.6 eV for back
metal work function, 4x1015cm-3 for the acceptor
concentration and 2250 nm for the CZTS thickness. The
optimized device showed a conversion efficiency of 15.68%.
ACKNOWLEDGMENTS
Joël Hervé Nkuissi Tchognia is grateful to the ICTP (The
Abdus Salam International Centre for Theoretical Physics)
and ANSOLE (African Network for Solar Energy) for
financial support within the framework of the Intra-African
Exchange (INEX) program. The authors acknowledge Dr.
Stephen Fonash’s group at the Pennsylvania state university
for providing AMPS-1D software.
REFERENCES
[1] D. Mitzi, O. Gunawan, T. Todorov, K. Wang and S. Guha “The path
towards a high-performance solution-processed kesterite solar cell”, Sol.
Energ. Mater. Sol. Cells Vol.95, pp. 1421-1436, June 2011.
[2] I. Repins, N. Vora, C. Beall, S.H .Wei, Y. Yan, M. Romero, G. Teeter, H.
Du, B. To, M. Young and R. Noufi, “Kesterites and chalcopyrites: a
comparison of close cousins” NREL/CP-5200-51286, pp. 1-12, 2011.
[3] M.A. Green, K. Emery, Y. Hishikawa and W. Warta, “Solar cell
efficiency tables (version 37)” Prog. Photovolt. Res. Appl. Vol.19, pp. 84-
92, 2011.
[4] K. Jimbo, R. Kimura, T. Kamimura, S.Yamada, W. S. Maw, H. Araki, K.
Oishi and H. Katagiri, “Cu2ZnSnS4-type thin film solar cells using
abundant material”, Thin Solid Films, Vol.515, pp. 5997–5999, 2007.
[5] H. Katagiri, K. Saitoh, T. Washio, H. Shinohara, T. Kurumadani and S.
Miyajima, “Development of thin film solar cell based on Cu2ZnSnS4 thin
films” Solar Energy Materials and Solar Cells, Vol.65, pp. 141-148,
2001.
[6] S. Chen, J. H. Yang, X. G. Gong, A. Walsh and S. H. Wei, “Intrinsic
point defects and complexes in the quaternary kesterite semiconductor
Cu2ZnSnS4” Phys. Rev. B, Vol.81, pp. 245204, 2010.
[7] W. Wang, M. T. Winkler, O. Gunawan, T. Gokmen, T. K. Todorov, Y.
Zhu and D. B. Mitzi, “Device characteristics of CZTSSe thin-film solar
cells with 12.6% efficiency”, Adv. Energy Mater. Vol.4 p. 1301465,
2014.
[8] I. D. Olekseyuk, I. V. Dudchak, and L. V. Piskach, “Phase equilibria in
the Cu2S-ZnS-SnS2 system,” Journal of Alloys and Compounds, Vol.368,
PP.135-143, 2004.
[9] M. Patel and A. Ray, “Enhancement of output performance of Cu2ZnSnS4
thin film solar cells-A numerical simulation approach and comparison to
experiments”, Physica B, Vol.407, pp. 4391–4397, 2012.
[10] H. Zhu, A. K. Kalkan, J. Y. Hou and S. J. Fonash, “Applications of
AMPS-1D for solar cell simulation”, Aip Conference Proceedings, Vol.
462, pp. 309-314, 1999.
[11] J. Arch, J. Hou, W. Howland, P. McElheny, A. Moquin, M. Rogosky, F.
Rubinelli, T. Tran, H. Zhu and S. J. Fonash, “A manual for AMPS-1D
BETA version 1.00”,1997.
[12] S. J. Fonash, Solar Cell Device Physics 2nd edition, Elsevier, 2010.
[13] M. Burgelman, J. Verschraegen, S. Degrave and P. Nollet “Modeling
Thin-film Pv Devices”, Prog.PV: Res. Appl. Vol.11, pp. 1-11, 2003.
[14] H. Katagiri, K. Jimbo, S. Yamada, T. Kamimura, W. S. Maw, T. Fukano,
T. Ito and T. Motohiro, “H2S Concentration Dependence of Properties of
Cu2ZnSnS4 Thin Film Prepared under Nonvacuum Condition” Appl.
Phys. Express, Vol.1, pp. 41201, 2008.
[15] S. Ahmed, K.B. Reuter, O. Gunawan, L. Guo, L.T. Romankiw and H.
Deligianni, “ A High Efficiency Electrodeposited Cu2ZnSnS4”, Adv.
Energy Mater. Vol.2, pp. 253-259, February, 2012.
[16] K. Wang, O. Gunawan, T. Todorov, B. Shin, S. J. Chey, N. A. Bojarczuk,
D. B. Mitzi and S. Guha, “Thermally evaporated Cu2ZnSnS4 solar cells”
Appl. Phys. Lett. Vol.97, pp. 143508, 2010.
[17] T. K. Todorov, K. B. Reuter and D. B. Mitzi, “High-efficiency solar cell
with Earth-abundant liquid-processed absorber”, Adv. Mater. Vol.22, pp.
E156-E159, 2010.
[18] B. A. Schubert, B. Marsen, S. Cinque, T. Unold, R. Klenk, S. Schorr and
W. H. Schock, “Cu2ZnSnS4 thin film solar cells by fast co-evaporation”,
Prog. Photovolt. Res. Appl. Vol.19, pp. 93-96, 2011.
[19] Q. Guo, H.W. Hillhouse and R. Agrawal, “Synthesis of Cu2ZnSnS4
nanocrystal ink and its use for solar cells” J. Am. Chem. Soc. Vol.131,
pp.11672-11673, August, 2009.
[20] W. Zhao, W. Zhou and X. Miao, “Numerical simulation of CZTS thin
film solar cell”, IEEE.Nano/Micro Engineered and Molecular Systems
(NEMS).Japan, Vol.7, PP. 502-505, March 2012.