Rev. Téc. Ing. Univ. Zulia. Vol. 44, Nº 2, May-August, 2021, 127-136
Rev. Téc. Ing. Univ. Zulia. Vol. 44, No. 2, May-August, 2021.
Design, Manufacture and Characterization of Solar Cells p-
CdTe / n-CdS with Thin Films
P. Grima-Gallardo
1,2,3*
, L. Ruiz
1,2
, B. Conquet
1,2
, E. Sánchez
1,2
, M.
Contreras
1,2
, L. Nieves
1,2,3
, A. Velásquez-Velasquez
1,
2
y W. González
3
1
National Center for Optical Technologies (CNTO). 5101 Mérida. Venezuela.
2
Astronomy Research Center (CIDA). 5101 Mérida. Venezuela.
3
Center for Semiconductor Studies (CES). Science Faculty. Physics Department. University of
The Andes. Mérida. Venezuela.
* Corresponding author: peg1952@gmail.com
https://doi.org/10.22209/rt.v44n2a07
Received: 12 June 2020 | Accepted: 30 March 2021 | Available: 30 April 2021
Abstract
Venezuela, due to its geographical location, richness in sources and natural resources, has a valuable
potential for the use of renewable energies, especially solar, a large part of its territory is characterized by an average
insolation above 3 kWh/m
2
, considered suitable for photovoltaic generation. CdTe is a well-known II-VI
semiconductor, which has established itself as a leader in the photovoltaic industry, has optimal properties for solar
cells. This research focused on design, manufacture and characterization of p-CdTe / n-CdS solar cells. The design
was carried out looking for the simplest and lowest cost model. The manufacturing was carried out in an Ortus-700
equipment, using the thermal evaporation technique. For the characterization of the cells, the techniques of X-ray
diffraction, optical absorption and current-voltage characteristic analysis were applied. Evidence of a typical
preferential orientation of a cubic structure of Zinc Blenda, direct energy gap values between 2.15 to 2.85 eV and
similar performance among prepared solar cells. The manufacture and characterization of nine p-CdTe / n-CdS cells
was achieved through the formation of a p-n junction; with efficiencies of ~ 3%.
Keywords: Efficiency, optical absorption, p-CdTe/n-CdS solar cells, thin films, X-ray diffraction.
Diseño, Fabricación y Caracterización de Celdas Solares p-
CdTe/n-CdS con Películas Delgadas
Resumen
Venezuela, debido a su ubicación geográfica, riqueza en fuentes y recursos naturales, tiene un valioso
potencial para el uso de energías renovables, especialmente la solar, gran parte de su territorio, se caracteriza por una
insolación media por encima de 3 kWh/m
2
, considerada apta para la generación fotovoltaica. El CdTe es un conocido
semiconductor II-VI, que se ha consolidado como líder en la industria fotovoltaica, posee propiedades óptimas para
las celdas solares. Esta investigación se enfocó en diseñar, fabricar y caracterizar celdas solares p-CdTe/n-CdS. El
diseño se realizó buscando el modelo más simple y de menor costo. La fabricación se llevó a cabo en un equipo
Ortus-700, usando la técnica de evaporación térmica. Para la caracterización de las celdas, se aplicaron las técnicas
de difracción de rayos-X, absorción óptica y análisis de característica corriente-voltaje. Evidenciándose una
orientación preferencial típica de una estructura cúbica de la Zinc Blenda, valores de brecha de energía directa entre
2,15 a 2,85 eV y un rendimiento similar entre las celdas solares preparadas. Se logró la fabricación y caracterización
de nueve celdas p-CdTe/n-CdS, mediante la formación de una unión p-n; con eficiencias de ~ 3 %.
Palabras clave: absorción óptica; celdas solares tipo p-CdTe/n-CdS; difracción de rayos X; eficiencia; películas
delgadas.
Solar cells p-CdTe / n-CdS with thin films 128
Rev. Téc. Ing. Univ. Zulia. Vol. 44, No. 2, May-August, 2021.
Introduction
Venezuela, due to its geographical location and richness in sources and natural resources, has valuable
potential for the use of renewable energy, especially solar. As can be seen in Figure 1, a large percentage of the
national territory is characterized by an average insolation of the order of 4-5 kWh / m
2
per day (Solargis, 2017).
Taking into account that values above 3 kWh / m
2
(Posso et al; 2013) are considered suitable for photovoltaic
generation, it can be concluded that Venezuela is a country where the exploitation of solar energy is very convenient.
In addition, it has been estimated that daily insolation in the territory is equivalent to 4.56 million barrels of oil, an
amount that exceeds the daily oil production of Venezuela at any time in its history (Posso et al; 2013).
Figure 1. Map of solar resources of Venezuela (Solargis, 2017).
There are many types of solar cells (Bagher et al; 2019), however, from the point of view of their industrial
production (Figure 2), they can be classified into three large groups: 1) polycrystalline silicon, 2) mono-crystalline
silicon, and 3) thin films ( Phillips and Warmuth, 2016). As can be seen in Figure 2, thin film technology has been
growing in recent years, particularly CdTe-based cells (Figure 3).
Figure 2. World photovoltaic production by technology (in GWp) (Phillips and Warmuth, 2016).
Grima-Gallardo et al . 129
Rev. Téc. Ing. Univ. Zulia. Vol. 44, No. 2, May-August, 2021.
Figure 3. Thin film technologies: global annual production of photovoltaic modules (Phillips and Warmuth, 2016).
CdTe (cadmium tellurium) is a well-known II-VI semiconductor (Strauss, 1977), which has optimal
properties for solar cells. CdTe thin films have a high optical absorption coefficient that generally exceeds 104 cm
–1
,
absorbing approximately 92% of visible light in a thickness of only 1 µm (Strauss, 1977; Rangel and Sobral, 2017);
this efficiency is well above crystalline silicon, which needs around 200 µm to reach the same absorption value
(Burgelman, 2006). In addition, it has a direct optical energy gap of 1.606 eV at liquid helium temperature and 1.529
eV at room temperature, optimal for coupling with the solar radiation spectrum (Strauss, 1977; Rangel and Sobral,
2017). Also, its constituent elements, Cd and Te, are still relatively abundant in the earth's crust; this means that the
commercial cost of CdTe powder in the international market is of the order of $ 245 / kg (4 N purity) and $ 280 / kg
(5 N) (Alibaba, (2017). In fact, CdTe has been consolidated as a leader in the thin film photovoltaic industry,
achieving efficiencies of up to 22% thanks to the optimization of the short-circuit current density (Jsc), which rose
from 26.1 to 31.69 mA / cm
2
(Green, 2018).
CdS is also a well-known semiconductor of the II-VI family (Bube, 2001), naturally n-type, with an optical
absorption coefficient, α of ~ 1.1x105 cm
-1
, transparent, with a gap of direct optical energy of 2.41 eV (Oliva et al.,
2001; Das and Pandey, 2011). Due to its optical characteristics and low cost ($ 50-100 / kg) (Alibaba, 2016) it is the
most used window material for solar cells with CdTe.
The objective of this research was to design, manufacture and characterize p-CdTe / n-CdS solar cells; using
a design, versatile and simple, so that it can be easily replicated, and at the lowest possible cost, thus bridging the gap
of the impossibility of easy access to most of the materials required for the manufacture of these photovoltaic
devices.
Experimental
Solar Cell Design
The design of the experimental cell used for the manufacture of the solar cell, in the present work, is shown
in Figure 4.
Figure 4. Details of the design of the p-CdTe / n-CdS cell. At the bottom, the experimental
arrangement used in this work. In the upper part, photograph of the cell already made.
Solar cells p-CdTe / n-CdS with thin films 130
Rev. Téc. Ing. Univ. Zulia. Vol. 44, No. 2, May-August, 2021.
Solar Cell Manufacturing
The equipment used for the manufacture of the nine p-CdTe / n-CdS cells was an Ortus-700 (from the
National Center for Optical Technologies adjunct to the Astronomy Research Center Mérida, Venezuela), which
allows obtaining thin films using four different techniques: thermal evaporation, electron gun, ion gun and radio
frequency. The substrate consisted of glass microscope slides, with dimensions of 25.4 x 76.2 x 1 mm. Before
deposition, the chamber was subjected to vacuum levels on the order of ~ 10-4 Torr; once achieved, the substrates
were subjected to an ABS (Assisting Beam Source) ion beam cleaning for 2 min, to remove impurities and achieve a
rough nanometric surface that facilitates the adhesion of materials.
For the CdTe deposition process, polycrystalline powder with a nominal purity of 5 N (99.999%) was used,
which was arranged in the form of granules with an approximate average value of 3 to 5 mm in diameter each piece,
within a container made of molybdenum (Mo), which in turn was connected to the pair of electrodes of the thermal
evaporation device inside the Ortus-700.
Films were made by gradually heating the CdTe inside the crucible until it evaporated. The vapor
molecules, leaving the source, were deposited on the surface of the substrate. The equipment has a support for the
substrates in the form of a spherical cap, which makes the distance between the heating source and the substrates
vary between 45 and 53 cm, depending on their location on said support (Figure 5). To obtain homogeneous films,
the mechanical rotation of said support was used, which was configured to rotate at 20 rpm, and the heating power
was adjusted until reaching deposition rates of 2-6 Å / s.
The nomenclature l, m and c in Figure 5 is arbitrary, allowing the subsequent evaluation of the difference in
thickness of the films as a function of position, since the concavity of the support makes the samples located in the
most eccentric part closer to each other and to the heater fountain.
Figure 5. Side view of the support of the substrates in the Ortus-700. The nomenclature l, m and c
is arbitrary and makes it possible to subsequently evaluate the difference in thickness of the films as
a function of said position (Modified to Izovac, 2014).
For the CdS deposition process, polycrystalline powder with a nominal purity of 5 N (99.999%) was used,
which was sintered into tablets with a diameter of 25 mm and a thickness of 0.5-0.8 mm. In a similar way to the
CdTe compound, fragments of said tablets were placed inside a container of Mo, connected to the electrodes of the
thermal evaporation device within the coating system. In this case, the substrate consisted of previously obtained
CdTe thin film samples. Before deposition, the chamber was also subjected to a vacuum of the order of ~ 10-4 Torr,
cleaning the CdTe samples by ABS ion beam for 2 min.
The thin films of CdS were also obtained by means of the resistive heating technique, where the substrate
during the deposition reached temperatures similar to those of CdTe. Mechanical rotation was set at 20 rpm and the
heating power was adjusted until reaching a deposition rate of 3 Å / s and a thickness of ~ 500 nm, measured in situ
by optical transmission, which corresponds to T = 50% for incident radiation with = 550 nm. Once the CdTe and
CdS layers were arranged on the glass substrate, the silver (Ag) finger-type contacts were fixed on them, which were
obtained by the electron beam deposition technique, reaching a thickness of about 350 nm.
Grima-Gallardo et al . 131
Rev. Téc. Ing. Univ. Zulia. Vol. 44, No. 2, May-August, 2021.
Characterization of Solar Cells
The first characterization carried out was the measurement of the films by X-ray diffraction (XRD), by
means of the powder technique, using a SIEMENS Brukers-AXS D5005 equipment. The second characterization
carried out was the measurement of the optical transmission, as a function of the incident wavelength for the films
obtained, in Cary-UV-Vis-NIR-5000 equipment at normal incidence.
The third characterization that corresponded to the measurement of the efficiency of the cells, using their
characteristic current-voltage density (J-V), is presented in Figure 6. The measurement circuit used is shown on the
left side. Suitable RL values can be estimated by the following procedure. First, Voc and Jsc were measured, and the
relationship of Voc, Jsc and RL was calculated. The value of RL is close to RL at the point of maximum power.
Resistance values less than or greater than RL, by about an order of magnitude or more were chosen to score points
in the JV characteristics towards Jsc or towards Voc, respectively, where more RL values should be chosen around
the maximum power point to increase the accuracy of the fill factor (FF) measurement, defined below (Dittrich,
2018). On the right side of Figure 6, the home station that was used to manipulate the electrical contacts to the solar
cell is also shown. Two bronze arms were used to hold the wire needed for each semiconductor material. Ohmic
contacts were used for an adequate response. The CdTe contact consisted of 0.5 mm copper wire that was soldered to
the surface of the thin film with silver paint, while a 0.2 mm diameter tungsten wire was used for the CdS. The
ohmic behavior of the electrodes was verified by measuring the current density-voltage characteristic.
Figure 6. Arrangement of the electrical circuit used for the measurement of the current and voltage characteristics
(J-V), using a potentiometer and two multimeters. R
int
: internal resistance, RL: variable resistance, A: amperemeter,
V: voltmeter.
Voltage and current readings were taken using two Keithley 2400 measuring instruments. J-V curves were
obtained point-to-point under illumination. The incident light was provided by a tungsten incandescent light bulb
(luminous efficacy 12.5-17.5 lumens / Watt), which was placed at a distance of about 1m above the solar cell sample
to avoid undesirable heating of the device; luminosity was measured with a commercial PCL-VDL 16I lux meter.
The efficiency (η) of a solar cell is determined as the fraction of incident energy that is converted into
electricity, and is defined as (Checa et al, 2015):







(1)
Where P
in
is the incident power, V
oc
is the open circuit voltage; J
sc
is the short-circuit current density and FF is the
fill factor, given by:








(2)
Where V
max
and J
ax
are the coordinates of the point max.
The incident power P
in
[W/cm
2
] is equal to the luminous flux Φ
V
[lumens/cm
2
], divided by the luminous efficacy L
ef
[lumens/W]:
P
in
= Φ
V
/L
ef
(3)
In this case, the incident power measured on the solar cell P
in
= 0.1 W / cm
2
.
Solar cells p-CdTe / n-CdS with thin films 132
Rev. Téc. Ing. Univ. Zulia. Vol. 44, No. 2, May-August, 2021.
Results and Discussion
Figure 7 shows the diffraction pattern obtained from the CdTe film. The diffraction pattern shows a
preferential orientation of plane 111, typical of a cubic crystalline structure, similar to Zinc Blenda (Datta et al.
2011), crystalline structure of CdTe. The CdS diffraction pattern is shown in Figure 8. Comparison of the
experimental pattern (red) with those calculated (blue and green,) unequivocally shows that the crystal structure
corresponds to the known Wurtzite hexagonal phase of CdS. It should be noted that the patterns correspond to the
films as they evaporated, without any subsequent heat treatment, due to the good crystallinity during their growth.
Figure 7. X-ray diffraction pattern of the CdTe thin film manufactured in this study. (°): degrees.
Figure 8. CdS thin film diffraction pattern (red line). For comparison, the calculated diffraction patterns of
hexagonal CdS (blue line: ICDD 01-074-9663) and cubic CdS (green line: ICDD 04-006-3897) are also shown. (°):
degrees.
Grima-Gallardo et al . 133
Rev. Téc. Ing. Univ. Zulia. Vol. 44, No. 2, May-August, 2021.
Figure 9 shows the measurements of the nine CdTe films, each color groups three different measurements
that practically overlap. The colors black, gray and blue, represent the positions l, m and c in Figure 6 and, therefore,
three different thicknesses of the films: 355, 220 and 90 nm, for the curves of black, gray and blue, respectively.
Figure 9. Optical transmission measurements as a function of the incident wavelength for the nine CdTe thin films.
The three colors (black, gray, and blue) correspond to positions l, m, and c in Figure 5, respectively. Notice that there
are three overlapping curves for each position. The numbers on the curve show the values of the two observed
absorption bands.
It is known that the value of the energy gap (Eg) depends on the thickness of the film (Oliva et al., 2001;
Goh et al., 2010; Khatri et al., 2010; Das and Pandey, 2011). In Figure 10 the values of Eg vs obtained from Figure
9 have been plotted with the values of the thicknesses of the CdTe films. An exponential type behavior is observed y
= Ax
-b
, with A = 5.68 eV and b = 0.195, which is coincident with that observed by other authors, for example for Ge
(Goh et al., 2010), InSbBi (Khatri et al., 2010) and CdS (Oliva et al., 2001; Das and Pandey, 2011).
Figura 10. Values of Eg vs (nm) thicknesses of the CdTe films. Experimentals values, 1.000 nm thicknesses
(Phixun et al; 2009).
Solar cells p-CdTe / n-CdS with thin films 134
Rev. Téc. Ing. Univ. Zulia. Vol. 44, No. 2, May-August, 2021.
The results of the nine CdS films are shown in Figure 11. In this case, all the curves overlap, so no change
in the energy gap is observed. This result indicates that the thickness of the films is the same: 500 nm, measured in
situ at the time of deposition. The value of Eg = 2.59 eV is within the range of the values obtained by other authors,
also in thin films: 2.15-2.25 eV (Demir and Gode, 2015); 2.85 eV (Imran et al., 2018) and 2.42 eV (Oliva et al.,
2001).
Figure 11. Optical transmission measurements as a function of incident wavelength for
the nine CdS thin films. The numbers on the curves are the values of the observed
absorption bands. The value of the direct energy gap (Eg) is given.
The current-voltage (J-V) characterization for the nine cells is shown in Figure 12; the current density J [mA
/ cm
2
] and the power output Pout [mW / cm
2
] are represented as a function of the applied voltage V [mV] for a
particular cell (cell number 1). In Table 1, the characteristic parameters of the nine cells are presented.
Figure 12. Current-voltage characteristic for the p-CdTe/n-CdS solar cell identified
with number 1 in Table 1. The black points are the experimental values of J vs V.
The red points are the output power. The yellow box represents the fill factor (FF).
Grima-Gallardo et al . 135
Rev. Téc. Ing. Univ. Zulia. Vol. 44, No. 2, May-August, 2021.
Table 1. Photovoltaic behavior of the p-CdTe / n-CdS solar cells manufactured in the present study.
Solar cell
number
V
max
[mV]
J
max
[mA/cm
2
]
V
oc
[mV]
J
sc
[mA/cm
2
]
η
[%]
1
88
3,6
160
6,2
3,2
2
95
3,5
191
5,0
3,3
3
92
3,5
178
5,8
3,2
4
93
3,4
175
5,9
3,2
5
101
3,1
181
5,4
3,1
6
99
3,1
178
5,6
3,1
7
99
3,5
169
5,9
3,5
8
97
3,6
165
6,1
3,5
9
91
3,4
163
6,2
3,1
V
max
: maximum voltage; J
max
: maximum current; V
oc
: open circuit voltage; J
sc
:
closed circuit current corriente FF: fill factor; ŋ: conversion efficiency.
As can be seen in Table 1, the performance of the solar cells prepared is very similar, this is because they were all
prepared following the same methodology and under the same experimental conditions.
Conclusions
The manufacture and characterization of the nine p-CdTe / n-CdS cells, by forming a p-n junction capable
of generating a potential difference when illuminated (photovoltaic effect). For this first prototype, a simple design
was applied with the lowest possible cost; achieving efficiencies of ~ 3%. The short-circuit current was relatively
low, probably due to a high value of the shunt resistance (Rs).
Acknowledgment
We thank the Crystallography Laboratory of the Universidad de Los Andes for the X-ray diffraction
measurements; to Fonacit for the financing of the project "Measurement and optimization of the efficiency of
photovoltaic solar cells manufactured in the National Center for Optical Technologies (CNTO)" (2018), and to the
Ministry of Popular Power for Science and Technology, for the financing of the POA 2019 project “Manufacture and
assembly of a photovoltaic module”.
References
Alibaba. Nanoparticles CdS powder price. <https://www.alibaba.com/product-detail/Nanoparticles-CdS-powder-
price-CAS-1306_60521158667.html.> [Accessed April 14, 2020].
Alibaba. price cadmium telluride. <https://spanish.alibaba.com/product-detail/pure-cdte-price-4n-5n-6n-in-stock-
with-good-price-cadmium-telluride-62416992602.html.> [Accessed April 7, 2020].
Bagher, A., Abadi, M. and Mohsen, M., 2019. Types of Solar Panels (2021) | GreenMatch.
Greenmatch.co.uk.<https://www.greenmatch.co.uk/blog/2015/09/types-of-solar-panels> [Accessed Abril 4, 2020].
Brus, L., 1994. Electron-electron and electron hole interactions in small semiconductor crystallites: The size
dependence of the lowest excited electronic state. J. Chem. Phys., 80(9).
Bube, R., 2001. Encyclopedia of materials. Amsterdam: Elsevier, pp.873-879.
Burgelman, M. (2006). Cadmium Telluride Thin Film Solar Cells: Characterization, Fabrication and Modeling.
In: Thin Film Solar Cells. Fabrication, Characterization and Applications. Poortmans, J.and Arkhipov, V.1st ed.
IMEC, Leuven, Belgium. John Wiley & Sons, Ltd, 277-324. ISBN: 0-470-09126-6.
Solar cells p-CdTe / n-CdS with thin films 136
Rev. Téc. Ing. Univ. Zulia. Vol. 44, No. 2, May-August, 2021.
Checa F., E., O. &amp, E. and Rosero E, E., 2015. Study of the relationship between irradiance and power
genarated by PV panel based on different silicon technologies in the city of pasto. Power Quality Aplications IEEE
Workshop.1-6.
Das, R. and Pandey, S., 2011. Comparison of optical properties of b and nano crystalline thin films of CdS using
different precursors. International Journal of Materials Science and Applications, 1(1), 35-40.
Demir, R. and Gode F., 2015. Structural, optical and electrical properties of nanocrystalline CdS thin films grown
by chemical bath deposition method. Chalcogenide Letters, 12(2), pp.43-50.
Datta, S., Saha- Dasgupta T. and Saima, S., 2011. Relatibe Stability of Zincblende and Wurtzite Structure in
Cdx(x=S, Se, Te) Serie-A NMTO Study. arXIV 1106.
Dittrich, T., 2018. Materials Concepts for Solar Cells. WorldScientific. Helmholtz Center Berlin for Materials and
Energy, 2, 27.
Goh, E., Chen, T., Sun, C. and Liu, Y., 2010. Thickness effect on the band gap and optical properties of germanium
thin films. Journal of Applied Physics, 107(2), p.024305.
Green, M., Hishikawa, Y., Dunlop, E., Levi, D., Hohl-Ebinger, J. y Ho-Baillie, A., 2017. Solar cell efficiency tables
(version 51). Progress in Photovoltaics: Research and Applications, 26(1). 3-12.
<https://onlinelibrary.wiley.com/doi/abs/10.1002/pip.2978> [Accessed Abril 19, 2020].
Hixum Ma, Z., Man Yu, K., Walukiewicz, W., Y. Yu, P. and S Mao, S., 2009. Stain relaxation of CdTe films
growing on lattice-mismatched sustrate. Appl Phys A., 96(379-384.).
Imran, A., Jiang, J., Eric, D., Zahid, M.N., Yousaf, M., Ahmad, A., Hassan, S.A. (2018). Efficiency enhancement
through flat intermediate band in Quantum dot solar cell. Results in Physics, 10, 241-247, ISSN 2211-3797,
https://doi.org/10.1016/j.rinp.2018.05.037.
Izovac Ltd.Vacuum Tecnological Equipament., 2014. RTUS-700. Intructional Manual, IM-0000033000. 1st ed.
Khatri, R., Vyas, S., Shah, D., Patel, P., Jani, M. and Pandya, G., 2010. Study of Thickness Dependence Optical
Bandgap of InSbBi Thin Films. International Journal of Physics and Applications, 2(3), pp.95-99.
Oliva, A., Solís-Canto, O., Castro-Rodríguez, R. and Quintana, P., 2001. Formation Of The Band Gap Of Cds Thin
films growth by different techniques. Modern Physics Letters B. 15.671-674.
Phillips, S. and Warmuth, W., 2016. “ Photovoltaics report”. [ebook] Fraunhofer Institute for Solar Energy Systems,
ISE with support of PSE Projects GmbH.
<https://www.ise.fraunhofer.de/content/dam/ise/de/documents/publications/studies/Photovoltaics-Report.pdf>
[Accessed Abril 19, 2020].
Posso, F., González, J., Guerra, F., Gómez, H. (2014). Estimación del potencial de energía solar en Venezuela
utilizando sistemas de información geográfica. Revista Geográfica Venezolana, 55(1), 27-43.
Rangel-Cardenas, J. and Sobral, H., 2017. Optical absorption enhancement in CdTe thin films by microstructuration
of the Silicon Substrate.. 1st ed. Materials (Basel).607.
Solargis. 2019. Solar resource maps of Venezuela. <https://solargis.com/es/maps-and-gis-data/download/venezuela>
[Accessed Abril 10, 2020].
Strauss, A., 1977. The physical properties of cadmium telluride. 12th ed. Revue Physique Appliquée.167-184.