Optica Applicata, Vol. XLI, No.

2, 2011

The role of antireflective coatings in silicon solar cells

– the influence on their electrical parameters
BARBARA SWATOWSKA1*, TOMASZ STAPINSKI1, KAZIMIERZ DRABCZYK2, PIOTR PANEK2

1
AGH – University of Science and Technology, Department of Electronics,
al. Mickiewicza 30, 30-059 Kraków, Poland

2
Polish Academy of Sciences, Institute of Metallurgy and Materials Science,
ul. Reymonta 25, 30-059 Kraków, Poland

*
Corresponding author: swatow@agh.edu.pl

In this work, the authors compared the properties of multicrystalline silicon solar cells which
depended on the kind of following antireflective layers: a-Si:C:H, a-Si:N:H and TiOx.
Current–voltage characteristics for multicrystalline silicon solar cells were measured by the use
of a computer controlled global spectrum sun simulator under an AM 1.5. The measurements
of I–V characteristics allow the determination of basic electrical parameters and efficiency using
the double exponential relationship from a two-diode solar cells model. Two key parameters:
refractive index and thickness of the film affect the final features of the antireflective coating.
Optimisation of these parameters and afterwards the experimental verification lead to
the minimalisation of the reflection coefficient that decides about the quality of the antireflective
layer. A high quality reflective layer can improve the efficiency of the solar cell even by 30%.

Keywords: antireflective coatings (ARC), solar cells, I–V characteristics.

1. Introduction
The basic factor that affects the efficiency of a solar cell is the reflection of light
from its front surface. The reflection coefficient can be reduced by covering the top
of the solar cell surface by antireflective coatings (ARC) – Fig. 1. The surface of
the silicon substrate can be covered with single or double antireflective coatings.
Various techniques can be used to deposit antireflective coatings: the chemical
vapour deposition (CVD), spray, spin-on or screen printing. The spin-on is the simplest
technique which is very efficient and does not need an expensive equipment [1, 2],
however this method can be used only on smooth, polished surfaces. For textured
surfaces, the thickness of the ARCs deposited by the spin-on method becomes
uneven. Another method – plasma enhanced chemical vapour deposition (PECVD) is
488 B. SWATOWSKA et al.

hν
Front electrode Front electrode
ARC
mc-Si n-type

mc-Si p-type

Back electrode

Fig. 1. Scheme of photovoltaic solar cell on the base of multicrystalline silicon (mc-Si) with antireflective
coating.

more expensive, but it enables the deposition of layers with very good parameters
(reflection coefficient, chemical contents) – very uniform and of controlled thickness.
The following materials are used as reflective coatings: TiO2 (refractive index n = 2.3),
Si3N4 (n = 1.9), Al2O3 (n = 1.8–1.9), SiO2 (n = 1.4–1.5), Ta2O5 (n = 2.1–2.3) [3].
Recently the amorphous layers a-SiNx:H [4] and a-Si:C:H [5] have been applied as
ARCs. In solar cell production process during diffusion, on the surface of silicon
substrate, phosphorous silicate glass (PSG) is formed which can also operate as a low
quality antireflective layer [6] which does not need any additional processes and costs.
For this reason it is removed just after the diffusion process with the use of H2F2 and
then the above mentioned antireflective layers are coated on the top surface of solar
cells.
The proper choice of thickness of the ARC is important for the metallisation
process of electrical contacts. Excessively thick ARC makes the process of pad
metallisation more difficult, and additionally decreases optical parameters and
efficiency of the solar cell.
The authors obtained a-Si:C:H and a-Si:N:H films by PECVD and applied them
as ARC in multicrystalline silicon solar cells. Films of TiOx are deposited by the use
of the spray technique.

2. Experiment
Amorphous a-Si:C:H and a-Si:N:H thin films were deposited by PECVD in an ultra
high vacuum reactor at 13.56 MHz in gaseous mixtures SiH4 + CH4 and SiH4 + NH3
with various CH4 and NH3 content in plasma [7]. In PECVD method, the probability
of dissociation of nitrogen is very low and the silane ammonia mixture should be
used to produce a-Si:N:H thin films. Other technological parameters: RF power of
5 W, substrate temperature of 180 °C and total pressure of 80 Pa were kept constant.
The TiOx films were deposited by the spray technique at 280 °C with tetraethyl-
orthotitanat (C2H5O)4Ti using purified air as a carrier gas [8]. Polished crystalline
(Cz-Si) and multicrystalline (mc-Si) silicon wafers were used as substrates. All these
films were deposited before contacts were screen-printed. Such procedure is important
The role of antireflective coatings in silicon solar cells... 489

for good soldering of the solar cells and good connecting them in a module. TiOx films
produced at the temperature below 300 °C are primarily amorphous like a-Si:C:H and
a-Si:N:H, which makes the metallisation of the front contacts easier.
Structural properties were detected by means of Fourier transform infrared
spectroscopy (FTIR – Biorad FTS-60V) and film morphology using scanning electron
microscope (SEM – NOVA NANO SEM 200, FEI Company). Optical measurements
of films deposited on silicon substrates were done by the use of Perkin–Elmer
Lambda 19 double beam spectrophotometer. The method of characterisation and
details of the experiment were thoroughly described in [5]. Current–voltage photo-
characteristics of solar cells with and without antireflective coatings were measured
using a computer controlled global spectrum sun simulator (I–V Curve Tracer For
Solar Cells Qualification, v 4.1.1) [9].

3. Results and discussion

SEM method makes possible a quantitative analysis of elemental composition of
a-Si:N:H and a-Si:C:H films. The obtained results confirm that the increase in
ammonia content in SiH4 + NH3 and methane content in SiH4 + CH4 causes
the increase in nitrogen wN and carbon wC content in a-Si:N:H and a-Si:C:H films
(wN = N/(Si + N); wC = C/(Si + C)), respectively. These two elements of the layers –
nitrogen and carbon – had the crucial influence on their optical properties: refractive
index, energy gap and effective reflectivity coefficient [7].
The presence of hydrogen in the films used for photovoltaic application is very
important for passivation of defects in multicrystalline silicon. These defects play
the role of recombination centers in the bulk and at the surfaces of mc-Si. FTIR
measurements, which were carried out in the spectral range from 400 to 4000 cm–1 in
an absorption mode, confirmed the presence of numerous diverse hydrogen bondings
both in a-Si:N:H as well as a-Si:C:H layers [7]. Exemplary bondings are: SiHn, NHn,
CHn, Si–CHn, H–SiNn, etc.
The measurement of optical transmission and reflection of films enabled us to
calculate a refractive index n, optical absorption α and Tauc optical energy gap Eg
using relations given in [10, 11]. The refractive index of a-Si:N:H apparently decreases
with an increase in nitrogen content in films, whereas the energy gap increases with
an increase in wN. For a-Si:C:H films we observed a similar dependence: n decreases
with an increase in carbon content, Eg increases with an increase in wC.
From the position of the reflectivity minimum λmin, the film thickness d can be
evaluated as [12]:
λmin
d = ---------------- (1)
4n
where n is the real part of light refractive index in ARC (k ≈ 0) and:

n2 = n0 n 2 (2)
490 B. SWATOWSKA et al.

where: n0 – refractive index of air which is equal to 1; n2 – refractive index of base

material; d – thickness of antireflective coating; λ – wavelength from the area of
maximum photosensitivity (400–1100 nm).
The refractive index for silicon for the wavelength λ = 600 nm is equal to
n2 = 3.6. It is worth to remind that the effective reflectivity coefficient Reff for
the monocrystalline silicon is in the range of 30–35%, while for the multicrystalline
silicon – in the range of 25–30%. For the wavelength equal to 600 nm and the refractive
index of silicon n2 = 3.6, the optimal thickness of the ARC is about 80 nm. As can be
seen from the Eq. (2), the optimal refractive index n for the ARC layer deposited on
a silicon solar cell should be about 1.9. The antireflective layer in dependence to its
chemical contents enables also the passivation of defects in multicrystalline silicon.
Multicrystalline silicon is a highly defected material with a large number of dangling
bonds at the boundaries of grains. Some technological steps, such as cutting of silicon
substrates and phosphorous diffusion, are the source of further defects. The presence
of hydrogen in ARC layers is in this case useful.
The total reflectivity in the wavelength range extending from 400 nm to 1100 nm
of mono- and multicrystalline silicon substrates with and without ARC films was
measured by the Perkin–Elmer spectrophotometer equipped with the integrating
sphere. From the spectral dependence of the reflectivity R(λ), the effective reflectivity
coefficient Reff was calculated [13].
TiOx layers are characterized by excellent optical properties as ARCs for silicon
solar cells due to the refractive index value of about 2.38 at the wavelength of 600 nm
and the extinction coefficient remaining near zero at the wavelength of 400 nm [14].
Main optical parameters of chosen antireflective coatings, obtained by described
methods, are collected in Tab. 1.

T a b l e 1. Main parameters of ARCs: d – film thickness, n – refractive index, Eg – energy gap,

Reff – effective reflectivity coefficient.

Parameters of ARC – best results

Kind of ARC d [nm] n Eg [eV] Reff [%]
a-Si:C:H 75 1.92 2.6 6.5
a-Si:N:H 80 1.97 2.7 8.3
TiOx 35 2.28 3.1 7.5

Solar cells parameters were calculated by fitting of a two diode model to the exper-
imental data [15]. Examined solar cells were made on multicrystalline silicon wafers
of 300 μm thickness, 1 Ωcm resistivity, p-type (boron doped). The measurements of
I–V characteristics allow us to determine the basic parameters like: ISC – short circuit
current, VOC – open circuit voltage, FF – fill factor and η – efficiency.
Each of tested layers has the beneficial influence on electric parameters of solar
cells on the base of the multicrystalline silicon – Tab. 2. In every case, an improvement
The role of antireflective coatings in silicon solar cells... 491

T a b l e 2. Main parameters of mc-Si solar cells of area 100 cm2: ISC – short circuit current, VOC – open
circuit voltage, FF – fill factor, η – efficiency; for different kinds of ARC (I–V measurements in standard
test conditions – STC).

Parameters of solar cells

Kind of solar cells ISC [A] VOC [mV] FF [%] η [%]
mc-Si solar cell without ARC 2.43 566.7 71.1 9.84
mc-Si solar cell with a-Si:C:H ARC 3.12 578.3 70.8 12.22
mc-Si solar cell with a-Si:N:H ARC 3.29 612.4 73.1 14.25
mc-Si solar cell with TiO2 ARC 3.24 593.4 72.7 14.00

in the efficiency of solar cells was observed. With the decrease in Reff , the increase in
the efficiency η and short circuit current were observed.
As can be seen in Tab. 2, most favourably a-Si:N:H layer influences the parameters
of solar cells, but very comparable results are being achieved through using a titanium
oxide layer. Probably, it is a result of additional influence of many hydrogen bondings,
present in the structure of a-Si:N:H antireflective coatings [13].
The output parameters of multicrystalline silicon solar cells with and without
antireflective coatings were simulated using a computer program PC1D and the results
compared with the measurement [7]. The obtained results are in good agreement with
the measured output parameters of analysed solar cells.

4. Conclusions
In common opinion it is not possible to produce efficient solar cells without any
antireflective coatings. The profitable influence of ARC is particularly exhibited by
the values of the short circuit current and the efficiency of cells. Moreover, the ARCs
have good protective properties of solar cells surface, before assembling them into
the panel.
PECVD is the most effective method of ARC preparation due to good reproducibility
of material properties. Additionally, this method guarantees the hydrogen presence
suitable for passivation of silicon defects, especially in multicrystalline substrates.
The numerical prediction of I–V characteristics of solar cells was confirmed by
the results of measurements.

Acknowledgements – The research was supported by EEA Financial Mechanism and the Norwegian
Financial Mechanism Grant PL0081.

References
[1] VAN OVERSTRAETEN R.J., MERTENS R.P., Heavy doping effects in silicon, Solid-State Electronics 30(11),
1987, pp. 1077–1087.
[2] FRISSON L., CHEEK G., MERTENS R.P., VAN OVERSTRAETEN R.J., Proceedings of 5th European
Photovoltaics, Adam Hilger, Bristol, 1986.
492 B. SWATOWSKA et al.

[3] GREEN M.A., Solar Cells. Operating Principles, Technology and System Applications,
The University of New South Wales, Kensington, 1986.
[4] LIPINSKI M., ZIEBA P., JONAS S., KLUSKA S., SOKOŁOWSKI M., CZTERNASTEK H., Optimisation of SiNx:H
layer for multicrystalline silicon solar cells, Opto-Electronics Review 12(1), 2004, pp. 41–44.
[5] SWATOWSKA B., CZTERNASTEK H., LIPINSKI M., STAPINSKI T., ZAKRZEWSKA K., Antireflective coatings
of a-Si:C:H on silicon, Proceedings of XXVIII International Conference of IMAPS Poland Chapter,
Wroclaw, September 26–29, 2004, pp. 385–388.
[6] CHAKRAVARTY B.C., VINOD P.N., SINGH S.N., CHAKRABORTY B.R., Design and simulation of
antireflection coating for application to silicon solar cells, Solar Energy Materials and Solar
Cells 73(1), 2002, pp. 59–66.
[7] STAPINSKI T., SWATOWSKA B., a-Si:C:H and a-Si:N:H thin films obtained by PECVD for applications
in silicon solar cells, Journal of Electronic Materials 37(6), 2008, pp. 905–911.
[8] PANEK P., LIPINSKI M., BELTOWSKA-LEHMAN E., DRABCZYK K., CIACH R., Industrial technology of
multicrystalline silicon solar cells, Opto-Electronics Review 11(4), 2003, pp. 269–275.
[9] GRANEK F., ZDANOWICZ T., Advanced system for calibration and characterization of solar cells,
Opto-Electronics Review 12(1), 2004, pp. 57–67.
[10] SWANEPOEL R., Determination of the thickness and optical constants of amorphous silicon, Journal
of Physics E: Scientific Instruments 16(12), 1983, pp. 1214–1222.
[11] TAUC J., Amorphous and Liquid Semiconductors, Plenum Press, New York, 1974.
[12] FOWLES G.R., Indroduction to Modern Optics, Rinehart and Winston, New York, 1968.
[13] SWATOWSKA B., STAPINSKI T., Amorphous hydrogenated silicon-nitride films for applications in solar
cells, Vacuum 82(10), 2008, pp. 942–946.
[14] RICHARDS B.S., COTTER J.E., HONSBERG C.B., WENHAM S.R., Novel uses of TiO2 in crystalline silicon
solar cells, Photovoltaic Specialists Conference, 2000. Conference Record of the Twenty-Eighth
IEEE, September 15–22, 2000, pp. 375–378.
[15] APPELBAUM J., CHAIT A., THOMPSON D., Parameter estimation and screening of solar cells, Progress
in Photovoltaics: Research and Applications 1(2), 1993, pp. 93–106.

Received September 25, 2010