Plasmonics (2013) 8:14851492

DOI 10.1007/s11468-013-9562-6

Fabrication of Ag Nanoparticles Embedded in Al:

ZnO as Potential Light-Trapping Plasmonic Interface
for Thin Film Solar Cells
Hisham Nasser & Zaki M. Saleh & Engin zkol &
Mete Gnoven & Alpan Bek & Rait Turan

Received: 25 January 2013 / Accepted: 11 April 2013 / Published online: 26 April 2013
# Springer Science+Business Media New York 2013

Abstract Incident photon conversion efficiency of the absorbing materials at either side of a thin film solar module
can be enhanced by integrating a plasmonic interface. Silver
nanoparticles represent a good candidate that can be integrated to a thin film solar cell for efficient light-trapping.
The aim of this work is to fabricate plasmonically active
interface consisting of Ag nanoparticles embedded in
Al:ZnO that has the potential to be used at the front surface
and at the back reflector of a thin film solar cell to enhance
light-trapping and increase the photoconversion efficiency.
We show that Ag can readily dewet the Al:ZnO surface
when annealed at temperatures significantly lower than the
melting temperature of Ag, which is beneficial for lowering
the thermal budget and cost in solar cell fabrication. We find
that such an interface fabricated by a simple dewetting
H. Nasser : Z. M. Saleh (*) : E. zkol : M. Gnoven : A. Bek :
R. Turan
Center for Solar Energy Research and Applications (GNAM),
Middle East Technical University, 06800, Ankara, Turkey
e-mail: zsaleh@aauj.edu
A. Bek : R. Turan
Department of Physics, Middle East Technical University,
06800, Ankara, Turkey
Z. M. Saleh
On leave of absence at GNAM from the Arab American
University-Jenin, Jenin, Palestine
H. Nasser : M. Gnoven
Micro and Nanotechnology Program of Graduate School of
Natural and Applied Sciences, Middle East Technical University,
06800, Ankara, Turkey
E. zkol
Department of Chemical Engineering, Middle East Technical
University, 06800, Ankara, Turkey

technique leads to plasmonic resonance in the visible and

near infrared regions of the solar spectrum, which is important in enhancing the conversion efficiency of thin film solar
cells.
Keywords Silver nanoparticles . Dewetting . Plasmonic
resonance . Light-trapping . Solar cells . Aluminum zinc oxide

Introduction
Thin film solar cells have gone through a long history of
development to improve the conversion efficiency and make
it a more competitive source of renewable energy. Optimizing
the properties of the basic components has reached a fairly
mature level, although drawbacks remain. [1, 2] While further
development in the basic components is still needed, there is a
clear need to harvest more of the solar spectrum available to
the cell. In fact, in a solar cell with a typical bandgap of1.8 eV,
over 50 % of the solar radiation is transmitted unutilized by the
cell. [3] Consequently, a great deal of todays research is
focused on light management schemes to optimize lighttrapping and maximize absorption in the active layer. These
schemes target almost total utilization of above-bandgap photons and more of the sub-bandgap ones. To maximize absorption, schemes involving cells with different bandgaps in tandem
and nanostructured materials with variable bandgap have been
implemented. [46] Efficient utilization of hot carriers through
the so-called multiple exciton generation has also been
reported. [7, 8] Antireflection coating to minimize reflective
losses, [9] highly reflective back contacts, [10] texturing of
surfaces and interfaces, [11] and using layers with matching
refractive index to preferentially scatter light into the active
layer [12] have all been used to optimize light absorption in
the active layer. However, many of these schemes involve high

1486

processing cost leaving a need for other less expensive lighttrapping schemes to improve efficiency at low cost.
More recently, considerable attention has been focused on
plasmonic materials to improve the light-trapping efficiency,
especially in thin film solar cells. [13, 14] Light-trapping by
plasmonic materials involves strong interaction of light with
the conduction electrons in metallic particles with appropriate
size and shape integrated to the surface. In this process,
incident light stimulates the oscillation of the conduction
electrons at interfaces containing metal nanoparticles or nanostructures of subwavelength size. When the natural frequency
of the collectively oscillating electrons matches that of the
incoming light, a localized surface plasmon resonance occurs.
At plasmonic resonance, light is preferentially scattered by the
nanoparticles decorating the surface of the solar cell into the
solar cells actively absorbing layer and improves free carrier
generation. [1517] Methods for detecting plasmonic resonance include a dip in the transmittance or a peak in the
absorbance/reflectance curves and several orders of magnitude enhancement of Raman scattering by these particles
(surface enhanced Raman scattering, SERS). [18] It has been
demonstrated that the local environment of nanoparticles can
modify the conversion efficiency of the incident photons. [19]
Developments of the nanostructured materials or patterns,
which convert sub-bandgap photons into above bandgap ones
usable by the cell, have also been introduced. [2022]
Theoretically, plasmonic light-trapping scheme dominate over
the conventional surface texturing light-trapping schemes.
[23] On the other hand, nanoparticles with undesired size or
shape may introduce parasitic losses, such as ohmic losses,
degrading the performance of the solar cell.
Silver is one of the best materials for solar cell applications due to its strong resonance and therefore large scattering cross-section but yet a low absorption cross-section in
the visible and near infrared region of the solar spectrum.
[24] Silver nanoparticles with mostly round shapes and
appropriate size can be fabricated by a simple technique
employing the dewetting of silver thin films. In this technique, a thin continuous metal film is deposited by
sputtering or other processes on a given substrate that becomes discontinuous and forms nanoparticles upon
annealing. [25, 26] The average particle size and distribution
are expected to depend on the starting film thickness, the
annealing temperature, and the underlying substrate properties such as surface roughness. However, a thorough study
of particle size distribution and shape dependences on these
parameters has not been published. [27] These parameters
are utilized in controlling the average particle size and
distribution which in turn is used to tune the plasmonic
scattering peak and promotes absorption. Particularly in
amorphous silicon solar cells, incident light with wavelength smaller than 550 nm is absorbed by the cell active
layer with ever weakening absorption in the remaining part

Plasmonics (2013) 8:14851492

of the solar spectrum towards longer wavelengths. The size

of silver nanoparticles can be controlled to have plasmonic
peaks that match the incident (front surface) or reflected
light (back reflector) to enhance the trapping of photons
with various energies including the longer wavelengths, so
that an increased effective optical path is achieved which
increases chances of absorption in the active layer.
The absorption and scattering cross-sections for spherical
nanoparticles depend strongly on the radius of the particle
and the dielectric constants of the particle and the medium,
in which it is embedded. It is therefore important to select a
medium with appropriate dielectric constant to produce the
desired plasmonic resonance. Furthermore, Ag is known to
oxidize and sulfurize rapidly in air at the operating temperatures of a solar cell, and it is therefore important to encapsulate them in a protective medium to control and stabilize
the capture cross sections and maximum absorption.
In this work, we investigate the formation mechanism
and controlling parameters of Ag nanoparticles embedded in
Al:ZnO layers and correlate their shape and size distribution
to the optical response and plasmonic peak position for their
potential use as light-trapping promoters to improve the
overall conversion efficiency of thin film solar cells.

Experimental Details
Layers of Ag nanoparticles were fabricated by the selfassembled method of dewetting of silver thin films
(15 nm), sputtered on top of an Al:ZnO layer. [2831]
The nanostructuring of the thin Ag film by dewetting was
achieved by annealing in nitrogen environment using a flow
rate of 150 sccm for 55 min in the temperature range of 200
500 C. The melting point and dielectric properties of silver
and the underlying Al:ZnO are taken into consideration to
ensure the appropriate formation of silver nanoparticles.
Immediately after dewetting process, scanning electron microscope (SEM) images were obtained (FEI, Model Quanta
400 F). The images were analyzed using image analysis
software (Gwyddion) to extract the particle mean radius
distribution and surface coverage. An optical setup
consisting of an 8-inch integrating sphere with 1-inch ports
(Oriel, Model no. 70679NS), a thermal light source, a
monochromator (Oriel Model no: 74100), and a UVenhanced silicon photodiode detector (Oriel, Model no.
70356) was used for optical measurements. A collimated
beam from a 100 W halogen lamp was directed towards the
surface of the sample which was mounted on a holder tilted
by 4 with the incident beam. The samples were mounted at
the front of the integrating sphere during transmittance
measurements and at the back of the integrating sphere
during reflectance measurements. Scattered and transmitted
light is left to passes through the monochromator before

Plasmonics (2013) 8:14851492

1487

Fig. 1 SEM images taken at different magnifications as indicated,

mean radius distributions, and optical response of Ag nanoparticles
formed in sample set A annealed in nitrogen at 200 (a, d, g), 300 (b, e,

h), and 500 C (c, f, i), and the average particle size and average size
distribution versus annealing temperature (j)

reaching the silicon detector. The wavelength range for both

reflection and transmission was varied between 300 and
1,100 nm. The raw data were processed using the dark
measurements and a calibration disc of a 100 % reflective
surface. The reflectance (R) and transmittance (T) were

measured and plotted as functions of wavelength, and we

define extinction (E 1  R  T) as the light either trapped
in Al:ZnO due to the scattering from the plasmonic interface
or absorbed. Therefore, maximum extinction is recorded
when either of the reflectance and/or the transmittance is at

1488

minimum. In the reflection mode measurements on opaque

samples, the maximum absorbance is recorded when the
reflectance is at its minimum. The plasmonic resonance of
the silver nanoparticles is calculated from the dip in the
transmittance curve for samples without a back reflector
and from the maximum in the reflectance curve in samples
with back reflectors, where no transmission is measured.
Five different sets of three samples each were prepared as
follows:
A Corning glass/300 nm, Al:ZnO/15 nm, and Ag/annealing
B Corning glass/300 nm; Al:ZnO/15 nm, Ag/annealing/60 nm
Al:ZnO
C Corning glass/300 nm; Al:ZnO/15 nm, Ag/60 nm,
Al:ZnO/annealing
D Corning glass/300 nm; Al:ZnO/15 nm, Ag at
150 C/60 nm, Al:ZnO/annealing
E Corning glass/90 nm, Ag/60 nm, Al:ZnO/15 nm,
Ag/annealing
Sample sets A through D are designed for a front
plasmonic interface, while set E is designed for a back
reflector, as it contains a 90 nm thick Ag film. The difference between sample sets B and C is that in set B, silver is
annealed to form the nanoparticles before the top Al:ZnO
was deposited. In set C, however, the top oxide layer was
deposited before the nanoparticles are formed to examine
their formation inside the oxide without exposure to air.
Sample set E was added to demonstrate the plasmonic effect
for light reflected from the back reflector.

Results
Figure 1 shows the SEM images, the size distribution, and
the optical response for the Ag nanoparticles obtained by
annealing samples in set A at 200, 300, and 500 C. A clear
increase in the mean particle radius with annealing temperature is noticed especially for the 500 C anneal. Since the
amount of Ag is conserved, the average spacing between the
nanoparticles also increases with growing nanoparticles as

Plasmonics (2013) 8:14851492

the annealing temperature increases. The median particle

diameter (called the particle size) is taken as the quantity
to compare particle sizes for this and subsequent structures.
As the distributions indicate, the median particle diameter
increases with annealing temperature.
The plasmonic resonance peak that occurs at 500 nm for
the film annealed at 200 C shifts to 630 nm for the film
annealed at 300 C and disappears for the 500 C anneal. This
red shift in resonance peak is due to the increase in the median
particle diameter from 35 (200 C anneal) to 190 nm
(300 C anneal). The sample annealed at 500 C, where the
median particle diameter exceeds 1 m, shows no plasmonic
resonance. The absence of plasmonic resonance peaks is
attributed to the large size and the poor surface coverage left
by the dispersed particles allowing incoming light to be mostly transmitted through the top layer. The increased absorption
below 520 nm is due to the 300 nm thick Al:ZnO film.
The larger contrast in plasmonic resonance signal between the 200 and 300 C anneals on the one hand and
the 300 and 500 C anneals on the other is consistent with
the conservation of silver mass. If we calculate the surface
coverage density by dividing the total volume occupying a
given area by the average particle size, we find a ratio of
5,000:30:1 for annealing temperatures of 200:300:500. It is
clear that the resonance signal contrast between 200 and
300 C anneals should be higher than that between 300 and
500 C anneals. This result qualitatively agrees with our
observed results as shown in Fig. 1 (g), (h), and (i).
To investigate the effect of top Al:ZnO layer on
plasmonic resonance and other optical properties of the
nanoparticles, a 60 nm thick Al:ZnO layer was sputtered
on top of the formed nanoparticles as indicated in sample set
B. Figure 2 shows the optical response of samples in set B
annealed at 200, 300, and 500 C. The particle formation in
this set is similar to that in similar set A, but the plasmonic
peaks are shifted due to the additional Al:ZnO top layer. As
shown in Fig. 2, the resonance peak shifts from 500 nm for
the sample in set A (no top oxide) to 550 nm for set B (with
top oxide) as can be calculated from the dips of the transmission curves in Figs. 1 and 2, respectively. This red shift

Fig. 2 Optical response of nanoparticles formed in samples in set B annealed at 200 (a), 300 (b), and 500 C (c), with a capping layer of 60 nm
Al:ZnO sputtered on top after annealing

Plasmonics (2013) 8:14851492

1489

Fig. 3 Optical response of nanoparticles formed from sample C annealed at 200 (a), 300 (b), and 500 C (c)

is well-known and documented. [13] It can be readily

explained from the implicit dependence of the scattering
cross-section on wavelength through its dependence on the
permittivities of the nanoparticle and surrounding medium.
Similar changes are not observed, however, for the
300 C anneal. This is probably because the particles formed
at this anneal are larger (150 nm), and the additional layer of
60 nm Al:ZnO sputtered on top of the annealed films is not
likely to cover them. The mean diameter of the Ag particles
is larger than 60 nm and, therefore, the second 60 nm thick
Al:ZnO layer sputtered on top will conform with the bottom
300 nm thick Al:ZnO forming a 360 nm thick Al:ZnO in the
nanoparticle-free surface areas and increase the absorption
as can be noticed in comparing Figs. 2 and 3 for 300 and
500 C.
To examine the formation and optical properties of the Ag
nanoparticles formed inside the Al:ZnO layer, we use the
structure in sample set C, where the anneals are performed
after the top oxide layer is sputtered. Figure 3 shows the
optical response of three samples in set C, where the 15 nm
Ag layer is encapsulated by the top Al:ZnO layer before
annealing at 200, 300, and 500 C. None of the samples show
any detectible plasmonic peaks. The absence of plasmonic
peaks for the 200 and 300 C anneals in this set is due to the
capping layer of 60 nm Al:ZnO that inhibits the dewetting
process and particle formation. As the temperature increases,
however, molten silver appears to flow, most likely through

cracks, up to the top of the Al:ZnO layer and form silver

spheroids.
In Fig. 4, we show SEM images of two samples in a
similar set annealed at 400 and 500 C along with an energy
dispersive X-ray spectroscopy scan for a formed particle at
500 C. The films were tilted by 90 to illustrate the diffusion of silver. At 400 C, the silver film appears to maintain
a continuous layer, where dewetting is inhibited by the top
oxide, even though some silver is beginning to diffuse. As
the temperature is increased to 500 C, the film appears to
diffuse and form particles that exceed 1.02 m in diameter
and consist of pure silver. These particles are not caused by
the dewetting process and, therefore, their shape and size are
not controlled properly. They are not embedded in the oxide
and cannot be compared with the samples in set B.
Furthermore, they are dispersed randomly and do not interact sufficiently with the incoming light both due to their size
compared to the wavelength of visible light and the large
spacing separating them.
Even though the dewetting process which involves
the exposure of Ag particles to atmospheric changes
appears to enhance absorption, it is still important to
develop a process in which the Ag nanoparticles are
embedded inside the dielectric medium any exposure to
atmosphere. Exposing silver nanoparticles to atmosphere
leads to atmospheric modifications of the silver nanoparticles
and their medium, which may influence their scattering cross-

Fig. 4 SEM images for samples in set B annealed at 400 (a) and
500 C (b), along with an EDX analysis (c) of a silver particle of
diameter 1.02 that is formed at the surface after the 500 C anneal.

The inset in part (c) shows a close-up image of the Ag nanoparticle

(marked by red arrow) from which the EDX data was acquired

1490

Plasmonics (2013) 8:14851492

Fig. 5 The optical response of three samples in set D annealed at 200 (a), 300 (b), and 500 C (c)

section. Since dewetting was not successful for the encapsulated Ag film, it was important to construct a structure in
which dewetting and the dielectric encapsulation are
performed in the same chamber without exposure to air. For
this purpose, sample set D was constructed with the 15 nm Ag
film sputtered at a substrate temperature of 150 C compared
to room temperature for sample set C to initiate the dewetting
process.
Figure 5 shows the optical response for three samples
in set D annealed at 200, 300, and 500 C as before.
The particle mean radius distribution is not possible to
obtain in this set because the top oxide layer inhibits its
measurements, but dips in absorption clearly indicate
that plasmonic resonance is occurring in these films.
The optical response from these samples shows clear
plasmonic peaks occurring at 709, 700, and 820 nm at
200, 300, and 500 C, respectively, suggesting that
nanoparticles from the dewetted film are probably taking place during the subsequent anneals. Increasing the
substrate temperature during Ag sputtering to 150 C
appears to enhance the dewetting of Ag film even in the
presence of the top oxide and prevents the film form
diffusing through the cracks in the top oxide layer as
was observed in sample set C. In Fig. 6, we show SEM
images of two 15 nm thin Ag films sputtered at room
temperature and at 150 C. The film sputtered at
150 C is rather discontinuous, indicating that dewetting
of Ag is beginning to take place, but nanoparticles are
not formed. The plasmonic peaks observed in Fig. 5
suggest that this discontinuity permits the formation of
nanoparticles even in the presence of the top oxide
layer at our standard annealing temperatures. Based on
preliminary data in an ongoing study, we strongly believe that sputtering Ag at 200 C or higher in vacuum
or inert gas would cause complete dewetting similar to
that of furnace annealing, eliminating the need for the
external furnace annealing step and enabling an in situ
process.
To demonstrate the particle formation and plasmonic
effects for the light scattered from the back reflector, sample
set E was prepared similar to set A except for the additional

90 nm Ag layer at the substrate. Figure 7 shows the SEM

images, the corresponding mean radius distribution, and
optical reflectance for the nanoparticles obtained by
annealing two samples in this set at 200 and 500 C. The
SEM images and particle size distribution is consistent with
the results of set A as expected, but the transmittance measurements in set D are naturally inhibited by the back
reflector. The plasmonic resonance peak occurs at
500 nm for the 200 C anneal, where the average particle
size is50 nm, and no plasmonic resonance peak is observed for the 500 C anneal, again due to the larger particle
size. The 90 nm thick Ag film in sample set C appears to
have no negative effect on the formation, size distribution,
and optical properties of the resulting Ag nanoparticles. This
confirms that the dewetting method can be used to facilitate
the light-trapping for both the top surface and the back
reflector of a solar cell device.

Conclusion
Silver nanoparticles of controllable shape and size embedded in Al:ZnO layer in a composite-like structure were
fabricated on top of Al:ZnO layers to correlate nanoparticle
shape, size and size distribution with the resulting plasmonic
peaks. It was found that the starting Ag film thickness, the
dewetting temperature, and the medium in which nanoparticles
are formed are important parameters in controlling the

Fig. 6 SEM images of two thin Ag films sputtered on top of Al:ZnO

film at room temperature (a) and at 150 C (b)

Plasmonics (2013) 8:14851492

1491

Fig. 7 SEM images, particle size distribution, and optical absorption for samples in set D annealed at 200 and 500 C

size and shape of the nanoparticles. Nanoparticles following a random distribution with a mean particle size
of 5060 nm in diameter are obtained by annealing a
sputtered 15 nm Ag layer at temperatures of about
200 C. Larger particles of120 nm in diameter were
obtained by increasing the annealing temperature to
300 C. We also found that total encapsulation of Ag
film in Al:ZnO inhibits dewetting as was demonstrated
in sample set C for anneals below 300 C. Increasing
the annealing temperature to over 400 C appears to
produce larger particles outside the Al:ZnO as was
shown in Fig. 4, but not through a conventional dewetting
process. The progressive increase in the size of particles that
appear to be emerging from the Al:ZnO at 400 and 500 C
(Fig. 4) suggests that the thin Ag film is flowing through
existing or induced cracks in the top Al:ZnO layer. Even
though the melting temperature of pure bulk silver is close
to 900 C, it is known that this temperature is drastically
reduced at the nanoscale. [32] Furthermore, the process may
also involve vertical bulk diffusion of silver through the top
Al:ZnO layer, followed by lateral surface diffusion leading to
the more spherical nanoparticles with narrower size distribution. Surface diffusion leads the emerging particles to reach a
more uniform size distribution at these elevated annealing
temperatures. The more uniform particle size and more spherical shapes are consistent with the results of the high temperature anneals of sample set A. To produce nanoparticles
embedded in the Al:ZnO, dewetting and particle

formation should be performed by annealing before

adding the capping Al:ZnO layer. This can be
performed with or without exposure to atmosphere
depending on the annealing temperature. An alternative
method to produce nanoparticles embedded in the
Al:ZnO without exposure to atmospheric agents is to
sputter the Ag film at higher temperature. Our results
show that sputtering Ag at 150 C appears to initiate
the dewetting process and allows for the fabrication of
encapsulated silver nanoparticles without exposure to
atmosphere upon subsequent annealing. The particle
sizes obtained at 200 and 300 C anneals produce
plasmonic peaks in the visible or near infrared regions
of the solar spectrum with expected strong light-trapping
effects that can be integrated to improve the efficiency
of thin film solar cells. Increasing the annealing temperature to over 400 C results in the formation of large
and pure Ag nanoparticles exceeding 1 m in diameter
and have no detectable peaks in the visible or near
infrared regions.
Acknowledgments This study has been supported by the Scientific
and Technological Research Council of Turkey (TUBITAK) under the
project Rainbow Energy, with the contract number 109R037. The
research leading to these results has received funding from the European Unions Seventh Framework Program FP7/2007-2013 under
grant agreement no. 270483. H. Nasser greatly acknowledges the
Scientific and Technological Research Council of Turkey (TUBITAK
BIDEB-2215) for financial support.

1492
Conflict of Interests
interests.

Plasmonics (2013) 8:14851492

The authors declare no competing financial

References
1. Staebler DL, Wronski CR (1977) Reversible conductivity changes
in dischargeproduced amorphous Si. Appl Phys Lett 31:292
2. Smirnov V, Reynolds S, Finger F, Carius R, Main C (2006)
Metastable effects in silicon thin films: atmospheric adsorption
and light-induced degradation. J Non-Cryst Solids 352:10751078
3. Atrel A C, Garca-Etxarri A, Alaeian H, Dionne JA (2012) Toward
high-efficiency solar upconversion with plasmonic nanostructures,
J. Opt. 14 024008
4. Kupich M, Grunsky D, Kumar P, Schroder B (2004) Preparation of
microcrystalline single junction and amorphousmicrocrystalline
tandem silicon solar cells entirely by hot-wire CVD. Sol Energy
Mater Sol Cells 81:141146
5. Zacharias M, Heitmann J, Scholz R, Kahler U, Schmidt M, Blasing
J (2004) Size-controlled highly luminescent silicon nanocrystals:
a-SiO/SiO2 superlattice approach. Appl Phys Lett 80:661663
6. Di D, Perez-Wurfl I, Conibeer G, Green MA (2010) Formation and
photoluminescence of Si quantum dots in SiO2/Si3N4 hybrid matrix
for all-Si tandem solar cells. Sol Energy Mater Sol Cells 94:22382243
7. Nozik AJ (2002) Quantum dot solar cells. Physica E 14:115120
8. Beard MC, Knutsen KP, Yu P, Luther JM, Song Q, Metzger WK,
Ellingson RJ, Nozik AJ (2007) Multiple exciton generation in
colloidal silicon nanocrystals. Nano Lett 7(8):25062512
9. Gangopadhyay U, Kim K, Mangalaraj D, Yi J (2004) Low cost
CBD ZnS antireflection coating on large area commercial monocrystalline silicon solar cells. Appl Surf Sci 230:364370
10. Soderstromn K, Haug FJ, Escarre J, Pahud C, Biron R, Ballif C
(2011) Highly reflective nanotextured sputtered silver back reflector for flexible high-efficiency nip thin-film silicon solar cells.
Sol Energy Mater Sol Cells 95:35853591
11. Ghannam MY, Abouelsaood AA, Alomar AS, Poortmans J (2010)
Analysis of thin-film silicon solar cells with plasma textured front
surface and multi-layer porous silicon back reflector. Sol Energy
Mater Sol Cells 94:850856
12. Fleischer K, Arca E, Shvets IV (2012) Improving solar cell efficiency with optically optimized TCO layers. Sol Energy Mater Sol
Cells 101:262269
13. Pillai S, Green MA (2010) Plasmonics for photovoltaic applications. Sol Energy Mater Sol Cells 94:14811486
14. Trupke T, Green MA, Wurfel P (2002) Improving solar cell efficiencies by up-conversion of sub-bandgap light. J Appl Phys 92:4117

15. Aberle AG (2001) Overview on SiN surface passivation of crystalline silicon solar cells. Sol Energy Mater Sol Cells 65:239248
16. Bohren CF, Huffman DR (2004) Absorption and scattering of light
by small particles. Wiley, Weinheim
17. Mirin NA, Halas NJ (2009) Light-bending nanoparticles. Nano
Lett 9:12559
18. Wang C, Fang J, Jin Y, Cheng M (2011) Fabrication and surfaceenhanced Raman scattering (SERS) of Ag/Au bimetallic films on
Si substrates. Appl Surf Sci 258:11441148
19. Kreibig U, Vollmer M (1995) Optical properties of metal clusters.
Springer, Berlin
20. Lahoz F, Perez-Rodrguez C, Hernandez SE, Martn IR, Lavn V,
Rodrguez-Mendoza UR (2011) Upconversion mechanisms in
rare-earth doped glasses to improve the efficiency of silicon solar
cells. Sol Energy Mater Sol Cells 95:16711677
21. Aubry A, Lei DY, Fernandez-Dominguez AI, Sonnefraud Y, Maier
SA, Pendry JB (2010) Plasmonic light-harvesting devices over the
whole visible spectrum. Nano Lett 10:25749
22. Guler U, Turan R (2010) Effect of particle properties and light
polarization on the plasmonic resonances in metallic nanoparticles.
Opt Express 18:1732217338
23. Beck FJ, Polman A, Catchpole KR (2009) Tunable light-trapping
for solar cells using localized surface plasmons. J Appl Phys
105:114310
24. Atwater HA, Polman A (2010) Plasmonics for improved photovoltaic devices. Nature Mater 9:20513
25. Gupta R, Dyer J, Weimer WA (2002) Preparation and characterization of surface plasmon resonance tunable gold and silver films.
J Appl Phys 92:526471
26. Suzuki Y, Ojima Y, Fukui Y, Fazyia H, Sagisaka K (2007)
Postannealing temperature dependence of infrared absorption enhancement of polymer on evaporated silver films. Thin Solid
Films 515:30738
27. Temple TL, Mahanama GDK, Reehal HS, Bagnall DM (2009)
Influence of localized surface plasmon excitation in silver
nanoparticles on the performance of silicon solar cells. Sol Energy
Mater Sol Cells 93:19781985
28. Heilman A (2003) Polymer films with embedded metal
nanoparticles. Springer, Berlin
29. Stuart HR, Hall DG (1998) Island size effects in nanoparticles
enhanced photo detectors. Appl Phys Lett 73:38157
30. Pillai S, Catchpole KR, Trupke T, Green MA (2007) Surface
plasmon enhanced silicon solar cells. J Appl Phys 101:093105
31. Moulin E, Sukmanowski J, Luo P, Carius R, Royer FX, Stiebig H
(2008) Improved light absorption in thin-film silicon solar cells by
integration of silver nanoparticles. J Non-Cryst Solids 354:248891
32. Xiao SF, Hu WY, Yang JY (2005) Melting behaviors of nanocrystalline Ag. J Phys Chem B109:20339