Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
art ic l e i nf o a b s t r a c t
Article history: In this paper, we investigate three different procedures for making modulated surface textures (MST) on
Received 30 July 2014 glass substrates and show the application of these substrates to microcrystalline silicon solar cells. MST
Received in revised form substrates have broader surface angle distribution (AD) than a single textured glass substrate and this
30 October 2014
distribution peaks at higher angles. Therefore, (i) the haze in reection of MST substrates covered with
Accepted 5 November 2014
an Ag layer is over 99% in the wavelength range above 350 nm, which is higher than in the case of a
single textured glass substrate; (ii) more reected light is scattered into higher angles, resulting in a
Keywords: broader reective angular intensity distribution (AID) in air. These properties lead to a higher Jsc for n-i-p
Modulated surface textures microcrystalline silicon solar cells deposited on the MST substrates compared to cells deposited on single
Glass texturing
textured glass substrate. A solar cell deposited on a MST glass substrate with smoother peaks shows the
Light trapping
highest electrical performance (Voc FF value), because of a lower density of defective regions. The best
Thin-lm silicon solar cells
cell achieved in this work has an active-area efciency of 10.49%, with a Voc of 0.532 V, a Jsc of 26.36 mA/cm2,
and a FF of 74.8% for a 3-m thick absorber layer.
& 2014 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.solmat.2014.11.013
0927-0248/& 2014 Elsevier B.V. All rights reserved.
G. Yang et al. / Solar Energy Materials & Solar Cells 133 (2015) 156162 157
commonly used nano-textures. In this study, we propose MST's in reference. The inuence of the surface morphology of the textured
which smooth and smaller features are added onto the micro- substrates on the mc-Si:H solar-cell performance is also discussed.
textures, with the aim to enhance the light trapping while keeping
the high Voc and FF in c-Si:H solar cells.
For the work presented in this paper we made MST substrates 2. Experimental
based on textured glass or a wet etched ZnO:Al coating layer on top
of the single textured glass substrate. The surface morphology of 2.1. Preparation and characterization of MST glass substrates
the MST substrates is presented. Then 250-nm thick Ag is evapo-
rated onto the rough surface of the substrate to form the back In this subsection we will discuss the methods that we have used to
reector (BR). In order to study the effect of the substrate morphol- make the MST structures. The MST structures were fabricated in a two-
ogy on the solar-cell performance, we deposit n-i-p single junction step process. First we describe the ITO Induced Texturing (IIT) process
c-Si:H substrate type solar cells on the rough surface of these BRs. that we used to make textured surfaces with larger features. Subse-
This implies that the light enters the solar cell via the silicon layers quently, three different methods were used to implement smaller
and not via the glass side of the solar cells. The optical properties of features onto the previously obtained larger features. These three
the obtained BRs and the performance of the substrate n-i-p mc-Si:H processes are on glass: (i) Aluminium Induced Texture (AIT), (ii) wet
solar cells are presented with single textured glass substrate as etching of ZnO:Al coating layer (EAZO), and (iii) ZnO:Al Induced Texture
Fig. 1. AFM scan of (a) IIT textured glass, (b) MST-AIT textured glass, (c) MST-EAZO textured substrate, (d) MST-ZIT textured glass; (e) arbitrary surface cross-section and
(f) Surface AngleDistribution (AD) of textured substrates.
158 G. Yang et al. / Solar Energy Materials & Solar Cells 133 (2015) 156162
(ZIT). Then a 250-nm thick Ag layer is evaporated onto the rough 2.1.5. Preparation of MST substrates
surface of these MST substrates to form the BRs for the following optical First, the IIT process is used to texture the glass to obtain a
measurements and deposition of substrate n-i-p mc-Si:H solar cell. rough surface with large features. Second, in order to obtain the
In order to get the MST substrate surface morphology para- smaller features that are embedded in the larger features, three
meters, e.g. the root mean square roughness (RMS), the textured different processes (AIT, EAZO and ZIT) are carried out on the
substrates are studied by an atomic force microscope (AFM, NT- rough glass surface obtained by IIT process. After the second
MDT). The optical properties of the 250-nm Ag coated BRs (i.e., processing procedure an MST is obtained. The AFM scan of MST
total reection, haze in reection, and angular intensity distribu- glass obtained by an AIT process following the IIT process on glass
tion (AIDair) of reected light) are measured in air using an (MST-AIT) is shown in Fig. 1(b), by an EAZO process following the
integrating sphere and the Angular Resolved Transmittance/ IIT process is shown in Fig. 1(c), and by a ZIT process following the
Reectance Analyzer in a PerkinElmer Lambda 950 spectrophot- IIT process is shown in Fig. 1(d).
ometer. Experimental details of the measurements can be found
elsewhere [24,25].
2.2. Preparation and characterization of solar cells
2.1.1. ITO Induced Textures (IIT) On IIT and MST substrates, we have deposited c-Si:H single
Corning Eagle XG glass covered with 200-nm thick In2O3:Sn junction substrate n-i-p solar cells in order to study the effect of the
(ITO) sacricial layer is wet etched with H2O2 and HF to achieve a substrate morphology on the solar-cell performance. The structure of
textured surface. The ITO layer is deposited with a magnetron the solar cells is as follows: textured substrate/Ag/n-SiOx:H/i-c-Si:H/
sputtering system at a substrate temperature of 115 1C, a sputter- p-SiOx:H/In2O3:Sn (ITO)/metal grid. The n-SiOx:H layer used here has
ing pressure of 2.6 mbar, a target-to-substrate distance of 110 mm, a similar reective index (2.3, in the wavelength range from
and a power density of 0.9 W/cm2. More detailed information on 400 nm to 1100 nm) compared to that for ZnO layer (2.0, in the
this process is published elsewhere [13,14]. The typical horizontal wavelength range from 400 nm to 1100 nm). It was shown before
feature size obtained with such IIT method is 15 m in average, as that the reectance for ZnO:Al/Ag BR and SiOx:H/Ag BR congura-
is shown in atomic force microscopy (AFM) picture in Fig. 1(a). tions is similar [27], which implies that the plasmonic absorption
losses are similar for these two BR congurations. Our experimental
results showed that the Voc of the solar cells on ZnO:Al/Ag BR is
2.1.2. Aluminium Induced Textures (AIT) slightly lower (not shown here) and therefore we did not use a ZnO:
AIT (Aluminium Induced Texturing) texturing of glass was Al layer between Ag and n-SiOx:H layer. The structure of the n-i-p c-
developed by Aberle et al. [9]. In this process 200-nm thick Si:H solar cells is schematically shown in Fig. 2. All the silicon layers
aluminium (Al) is evaporated on a glass substrate. Then this are deposited in a multi-chamber PECVD system: n-SiOx:H (50 nm)
structure is annealed at 600 1C in vacuum (at a background and p-SiOx:H (20 nm) layers are deposited with rf-PECVD, while the
pressure of 10 7 mbar) for 3 h before etching with HNO3 and HF. intrinsic c-Si:H layer is deposited with VHF-PECVD at 40.68 MHz
The feature size of textured glass obtained by the AIT technique with SiH4 and H2 as precursor gases [14]. The deposition rate is 6.7 /s
can be tuned by changing the Al thickness, annealing temperature, and the i-layer thickness varies from 2.0 m to 3.0 m. Then a 75-nm
and annealing duration. The obtained feature size in this work is thick ITO layer is sputtered, onto which a metallic grid (Ag-100 nm/Cr-
23 m. 30 nm/Al-300 nm) is evaporated to form the solar-cell device with an
area of 0.25 cm2. Before the measurements, the solar cells were
annealed for 2 h at 170 1C.
2.1.3. ZnO Induced Textures (ZIT) The Voc and FF of the solar cells are determined from JV
A 300-nm thick ZnO:Al (AZO) sacricial layer is sputtered on measurements under 100 mW/cm2 illumination (PASAN). The results
glass and followed by etching with HNO3 and HF to obtain a reported in this paper are the average from the 10 best performing
textured surface. ZnO:Al lm is sputtered using a magnetron cells out of 24 cells on one substrate stripe. The Jsc is calculated from
sputtering system at substrate temperature of 200 1C, a pressure the external quantum efciency (EQE) measured with an in-house
of 2.5 mbar, and a sputtering plasma power density of 1.25 W/cm2, built setup under short-circuit condition in the wavelength range
and using 2 wt% Al2O3 doped ceramic ZnO:Al as target. The from 300 nm to 1100 nm using the AM1.5G spectrum [28]. In this
etchant ratio of [HNO3] (68%):[HF] (40%) is 8:1. The obtained way the overestimation of the Jsc due to the lateral current collection
feature size can be tuned by varying the ZnO:Al layer thickness, and inaccuracy in dening the active area of the solar cells in JV
ZnO:Al layer crystal properties (by adjusting the sputtering con- measurements is excluded [29].
ditions), and the etchant composition. Details on how to prepare
ZIT textured glass is published elsewhere [23]. The feature size
obtained in this work is 2 m on average.
3. Results
AFM scans of the IIT glass together with three different MST
substrates fabricated are shown in Fig. 1. Compared to IIT glass, the
RMS MST-AIT and MST-EAZO are higher, but MST-ZIT becomes less
rough after the ZIT process giving a smaller RMS. This difference is
mainly due to the difference in texturing mechanisms:
Table 1
Performance of solar cells deposited on textured substrates. The standard deviation (SD) for each solar cell parameter calculated from the 10 best performing cells on each
substrate is similar: SD (Voc) E 0.003 V, SD (FF) E0.004, SD (Jsc) E0.42 mA/cm2, SD (Eff.)E 0.13%, SD (Voc FF) E0.005 (V).
Substrate Thickness (m) Voc (V) FF (-) Jsc (mA/cm2) Eff. (%) Voc FF (V)
a
The variation of Jsc for solar cells deposited on MST substrates compared to the solar cell deposited on IIT substrate.
b
The variation of the efciency for solar cells deposited on MST substrates compared to IIT substrate.
c
The variation of Voc FF products when the solar cell i-layer thickness is increased from 2 m to 3 m.
Fig. 5. External quantum efciency (EQE) for solar cells with 3.0-m thick c-Si:H
i-layer deposited on different textured substrates.
4. Conclusion
Acknowledgement
reduce the solar-cell electrical performance, resulting in a lower
Voc and FF. These ndings are in agreement with results reported The authors thank Martijn Tijssen and Stefaan Heirman for
by other groups [21,30,34]. The increase in the density of defective their technical support, DoYun Kim for the useful discussion. G.
regions for the cell deposited on the MSTEAZO substrate is also Yang thanks the China Scholarship Council for the nancial
one of the main reasons for the rapid drop of the Voc FF product, support.
by 9.3%, when the i-layer thickness increases from 2 m to 3 m
(see Table 1).
References
electrical performance in thin-lm nanocrystalline silicon solar cells, Appl. [26] G. Yang, R.A.C.M.M. van Swaaij, S. Dobrovolskiy, M. Zeman, Textured substrate
Phys. Lett. 103 (2013) 173905. for high-efciency n-i-p mc-Si:H solar cells, in: Proceedings of the 39th IEEE
[15] M. Boccard, C. Battaglia, S. Hanni, K. So derstro m, J. Escarre, S. Nicolay, PVSC, Tampa, FL, 2013.
F. Meillaud, M. Despeisse, C. Ballif, Multiscale transparent electrode architec- [27] V. Demontis, C. Sanna, J. Melskens, R. Santbergen, A.H.M. Smets, A. Damiano,
ture for efcient light management and carrier collection in solar cells, Nano M. Zeman, The role of oxide interlayers in back reector congurations for
Lett. 12 (3) (2012) 13441348. amorphous silicon solar cells, J. Appl. Phys. 113 (2013) 064508.
[16] Q.J. Jiang, J.G. Lu, J. Zhang, Y.L. Yuan, H. Cai, L. Hu, L.S. Feng, B. Lu, X.H. Pan [28] http://rredc.nrel.gov/solar/spectra/am1.5/.
Z.Z. Ye, Texture-etched broad surface features of double-layered ZnO:Al [29] H. Tan, R. Santbergen, A.H.M. Smets, M. Zeman, Plasmonic light trapping in
transparent conductive lms for high haze values, J. Alloys Compd. 596 thin-lm silicon solar cells with improved self-assembled silver nanoparticles,
(2014) 107112. Nano Lett. 12 (8) (2012) 40704076.
[17] A. Gordijn, L. Hodakova, J. Rath, R. Schropp, Inuence on cell performance of [30] M. Python, E. Vallat-Sauvain, J. Bailat, D. Domine, L. Fesquet, A. Shah, C. Ballif,
bulk defect density in microcrystalline silicon grown by VHF PECVD, J. Non- Relation between substrate surface morphology and microcrystalline silicon
Cryst. Solids 352 (2006) 18681871. solar cell performance, J. Non-Cryst. Solids 354 (2008) 22582262.
[18] G. Bugnon, G. Parascandolo, T. Soderstrom, P. Cuony, M. Despeisse, S. Hanni, [31] M. Sever, B. Lipovek, J. Kr, A. ampa, G.S. Plaza, F. Haug, M. Duchamp,
J. Holovsky, F. Meillaud, C. Ballif, A new view of microcrystalline silicon: the W. Soppe, M. Topi, Combined model of non-conformal layer growth for
role of plasma processing in achieving a dense and stable absorber material accurate optical simulation of thin-lm silicon solar cells, Sol. Energy Mater.
for photovoltaic applications, Adv. Funct. Mater. 22 (2012) 36653672.
Sol. Cells 119 (2013) 5966.
[19] H. Sai, H. Fujiwara, M. Kondo, Y. Kanamori, Enhancement of light trapping in
[32] J. Springer, A. Poruba, L. Mullerova, M. Vanecek, O. Kluth, B. Rech, Absorption
thin-lm hydrogenated microcrystalline Si solar cells using back reectors
loss at nano-rough silver back reector of thin-lm silicon solar cells, J. Appl.
with self-ordered dimple pattern, Appl. Phys. Lett. 93 (2008) 143501.
Phys. 95 (2004) 14271429.
[20] H. Sai, H. Jia, M. Kondo, Impact of front and rear texture of thin-lm
[33] M. Python, O. Madani, D. Domine, F. Meillaud, E. Vallat-Sauvain, C. Ballif,
microcrystalline silicon solar cells on their light trapping properties, J. Appl.
Inuence of the substrate geometrical parameters on microcrystalline silicon
Phys. 108 (2010) 044505.
[21] H. Sai, K. Saito, N. Hozuki, M. Kondo, Relationship between the cell thickness growth for thin-lm solar cells, Sol. Energy Mater. Sol. Cells 93 (2009)
and the optimum period of textured back reectors in thin-lm microcrystal- 17141720.
line silicon solar cells, Appl. Phys. Lett. 102 (2013) 053509. [34] H.T. Li, R.H. Franken, J.K. Rath, R.E.I. Schropp, Structural defects caused by a
[22] B. Yan, G. Yue, L. Sivec, J. Owens-Mawson, J. Yang, S. Guha, Correlation of rough substrate and their inuence on the performance of hydrogenated
texture of Ag/ZnO back reector and photocurrent in hydrogenated nano- nano-crystalline silicon n-i-p solar cells, Sol. Energy Mater. Sol. Cells 93 (2009)
crystalline silicon solar cells, Sol. Energy Mater. Sol. Cells 104 (2012) 1317. 338349.
[23] G. Yang, R.A.C.M.M. van Swaaij, O. Isabella, M. Zeman, A novel way of texturing [35] R. Kind, R.A.C.M.M. van Swaaij, F.A. Rubinelli, S. Solntsev, M. Zeman, Thermal
glass for microcrystalline Si:H thin lm solar cells application, Prog. Photo- ideality factor of hydrogenated amorphous silicon p-i-n solar cells, J. Appl.
volt.: Appl. Res. (2014), http://dx.doi.org/10.1002/pip.2550. Phys. 110 (2011) 104512.
[24] K. Jger, M. Fischer, R.A.C.M.M. van Swaaij, M. Zeman, A scattering model for [36] B.E. Pieters, H. Stiebig, M. Zeman, R.C.M.M. van Swaaij, Determination of the
nano-textured interface and its application in opto-electrical simulations of mobility gap of intrinsic c-Si:H in p-i-n solar cells, J. Appl. Phys. 105 (2009)
thin lm solar cells, J. Appl. Phys. 111 (2012) 081008. 044502.
[25] K. Jger, O. Isabella, R.A.C.M.M. van Swaaij, M. Zeman, Angular resolved [37] O. Vetterl, A. Lambertz, A. Dasgupta, F. Finger, B. Rech, O. Kluth, H. Wagner,
scattering measurements of nano-textured substrates in a broad wavelength Thickness dependence of microcrystalline silicon solar cell properties, Sol.
range, Meas. Sci. Technol. 22 (2011) 105601. Energy Mater. Sol. Cells 66 (2001) 345351.