§ 3-1 Introduction
InGaN-based solar cells have recently drawn much research attention due to
many desired photovoltaic (PV) characteristics, including the potential to realize
nearly full absorption of solar spectrum, high absorption coefficients, and high
mobilities.[1-3] In addition, InGaN alloys have superior radiation resistance, thermal
stability, and chemical tolerance.[3] In consideration of the structure design for active
regions, multiple quantum wells (MQWs) consisting of the InGaN layers thinner than
the critical thickness on GaN not only prevent the formation of undesired crystal
defects, but also provide an additional control of light absorption through the
quantized energy levels.[4]
A number of methods have been developed to improve the performances of
InGaN-based MQW solar cells.[5-8]For example, Ag nanoparticles have been utilized
to induce plasmonic resonance on the device surface to enhance light absorption in
the InGaN active region.[7] Kuwahara et al. reported that the PV performances of
InGaN solar cells with supperlattice active layers can be enhanced by optimizing the
barrier thickness.[8]
Recently, it is reported that subwavelength nanostructures exhibit superior
antireflective (AR) performances to their quarter-wavelength thin-film
counterparts.[9,10] It can be clarified by the light trapping effect at short wavelength
regions and the effective medium effect at long wavelength regions [11,12], where the
nanostructures can be regarded as an effective medium with the effective refractive
index (neff) increasing gradually from air to device surfaces. As the incident light is
reflected at different depths in nanostructured layers, the suppression of the reflection
over a wide range of wavelengths can occur through destructive interferences among
the reflected waves. Periodic subwavelength nanostructures have been fabricated with
various techniques, such as the nanosphere lithography[13,14]and the anodic aluminum
oxide templates.[15,16] For the practical applications, the interests in the AR
nanostructures have been extended to disordered nanostructures using a simple yet
scalable method with low cost.[6,9,17,18]
In this study, we fabricated AR SiO2 nanorod arrays (NRAs) utilizing
self-assembled Ag nanoparticles as a mask and reactive ion etching (RIE) techniques
with photoresist-free and wafer-scale-uniformity capabilities. SiO2 NRAs effectively
reduce surface reflections over a wide wavelength range and thus InGaN MQW solar
cells employing the SiO2 NRAs generate additional photocurrents, corresponding to
the conversion efficiency enhancement of 21 % due to increased short-circuit current
(Jsc) from 0.71 to 0.76 mA/cm2. Simulation results based on finite-difference
time-domain (FDTD) analysis also indicate that the improved device performance is
due to the enhanced optical absorption in the MQW layers upon the application of
SiO2 NRAs. The proposed concept in this study is applicable to other optoelectronic
devices.
§ 3-2 Experiment
The MQW solar cells were grown by metal-organic chemical vapor deposition
on c-plane sapphire substrates. The layer structures consist of nine periods of
intentionally undoped In0.3Ga0.7N (3 nm)/GaN (17 nm) MQWs, sandwiched by a
2.5-μm n-type and a 0.2-μm p-type GaN layer. In content in the MQWs determined by
x-ray diffraction is around 30 %. Following the growth, transparent Ohmic contacts to
p-GaN were formed with indium tin oxide (ITO) deposited by the electron beam
evaporation. The 1×1 mm2diode mesas were then defined by chlorine-based plasma
etching. The contact scheme consists of fingered Ti/Al/Ni/Au metal grids deposited
on the ITO and the n-GaN.
Fig. 3.1 describes the fabrication procedure for SiO2NRAs. First, the SiO2/Ag
layers with the thickness of 300/15 nm were deposited on the ITO layer of
InGaN/GaN MQW solar cells by the electron beam evaporation [Fig. 3.1(a)]. For
forming Ag nanoparticles, a thermal annealing process was carried out in the furnace
at 270 ℃ for 2 min [Fig. 3.1(b)]. Using these Ag nanoparticles as etching masks, the
SiO2layer was patterned by the RIE process with CHF3gas (30 SCCM) and rf power
of 90 W for 9 min [Fig. 3.1(c)]. After removing the remaining Ag nanoparticles by
HNO3, the SiO2NRAs were obtained, as shown in Fig. 3.1(d). It should be mentioned
that we had controlled Ag thickness from 5 nm to 30 nm, annealing temperatures from
250 ℃ to 350 ℃, and annealing times from 1 min to 30 min, which contribute to the
different geometries of self-assembled Ag nanoparticles. Changing etching times from
5 min to 10 min of RIE increased nanorod lengths. After experimental parameter
optimization, the best photovoltaic conversion efficiency of our solar cells devices can
be obtained using 15-nm-thick Ag films combining an annealing process at 270 ℃ for
2 min and subsequent RIE process for 9 min.
Figure 3.1 Schematic of fabrication procedures for the antireflective nanostructures forming Ag nanoparticles, a thermal annealing process was carried out in the furnace
at 270 ℃ for 2 min [Fig. 3.1(b)]. Using these Ag nanoparticles as etching masks, the
SiO2layer was patterned by the RIE process with CHF3gas (30 SCCM) and rf power
of 90 W for 9 min [Fig. 3.1(c)]. After removing the remaining Ag nanoparticles by
HNO3, the SiO2NRAs were obtained, as shown in Fig. 3.1(d). It should be mentioned
that we had controlled Ag thickness from 5 nm to 30 nm, annealing temperatures from
250 ℃ to 350 ℃, and annealing times from 1 min to 30 min, which contribute to the
different geometries of self-assembled Ag nanoparticles. Changing etching times from
5 min to 10 min of RIE increased nanorod lengths. After experimental parameter
optimization, the best photovoltaic conversion efficiency of our solar cells devices can
be obtained using 15-nm-thick Ag films combining an annealing process at 270 ℃ for
2 min and subsequent RIE process for 9 min.
Figure 3.1 Schematic of fabrication procedures for the antireflective nanostructures forming Ag nanoparticles, a thermal annealing process was carried out in the furnace
at 270 ℃ for 2 min [Fig. 3.1(b)]. Using these Ag nanoparticles as etching masks, the
SiO2layer was patterned by the RIE process with CHF3gas (30 SCCM) and rf power
of 90 W for 9 min [Fig. 3.1(c)]. After removing the remaining Ag nanoparticles by
HNO3, the SiO2NRAs were obtained, as shown in Fig. 3.1(d). It should be mentioned
that we had controlled Ag thickness from 5 nm to 30 nm, annealing temperatures from
250 ℃ to 350 ℃, and annealing times from 1 min to 30 min, which contribute to the
different geometries of self-assembled Ag nanoparticles. Changing etching times from
5 min to 10 min of RIE increased nanorod lengths. After experimental parameter
optimization, the best photovoltaic conversion efficiency of our solar cells devices can
be obtained using 15-nm-thick Ag films combining an annealing process at 270 ℃ for
2 min and subsequent RIE process for 9 min.
Figure 3.1 Schematic of fabrication procedures for the antireflective nanostructures
on InGaN MQW solar cells.
§ 3-3 Results and Discussion
Fig. 3.2(a)-(b) respectively reveal the top-view and 45-degree cross-sectional
images of the fabricated MQW device with SiO2 NRA surfaces. With Ag
nanoparticles as etching masks, the underlying SiO2 layer was selectively etched
using a CHF3 RIE. Because the metal Ag was slowly eroded away during the RIE,
slightly tapered NRAs can be created on the MQW solar cells through prolonged
etching. Using the CHF3 gas for RIE maintains a smoth surface finish. The average
lengths of the NRAs are around 230 nm, and the diameters are in the range of 50-100
nm, which are controlled by annealing times and temperatures for Ag thin films.
According to Fig. 3.2(a), the area density of the NRAs, defined as the number of
nanorods per unit area, is approximately 1.5×1010 cm-2. The coverage of NRAs is
characterized by the fill factor, which is defined by the area ratio of NRAs to the
entire substrate surfaces using the top-view SEM images of Fig. 3.2(a). The fill factor
of the NRA surface is 0.68. The determined fill factor will be used later for
calculating effective refractive index (neff). It is worth mentioning that the dependence
of AR properties on geometric features for nanorods has been widely studied.[9,19,20]In
particular, it is found that increasing the diameter and the length of the naorods leads
to reduced surface reflectance and improved omnidirectionality. However, it is also
found that the over-lengthened nanorods may lead to the degraded Jsc despite their
lowest reflection, which is attributed to the excessive absorption of the lengthened
nanorods.[6] Accordingly, an appropriate length of nanorods is needed for maximizing
the Jscenhancement. The robustness of the SiO2nanorods can be confirmed during the
fabrication process. For the removal of Ag nanoparticles after etching, the SiO2
nanorods were immersed in 70 % HNO3solvent at 40 ℃－50 ℃ for a few minutes.
Fig. 3.2 shows that the geometries of the overall NRAs are still well preserved after
HNO3 treatment. These results indicate that the rod-like nanostructure can be
fabricated with controllable manner and good yield for achieving low-cost solar
devices.
Figure 3.2 (a) Top-view and (b) 45-degree tilted-view SEM images of the SiO2
NRAs.
Fig. 3.3 is the specular reflection spectra with the wavelengths ranging from 330
to 570 nm. The surface reflectance of the SiO2NRA surface is much lower than that
on the bare surface for the entire studied wavelengths. The significantly low
reflectance at 330-450 nm is particularly important to the efficiency enhancement of
the MQW solar cell, which will be shown in the EQE measurements later. The
suppressed reflection by NRA structures is attributed to several effects. As the
incident wavelength is much higher than the geometric size of NRAs at the long
wavelength region, the reduced reflectance can be explained by the effective medium
theory. Due to the subwavelength dimensions and the refractive index (around 1.56 at
400 nm) of SiO2 [21]
, the SiO2 NRAs behave like an effective medium with the neff
between the refractive indices of air (~1) and ITO (~2.3) layer. neff of the NRAs can
be estimated by the equation[22]:
neff={fnSiOq
2+ [1- fnairq ]}1/q (1)
where q is 2/3,
SiO2
n and nair are respectively the refractive indices of SiO2 and air,
and f is the fill factor. The calculated neffis 1.37 with f = 0.68. Such low neff is rarely
found in natural materials, but is of great importance to achieve broadband AR at the
air/device interface. For the short incident wavelength region, the gaps between NRAs
lead to the light trapping effect. As the light impinges on the nanostructrued surface, it
diffracts to several beams with different diffraction angles, and then re-bounces
between NRAs until the light propagates into MQW regions at a high angle from the
normal, which increases the opportunity of optical absorption by the MQWs due to
the increase of light propagation paths in MQW regions. This speculation will be
further demonstrated by PV measurements and simulations later. We note that the
subwavelength nanostructures result in the suppressed reflectance not only over a
wide range of wavelengths, but also a wide range of incident angles, which can be
clarified by the effective medium theory and the light trapping effect.[23-25]
Figure 3.3 Specular reflectance measured on the MQW solar cells with bare and SiO2
NRA surfaces.
In order to reveal the light propagation across the interfaces, the distributions of
electromagnetic fields within the device structures were simulated by FDTD analysis.
We modeled the devices with bare and SiO2NRA surfaces. Fig. 3.4(a)-(b) visualize
the time-averaged TE-polarized electric field intensity distributions, |Ey|, for the
MQW solar cells with two surface conditions at 380 nm as considering n and k of all
materials.[26] All of the calculated values are normalized to the ones of the excitation
source. It can be seen that the field intensities inside the InGaN MQW region are
enhanced with SiO2NRAs. In the inset of Fig. 3.4(b), where the region of SiO2NRAs
is enlarged, one can see strong field intensity between nanorods, indicating that the
nanorods behave as effective scattering centers. The strong scattering within the
NRAs therefore prevents the incident waves bouncing back to the air and prolongs the
optical path, giving rise to the light-trapping effect. Fig. 3.4(c) shows the normalized
optical power, integrated over the MQW region, as a function of times for bare and
NRA surfaces. The steady-state power values for the cells with bare and NRA
surfaces are 0.744 and 0.824, respectively. The results indicate that the NRAs increase
the number of the photons reaching the MQW region, which benefits the conversion
efficiency of the MQW solar cell.
Figure 3.4 Time-averaged, normalized TE electric field distribution, |Ey|, simulated by FDTD analysis for the MQW solar cells with (a) bare and (b) SiO2NRA surfaces with a 380-nm incident light. (c) Normalized optical power, integrated over the MQW region, as a function of times for two kinds of solar cells.
Fig. 3.5(a) presents the spectra of the EQE for the MQW solar cells with two
kinds of surface conditions, showing the influence of the light-harvesting SiO2NRAs
on PV performances. The EQE for MQW solar cells with SiO2NRAs is improved at
370-440 nm, which agrees with the reduced reflectance seen in Fig. 3.3. Fig. 3.5(b)
shows the J-V curves of the two kinds of solar cells. The PV data summarized from Figure 3.4 Time-averaged, normalized TE electric field distribution, |Ey|, simulated by FDTD analysis for the MQW solar cells with (a) bare and (b) SiO2NRA surfaces with a 380-nm incident light. (c) Normalized optical power, integrated over the MQW region, as a function of times for two kinds of solar cells.
Fig. 3.5(a) presents the spectra of the EQE for the MQW solar cells with two
kinds of surface conditions, showing the influence of the light-harvesting SiO2NRAs
on PV performances. The EQE for MQW solar cells with SiO2NRAs is improved at
370-440 nm, which agrees with the reduced reflectance seen in Fig. 3.3. Fig. 3.5(b)
shows the J-V curves of the two kinds of solar cells. The PV data summarized from Figure 3.4 Time-averaged, normalized TE electric field distribution, |Ey|, simulated by FDTD analysis for the MQW solar cells with (a) bare and (b) SiO2NRA surfaces with a 380-nm incident light. (c) Normalized optical power, integrated over the MQW region, as a function of times for two kinds of solar cells.
Fig. 3.5(a) presents the spectra of the EQE for the MQW solar cells with two
kinds of surface conditions, showing the influence of the light-harvesting SiO2NRAs
on PV performances. The EQE for MQW solar cells with SiO2NRAs is improved at
370-440 nm, which agrees with the reduced reflectance seen in Fig. 3.3. Fig. 3.5(b)
shows the J-V curves of the two kinds of solar cells. The PV data summarized from
the J-V curve is listed in Table 3.1. The agreement between the results in Fig. 3.3, Fig.
3.4, and Fig. 3.5 indicates that the light-harvesting SiO2 NRAs increase the optical
transmission and light propagation paths through the device surface, and hence
enhance the light absorption in the MQW region, giving rise to the additional
photocurrent of the MQW devices with NRA surfaces. The enhanced Jschence boosts
the conversion efficiency from 0.37 % to 0.45 %, which is an efficiency improvement
of ∼21 %.
Figure 3.5 (a) EQEs and (b) J–V characteristics measured on the MQW solar cells with bare and SiO2NRA surfaces.
(a) (b)
Table 3.1 Device characteristics of the MQW solar cells with bare and NRAs surfaces.
Note that FF stands for the fill factor of solar cells, which is defined as the ratio of the actual maximum obtainable power, to the product of Jsc and open circuit voltage (Voc).
AR Layers Jsc(mA/cm2) Voc(V) FF (%) η (%)
Bare 0.71 1.95 27.28 0.37
SiO2NARs 0.76 1.93 31.04 0.45
§ 3-4 Summary
Using self-assembled Ag nanoparticles as an etching mask and RIE method, we
successfully fabricated low-cost light-trapping SiO2 NRAs to improve the optical
absorption of InGaN MQW solar cells. The light-trapping SiO2NRA layer increases
the EQE of the solar cell mostly at 370-440 nm, corresponding to the improvement of
conversion efficiency by up to ~21 %. The superior AR performance of the SiO2
NRAs is attributed to the subwavelength dimensions and the nanorod-structured
geometry, effectively suppressing the surface reflection at the wavelengths from
330–570 nm via the light trapping effect and the graded neff, which have been
demonstrated by PV measurements and simulations. Presented concepts and
manufacturing techniques for light-harvesting nanostructures would be a viable way
to boost the efficiency for a variety of PV devices.
§ 3-5 References
[1] K. Y. Lai, G. J. Lin, Y.-L. Lai, Y. F. Chen, J. H. He, Effect of indium fluctuation on the photovoltaic characteristics of InGaN/GaN multiple quantum well solar cells, Appl. Phys. Lett. 96 (2010) 081103.
[2] K. Y. Lai, G. J. Lin, Y.-L. Lai, J. H. He, Origin of hot carriers in InGaN-based quantum-well solar cells, IEEE Electron Dev. Lett. 32 (2011) 179–181.
[3] J. Wu, W. Walukiewicz, K. M. Yu, W. Shan, J. W. Ager, E. E. Haller, H. Lu, W.
J. Schaff, W. K. Metzger, S. Kurtz, Superior radiation resistance of In1−xGaxN alloys: full-solar-spectrum photovoltaic material system, J. Appl. Phys. 94, (2003) 6477–6482.
[4] K. W. J. Barnham, G. Duggan, A new approach to high-efficiency multi-band-gap solar cells, J. Appl. Phys. 67 (1990) 3490–3493.
[5] E. Matioli, C. Neufeld, M. Iza, S. C. Cruz, A. A. Al-Heji, X. Chen, R. M.
Farrell, S. Keller, S. DenBaars, U. Mishra, S. Nakamura, J. Speck, C. Weisbuch, High internal and external quantum efficiency InGaN/GaN solar cells, Appl.
Phys. Lett. 98 (2011) 021102.
[6] G. J. Lin, K. Y. Lai, C. A. Lin, Y.-L. Lai, J. H. He, Efficiency enhancement of InGaN-based multiple quantum well solar cells employing antireflective ZnO nanorod arrays, IEEE Electron Dev. Lett. 32 (2011) 1104–1106.
[7] I. M. Pryce, D. D. Koleske, A. J. Fischer, H. A. Atwater, Plasmonic nanoparticle enhanced photocurrent in GaN/InGaN/GaN quantum well solar cells, Appl. Phys. Lett. 96 (2010) 153501.
[8] Y. Kuwahara, T. Fujii, T. Sugiyama, D. Iida, Y. Isobe, Y. Fujiyama, Y. Morita, M. Iwaya, T. Takeuchi, S. Kamiyama, I. Akasaki, H. Amano, GaInN-based solar cells using strained-layer GaInN/GaInN superlattice active layer on a
freestanding GaN substrate, Appl. Phys. Express 4 (2011) 021001.
[9] Y. C. Chao, C. Y. Chen, C. A. Lin, Y. A. Dai, J. H. He, Antireflection effect of ZnO nanorod arrays, J. Mater. Chem. 20 (2010) 8134–8138.
[10] K. Y. Lai, Y. R. Lin, H. P. Wang, J. H. He, Synthesis of anti-reflective and hydrophobic Si nanorod arrays by colloidal lithography and reactive ion etching, Cryst. Eng. Comm. 13 (2011) 1014–1017.
[11] P. Beckman, A. Spizzichno, The scattering of electromagnetic waves from rough surfaces, Pergamon, Oxford, 1963.
[12] P. B. Clapham, M. C. Hutley, Reduction of lens reflexion by the ‘moth eye’
principle, Nature 244 (1973) 281–282.
[13] L. Li, T. Y. Zhai, H. B. Zeng, X. S. Fang, Y. Bando, D. Golberg, Polystyrene sphere-assisted one-dimensional nanostructure arrays: synthesis and applications, J. Mater. Chem. 21 (2011) 40–56.
[14] X. D. Wang, E. Graugnard, J. S. King, Z. L. Wang, Large-scale fabrication of ordered nanobowl arrays, Nano Lett. 4 (2004) 2223–2226.
[15] Z. Fan, J. C. Ho, Self-assembly of one-dimensional nanomaterials for cost-effective photovoltaics, Int. J. Nanoparticles 4 (2011) 164–183.
[16] Z. Fan, R. Kapadia, P. W. Leu, X. Zhang, Y.-L. Chueh, K. Takei, K. Yu, A.
Jamshidi, A. A. Rathore, D. J. Ruebusch, M. Wu, A. Javey, Ordered arrays of dual-diameter nanopillars for maximized optical absorption, Nano Lett. 10 (2010) 3823–3827.
[17] Y. J. Lee, D. S. Ruby, D. W. Peters, B. B. McKenzie, J. W. P. Hsu, ZnO nanostructures as efficient antireflection layers in solar cells, Nano Lett. 8 (2008) 1501–1505.
[18] H. C. Chang, K. Y. Lai, Y. A. Dai, H. H. Wang, C. A. Lin and J. H. He,
Nanowire arrays with controlled structure profiles for maximizing optical collection efficiency, Energy Environ. Sci. 4 (2011) 2863–2869.
[19] Y. R. Lin, H. P. Wang, C. A. Lin and J. H. He, "Surface profile-controlled close-packed Si nanorod arrays for self-cleaning antireflection coatings," J.
Appl. Phys. 106 (2009) 114310.
[20] Y. A. Dai, H. J. Chang, K. Y. Lai, C. A. Lin, R. J. Chung, G. R. Lin and J. H. He,
"Subwavelength Si nanowire arrays for self-cleaning antireflection coatings," J.
Mater. Chem. 20 (2010) 10924–10930.
[21] E. D. Palik, Handbook of Optical Constants of Solids, Academic Press.
[22] D. G. Stavenga, S. Foletti, G. Palasantzas, K. Arikawa, Light on the moth-eye corneal nipple array of butterflies, Proc. R. Soc. B 273 (2006) 661–667.
[23] J. Q. Xi, M. F. Schubert, J. K. Kim, E. F. Schubert, M. Chen, S. Y. Lin, W. Liu, J. A. Smart, Optical thin-film materials with low refractive index for broadband elimination of Fresnel reflection, Nat. Photonics 1 (2007) 176–179.
[24] J. Q. Xi, J. K. Kim, E. F. Schubert, Silica nanorod-array films with very low refractive indices, Nano Lett. 5 (2005) 1385–1387.
[25] S. L. Diedenhofen, G. Vecchi, R. E. Algra, A. Hartsuiker, O. L. Muskens, G.
Immink, E. P. A. M. Bakkers, W. L. Vos, J. G. Rivas, Broad-band and omnidirectional antireflection coatings based on semiconductor nanorods, Adv.
Mater. 21 (2009) 973–978.
[26] S. Laux, N. Kaiser, A. Zoller, R. Gotzelmann, H. Lauth, H. Bernitzki, Room-temperature deposition of indium tin oxide thin films with plasma ion-assisted evaporation, Thin Solid Films 335 (1998) 1–5.