Role of surface recombination in affecting the efficiency of nanostructured thin-film solar cells Yun Da1 and Yimin Xuan1,2,* 1
School of Energy and Power Engineering, Nanjing University of Science and Technology, Nanjing 210094, China School of Energy and Power Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing 210016, China * [email protected]
2
Abstract: Nanostructured light trapping is a promising way to improve the efficiency in thin-film solar cells recently. In this work, both the optical and electrical properties of thin-film solar cells with 1D periodic grating structure are investigated by using photoelectric coupling model. It is found that surface recombination plays a key role in determining the performance of nanostructured thin-film solar cells. Once the recombination effect is considered, the higher optical absorption does not mean the higher conversion efficiency as most existing publications claimed. Both the surface recombination velocity and geometric parameters of structure have great impact on the efficiency of thin-film solar cells. Our simulation results indicate that nanostructured light trapping will not only improve optical absorption but also boost the surface recombination simultaneously. Therefore, we must get the tradeoffs between optical absorption and surface recombination to obtain the maximum conversion efficiency. Our work makes it clear that both the optical absorption and electrical recombination response should be taken into account simultaneously in designing the nanostructured thin-film solar cells. ©2013 Optical Society of America OCIS codes: (040.5350) Photovoltaic; (050.0050) Diffraction and gratings; (310.0310) Thin films; (040.6040) Silicon; (290.1990) Diffusion.
References and links 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.
Y. Humakawa, “Thin-Film Solar Cells, Next Generation Photovoltaics and Its Applications. (Springer, 2004). S. J. Jang, Y. M. Song, C. I. Yeo, C. Y. Park, J. S. Yu, and Y. T. Lee, “Antireflective property of thin film a-Si solar cell structures with graded refractive index structure,” Opt. Express 19(S2 Suppl 2), A108–A117 (2011). L. Yang, Y. Xuan, and J. Tan, “Efficient optical absorption in thin-film solar cells,” Opt. Express 19(S5 Suppl 5), A1165–A1174 (2011). Y. A. Akimov, W. S. Koh, and K. Ostrikov, “Enhancement of optical absorption in thin-film solar cells through the excitation of higher-order nanoparticle plasmon modes,” Opt. Express 17(12), 10195–10205 (2009). Y. A. Akimov and W. S. Koh, “Design of plasmonic nanoparticles for efficient subwavelength light trapping in thin-film solar cells,” Plasmonics 6(1), 155–161 (2011). H. A. Atwater and A. Polman, “Plasmonics for improved photovoltaic devices,” Nat. Mater. 9(3), 205–213 (2010). R. A. Pala, J. White, E. Barnard, J. Liu, and M. L. Brongersma, “Design of plasmonic thin-film solar cells with broadband absorption enhancement,” Adv. Mater. 21(34), 3504–3509 (2009). V. E. Ferry, M. A. Verschuuren, H. B. T. Li, E. Verhagen, R. J. Walters, R. E. I. Schropp, H. A. Atwater, and A. Polman, “Light trapping in ultrathin plasmonic solar cells,” Opt. Express 18(S2 Suppl 2), A237–A245 (2010). M. A. Sefunc, A. K. Okyay, and H. V. Demir, “Volumetric plasmonic resonator architecture for thin-film solar cells,” Appl. Phys. Lett. 98(9), 093117 (2011). J. Buencuerpo, L. E. Munioz-Camuniez, M. L. Dotor, and P. A. Postigo, “Optical absorption enhancement in a hybrid system photonic crystal - thin substrate for photovoltaic applications,” Opt. Express 20(S4 Suppl 4), A452–A464 (2012). Y. Park, E. Drouard, O. El Daif, X. Letartre, P. Viktorovitch, A. Fave, A. Kaminski, M. Lemiti, and C. Seassal, “Absorption enhancement using photonic crystals for silicon thin film solar cells,” Opt. Express 17(16), 14312– 14321 (2009). W. Wu and R. Magnusson, “Total absorption of TM polarized light in a 100 nm spectral band in a nanopatterned thin a-Si film,” Opt. Lett. 37(11), 2103–2105 (2012). E. R. Martins, J. Li, Y. Liu, J. Zhou, and T. F. Krauss, “Engineering gratings for light trapping in photovoltaics: The supercell concept,” Phys. Rev. B. 86(4), 041404 (2012).
#197209 - $15.00 USD Received 6 Sep 2013; revised 15 Oct 2013; accepted 17 Oct 2013; published 25 Oct 2013 (C) 2013 OSA 4 November 2013 | Vol. 21, No. S6 | DOI:10.1364/OE.21.0A1065 | OPTICS EXPRESS A1065
14. M. G. Deceglie, V. E. Ferry, A. P. Alivisatos, and H. A. Atwater, “Design of nanostructured solar cells using coupled optical and electrical modeling,” Nano Lett. 12(6), 2894–2900 (2012). 15. X. Li, N. P. Hylton, V. Giannini, K. Lee, N. J. Ekinsdaukes, and S. A. Maier, “Bridging electromagnetic and carrier transport calculations for three-dimensional modeling of plasmonic solar cells,” Opt. Express 19(S4), A888–A896 (2011). 16. A. Deinega, S. Eyderman, and S. John, “Coupled optical and electrical modeling of solar cell based on conical pore silicon photonic crystals,” J. Appl. Phys. 113(22), 224501 (2013). 17. S. Yu, F. Roemer, and B. Witzigmann, “Analysis of surface recombination in nanowire array solar cells,” J. Photon. Energy 2(1), 028002 (2012). 18. G. Gomard, X. Meng, E. Drouard, K. E. Hajjam, E. Gerelli, R. Peretti, A. Fave, R. Orobtchouk, M. Lemiti, and C. Seassal, “Light harvesting by planar photonic crystals in solar cells: the case of amorphous silicon,” J. Opt. 14(2), 024011 (2012). 19. A. Taflove and S. C. Hagness, Computational Electrodynamics: the Finite-Difference Time-Domain Method (Artech, 2005). 20. AM1, 5 solar spectrum irradiance data: http://rredc.nrel.gov/solar/spectra/am1.5. 21. S. Selberherr, Analysis and Simulation of Semiconductor Devices (Springer, 1984). 22. J. Nelson, The Physics of Solar Cells (Imperial College, 2003). 23. W. E. I. Sha, W. C. H. Choy, Y. Wu, and W. C. Chew, “Optical and electrical study of organic solar cells with a 2D grating anode,” Opt. Express 20(3), 2572–2580 (2012). 24. S. Chuang, Physics of Optoelectronic Devices (Wiley, 1995). 25. E. D. Palik, Handbook of Optical Constants of Solids (Academic Press, New York, 1985). 26. L. G. Jeffery, Handbook of Photovoltaic Science and Engineering (Antonio Luque, 2004). 27. T. Markvart and L. Castaner, Practical Handbook of Photovoltaics: Fundamentals and Applications (Elsevier Advanced Technology, 2003). 28. F. Wang, H. Yu, J. Li, S. Wong, X. W. Sun, X. Wang, and H. Zheng, “Design guideline of high efficiency crystalline Si thin film solar cell with nanohole array textured surface,” J. Appl. Phys. 109(8), 084306 (2011). 29. O. Demichel, V. Calvo, A. Besson, P. Noé, B. Salem, N. Pauc, F. Oehler, P. Gentile, and N. Magnea, “Surface recombination velocity measurements of efficiently passivated gold-catalyzed silicon nanowires by a new optical method,” Nano Lett. 10(7), 2323–2329 (2010).
1. Introduction Thin-film solar cells are nowadays considered as one of the most promising candidates for future generation of photovoltaics due to their lower production costs and a better adaptability to a wide range of structures and equipments compared to the bulk solar cells [1]. Unfortunately, a primarily unavoidable drawback of thin-film solar cells is the poor optical absorption, which limits the conversion efficiencies of the cells. This limitation has motivated the researchers to develop light trapping technology, which becomes the hotspots in solar cell research and concerns both the inorganic and organic solar cells. Up to date, much effort has been devoted to designing light trapping structures in the realm of thin-film solar cells. Various novel structures have being proposed to enhance light harvesting in solar cells including graded refractive index antireflective coating [2], metallic nanoparticles [3–6], periodic metallic gratings [7–9], photonic crystals [10, 11] and so on. Among these promising nanophotonic light trapping structures, patterning the active layer itself as a one-dimensional (1D) periodic grating is one of the most popular approaches because it is simple and easy to be fabricated. Such nanophotonic approaches can offer the ability to improve light trapping through antireflection effects and excited resonance effects that originate from the excitation of guided or leaky modes inside the active layer [12, 13]. However, most theoretical literatures of light trapping methods take into account only for optical effects. The carrier recombination processes are ignored. It is important to note that solar cell is a photoelectric coupling device, the electrical effect is as important as optical effect. Thus, it is necessary to consider both the optical absorption of the device and the electrical device response in designing the thin-film solar cells [14–17]. In this work, we use optical simulations coupled with electrical device simulations by solving the Maxwell’s equations and semiconductor equations (Poisson, continuity, and driftdiffusion equations) to enable the simultaneous consideration of optical and electrical effect when investigating the performance of 1D periodic grating structure thin-film solar cells. For periodic grating structure, surface recombination becomes a dominant concern due to its large surface to volume ratio. Hence, the focus of this article is to investigate the effect of the surface recombination on the efficiency of thin-film solar cells. At the same time, radiative recombination, Shockley-Read-Hall recombination and Auger recombination are also #197209 - $15.00 USD Received 6 Sep 2013; revised 15 Oct 2013; accepted 17 Oct 2013; published 25 Oct 2013 (C) 2013 OSA 4 November 2013 | Vol. 21, No. S6 | DOI:10.1364/OE.21.0A1065 | OPTICS EXPRESS A1066
included in calculating the conversion efficiency of solar cells. The objective of this article is to demonstrate that the higher optical absorption does not mean higher conversion efficiency for nanostructured thin-film solar cells in consideration of recombination effect. In other words, it is insufficient for one to predict the performance of nanostructured thin-film solar cells when taking account only for the light trapping effect. In their work [18], Gomard et al were aware of such an effect. They demonstrated that different configurations possessing the same integrated absorption do not present the same power conversion efficiency due to Shockley-Read-Hall recombination even when the surface recombination was ignored. Here we demonstrate that even a configuration possessing a higher integrated absorption can finally end with a lower conversion efficiency once the surface recombination is introduced. Therefore, the article is structured as follows. The brief description of our theoretical model is given in section 2. Based on the photoelectric coupling simulation method, the results and discussion are presented in section 3. The impact of surface recombination velocity, geometric parameters of the structure is discussed in detail. The final conclusion is given in section 4. 2. Theoretical model 2.1 Optical simulation The optical absorption of thin-film solar cells is evaluated by solving Maxwell’s equations through the Finite Difference Time Domain (FDTD) algorithm, which is widely adopted for rigorous electromagnetic propagation calculation [19]. In the simulation, plane waves are normally incident on the cells. The perfectly matched layer (PML) absorbing boundary condition is implemented on the top and bottom boundaries while the periodic boundary condition is imposed on the left and right boundaries within the simulation unit. After obtaining the spatial distribution of the electromagnetic field, the optical generation rate can be expressed as Gopt (r ) =
λmax
ε " E (r , λ ) 2
λmin
2
I AM 1.5 (λ )d λ
(1)
where ε " is the imaginary part of the permittivity of the semiconductor material, r is the space coordinates, λmax is the maximum wavelength corresponding to the bandgap of semiconductor, λmin is the minimum wavelength of the solar radiation, E is the electric field, is the reduced Planck constant and I AM 1.5 is the global standard solar irradiance spectrum of AM1.5G [20]. The sunlight is un-polarized and can be treated as the superposition of two orthogonally polarized waves. Hence, the optical absorption and the optical generation rate is the average value for TE and TM polarizations. In the existing publications, in order to quantitatively illustrate the overall optical light trapping performance of the thin-film solar cells, the maximum achievable photocurrent density is widely used and defined as [10] J sc max =
λmax
λmin
e
λ hc
A(λ ) I AM 1.5 (λ )d λ
(2)
where h is the Planck constant, c is the speed of light in vacuum and A(λ ) is the optical absorption of the active layer. Obviously, the carrier recombination processes are not taken into account. This expression quantitatively reveals the response of light trapping effect.
2.2 Electrical simulation The electrical simulation is conducted based on the semiconductor equations, consisting of Poisson, continuity, and drift-diffusion equations [21–24]. All these equations would be solved simultaneously through finite element method and can be described as
#197209 - $15.00 USD Received 6 Sep 2013; revised 15 Oct 2013; accepted 17 Oct 2013; published 25 Oct 2013 (C) 2013 OSA 4 November 2013 | Vol. 21, No. S6 | DOI:10.1364/OE.21.0A1065 | OPTICS EXPRESS A1067
∇ ⋅ (ε∇φ ) = q( p − n + N D − N A )
(3)
∇ ⋅ J n = − q(G − R )
(4)
∇ ⋅ J p = q (G − R )
(5)
J n = − q μn n∇φ + qDn ∇n
(6)
J p = − q μ p p∇φ − qD p ∇p
(7)
where ε is the dielectric constant of semiconductor, q is the electron charge, φ is the electrical potential, n( p) is the electron (hole) concentration, N D ( N A ) is the donor (acceptor) doping concentration, J n ( J p ) is the current density of electron (hole), G is the optical generation rate obtained from optical simulation and R is the carrier recombination rate including radiative recombination, Shockley-Read-Hall recombination, Auger recombination and surface recombination. In addition, μn ( μ p ) is the electron (hole) mobility and Dn ( D p ) is the electron (hole) diffusion coefficient abided by Einstein relations with mobility: k BT Dn = q μ n D = k BT μ p p q
(8)
where k B is the Boltzmann constant and T is the operating temperature. The total carrier recombination rate ( R ) can be divided into four types including radiative recombination ( Rrad ), Shockley-Read-Hall recombination ( RSRH ), Auger recombination ( RAug ) and surface recombination ( Rsurf ). For each recombination mechanism, the corresponding models are described in the following equations. R = Rrad + RSRH + RAug + Rsurf
(9)
Rrad = B (np − ni2 )
(10)
np − ni2 τ p (n + n1 ) + τ n ( p + p1 )
(11)
RAug = (Cn 0 n + C p 0 p )(np − ni2 )
(12)
RSRH =
Rsurf =
np − ni2 1 1 (n + n1s ) + ( p + p1s ) Sp Sn
(13)
where B is the radiative recombination coefficient, τ n (τ p ) is the electron (hole) lifetime, n1 ( p1 ) is the electron (hole) concentration in the trap states, Cn 0 (C p 0 ) is the electron Auger
recombination coefficient, ni is the intrinsic carrier concentration and S n ( S p ) is the surface recombination velocity of electron (hole). It is clear to find that the model of surface recombination rate is similar to Shockley-Read-Hall recombination but differs slightly since the surface recombination just occurs on the cell surface. For the electrical simulation, we #197209 - $15.00 USD Received 6 Sep 2013; revised 15 Oct 2013; accepted 17 Oct 2013; published 25 Oct 2013 (C) 2013 OSA 4 November 2013 | Vol. 21, No. S6 | DOI:10.1364/OE.21.0A1065 | OPTICS EXPRESS A1068
focus on the region of the semiconductor. The surface recombination is implemented on the upper and lower interface of the semiconductor. The right and left interface is truncated by adopting the Neumann boundary condition, which is widely used in the semiconductor device simulation [23]. 2.3 Description of simulation structures and parameters
Fig. 1. Schematics of the simulation structure of thin-film solar cells. (a) First reference cell with planar surface. (b) Second reference cell with an optimal antireflection coating. (c) 1D periodic grating of light trapping structure. The value of the thickness is fixed as d1 = 0.4 µm.
The simulation structure of thin-film solar cells is schematically shown in Fig. 1. The active layer is composed of crystalline silicon (c-Si), which possesses the advantages of nontoxicity, abundance and mature processing. To reduce the loss of the incident light transmission, the Ag back reflector is widely used. The thickness of the c-Si film d1 and Ag back reflector layer thickness d 2 are fixed as 0.4 μm and 0.1 μm respectively. The planar cell and the cell with antireflection coating are considered as the reference for comparison. For the cell with antireflection coating, silica is used as the antireflection coating with an optimal thickness t = 0.08 μm. For 1D periodic grating structure, the design parameters include the grating period Λ , filling factor of the semiconductor f and the grating depth h . These geometric parameters not only affect the optical absorption but also influence the electrical properties. For investigating the optical properties of thin-film solar cells, the optical parameters of material are referenced from the literature [25]. The incident wavelength range is selected from 300 nm to 1100 nm, where the upper limit corresponding to the band-gap of c-Si. In terms of electrical aspect, the carrier mobility μn (μ p ) for electron (holes) is considered as the function of doping level and can be well approximated by [26]
μn = 92 +
μ p = 54.3 +
1268 cm 2 /Vs N D + N A 0.91 1+ ( ) 1.3 × 1017 406.9 cm 2 /Vs N D + N A 0.88 1+ ( ) 2.35 × 1017
(14)
(15)
The parameters of recombination are referenced from literature [27] and listed in Table 1. In addition, Table 1 includes the key parameters for electrical simulation as well. It is wellknown that the device configuration plays a key role in an accurate semiconductor device simulation. In our work, a planar junction configuration is adopted with an n-type emitter and
#197209 - $15.00 USD Received 6 Sep 2013; revised 15 Oct 2013; accepted 17 Oct 2013; published 25 Oct 2013 (C) 2013 OSA 4 November 2013 | Vol. 21, No. S6 | DOI:10.1364/OE.21.0A1065 | OPTICS EXPRESS A1069
p-type base as found as in reference [28]. The emitter width is fixed at 40 nm and the doping concentrations N D and N A are respectively fixed at 1019 and 1016 cm -3 . Table 1. Typical parameters for silicon thin-film solar cells in the simulation Parameters
ε
Values 11.7
Unit
Eg
1.12
eV
NA ND
16
10
cm -3
1019
cm -3
μn μp
115.886
cm 2 /Vs
68.7524
cm 2 /Vs
B
1.8 × 10−15 3.3 × 10 −7
cm 3 /s s
8.2 × 10−5
s
τn τp Cn 0 Cp0
−31
cm 6 /s
9.9 × 10 −32
cm 6 /s
Sn , S p
102 − 105
T
300
cm/s K
2.8 × 10
3. Results and discussion
3.1 Impact of surface recombination velocity Figure 2(a) displays the spectral absorption for the grating structured thin-film solar cells in comparison with two reference cells including planar cell and antireflective cell. The geometric parameters of the grating structure with maximum integrated optical absorption are Λ = 0.5 μm, h = 0.16 μm and f = 0.5 . It is clear that the optical absorption of grating structure increases remarkably in the long wavelength range from 550 nm to 1100 nm. Figure 2(b) shows the maximum achievable photocurrent density to show light trapping effect for all the investigated structures. It is obvious that the grating structure exhibits the best light trapping performance. According to traditional viewpoint, the higher optical absorption means the higher conversion efficiency for the thin-film solar cells. In other words, the grating structure thin-film solar cells have higher conversion efficiency comparing with the reference cells. However, the carrier recombination processes are not taken into account in the preceding discussion. As is known to all, nanostructured thin-film solar cells have much higher surface to volume ratio compared to bulk solar cells. Thus, surface recombination plays an extremely important role in determining the performance of thin-film solar cells. The impact of the surface recombination velocity will be discussed in the following.
#197209 - $15.00 USD Received 6 Sep 2013; revised 15 Oct 2013; accepted 17 Oct 2013; published 25 Oct 2013 (C) 2013 OSA 4 November 2013 | Vol. 21, No. S6 | DOI:10.1364/OE.21.0A1065 | OPTICS EXPRESS A1070
Fig. 2. Results of three structure c-Si thin-film solar cells. (a) Absorption spectra in the c-Si layer. (b) Maximum achievable photocurrent density to show light trapping effect. (c) Optical generation rate. (d) Simulated J-V and P-V characteristics with different surface recombination velocity. The blue, red and black lines correspond to the results for grating structure, antireflective coating structure and planar structure, respectively.
Figure 2(c) shows the optical generation rate of three investigated structures, which acts as the input profile in the electrical simulation. It is noticeable that the optical generation rate calculated by FDTD departs significantly from that obtained with Beer-Lambert law. The current-voltage (J-V) curves of three investigated structures together with power-voltage (PV) curves under different surface recombination velocity are shown in Fig. 2(d). As shown in
#197209 - $15.00 USD Received 6 Sep 2013; revised 15 Oct 2013; accepted 17 Oct 2013; published 25 Oct 2013 (C) 2013 OSA 4 November 2013 | Vol. 21, No. S6 | DOI:10.1364/OE.21.0A1065 | OPTICS EXPRESS A1071
Fig. 2(d), the open-circuit voltage ( Voc ) and the short-circuit current density ( J sc ) are reduced with the increase of surface recombination velocity (SRV) for all the investigated structures. In addition, the difference of Voc and J sc between the grating structure and the reference cells becomes smaller with the increase of SRV. For the grating structure, for example, as the SRV S n = S p varies from 10 cm/s to 103 cm/s, the value of Voc and J sc varies from 0.700 V to 0.564 V and from 15.60 mA/cm2 to 5.64 mA/cm2, respectively. While for the antireflective cell, the value of Voc and J sc varies from 0.695 V to 0.565 V and from 11.71 mA/cm2 to 5.43 mA/cm2 if the SRV S n = S p varies from 10 cm/s to 103 cm/s. It validates that the SRV has great impact on Voc and J sc for nanostructured thin-film solar cells as expected. The surface area of grating structure is larger than that of the reference cells and so is the surface recombination. The boosted surface recombination has suppressed the influence of the optical absorption enhancement. Both Voc and J sc are the key parameters of determining the power conversion efficiency of the solar cells. Therefore, under the condition of strong surface recombination velocity, although the optical absorption of grating structure is higher, the power conversion efficiency (PCE) is probably smaller comparing to the reference cells.
Fig. 3. PCE of different configurations for c-Si thin-film solar cells with varied surface recombination velocity.
Here the PCE of different configurations as a function of the SRV is further studied in order to make this statement more clear. As shown in Fig. 3, it is obvious that the PCE of all the investigated structures is decreased when the SRV is increased. The larger SRV value always results in the lower PCE for the nanostructured thin-film solar cells. Moreover, it is found that the structure with a higher optical absorption does not always have a higher PCE when the SRV is considered, although all the investigated grating structures have a higher integrated optical absorption than the planar surfaces. For example, when the SRV S n = S p is
#197209 - $15.00 USD Received 6 Sep 2013; revised 15 Oct 2013; accepted 17 Oct 2013; published 25 Oct 2013 (C) 2013 OSA 4 November 2013 | Vol. 21, No. S6 | DOI:10.1364/OE.21.0A1065 | OPTICS EXPRESS A1072
10 cm/s, the value of PCE for the grating structure with geometric parameters Λ = 0.8 μm, h = 0.2 μm, f = 0.5 and the antireflective structure respectively corresponds to 7.07% and
6.35%, which is in accord with their optical absorption. Once the SRV S n = S p changes to 102 cm/s, however, the value of the PCE of the above-mentioned structured cells respectively changes to 4.74% and 5.10%, which is in contrary to their optical absorption. This phenomenon can be explained as follows. In the former case, the effect of optical absorption enhancement is greater than that of the surface recombination, which leads to increase in the PCE. For the later case, the effect of surface recombination is greater than that of optical absorption enhancement, leading to the drop of PCE. Thus, the value of the SRV plays an extremely important role in determining the PCE of the nanostructured thin-film solar cells. In addition, the grating structure with geometric parameters Λ = 0.4 μm, h = 0.1 μm, and f = 0.5 has the same surface area compared to the grating structure with geometric
parameters Λ = 0.8 μm, h = 0.2 μm, and f = 0.5 , so that they have the same surface recombination. The PCE of the former is always higher than that of the latter under the different values of SRV, corresponding to its higher optical absorption. The PCE of the grating structure with geometric parameters Λ = 0.5 μm, h = 0.16 μm, and f = 0.5 is the same as that of the grating structure with geometric parameters Λ = 0.4 μm, h = 0.1 μm, and f = 0.5 in the case of S n = S p = 10 2 cm/s. It can be attributed to the balance of higher optical absorption and higher surface recombination. Therefore, it can be concluded that the surface morphology (dominating the surface area and optical absorption) and surface passivation (determining the value of SRV) have great influence in PCE for nanostructured thin-film solar cells. It is emphasized that the higher optical absorption does not always mean the higher conversion efficiency for nanostructured thin-film solar cells in consideration of the recombination effect. The tradeoffs between optical absorption and surface recombination should be obtained in designing the nanostructured thin-film solar cells. As a key factor of determining the surface recombination, the accurate value of SRV is necessary in practice in designing the patterned nanostructured thin-film solar cells.
#197209 - $15.00 USD Received 6 Sep 2013; revised 15 Oct 2013; accepted 17 Oct 2013; published 25 Oct 2013 (C) 2013 OSA 4 November 2013 | Vol. 21, No. S6 | DOI:10.1364/OE.21.0A1065 | OPTICS EXPRESS A1073
3.2 Impact of geometric parameters of the structure
Fig. 4. Results of grating structure thin-film solar cells with different geometric parameters. (a) Absorption spectra. (b) Maximum achievable photocurrent density to show light trapping effect. (c) Simulated J-V and P-V characteristics under ideal condition. (d) Simulated J-V and P-V characteristics under considering recombination condition. (e) The contribution of each recombination process to the PCE of the nanostructured thin-film solar cells.
To further demonstrate our viewpoint, we choose two cases of grating structure to investigate their optical and electrical characteristics. The geometric parameters of Case1 are Λ = 0.5 μm, h = 0.2 μm, and f = 0.5 . While for Case2, the geometric parameters are Λ = 0.6 μm, h = 0.1 μm, and f = 0.5 . The SRV S n = S p = 100 cm/s is selected for investigating their electrical properties [16]. The optical properties including spectral absorption and maximum achievable photocurrent density are shown in Fig. 4(a) and Fig. 4(b) respectively. It is obvious that the Case1 manifests a better effect of optical absorption #197209 - $15.00 USD Received 6 Sep 2013; revised 15 Oct 2013; accepted 17 Oct 2013; published 25 Oct 2013 (C) 2013 OSA 4 November 2013 | Vol. 21, No. S6 | DOI:10.1364/OE.21.0A1065 | OPTICS EXPRESS A1074
than Case2. If the impact of recombination effect is ignored, the Case1 also shows a better performance of PCE than Case2 corresponding to its higher optical absorption as shown in Fig. 4(c). However, as shown in Fig. 4(d), when the recombination effect is considered, the Case2 shows a better performance of PCE than Case1 in contrary to its lower optical absorption. This interesting phenomenon which seems on the opposite of intuition can mainly attribute to surface recombination effect. The surface area of Case1 is larger than that of Case2 within the unit cell and so is the surface recombination. Although the optical absorption of thin-film solar cells increases, the recombination increases as well. As a result, the PCE of Case1 is lower than that of Case1 because the impact of surface recombination is stronger than that of the enhancement of optical absorption. Therefore, we should get tradeoffs between improving optical absorption and reducing surface recombination in designing nanostructured thin-film solar cells. In order to further investigate the contribution of each recombination process ( Rrad , RAug , RSRH , and Rsurf ) to the PCE of the solar cells, it is separately considered in the simulation. As illustrated in Fig. 4(e), it is obvious that the radiative recombination ( Rrad ) and Auger recombination ( RAug ) could be neglected in our computations for nanostructured thinfilm solar cells because they make almost no contribution to the PCE. Once the ShockleyRead-Hall recombination ( RSRH ) is considered separately, the value of PCE decreases from 12.69% to 10.08% for Case1 and from 11.75% to 9.05% for Case2. It is obvious that the PCE of Case1 is higher than that of Case2, corresponding to its higher optical absorption. However, the value of PCE is decreased from 12.69% to 5.97% for Case1 and from 11.75% to 6.38% for Case2 when surface recombination ( Rsurf ) is considered separately. It is clear that the Case2 shows the better performance of PCE than Case1 in contrary to its lower optical absorption. When all the recombination processes are considered simultaneously, the value of PCE is 5.64% for Case1 and 5.92% for Case2, corresponding to the results of considering Rsurf separately. Therefore, this can demonstrate that the surface recombination is the dominating recombination process for the nanostructured thin-film solar cells. The higher optical absorption does not always mean the higher power conversion efficiency for nanostructured thin-film solar cells due to the inevitable surface recombination.
#197209 - $15.00 USD Received 6 Sep 2013; revised 15 Oct 2013; accepted 17 Oct 2013; published 25 Oct 2013 (C) 2013 OSA 4 November 2013 | Vol. 21, No. S6 | DOI:10.1364/OE.21.0A1065 | OPTICS EXPRESS A1075
3.3 Optimization of the light trapping structure
Fig. 5. Parametric analysis of the optimized process for light trapping structure. (a) Maximum achievable photocurrent density and PCE vs Λ . (b) Maximum achievable photocurrent density and PCE vs h . (c) Maximum achievable photocurrent density and PCE vs f . (d) Electrically optimized J-V and P-V characteristics.
Since the optical and electrical performance of nanostructured thin-film solar cells is sensitive to geometric parameters, the optimization of light trapping structure becomes important. The parametric analysis of the optimized process for light trapping structure is shown in Fig. 5. During the process of the optimization, each geometric parameter is scanned in a predetermined range while the other geometric parameters are kept constant. According to the references [16, 29], the surface recombination velocity can be reduced to 100 cm/s after the passivation process, that is why we selected this typical value for our next calculations. For each parameter, both the maximum achievable photocurrent density and power conversion efficiency are plotted as a function of the geometric parameter value. The first parameter of analysis is the period Λ of the grating structure, which is varied from 0.2 μm to 0.8 μm. The other invariant parameters are h = 0.1 μm and f = 0.5 . As depicted in Fig. 5(a), the peak of maximum achievable photocurrent density is at Λ = 0.45 μm while the PCE peaks at Λ = 0.5 μm. The second parameter we analyzed is the depth h of the grating structure with the variety from 0.06 μm to 0.22 μm. The period Λ and filling factor f are kept constant as 0.5 μm and 0.5 respectively. It can be seen from Fig. 5(b) that the maximum achievable photocurrent density peaks at h = 0.16 μm while the peak of PCE is at h = 0.1 μm. Figure 5(a) agrees well with Fig. 5(b) that maximum optical absorption point is not consistent with the maximum power conversion efficiency point. This interesting phenomenon can be attributed to surface recombination effect. The maximum optical absorption point has larger surface area within the unit cell comparing to that of maximum power conversion efficiency point. Therefore, the maximum optical absorption and
#197209 - $15.00 USD Received 6 Sep 2013; revised 15 Oct 2013; accepted 17 Oct 2013; published 25 Oct 2013 (C) 2013 OSA 4 November 2013 | Vol. 21, No. S6 | DOI:10.1364/OE.21.0A1065 | OPTICS EXPRESS A1076
relatively strong surface recombination lead to the results that the PCE is not the biggest. It can be concluded that the surface recombination plays a key role in determining the performance of nanostructured thin-film solar cells. The third parameter we analyzed is the filling factor f . The filling factor f is varied from 0.2 to 0.7 while the period Λ and depth h are kept constant as 0.5 μm and 0.1 μm respectively. It can be seen from Fig. 5(c) that the maximum achievable photocurrent density peaks at f = 0.55 as well as the PCE. When the filling factor is varied, the surface area within the unit cell remains the same. As a result, maximum optical absorption point is consistent with the maximum power conversion efficiency point during the same surface recombination. Figure 5(d) shows the electrically optimized J-V and P-V characteristics of 1D periodic grating structure thin-film solar cells. The maximum PCE of the grating structure reached is 6.28% after the optimization, exhibiting 23% enhancement compared to the reference cell with the antireflective coating. 4. Conclusions
In conclusion, both the optical and electrical properties of thin-film cells with 1D periodic grating structure have been studied. The framework of photoelectric coupling model is developed to optimize the light trapping structures for thin-film solar cells. In contrast to most existing investigations, our simulation results indicate that the higher optical absorption does not always mean the higher conversion efficiency for nanostructured thin-film solar cells in consideration of recombination effect. In particular, surface recombination is the dominant recombination process and plays an important role in affecting the efficiency of nanostructured thin-film solar cells. The key parameters of nanostructured solar cells such as Voc and J sc are significantly affected by the surface recombination. It is critical to get the tradeoffs between enhancing optical absorption and reducing the surface area in designing nanostructured thin-film solar cells when using the same surface passivation technology. This work makes it clear that it may be incomprehensive for one to account only for the light trapping effect for designing high-performance solar cells and it will be helpful for designing high-performance nanostructured thin-film solar cells. Acknowledgment
We are grateful to the financial support from the National Natural Science Foundation of China (Grant No. 51336003).
#197209 - $15.00 USD Received 6 Sep 2013; revised 15 Oct 2013; accepted 17 Oct 2013; published 25 Oct 2013 (C) 2013 OSA 4 November 2013 | Vol. 21, No. S6 | DOI:10.1364/OE.21.0A1065 | OPTICS EXPRESS A1077
School of Energy and Power Engineering, Nanjing University of Science and Technology, Nanjing 210094, China School of Energy and Power Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing 210016, China * [email protected]
2
Abstract: Nanostructured light trapping is a promising way to improve the efficiency in thin-film solar cells recently. In this work, both the optical and electrical properties of thin-film solar cells with 1D periodic grating structure are investigated by using photoelectric coupling model. It is found that surface recombination plays a key role in determining the performance of nanostructured thin-film solar cells. Once the recombination effect is considered, the higher optical absorption does not mean the higher conversion efficiency as most existing publications claimed. Both the surface recombination velocity and geometric parameters of structure have great impact on the efficiency of thin-film solar cells. Our simulation results indicate that nanostructured light trapping will not only improve optical absorption but also boost the surface recombination simultaneously. Therefore, we must get the tradeoffs between optical absorption and surface recombination to obtain the maximum conversion efficiency. Our work makes it clear that both the optical absorption and electrical recombination response should be taken into account simultaneously in designing the nanostructured thin-film solar cells. ©2013 Optical Society of America OCIS codes: (040.5350) Photovoltaic; (050.0050) Diffraction and gratings; (310.0310) Thin films; (040.6040) Silicon; (290.1990) Diffusion.
References and links 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.
Y. Humakawa, “Thin-Film Solar Cells, Next Generation Photovoltaics and Its Applications. (Springer, 2004). S. J. Jang, Y. M. Song, C. I. Yeo, C. Y. Park, J. S. Yu, and Y. T. Lee, “Antireflective property of thin film a-Si solar cell structures with graded refractive index structure,” Opt. Express 19(S2 Suppl 2), A108–A117 (2011). L. Yang, Y. Xuan, and J. Tan, “Efficient optical absorption in thin-film solar cells,” Opt. Express 19(S5 Suppl 5), A1165–A1174 (2011). Y. A. Akimov, W. S. Koh, and K. Ostrikov, “Enhancement of optical absorption in thin-film solar cells through the excitation of higher-order nanoparticle plasmon modes,” Opt. Express 17(12), 10195–10205 (2009). Y. A. Akimov and W. S. Koh, “Design of plasmonic nanoparticles for efficient subwavelength light trapping in thin-film solar cells,” Plasmonics 6(1), 155–161 (2011). H. A. Atwater and A. Polman, “Plasmonics for improved photovoltaic devices,” Nat. Mater. 9(3), 205–213 (2010). R. A. Pala, J. White, E. Barnard, J. Liu, and M. L. Brongersma, “Design of plasmonic thin-film solar cells with broadband absorption enhancement,” Adv. Mater. 21(34), 3504–3509 (2009). V. E. Ferry, M. A. Verschuuren, H. B. T. Li, E. Verhagen, R. J. Walters, R. E. I. Schropp, H. A. Atwater, and A. Polman, “Light trapping in ultrathin plasmonic solar cells,” Opt. Express 18(S2 Suppl 2), A237–A245 (2010). M. A. Sefunc, A. K. Okyay, and H. V. Demir, “Volumetric plasmonic resonator architecture for thin-film solar cells,” Appl. Phys. Lett. 98(9), 093117 (2011). J. Buencuerpo, L. E. Munioz-Camuniez, M. L. Dotor, and P. A. Postigo, “Optical absorption enhancement in a hybrid system photonic crystal - thin substrate for photovoltaic applications,” Opt. Express 20(S4 Suppl 4), A452–A464 (2012). Y. Park, E. Drouard, O. El Daif, X. Letartre, P. Viktorovitch, A. Fave, A. Kaminski, M. Lemiti, and C. Seassal, “Absorption enhancement using photonic crystals for silicon thin film solar cells,” Opt. Express 17(16), 14312– 14321 (2009). W. Wu and R. Magnusson, “Total absorption of TM polarized light in a 100 nm spectral band in a nanopatterned thin a-Si film,” Opt. Lett. 37(11), 2103–2105 (2012). E. R. Martins, J. Li, Y. Liu, J. Zhou, and T. F. Krauss, “Engineering gratings for light trapping in photovoltaics: The supercell concept,” Phys. Rev. B. 86(4), 041404 (2012).
#197209 - $15.00 USD Received 6 Sep 2013; revised 15 Oct 2013; accepted 17 Oct 2013; published 25 Oct 2013 (C) 2013 OSA 4 November 2013 | Vol. 21, No. S6 | DOI:10.1364/OE.21.0A1065 | OPTICS EXPRESS A1065
14. M. G. Deceglie, V. E. Ferry, A. P. Alivisatos, and H. A. Atwater, “Design of nanostructured solar cells using coupled optical and electrical modeling,” Nano Lett. 12(6), 2894–2900 (2012). 15. X. Li, N. P. Hylton, V. Giannini, K. Lee, N. J. Ekinsdaukes, and S. A. Maier, “Bridging electromagnetic and carrier transport calculations for three-dimensional modeling of plasmonic solar cells,” Opt. Express 19(S4), A888–A896 (2011). 16. A. Deinega, S. Eyderman, and S. John, “Coupled optical and electrical modeling of solar cell based on conical pore silicon photonic crystals,” J. Appl. Phys. 113(22), 224501 (2013). 17. S. Yu, F. Roemer, and B. Witzigmann, “Analysis of surface recombination in nanowire array solar cells,” J. Photon. Energy 2(1), 028002 (2012). 18. G. Gomard, X. Meng, E. Drouard, K. E. Hajjam, E. Gerelli, R. Peretti, A. Fave, R. Orobtchouk, M. Lemiti, and C. Seassal, “Light harvesting by planar photonic crystals in solar cells: the case of amorphous silicon,” J. Opt. 14(2), 024011 (2012). 19. A. Taflove and S. C. Hagness, Computational Electrodynamics: the Finite-Difference Time-Domain Method (Artech, 2005). 20. AM1, 5 solar spectrum irradiance data: http://rredc.nrel.gov/solar/spectra/am1.5. 21. S. Selberherr, Analysis and Simulation of Semiconductor Devices (Springer, 1984). 22. J. Nelson, The Physics of Solar Cells (Imperial College, 2003). 23. W. E. I. Sha, W. C. H. Choy, Y. Wu, and W. C. Chew, “Optical and electrical study of organic solar cells with a 2D grating anode,” Opt. Express 20(3), 2572–2580 (2012). 24. S. Chuang, Physics of Optoelectronic Devices (Wiley, 1995). 25. E. D. Palik, Handbook of Optical Constants of Solids (Academic Press, New York, 1985). 26. L. G. Jeffery, Handbook of Photovoltaic Science and Engineering (Antonio Luque, 2004). 27. T. Markvart and L. Castaner, Practical Handbook of Photovoltaics: Fundamentals and Applications (Elsevier Advanced Technology, 2003). 28. F. Wang, H. Yu, J. Li, S. Wong, X. W. Sun, X. Wang, and H. Zheng, “Design guideline of high efficiency crystalline Si thin film solar cell with nanohole array textured surface,” J. Appl. Phys. 109(8), 084306 (2011). 29. O. Demichel, V. Calvo, A. Besson, P. Noé, B. Salem, N. Pauc, F. Oehler, P. Gentile, and N. Magnea, “Surface recombination velocity measurements of efficiently passivated gold-catalyzed silicon nanowires by a new optical method,” Nano Lett. 10(7), 2323–2329 (2010).
1. Introduction Thin-film solar cells are nowadays considered as one of the most promising candidates for future generation of photovoltaics due to their lower production costs and a better adaptability to a wide range of structures and equipments compared to the bulk solar cells [1]. Unfortunately, a primarily unavoidable drawback of thin-film solar cells is the poor optical absorption, which limits the conversion efficiencies of the cells. This limitation has motivated the researchers to develop light trapping technology, which becomes the hotspots in solar cell research and concerns both the inorganic and organic solar cells. Up to date, much effort has been devoted to designing light trapping structures in the realm of thin-film solar cells. Various novel structures have being proposed to enhance light harvesting in solar cells including graded refractive index antireflective coating [2], metallic nanoparticles [3–6], periodic metallic gratings [7–9], photonic crystals [10, 11] and so on. Among these promising nanophotonic light trapping structures, patterning the active layer itself as a one-dimensional (1D) periodic grating is one of the most popular approaches because it is simple and easy to be fabricated. Such nanophotonic approaches can offer the ability to improve light trapping through antireflection effects and excited resonance effects that originate from the excitation of guided or leaky modes inside the active layer [12, 13]. However, most theoretical literatures of light trapping methods take into account only for optical effects. The carrier recombination processes are ignored. It is important to note that solar cell is a photoelectric coupling device, the electrical effect is as important as optical effect. Thus, it is necessary to consider both the optical absorption of the device and the electrical device response in designing the thin-film solar cells [14–17]. In this work, we use optical simulations coupled with electrical device simulations by solving the Maxwell’s equations and semiconductor equations (Poisson, continuity, and driftdiffusion equations) to enable the simultaneous consideration of optical and electrical effect when investigating the performance of 1D periodic grating structure thin-film solar cells. For periodic grating structure, surface recombination becomes a dominant concern due to its large surface to volume ratio. Hence, the focus of this article is to investigate the effect of the surface recombination on the efficiency of thin-film solar cells. At the same time, radiative recombination, Shockley-Read-Hall recombination and Auger recombination are also #197209 - $15.00 USD Received 6 Sep 2013; revised 15 Oct 2013; accepted 17 Oct 2013; published 25 Oct 2013 (C) 2013 OSA 4 November 2013 | Vol. 21, No. S6 | DOI:10.1364/OE.21.0A1065 | OPTICS EXPRESS A1066
included in calculating the conversion efficiency of solar cells. The objective of this article is to demonstrate that the higher optical absorption does not mean higher conversion efficiency for nanostructured thin-film solar cells in consideration of recombination effect. In other words, it is insufficient for one to predict the performance of nanostructured thin-film solar cells when taking account only for the light trapping effect. In their work [18], Gomard et al were aware of such an effect. They demonstrated that different configurations possessing the same integrated absorption do not present the same power conversion efficiency due to Shockley-Read-Hall recombination even when the surface recombination was ignored. Here we demonstrate that even a configuration possessing a higher integrated absorption can finally end with a lower conversion efficiency once the surface recombination is introduced. Therefore, the article is structured as follows. The brief description of our theoretical model is given in section 2. Based on the photoelectric coupling simulation method, the results and discussion are presented in section 3. The impact of surface recombination velocity, geometric parameters of the structure is discussed in detail. The final conclusion is given in section 4. 2. Theoretical model 2.1 Optical simulation The optical absorption of thin-film solar cells is evaluated by solving Maxwell’s equations through the Finite Difference Time Domain (FDTD) algorithm, which is widely adopted for rigorous electromagnetic propagation calculation [19]. In the simulation, plane waves are normally incident on the cells. The perfectly matched layer (PML) absorbing boundary condition is implemented on the top and bottom boundaries while the periodic boundary condition is imposed on the left and right boundaries within the simulation unit. After obtaining the spatial distribution of the electromagnetic field, the optical generation rate can be expressed as Gopt (r ) =
λmax
ε " E (r , λ ) 2
λmin
2
I AM 1.5 (λ )d λ
(1)
where ε " is the imaginary part of the permittivity of the semiconductor material, r is the space coordinates, λmax is the maximum wavelength corresponding to the bandgap of semiconductor, λmin is the minimum wavelength of the solar radiation, E is the electric field, is the reduced Planck constant and I AM 1.5 is the global standard solar irradiance spectrum of AM1.5G [20]. The sunlight is un-polarized and can be treated as the superposition of two orthogonally polarized waves. Hence, the optical absorption and the optical generation rate is the average value for TE and TM polarizations. In the existing publications, in order to quantitatively illustrate the overall optical light trapping performance of the thin-film solar cells, the maximum achievable photocurrent density is widely used and defined as [10] J sc max =
λmax
λmin
e
λ hc
A(λ ) I AM 1.5 (λ )d λ
(2)
where h is the Planck constant, c is the speed of light in vacuum and A(λ ) is the optical absorption of the active layer. Obviously, the carrier recombination processes are not taken into account. This expression quantitatively reveals the response of light trapping effect.
2.2 Electrical simulation The electrical simulation is conducted based on the semiconductor equations, consisting of Poisson, continuity, and drift-diffusion equations [21–24]. All these equations would be solved simultaneously through finite element method and can be described as
#197209 - $15.00 USD Received 6 Sep 2013; revised 15 Oct 2013; accepted 17 Oct 2013; published 25 Oct 2013 (C) 2013 OSA 4 November 2013 | Vol. 21, No. S6 | DOI:10.1364/OE.21.0A1065 | OPTICS EXPRESS A1067
∇ ⋅ (ε∇φ ) = q( p − n + N D − N A )
(3)
∇ ⋅ J n = − q(G − R )
(4)
∇ ⋅ J p = q (G − R )
(5)
J n = − q μn n∇φ + qDn ∇n
(6)
J p = − q μ p p∇φ − qD p ∇p
(7)
where ε is the dielectric constant of semiconductor, q is the electron charge, φ is the electrical potential, n( p) is the electron (hole) concentration, N D ( N A ) is the donor (acceptor) doping concentration, J n ( J p ) is the current density of electron (hole), G is the optical generation rate obtained from optical simulation and R is the carrier recombination rate including radiative recombination, Shockley-Read-Hall recombination, Auger recombination and surface recombination. In addition, μn ( μ p ) is the electron (hole) mobility and Dn ( D p ) is the electron (hole) diffusion coefficient abided by Einstein relations with mobility: k BT Dn = q μ n D = k BT μ p p q
(8)
where k B is the Boltzmann constant and T is the operating temperature. The total carrier recombination rate ( R ) can be divided into four types including radiative recombination ( Rrad ), Shockley-Read-Hall recombination ( RSRH ), Auger recombination ( RAug ) and surface recombination ( Rsurf ). For each recombination mechanism, the corresponding models are described in the following equations. R = Rrad + RSRH + RAug + Rsurf
(9)
Rrad = B (np − ni2 )
(10)
np − ni2 τ p (n + n1 ) + τ n ( p + p1 )
(11)
RAug = (Cn 0 n + C p 0 p )(np − ni2 )
(12)
RSRH =
Rsurf =
np − ni2 1 1 (n + n1s ) + ( p + p1s ) Sp Sn
(13)
where B is the radiative recombination coefficient, τ n (τ p ) is the electron (hole) lifetime, n1 ( p1 ) is the electron (hole) concentration in the trap states, Cn 0 (C p 0 ) is the electron Auger
recombination coefficient, ni is the intrinsic carrier concentration and S n ( S p ) is the surface recombination velocity of electron (hole). It is clear to find that the model of surface recombination rate is similar to Shockley-Read-Hall recombination but differs slightly since the surface recombination just occurs on the cell surface. For the electrical simulation, we #197209 - $15.00 USD Received 6 Sep 2013; revised 15 Oct 2013; accepted 17 Oct 2013; published 25 Oct 2013 (C) 2013 OSA 4 November 2013 | Vol. 21, No. S6 | DOI:10.1364/OE.21.0A1065 | OPTICS EXPRESS A1068
focus on the region of the semiconductor. The surface recombination is implemented on the upper and lower interface of the semiconductor. The right and left interface is truncated by adopting the Neumann boundary condition, which is widely used in the semiconductor device simulation [23]. 2.3 Description of simulation structures and parameters
Fig. 1. Schematics of the simulation structure of thin-film solar cells. (a) First reference cell with planar surface. (b) Second reference cell with an optimal antireflection coating. (c) 1D periodic grating of light trapping structure. The value of the thickness is fixed as d1 = 0.4 µm.
The simulation structure of thin-film solar cells is schematically shown in Fig. 1. The active layer is composed of crystalline silicon (c-Si), which possesses the advantages of nontoxicity, abundance and mature processing. To reduce the loss of the incident light transmission, the Ag back reflector is widely used. The thickness of the c-Si film d1 and Ag back reflector layer thickness d 2 are fixed as 0.4 μm and 0.1 μm respectively. The planar cell and the cell with antireflection coating are considered as the reference for comparison. For the cell with antireflection coating, silica is used as the antireflection coating with an optimal thickness t = 0.08 μm. For 1D periodic grating structure, the design parameters include the grating period Λ , filling factor of the semiconductor f and the grating depth h . These geometric parameters not only affect the optical absorption but also influence the electrical properties. For investigating the optical properties of thin-film solar cells, the optical parameters of material are referenced from the literature [25]. The incident wavelength range is selected from 300 nm to 1100 nm, where the upper limit corresponding to the band-gap of c-Si. In terms of electrical aspect, the carrier mobility μn (μ p ) for electron (holes) is considered as the function of doping level and can be well approximated by [26]
μn = 92 +
μ p = 54.3 +
1268 cm 2 /Vs N D + N A 0.91 1+ ( ) 1.3 × 1017 406.9 cm 2 /Vs N D + N A 0.88 1+ ( ) 2.35 × 1017
(14)
(15)
The parameters of recombination are referenced from literature [27] and listed in Table 1. In addition, Table 1 includes the key parameters for electrical simulation as well. It is wellknown that the device configuration plays a key role in an accurate semiconductor device simulation. In our work, a planar junction configuration is adopted with an n-type emitter and
#197209 - $15.00 USD Received 6 Sep 2013; revised 15 Oct 2013; accepted 17 Oct 2013; published 25 Oct 2013 (C) 2013 OSA 4 November 2013 | Vol. 21, No. S6 | DOI:10.1364/OE.21.0A1065 | OPTICS EXPRESS A1069
p-type base as found as in reference [28]. The emitter width is fixed at 40 nm and the doping concentrations N D and N A are respectively fixed at 1019 and 1016 cm -3 . Table 1. Typical parameters for silicon thin-film solar cells in the simulation Parameters
ε
Values 11.7
Unit
Eg
1.12
eV
NA ND
16
10
cm -3
1019
cm -3
μn μp
115.886
cm 2 /Vs
68.7524
cm 2 /Vs
B
1.8 × 10−15 3.3 × 10 −7
cm 3 /s s
8.2 × 10−5
s
τn τp Cn 0 Cp0
−31
cm 6 /s
9.9 × 10 −32
cm 6 /s
Sn , S p
102 − 105
T
300
cm/s K
2.8 × 10
3. Results and discussion
3.1 Impact of surface recombination velocity Figure 2(a) displays the spectral absorption for the grating structured thin-film solar cells in comparison with two reference cells including planar cell and antireflective cell. The geometric parameters of the grating structure with maximum integrated optical absorption are Λ = 0.5 μm, h = 0.16 μm and f = 0.5 . It is clear that the optical absorption of grating structure increases remarkably in the long wavelength range from 550 nm to 1100 nm. Figure 2(b) shows the maximum achievable photocurrent density to show light trapping effect for all the investigated structures. It is obvious that the grating structure exhibits the best light trapping performance. According to traditional viewpoint, the higher optical absorption means the higher conversion efficiency for the thin-film solar cells. In other words, the grating structure thin-film solar cells have higher conversion efficiency comparing with the reference cells. However, the carrier recombination processes are not taken into account in the preceding discussion. As is known to all, nanostructured thin-film solar cells have much higher surface to volume ratio compared to bulk solar cells. Thus, surface recombination plays an extremely important role in determining the performance of thin-film solar cells. The impact of the surface recombination velocity will be discussed in the following.
#197209 - $15.00 USD Received 6 Sep 2013; revised 15 Oct 2013; accepted 17 Oct 2013; published 25 Oct 2013 (C) 2013 OSA 4 November 2013 | Vol. 21, No. S6 | DOI:10.1364/OE.21.0A1065 | OPTICS EXPRESS A1070
Fig. 2. Results of three structure c-Si thin-film solar cells. (a) Absorption spectra in the c-Si layer. (b) Maximum achievable photocurrent density to show light trapping effect. (c) Optical generation rate. (d) Simulated J-V and P-V characteristics with different surface recombination velocity. The blue, red and black lines correspond to the results for grating structure, antireflective coating structure and planar structure, respectively.
Figure 2(c) shows the optical generation rate of three investigated structures, which acts as the input profile in the electrical simulation. It is noticeable that the optical generation rate calculated by FDTD departs significantly from that obtained with Beer-Lambert law. The current-voltage (J-V) curves of three investigated structures together with power-voltage (PV) curves under different surface recombination velocity are shown in Fig. 2(d). As shown in
#197209 - $15.00 USD Received 6 Sep 2013; revised 15 Oct 2013; accepted 17 Oct 2013; published 25 Oct 2013 (C) 2013 OSA 4 November 2013 | Vol. 21, No. S6 | DOI:10.1364/OE.21.0A1065 | OPTICS EXPRESS A1071
Fig. 2(d), the open-circuit voltage ( Voc ) and the short-circuit current density ( J sc ) are reduced with the increase of surface recombination velocity (SRV) for all the investigated structures. In addition, the difference of Voc and J sc between the grating structure and the reference cells becomes smaller with the increase of SRV. For the grating structure, for example, as the SRV S n = S p varies from 10 cm/s to 103 cm/s, the value of Voc and J sc varies from 0.700 V to 0.564 V and from 15.60 mA/cm2 to 5.64 mA/cm2, respectively. While for the antireflective cell, the value of Voc and J sc varies from 0.695 V to 0.565 V and from 11.71 mA/cm2 to 5.43 mA/cm2 if the SRV S n = S p varies from 10 cm/s to 103 cm/s. It validates that the SRV has great impact on Voc and J sc for nanostructured thin-film solar cells as expected. The surface area of grating structure is larger than that of the reference cells and so is the surface recombination. The boosted surface recombination has suppressed the influence of the optical absorption enhancement. Both Voc and J sc are the key parameters of determining the power conversion efficiency of the solar cells. Therefore, under the condition of strong surface recombination velocity, although the optical absorption of grating structure is higher, the power conversion efficiency (PCE) is probably smaller comparing to the reference cells.
Fig. 3. PCE of different configurations for c-Si thin-film solar cells with varied surface recombination velocity.
Here the PCE of different configurations as a function of the SRV is further studied in order to make this statement more clear. As shown in Fig. 3, it is obvious that the PCE of all the investigated structures is decreased when the SRV is increased. The larger SRV value always results in the lower PCE for the nanostructured thin-film solar cells. Moreover, it is found that the structure with a higher optical absorption does not always have a higher PCE when the SRV is considered, although all the investigated grating structures have a higher integrated optical absorption than the planar surfaces. For example, when the SRV S n = S p is
#197209 - $15.00 USD Received 6 Sep 2013; revised 15 Oct 2013; accepted 17 Oct 2013; published 25 Oct 2013 (C) 2013 OSA 4 November 2013 | Vol. 21, No. S6 | DOI:10.1364/OE.21.0A1065 | OPTICS EXPRESS A1072
10 cm/s, the value of PCE for the grating structure with geometric parameters Λ = 0.8 μm, h = 0.2 μm, f = 0.5 and the antireflective structure respectively corresponds to 7.07% and
6.35%, which is in accord with their optical absorption. Once the SRV S n = S p changes to 102 cm/s, however, the value of the PCE of the above-mentioned structured cells respectively changes to 4.74% and 5.10%, which is in contrary to their optical absorption. This phenomenon can be explained as follows. In the former case, the effect of optical absorption enhancement is greater than that of the surface recombination, which leads to increase in the PCE. For the later case, the effect of surface recombination is greater than that of optical absorption enhancement, leading to the drop of PCE. Thus, the value of the SRV plays an extremely important role in determining the PCE of the nanostructured thin-film solar cells. In addition, the grating structure with geometric parameters Λ = 0.4 μm, h = 0.1 μm, and f = 0.5 has the same surface area compared to the grating structure with geometric
parameters Λ = 0.8 μm, h = 0.2 μm, and f = 0.5 , so that they have the same surface recombination. The PCE of the former is always higher than that of the latter under the different values of SRV, corresponding to its higher optical absorption. The PCE of the grating structure with geometric parameters Λ = 0.5 μm, h = 0.16 μm, and f = 0.5 is the same as that of the grating structure with geometric parameters Λ = 0.4 μm, h = 0.1 μm, and f = 0.5 in the case of S n = S p = 10 2 cm/s. It can be attributed to the balance of higher optical absorption and higher surface recombination. Therefore, it can be concluded that the surface morphology (dominating the surface area and optical absorption) and surface passivation (determining the value of SRV) have great influence in PCE for nanostructured thin-film solar cells. It is emphasized that the higher optical absorption does not always mean the higher conversion efficiency for nanostructured thin-film solar cells in consideration of the recombination effect. The tradeoffs between optical absorption and surface recombination should be obtained in designing the nanostructured thin-film solar cells. As a key factor of determining the surface recombination, the accurate value of SRV is necessary in practice in designing the patterned nanostructured thin-film solar cells.
#197209 - $15.00 USD Received 6 Sep 2013; revised 15 Oct 2013; accepted 17 Oct 2013; published 25 Oct 2013 (C) 2013 OSA 4 November 2013 | Vol. 21, No. S6 | DOI:10.1364/OE.21.0A1065 | OPTICS EXPRESS A1073
3.2 Impact of geometric parameters of the structure
Fig. 4. Results of grating structure thin-film solar cells with different geometric parameters. (a) Absorption spectra. (b) Maximum achievable photocurrent density to show light trapping effect. (c) Simulated J-V and P-V characteristics under ideal condition. (d) Simulated J-V and P-V characteristics under considering recombination condition. (e) The contribution of each recombination process to the PCE of the nanostructured thin-film solar cells.
To further demonstrate our viewpoint, we choose two cases of grating structure to investigate their optical and electrical characteristics. The geometric parameters of Case1 are Λ = 0.5 μm, h = 0.2 μm, and f = 0.5 . While for Case2, the geometric parameters are Λ = 0.6 μm, h = 0.1 μm, and f = 0.5 . The SRV S n = S p = 100 cm/s is selected for investigating their electrical properties [16]. The optical properties including spectral absorption and maximum achievable photocurrent density are shown in Fig. 4(a) and Fig. 4(b) respectively. It is obvious that the Case1 manifests a better effect of optical absorption #197209 - $15.00 USD Received 6 Sep 2013; revised 15 Oct 2013; accepted 17 Oct 2013; published 25 Oct 2013 (C) 2013 OSA 4 November 2013 | Vol. 21, No. S6 | DOI:10.1364/OE.21.0A1065 | OPTICS EXPRESS A1074
than Case2. If the impact of recombination effect is ignored, the Case1 also shows a better performance of PCE than Case2 corresponding to its higher optical absorption as shown in Fig. 4(c). However, as shown in Fig. 4(d), when the recombination effect is considered, the Case2 shows a better performance of PCE than Case1 in contrary to its lower optical absorption. This interesting phenomenon which seems on the opposite of intuition can mainly attribute to surface recombination effect. The surface area of Case1 is larger than that of Case2 within the unit cell and so is the surface recombination. Although the optical absorption of thin-film solar cells increases, the recombination increases as well. As a result, the PCE of Case1 is lower than that of Case1 because the impact of surface recombination is stronger than that of the enhancement of optical absorption. Therefore, we should get tradeoffs between improving optical absorption and reducing surface recombination in designing nanostructured thin-film solar cells. In order to further investigate the contribution of each recombination process ( Rrad , RAug , RSRH , and Rsurf ) to the PCE of the solar cells, it is separately considered in the simulation. As illustrated in Fig. 4(e), it is obvious that the radiative recombination ( Rrad ) and Auger recombination ( RAug ) could be neglected in our computations for nanostructured thinfilm solar cells because they make almost no contribution to the PCE. Once the ShockleyRead-Hall recombination ( RSRH ) is considered separately, the value of PCE decreases from 12.69% to 10.08% for Case1 and from 11.75% to 9.05% for Case2. It is obvious that the PCE of Case1 is higher than that of Case2, corresponding to its higher optical absorption. However, the value of PCE is decreased from 12.69% to 5.97% for Case1 and from 11.75% to 6.38% for Case2 when surface recombination ( Rsurf ) is considered separately. It is clear that the Case2 shows the better performance of PCE than Case1 in contrary to its lower optical absorption. When all the recombination processes are considered simultaneously, the value of PCE is 5.64% for Case1 and 5.92% for Case2, corresponding to the results of considering Rsurf separately. Therefore, this can demonstrate that the surface recombination is the dominating recombination process for the nanostructured thin-film solar cells. The higher optical absorption does not always mean the higher power conversion efficiency for nanostructured thin-film solar cells due to the inevitable surface recombination.
#197209 - $15.00 USD Received 6 Sep 2013; revised 15 Oct 2013; accepted 17 Oct 2013; published 25 Oct 2013 (C) 2013 OSA 4 November 2013 | Vol. 21, No. S6 | DOI:10.1364/OE.21.0A1065 | OPTICS EXPRESS A1075
3.3 Optimization of the light trapping structure
Fig. 5. Parametric analysis of the optimized process for light trapping structure. (a) Maximum achievable photocurrent density and PCE vs Λ . (b) Maximum achievable photocurrent density and PCE vs h . (c) Maximum achievable photocurrent density and PCE vs f . (d) Electrically optimized J-V and P-V characteristics.
Since the optical and electrical performance of nanostructured thin-film solar cells is sensitive to geometric parameters, the optimization of light trapping structure becomes important. The parametric analysis of the optimized process for light trapping structure is shown in Fig. 5. During the process of the optimization, each geometric parameter is scanned in a predetermined range while the other geometric parameters are kept constant. According to the references [16, 29], the surface recombination velocity can be reduced to 100 cm/s after the passivation process, that is why we selected this typical value for our next calculations. For each parameter, both the maximum achievable photocurrent density and power conversion efficiency are plotted as a function of the geometric parameter value. The first parameter of analysis is the period Λ of the grating structure, which is varied from 0.2 μm to 0.8 μm. The other invariant parameters are h = 0.1 μm and f = 0.5 . As depicted in Fig. 5(a), the peak of maximum achievable photocurrent density is at Λ = 0.45 μm while the PCE peaks at Λ = 0.5 μm. The second parameter we analyzed is the depth h of the grating structure with the variety from 0.06 μm to 0.22 μm. The period Λ and filling factor f are kept constant as 0.5 μm and 0.5 respectively. It can be seen from Fig. 5(b) that the maximum achievable photocurrent density peaks at h = 0.16 μm while the peak of PCE is at h = 0.1 μm. Figure 5(a) agrees well with Fig. 5(b) that maximum optical absorption point is not consistent with the maximum power conversion efficiency point. This interesting phenomenon can be attributed to surface recombination effect. The maximum optical absorption point has larger surface area within the unit cell comparing to that of maximum power conversion efficiency point. Therefore, the maximum optical absorption and
#197209 - $15.00 USD Received 6 Sep 2013; revised 15 Oct 2013; accepted 17 Oct 2013; published 25 Oct 2013 (C) 2013 OSA 4 November 2013 | Vol. 21, No. S6 | DOI:10.1364/OE.21.0A1065 | OPTICS EXPRESS A1076
relatively strong surface recombination lead to the results that the PCE is not the biggest. It can be concluded that the surface recombination plays a key role in determining the performance of nanostructured thin-film solar cells. The third parameter we analyzed is the filling factor f . The filling factor f is varied from 0.2 to 0.7 while the period Λ and depth h are kept constant as 0.5 μm and 0.1 μm respectively. It can be seen from Fig. 5(c) that the maximum achievable photocurrent density peaks at f = 0.55 as well as the PCE. When the filling factor is varied, the surface area within the unit cell remains the same. As a result, maximum optical absorption point is consistent with the maximum power conversion efficiency point during the same surface recombination. Figure 5(d) shows the electrically optimized J-V and P-V characteristics of 1D periodic grating structure thin-film solar cells. The maximum PCE of the grating structure reached is 6.28% after the optimization, exhibiting 23% enhancement compared to the reference cell with the antireflective coating. 4. Conclusions
In conclusion, both the optical and electrical properties of thin-film cells with 1D periodic grating structure have been studied. The framework of photoelectric coupling model is developed to optimize the light trapping structures for thin-film solar cells. In contrast to most existing investigations, our simulation results indicate that the higher optical absorption does not always mean the higher conversion efficiency for nanostructured thin-film solar cells in consideration of recombination effect. In particular, surface recombination is the dominant recombination process and plays an important role in affecting the efficiency of nanostructured thin-film solar cells. The key parameters of nanostructured solar cells such as Voc and J sc are significantly affected by the surface recombination. It is critical to get the tradeoffs between enhancing optical absorption and reducing the surface area in designing nanostructured thin-film solar cells when using the same surface passivation technology. This work makes it clear that it may be incomprehensive for one to account only for the light trapping effect for designing high-performance solar cells and it will be helpful for designing high-performance nanostructured thin-film solar cells. Acknowledgment
We are grateful to the financial support from the National Natural Science Foundation of China (Grant No. 51336003).
#197209 - $15.00 USD Received 6 Sep 2013; revised 15 Oct 2013; accepted 17 Oct 2013; published 25 Oct 2013 (C) 2013 OSA 4 November 2013 | Vol. 21, No. S6 | DOI:10.1364/OE.21.0A1065 | OPTICS EXPRESS A1077
