Numerical simulation of both single and double absorbing layers in amorphous silicon thin film solar cell is performed with the use of AFORS-HET. A single absorbing layer solar cell with both a-SiH and a-SiGeH is designed and compared with a tandem heterojunction solar cell, a-SiC/a-SiH/a-Si(i)/a-SiGeH. Design parameters are investigated, compared and optimized. The maximum efficiency for each single absorbing layer and for a tandem heterojunction thin film solar cell, a-SiC/a-SiH/a-Si(i)/a-SiGeH, is predicted. The results are validated by comparing with two different method of analysis.
© 2018 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
In recent years, the hydrogenated amorphous silicon (a-SiH) thin film solar cell has attracted lots of attention due to its low cost . Today, the most economical technology is Si based technology. However, it undergoes a serious light-induced degradation effect, the (S-W) effect , which becomes a bottleneck for the application of a-SiH thin film solar cell. Despite its limitations, including the degradation effect (S-W), and its low efficiency , silicon is still a promising cost-effective source due to its abundance . Some methods that can resolve these limitations are the following. (a) Using tandem technology, the S-W effect in a-SiH thin film can be improved with amorphous hydrogenated silicon germanium (a-SiGeH) . The S-W effect is not obvious in a-SiGeH. An alloy of s-SiH with Ge reduces the optical gap and allows band gap tuning by varying Ge content. (b) Inserting a-SiC as a window layer [4, 5] helps to improve the efficiency of a-SiH thin film solar cell. Amorphous hydrogenated silicon carbide (a-SiC) thin layer has been widely used as a window layer in tandem technology to achieve better optical transparency . It does so by widening the band gap of material [6, 7]. (c) Inserting a thin buffer layer or back surface layer of intrinsic amorphous silicon a-Si(i) at the interface helps to reduce recombination of carriers. Thus, the tandem thin film solar cell, a-SiH/a-SiGeH, is fabricated by stacking sub-cells within the band gaps of 1.75eV to 1.55eV respectively. The top, middle, and bottom of a tandem cell is designed to absorb blue, green, and red photons [5, 8].
The technique of tandem design is used to minimize the contact recombination, by different semiconductors with different band gaps, which are used to form solar absorbing layers for a collection of photons. The generated excess carriers are separated by an electric field, and attracted towards the outer contact. These selective contacts are designed by adjusting the doping concentrations at the edges of the solar cell. As a result, it creates an internal electric field by which the excess carriers are separated and collected at different contacts.
Amorphous silicon is considered the cheapest source in photovoltaic (PV) design giving researchers a wide range of options to improve its efficiency using a-Si alloys. The potential of this technology has been demonstrated by Sanyo by reaching 22.80% cell efficiency in lab  and 15.00% in a commercial module . Recently, The Advance Institute of Industrial Science and Technology (AIST) designed high efficiency (11.00%) microcrystalline silicon thin film solar cells using superior light trapping capabilities (honeycomb textured substrates) . Some other groups also have demonstrated efficiencies close to 20% [12–16].
The main objective of this study is to use simulation to simply show the improvement in efficiency when using double absorbing layers versus single absorbing layers and suggest optimal device layer thickness to reduce recombination losses in the bulk as well as the interfaces. To address these objectives, first, numerical simulations of both single and double absorbing layers of amorphous silicon thin film solar cell are performed. Single absorbing layer solar cells with a-SiH and a-SiGeH are designed and compared with a tandem heterojunction solar cell comprised of both a-SiH and a-SiGeH absorbing layers. Then, the design parameters are investigated, compared and optimized to predict the maximum efficiency. The simulation is performed using AFORS-HET (simulation program for hetero-junction solar cells) . Moreover, the effect of back surface field (BSF) on the parameters of a-Si solar cell was also investigated. Results are validated using two different analysis methods.
2. Theoretical model
The incoming spectral photon flux Ф0(λ, t), is measured with respect to the front contact absorption and reflectance. The photon flux incoming through the first semiconductor layer is calculated by all three terms as Ф0(λ, t) R(λ) A(λ). The theoretical model uses AFORS-HET for the tracing of light extended path caused by textured surface, is needed to be adjust the incident angle δ of incoming light. For a silicon textured wafer with <111> direction the angle is δ = 54.74°, and for normal plane δ = 0°. The angle γ by which the light travels through the layer stack depends on the wavelength of the incoming light and is calculated by Snell’s law.
The doping densities ND and NA at position x of fixed donor or acceptor state within any cell is supposed to always be completely ionized and defects Ntrap at any position with specific energy E within the band gap of the semiconductor can be locally charged or uncharged within the system. These defects can be defined like acceptor-like Shockley-Read-Hall defects, donor-like Shockley-Read-Hall defects or dangling bond defects. The Poisson equation, which is to be solved within each layer as given
The one-dimensional equations of continuity and transport for electron and holes, which have to be solved with in each layer,
The super band gap generation rates Gn and Gp are determined by optical modeling and the recombination rates Rn or Rp. The electron and hole currents Jn and Jp are derived by the gradient of the corresponding quasi Fermi energy EFn and EF. The position dependent Fermi energies and the corresponding local electron and hole current are defined with the help of Maxwell Boltzmann approximation for Fermi-Dirac distribution function.
The radiative band to band rate constant r has to be defined in order to balance the radiative band to band recombination rates Rn,p; as the result the electrons and holes are always equal.
According to Krichen, and Arab , the effect of back surface field (BSF) in terms of surface recombination velocity ERV (Sbsf) in the vicinity of back surface junction is defined as:
3. Computer model
Figure 1 shows the schematics diagram for the solar cells simulated using AFORS–HET. For all cells, the top p+ layer (Layer 2) and the bottom n+ layer (Layer 6 and 8) have a 5nm thickness with doping concentrations of and , respectively. The band gap of the top p+ layer (layer 2) is 2eV, and the bottom, n+ layer is 1.8eV.
The single absorbing layer cells: both a–SiH and a-SiGeH are shown in Fig. 1(a). In both cells, layer 4 is the absorbing layer, and layer 5 is the mid intrinsic absorbing region. The tandem heterojunction (with double absorbing layers) solar cell is designed using 100nm a-SiH absorbing layer and 80nm a-SiGeH absorbing layer separated by 10nm mid intrinsic absorbing region, layer 5. The schematics of the tandem heterojunction solar cell is shown in Fig. 1(b). For the simulations, the a-SiH absorbing layer has a graded band gap varying from 1.75eV to 1.65eV. The a-SiGeH absorbing layer has a fixed band gap of 1.55eV. The band gap diagram for double layer tandem heterojunction solar cell is shown in Fig. 2.
An n-type 80nm indium doped tin oxide (ITO) layer is used at the top surface. A 200nm thick region of Ag is used as a bottom contact. For the ITO and Ag regions, the default optical constant models for n(E) and k(E) available in the database of AFORS-HET software are used. A 10nm intrinsic a-SiC with 1.85eV band gap is used as window layer. The a-Si(i) buffer layer, layer 7, acts as back surface field (BSF) layer. A band gap of 1.7eV is used for all of the intrinsic silicon layers.
Cells are illuminated with AM 1.5 radiation. For front and back contacts, MS Shockley model is applied. The Drift –diffusion model and thermionic emission model are used for the interfaces. The Beer-Lambert absorption model is applied for the optical bulk layer.
4. Result and discussion
The width of the intrinsic region and diffusion length of both electrons and holes determine the short circuit current (Jsc) of the cell. To reduce the probability of recombination of excess carriers, simulations show the highly doped p+ and n+ regions need to have a width less than five percent of the total intrinsic layer.
Amorphous silicon materials such as a-SiH, a-SiGeH and a-SiCH contain a large number of defect states within the band gap. To accurately model the device, the continuous density of defect states in a-Si and in all of its alloys is considered. The density of states (DOS) is described as a combination of exponential decaying band tail states (for both donor like states and acceptors like states) and Gaussian distribution of mid gap states as shown in Fig. 3. The a-Si characteristics energies of the conduction band and valence band tails are 0.06eV and 0.03eV. Similarly, the peak energy distributions of acceptor and donor type Gaussian states are 1.2eV and 0.56eV with a deviation of 0.08eV. The concept of Gaussian distribution of defect states in amorphous silicon material reflects its disordered structure and disorder of states at a particular energy level. This distribution of density of defect states is defined for all layers including a-SiC and a-Si(i) layers using conduction and valence tail states and two Gaussian states  as shown in Fig. 3.
The efficiency of a-SiH is 9.64% with FF = 83.07%, Voc = 1069 mV, and Jsc = 10.86 mA/cm2. The Voc is highest as compare to the other designs due to thickness of the layer. The short circuit current is nearly comparable to the tandem design. It is because of the high band gap of a-Si as compare to a-SiGeH. The a-Si layer is sensitive to thermal generation, there is more generation rate as compare to other layers but it has high recombination. The Fig. 4 shows the efficiencies of a-SiH cell.
The efficiency of a-SiGeH is 5.40% with FF = 81.85%, Voc = 1026 mV, and Jsc = 6.43 mA/cm2. The short circuit current and efficiency of a-SiGeH is lower because of lower band gap and thickness. The a-SiGeH is supposed to have high reliability as compared to a-SiH by having less defect density. The Fig. 5 shows the efficiencies of a-SiGeH cell. The efficiency of a-SiGeH is lower than the efficacy of a-Si, obviously, the short circuit current is lower because of lower band gap and thickness of the layer but it has high reliability as compared to a-SiH by having less defect density.
The IV characteristics for tandem solar cell is shown in Fig. 6. The short circuit current (Jsc), open circuit voltage (Voc), fill factor (FF), and efficiencies (η) of the single absorbing layer solar cells and tandem heterojunction solar cell are shown in Table 1.
The buffer layer, layer (5), helps to reduce the interface recombination and improves the interface quality with better surface passivation . The thicknesses of buffer layers a-Si(i) are varied from 5nm to 20nm. By increasing thinness of mid a-Si(i) buffer layer, layer 5, Voc decreases, Joc increases, and FF and efficiency also increase. The 10nm thinness is considered the optimum thickness for layer 5 after investigating the thickness versus efficiency contributed by this layer with overall efficiency for tandem heterojunction solar cell is 9.46%.
The layer 7 acts as back surface field layer. The effect of varying the band gap of amorphous silicon layer, after removing the BSF (layer 7), on its efficacy (η), is shown in Fig. 7(c). Overall, 7.17% increase in efficiency with addition of BSF, which is contributed by back surface layer (layer 7).
The Figs. 7(a)-7(c) show the variation in open circuit voltages, short circuit current, and efficiency as a function of layer thickness and band gap of absorbing layer (layer 4), calculated for the tandem heterojunction solar cell after removing BSF layer (layer 7) from the design. However, after inserting the BSF layer (layer 7) the efficiency increases with increasing the thickness and band gap of the same layer. The improvement in open circuit voltage can be noted decreasing with increase band gap and with decrease of thickness. It can also be noted that the improvement in open circuit voltage is particularly important when the surface recombination velocity (Sbsf) reduces. However, when back surface recombination declines, the reverse saturation current density in the layer decreases. Consequently, the improvement observed in open circuit voltage, when Sbsf decreases can be attributed to reverse saturated current density in the base region, coupled with an increase in short circuit current. Back surface field layer band gap varied from 1.20 to 1.80eV. The increase in band gap causes increase in Jsc. As the band gap broadens, it interferes as a barrier for the majority carriers. The optimum band gap selected for the BSF layer is 1.5-1.7 eV. Similarly, eliminating both buffer layers (layer 5 and 7), but keeping the same design and all parameters of the solar cell, the efficiency reduces to 9.71% with Voc = 1049mA, Joc = 10.83mA/cm2.
The spectral response for the tandem heterojunction solar cell is shown in Fig. 8. The internal quantum efficiency (IQE) proceeds to zero when the incident photon has less energy than the band gap. This was observed at a wavelength of 766 nm for a-SiH solar cell and at 700 nm for a-SiGeH. The internal quantum efficiency for the tandem heterojunction solar cell proceeds to zero at 820 nm. The least absorbing layer is a-SiGeH, whereas the combination ofboth individual absorbing layers (a-SiH and a-SiGeH) results in an increase in the internal quantum efficiency (IQE) and external quantum efficiency (EQE) of the tandem heterojunction solar cell. The internal quantum efficiency should be always greater than the external quantum efficiency since the IQE governs the generation of electrons via absorption of each photon. The external efficiency depends on both the absorption and collection of charges (Table 2).
The AFORS-HET analysis model is based on the density of states; therefore, the same design of the a-Si tandem heterojunction solar cell with the Lumerical FDTD/Device 4.6  was investigated in order to confirm the findings. With the use of the Lumarical FDTD/Device 4.6, similar results were observed for all three designs as when had been done using AFORS-HET. Results with Lumerical FDTD/Device 4.6 indicated the fill factor and the efficiency for a-SiGeH (FF = 78.98% and η = 6.03%), a-SiH (FF = 84.27% and η = 7.06%), and Tendon design (FF = 81.90% and η = 9.84%) respectively. These findings are recorded in Table 1.
The numerical analysis of amorphous silicon thin film solar cell was investigated with the use of AFORS-HET. The efficiencies of both single absorbing layers a-SiH and a-SiGeH were compared with tandem heterojunction thin film solar cell (a-SiH + a-SiGeH). The efficiencies for single absorbing layers were observed to be 9.64% for a-SiH and 5.40% for a-SiGeH respectively. The overall efficiency of the tandem heterojunction cell was recorded as 10.46%. The buffer layers contributed to a 7.17% increase in efficiency of the tandem solar cell. In order to validate these findings, the results were compared using the Lumerical FDTD/Device 4.6. Similar results were observed by both models of analysis as seen in Table 1.
References and links
1. S. Ke, C. Wang, T. Pan, J. Yang, and Y. Yang, “Numerical simulation of the performance of the a-Si: H/a-SiGeH/a-SiGeH tandem solar cell,” J. Semiconductors 35(3), 034013 (2014). [CrossRef]
2. D. Staebler and C. Wronski, “Reversible conductivity changes in discharge‐produced amorphous Si,” Appl. Phys. Lett. 31(4), 292–294 (1977). [CrossRef]
3. M. Nawaz, “Computer analysis of thin-film amorphous silicon heterojunction solar cells,” J. Phys. D Appl. Phys. 44(14), 145105 (2011). [CrossRef]
4. Y. Kim, B. Hong, G. Jang, S. Suh, and D. Yoon, “Characterization of a‐SiCH films deposited by RF plasma CVD,” Crystal Research and Technology: Journal of Experiment and Industrial Crystallography 37(3), 219–224 (2002).
5. S. B. Patil, A. A. Kumbhar, S. Saraswat, and R. Dusane, “Preliminary results on a-SiCH based thin film light emitting diode by hot wire CVD,” Thin Solid Films 430(2), 257–260 (2003). [CrossRef]
6. J. Yang, A. Banerjee, and S. Guha, “Amorphous silicon based photovoltaics—from earth to the final frontier,” Sol. Energy Mater. Sol. Cells 78(4), 597–612 (2003). [CrossRef]
7. J. Becker, D. Pysch, A. Leimenstoll, M. Hermle, and S. Glunz, “Wet-chemical pre-treatment of c-Si substrates enhancing the performance of a-Si: H/c-Si hetero-junction solar cells,” 24th European PV Solar Energy Conference and Exhibition, Hamburg, Germany, Sept. 2009.
8. T. Matsui, H. Jia, and M. Kondo, “Thin film solar cells incorporating microcrystalline Si1–xGex as efficient infrared absorber: an application to double junction tandem solar cells,” Prog. Photovolt. Res. Appl. 18(1), 48–53 (2010). [CrossRef]
9. T. Mishima, M. Taguchi, H. Sakata, and E. Maruyama, “Development status of high-efficiency HIT solar cells,” Sol. Energy Mater. Sol. Cells 95(1), 18–21 (2011). [CrossRef]
10. R. Grigorovici, R. Croitoru, N. Marina, and I. Natasi, “Heterojunctions between amorphous Si and Si single crystals,” Rev. Roum. Physiol. 13(4), 317–325 (1968).
11. H. Sai, T. Matsui, K. Matsubara, M. Kondo, and I. Yoshida, “11.0 percent efficient thin-film microcrystalline silicon solar cells with honeycomb textured substrates,” IEEE J. Photovoltaics 4(6), 1349–1353 (2014). [CrossRef]
12. Z. C. Holman, A. Descoeudres, L. Barraud, F. Z. Fernandez, J. P. Seif, S. De Wolf, and C. Ballif, “Current losses at the front of silicon heterojunction solar cells,” IEEE J. Photovoltaics 2(1), 7–15 (2012). [CrossRef]
13. D. Muñoz, T. Desrues, A. Ozanne, S. De Vecchi, S. M. de Nicolás, F. Jay, F. Souche, N. Nguyen, C. Denis, and C. Arnal, “Key aspects on development of high efficiency heterojunction and IBC heterojunction solar cells: Towards 22 percent efficiency on industrial size,” in Proc. 27th European Photovoltaic Solar Energy Conference and Exhibition 576 (2012).
14. S. De Wolf, A. Descoeudres, Z. C. Holman, and C. Ballif, “High-efficiency silicon heterojunction solar cells: a review,” Green 2(1), 7–24 (2012). [CrossRef]
15. G. A. Armin, B. B. Matthew, H. Bram, and M. Thomas, “Industrial silicon wafer solar cells–status and trends,” Green 2(4), 135–148 (2012). [CrossRef]
16. D. Bätzner, Y. Andrault, L. Andreetta, A. Büchel, W. Frammelsberger, C. Guerin, N. Holm, D. Lachenal, J. Meixenberger, and P. Papet, “Properties of high efficiency silicon heterojunction cells,” Energy Procedia (Elsevier Ltd., 2011), pp. 153–159.
17. R. Stangl, C. Leendertz, and J. Haschke, “Numerical simulation of solar cells and solar cell characterization methods: the open-source on demand program AFORS-HET (INTECH Croatia,” Sol. Energy (2010).
18. W. Lisheng, C. Fengxiang, and A. Yu, “Simulation of high efficiency heterojunction solar cells with AFORS-HET,” J. Phys. Conf. Ser. 276(1), 012177 (2011). [CrossRef]
19. M. Krichen and A. B. Arab, “Analysis and study of a-Si thin heterojunction solar cells with back surface field,” J. Comput. Electron. 15(1), 269–276 (2016). [CrossRef]
20. F. D. T. D. Solutions, Lumerical, Device and FDTD, A TCAD device simulation for design, analysis, & optimization of semiconductor optoelectronic devices, (Lumerical Solutions. Inc., 2016). https://www.lumerical.com/tcad-products/fdtd2/