1290
OPTICS LETTERS / Vol. 40, No. 7 / April 1, 2015
Performance enhancement in a-Si:H/μc-Si:H tandem solar cells with periodic microstructured surfaces Xiangqian Shen, Qingkang Wang,* Peihua Wangyang, Kun Huang, Le Chen, and Daiming Liu Key Laboratory for Thin Film and Microfabrication Technology of Ministry of Education, School of Electronic Information and Electrical Engineering, Shanghai Jiao Tong University, Shanghai 200240, China *Corresponding author: [email protected] Received December 17, 2014; revised February 9, 2015; accepted February 10, 2015; posted February 11, 2015 (Doc. ID 230934); published March 20, 2015 Here we report on an efficient light-coupling scheme with a periodic microstructured surface to enhance the performance of thin film silicon solar cells. The centerpiece of the surface structure is the hemispherical pit arrays (HPAs), which are fabricated using an inexpensive and scalable process. The integration of HPAs into micromorph tandem thin film silicon solar cells leads to superior broadband reflection suppression properties. With this design, the reflection losses of the tandem cell are reduced to only 1.5%. We demonstrate an efficiency increase from 11.67% to 12.23% compared to a conventional cell with a flat surface, with a 4.6% increase in short circuit current density. The surface microstructures reported here can be applied to a variety of photovoltaic devices to further improve their performance. © 2015 Optical Society of America OCIS codes: (350.6050) Solar energy; (290.5850) Scattering, particles; (160.4760) Optical properties. http://dx.doi.org/10.1364/OL.40.001290
In the area of low-cost photovoltaic applications, amorphous/microcrystalline silicon (a-Si:H/μc-Si:H) tandem solar cells are one of the more promising candidates, considering both their efficiency and stability [1,2]. Moreover, silicon (Si) is an environment-friendly material that is abundant in nature, is nontoxic, and is a wellestablished technology [3]. Ideally, a solar cell should absorb all incident photons with energies that are larger than their active materials’ band gap. However, the quantum efficiency of the solar cell cannot reach 100%, due to optical losses. Optical reflection at the interface due to refractive index mismatch (Si ∼3.5, transparent conductive oxide (TCO) ∼2.0, glass ∼1.5, air ∼1.0) is one of the major loss mechanisms [4–6] that seriously affects the solar cell efficiency. Therefore, antireflection coatings play an important role in enhancing the efficiency of solar cells [7]. Lots of work has been done to suppress reflection from the interface [8–10]. Conventionally, there exist two mechanisms to reduce the reflectance. One is by the destructive interference between the incoming and the reflected light [8]. According to the light interference principle, the film with an optical thickness equal to a quarter of the wavelength can produce destructive interference. Silicon nitride Si3 N4 is the standard industrial material that is used as a quarter antireflection coating [5]. However, with this method, the antireflection effect is only for one specific wavelength and for one specific angle of the incoming light. Another mechanism to minimize reflectance is by gradually changing the refractive index in the interface [9,10]. The reflection R at the interface is governed by the formula R n1 − n2 2 ∕n1 n2 2 , where n1 and n2 are the refractive indices of the two materials. From the above equation, one can see that the closer the values of n1 and n2 are, the less optical loss there is. A graded-index structure is therefore a good method to use in the reduction of reflection at the interface. The graded-index structure can be realized by inserting a transparent material with high refractive index, such as titanium dioxide TiO2 [11], or by fabricating a nanostructure in the interface, such as textured roughness [12], nanowires [13], or nano-cones [14–16], 0146-9592/15/071290-04$15.00/0
which can form a lateral graded index at the interface. Despite the work has been done in the antireflection field, about 10% reflection losses still exist. Even for the high-efficiency micromorph tandem solar cells, the value is about 6.5%, according to the results we measured. The reason for this is that most efforts have been focused on the internal layers, whereas the mismatch between air and the surface of solar cells has attracted less attention. Very recently, Escarré et al. proposed a new way to reduce the primary reflectance at the air/glass interface [2]. In this approach, the pyramidal textures were imprinted on the front of the glass substrates that were used for suppressing the reflection. For high-efficiency micromorph solar cells, a respectable total current gain has been achieved by applying such structures. However, the feature size of pyramidal textures depends on the selective anisotropic etching of silicon, which may prevent the cell device from reaching its optimum performance potential [5]. Moreover, the mechanical stability of the imprinted texture, such as its adhesion to glass, should be taken into consideration in the case of module application. In order to circumvent this key limitation, a geometric light-coupling scheme that uses micrometer-scale hemispherical pit arrays (HPAs) is presented in this Letter. Compared to the pyramidal textures, one major advantage of this approach is that the hemispherical microstructures can be directly patterned on the glass substrates. The shape, size, and array pitch of the periodic structure can be easily controlled for the cell device application. Here, a design of HPAs is demonstrated to have both broadband and an omnidirectional antireflective effect, and we show a dramatic photocurrent enhancement of micromorph tandem cells by applying such microstructure arrays. HPAs can be fabricated on quartz glass substrates via a simple and low-cost wet etching method. The preparation flow includes three pattern transfers. First, a 100 nm chromium (Cr) was deposited on the quartz glass. Then, the glass was coated with a photoresist layer. Afterwards, the sample was exposed to ultraviolet light under a © 2015 Optical Society of America
April 1, 2015 / Vol. 40, No. 7 / OPTICS LETTERS
quartz glass mask with specific hole arrays. After developing, the patterns were transferred onto the photoresist layer. The Cr layer without the photoresist protection was first chemically etched in a de-chroming solution, and then the patterns were transferred onto the Cr layer. Finally, the quartz glass sample was dipped into a buffered hydrofluoric acid solution, and the HPA surface was obtained through isotropic etching. The period and depth of the hemispheric pit can be optimized by varying the mask size and the isotropic etching time in the buffered hydrofluoric acid solution. In this Letter, a mask with a pitch of 10 μm and a diameter of 5 μm was used, since textures with feature sizes of 10 μm are required in accordance with the spectral range of solar interest [2]. The time spent in the acid solutions was set at 5 min. To confirm the antireflection effect of the HPAs for practical applications in solar cells, micromorph tandem thin film silicon solar cells in a superstrate configuration were grown on two different substrates: a HPAs glass substrate and a bare glass substrate. Here, the bare glass served as a reference. The layers of the devices were deposited using a multi-chamber plasma-enhanced chemical vapor deposition system, and with following structure [see Fig. 1(d)]: glass (HPAs/bare)/ZnO (texture etched)/a-Si(p-i-n)/μc-Si (p-i-n)/ZnO:Al/Ag. The thicknesses of the a-Si top cell and the μc-Si bottom cell are 0.26 μm and 1.2 μm, respectively. In this cell configuration, the transparent ZnO film and the ZnO:Al/Ag layer are used as front and rear electrodes, respectively. As presented in Fig. 1(d), the electrodes are extracted with a conductive silver paste and connected by the wires for the purposes of testing. Figure 1(a) is an entity picture of the corresponding micromorph tandem solar cell (20 mm × 20 mm). The surface morphology of the micrometer-sized hemispherical pit textures, as shown in Fig. 1(b), was examined by means of a field emission scanning electron microscope. The reflectance of the cells was measured using a UV/VIS/NIR spectrophotometer equipped with an integrating sphere (PerkinElmer, Lambda 950). The external quantum efficiency (EQE) of the top cell (a-Si) and bottom cell (μc-Si) was measured under red and blue bias light illumination, respectively (Bentham PVE 300). The electrical
Fig. 1. (a) The photograph of a a-Si:H/μc-Si:H tandem solar cell incorporated into the HPAs glass substrate. (b) The SEM image of the fabricated HPAs. (c) The details of a hemisphere pit. The lines drawn in the graph are guides for the eyes. (d) Schematic representation of the micromorph tandem thin film silicon solar cells with HPAs. At the cell edges, a portion of Si material (the dashed yellow rectangle box, ∼1 mm in width) is used as a sacrificial layer to avoid a short circuit between the front and rear electrodes.
1291
properties of the current-voltage curves were characterized by a solar simulator under standard test conditions (AM 1.5 G, 100 mW∕cm2 at 25°C). Figure 2 shows the measured performance of the a-Si: H/μc-Si:H tandem solar cells with two different surface structures. From Fig. 2(a), it is obvious that the total reflectance from the cell device with HPAs (the red line) decreased dramatically compared to the device with the flat glass substrate (the dark line). Over the broadband wavelength, the reflectance reduced from about 6.5% to less than 1.5%. This superior reflection reduction phenomenon demonstrates that HPAs surface structure increased the optical absorption of the solar cells. The antireflection characteristics of the periodic air/glass interface can be explained by the following reasons. When the light strikes an interface, one part of the light is transmitted, and the other is reflected. For the flat air/glass interface, the proportion of the reflected light is about 5%. However, when the light is reflected from the HPAs surface, as opposed to the bare surface, the large-scale hemispherical pit allows the reflected light to have a second opportunity to interact with the surface of the glass and to be refracted into the absorbing layer of the device. The light experiences double rebounds in the hemispherical pit, decreasing its reflectance from R to R2 . Figure 3(a) shows how the incoming light interacts with the hemispherical surface structure. Considering the normal incidence case, there exists a critical angle, as shown in Fig. 3(a1), which makes a distinction between region S 1 (gray area) and S 2 (yellow area). When the incident
Fig. 2. Performance of a-Si:H/μc-Si:H solar cells with two different surface structures. (a) Total reflectance from the cell devices’ surface. (b) Quantum efficiency curves for the top and bottom cell components. (c) The current-voltage curves under the AM 1.5 spectrum.
1292
OPTICS LETTERS / Vol. 40, No. 7 / April 1, 2015
light hits the area of S 1 [see Fig. 3(a2)], the corresponding reflected light escapes directly from the cell devices’ surface. In contrast, the incident light interacting with the area of S 2 allows the corresponding reflected light to satisfy the condition of having a double rebound in the hemispherical pit [see Fig. 3(a3)]. Therefore, the proportion P of the reflected light that experienced double rebounds is P
AreaS 2 AreaS 1 1− : AreaS 1 S 2 AreaS 1 S 2
Since AreaS 1 πl2OA , AreaS 1 S 2 πl2OB lOB lOP , one can deduce that P 1−
πl2OA l2OA 1 − 1 − sin2 θ; πl2OB l2OP
(1) and
(2)
where lOA , lOB , and lOP are the lengths of OA, OB, and OP, respectively, and θ is the critical angle. According to the simple geometric relation, one can obtain the value of the critical angle, which is 24.29°. This means that, for the HPAs surface, the value of P is 0.83 and about 83% part of incoming light will hit a second hemispherical facet, thereby reducing the optical reflection. The performance enhancement also comes from the diffraction scattering, since the diffracted light with large angles can be well confined inside the absorbing layers. We investigated the diffraction characteristics of a glass (surface texturing)/ ZnO (1.4 μm) substrate via simulation by applying a rigorous coupled-wave analysis method. Figure 3(b) illustrates the diffuse transmission and haze of the glass/ZnO with different surface texturing. It is seen that the diffuse transmission is significantly enhanced as the hemisphere size increases, especially in the red and near infrared regions. When the array periods were designed to be 10 μm (the size we fabricated in our experiments), the
Fig. 3. (a) Schematic representation of how light interacts with the hemispherical pit texture under different incidents on case. (b) The simulated diffuse transmission and haze as a function of the wavelength for the glass/ZnO with different surface texturing.
hemisphere geometry exhibits superior diffuse transmission and haze performance to the typical pyramidal texture (10 μm in lateral size, 5 μm in depth, and a 54.7° inclined plane). This proves its potential advantages as an excellent light-trapping device. The absorption enhancement in the solar cell is further confirmed by the EQE curves, as shown in Fig. 2(b). For the top cell component, the EQE curves show a pronounced improvement in the spectral range around 480 nm in the case of the cell with the HPAs surface. However, in the wavelength region shorter than 400 nm, the increment is not as high as expected. The optical losses, in which the enhanced light comes from the antireflection, could be attributed to the type of p absorbing layer. For the bottom cell component, interestingly, we observed that the interference fringes became weak with the HPAs, which is due to the double rebounds. When accompanied by double rebounds, the interference between the reflected and incident light was broken. The significant conclusion we can infer from the EQE spectrum is that, compared to the bare glass solar cell, both the top and bottom component cells with HPAs have a higher current. For the tandem solar cells, the output J sc is determined by the lowest current between the two component cells. That is, to achieve a better solar cell performance, the currents of the top and bottom cells should be enhanced simultaneously. The value of J sc deduced from the EQE spectrum in Fig. 2(b) shows that, with the HPAs glass substrate, the current increases from 11.23 to 11.61 mA∕cm2 for the top cell, and from 10.98 to 11.49 mA∕cm2 for the bottom cell. The relative increases for the top and bottom cells are 3.38% and 4.64%, respectively. Figure 2(c) illustrates the current-voltage characteristics of the two sample cell devices. The value of J sc of the HPAs cell is 4.7% higher than that of the bare substrate. These results are in excellent agreement with the EQE calculations discussed above, which leads to a cell efficiency increase from 11.67% to 12.23%. In contrast, there is slight variation observed in the fill factor (FF) and open circuit voltage (V oc ). That is because, compared to the bare case, the method discussed in this Letter improved cell efficiency by decreasing reflectance at the air/glass interface, but invited no modifications in the silicon layers and the preparation process. To further investigate the angle-independent optical properties of the HPAs surface structure, the efficiency of the solar cells was measured under different angles of incident light. Figure 4 shows the efficiency of the two different cell devices as a function of the incident angles. For all the measured angles (0–85°), the efficiency of the
Fig. 4. Cell efficiency of the a-Si:H/μc-Si:H solar cells as a function of the incident angle.
April 1, 2015 / Vol. 40, No. 7 / OPTICS LETTERS
cell devices with HPAs remain higher than that of the bare one. Beside, its efficiency degrades remarkably at the angle of 60°, whereas the angle for the reference is about 45°. This observation demonstrates that the influence of the incident angle of light on the HPAs device is weaker than that on the bare device. The independent angle is significant for commercial applications, since the solar light is incident on the cell devices over a wide range of angles during the whole day. In conclusion, we have demonstrated a simple and novel periodic microstructured surface to enable broadband absorption enhancement in a-Si:H/μc-Si:H tandem solar cells. The HPAs surface structures were fabricated on quartz glass substrates via the wet chemical method, and incorporated into thin film solar cells. With this design, the cell devices displayed broadband reflection suppression from 400 to 1200 nm, resulting in a reflectance as low as 1.5%. The performance improvement in the cell was confirmed by the results measured from the EQE and current-voltage curves. Compared to the bare substrate, the cell efficiency of the HPAs glass substrate increased from 11.67% to 12.23%, without any deterioration in V oc or FF. The excellent cell performance is a result of the antireflection provided by the periodic interface. Another advantage of the HPAs surface structures for solar cell application is the angle-independent property, which is crucial considering the wide range of angle incidences contained in sunlight. This approach to enhancing absorption is not strictly limited to a-Si:H/μc-Si:H tandem solar cells, but provides an efficient light management strategy for a wide variety of photovoltaic devices. This research was supported by the State High-Tech Development Plan of China (863 Program, Grant No. 2011AA050518)
1293
References 1. A. Bessonov, Y. Cho, S.-J. Jung, E.-A. Park, E.-S. Hwang, J.-W. Lee, M. Shin, and S. Lee, Sol. Energy Mater. Sol. Cells 95, 2886 (2011). 2. J. Escarre, K. Soderstrom, M. Despeisse, S. Nicolay, C. Battaglia, G. Bugnon, L. Ding, F. Meillaud, F. Haug, and C. Ballif, Sol. Energy Mater. Sol. Cells 98, 185 (2012). 3. S. Pillai, K. R. Catchpole, T. Trupke, and M. A. Green, J. Appl. Phys. 101, 093105 (2007). 4. C. Das, A. Lambertz, J. Huepkes, W. Reetz, and F. Finger, Appl. Phys. Lett. 92, 053509 (2008). 5. C. Sun, P. Jiang, and B. Jiang, Appl. Phys. Lett. 92, 061112 (2008). 6. Y. Lee, D. S. Ruby, D. W. Peters, B. B. McKenzie, and J. W. P. Hsu, Nano Lett. 8, 1502 (2008). 7. S. Jang, Y. Song, J. Yu, C. I. Yeo, and Y. Lee, Opt. Lett. 36, 253 (2011). 8. Y. Song, S. Jang, J. Yu, and Y. Lee, Small 6, 984 (2010). 9. S. Jeong, S. Wang, and Y. Cui, J. Vac. Sci. Technol. 30, 060801 (2012). 10. Y. Song, J. Yu, and Y. Tak Lee, Opt. Lett. 35, 276 (2010). 11. T. Fujibayashi, T. Matsui, and M. Kondo, Appl. Phys. Lett. 88, 183508 (2006). 12. C. Battaglia, K. Söderström, J. Escarré, F. Haug, D. Dominé, P. Cuony, M. Boccard, G. Bugnon, C. Denizot, M. Despeisse, A. Feltrin, and C. Ballif, Appl. Phys. Lett. 96, 213504 (2010). 13. J. Zhu, Z. Yu, S. Fan, and Y. Cui, Mat. Sci. Eng. R 70, 330 (2010). 14. Q. Du, C. Kam, H. Demir, H. Yu, and X. Sun, Opt. Lett. 36, 1713 (2011). 15. C. Hsu, C. Battaglia, C. Pahud, Z. Ruan, F. Haug, S. Fan, C. Ballif, and Y. Cui, Adv. Energy Mater. 2, 628 (2012). 16. L. Zhou, X. Yu, and J. Zhu, Nano Lett. 14, 1093 (2014).
OPTICS LETTERS / Vol. 40, No. 7 / April 1, 2015
Performance enhancement in a-Si:H/μc-Si:H tandem solar cells with periodic microstructured surfaces Xiangqian Shen, Qingkang Wang,* Peihua Wangyang, Kun Huang, Le Chen, and Daiming Liu Key Laboratory for Thin Film and Microfabrication Technology of Ministry of Education, School of Electronic Information and Electrical Engineering, Shanghai Jiao Tong University, Shanghai 200240, China *Corresponding author: [email protected] Received December 17, 2014; revised February 9, 2015; accepted February 10, 2015; posted February 11, 2015 (Doc. ID 230934); published March 20, 2015 Here we report on an efficient light-coupling scheme with a periodic microstructured surface to enhance the performance of thin film silicon solar cells. The centerpiece of the surface structure is the hemispherical pit arrays (HPAs), which are fabricated using an inexpensive and scalable process. The integration of HPAs into micromorph tandem thin film silicon solar cells leads to superior broadband reflection suppression properties. With this design, the reflection losses of the tandem cell are reduced to only 1.5%. We demonstrate an efficiency increase from 11.67% to 12.23% compared to a conventional cell with a flat surface, with a 4.6% increase in short circuit current density. The surface microstructures reported here can be applied to a variety of photovoltaic devices to further improve their performance. © 2015 Optical Society of America OCIS codes: (350.6050) Solar energy; (290.5850) Scattering, particles; (160.4760) Optical properties. http://dx.doi.org/10.1364/OL.40.001290
In the area of low-cost photovoltaic applications, amorphous/microcrystalline silicon (a-Si:H/μc-Si:H) tandem solar cells are one of the more promising candidates, considering both their efficiency and stability [1,2]. Moreover, silicon (Si) is an environment-friendly material that is abundant in nature, is nontoxic, and is a wellestablished technology [3]. Ideally, a solar cell should absorb all incident photons with energies that are larger than their active materials’ band gap. However, the quantum efficiency of the solar cell cannot reach 100%, due to optical losses. Optical reflection at the interface due to refractive index mismatch (Si ∼3.5, transparent conductive oxide (TCO) ∼2.0, glass ∼1.5, air ∼1.0) is one of the major loss mechanisms [4–6] that seriously affects the solar cell efficiency. Therefore, antireflection coatings play an important role in enhancing the efficiency of solar cells [7]. Lots of work has been done to suppress reflection from the interface [8–10]. Conventionally, there exist two mechanisms to reduce the reflectance. One is by the destructive interference between the incoming and the reflected light [8]. According to the light interference principle, the film with an optical thickness equal to a quarter of the wavelength can produce destructive interference. Silicon nitride Si3 N4 is the standard industrial material that is used as a quarter antireflection coating [5]. However, with this method, the antireflection effect is only for one specific wavelength and for one specific angle of the incoming light. Another mechanism to minimize reflectance is by gradually changing the refractive index in the interface [9,10]. The reflection R at the interface is governed by the formula R n1 − n2 2 ∕n1 n2 2 , where n1 and n2 are the refractive indices of the two materials. From the above equation, one can see that the closer the values of n1 and n2 are, the less optical loss there is. A graded-index structure is therefore a good method to use in the reduction of reflection at the interface. The graded-index structure can be realized by inserting a transparent material with high refractive index, such as titanium dioxide TiO2 [11], or by fabricating a nanostructure in the interface, such as textured roughness [12], nanowires [13], or nano-cones [14–16], 0146-9592/15/071290-04$15.00/0
which can form a lateral graded index at the interface. Despite the work has been done in the antireflection field, about 10% reflection losses still exist. Even for the high-efficiency micromorph tandem solar cells, the value is about 6.5%, according to the results we measured. The reason for this is that most efforts have been focused on the internal layers, whereas the mismatch between air and the surface of solar cells has attracted less attention. Very recently, Escarré et al. proposed a new way to reduce the primary reflectance at the air/glass interface [2]. In this approach, the pyramidal textures were imprinted on the front of the glass substrates that were used for suppressing the reflection. For high-efficiency micromorph solar cells, a respectable total current gain has been achieved by applying such structures. However, the feature size of pyramidal textures depends on the selective anisotropic etching of silicon, which may prevent the cell device from reaching its optimum performance potential [5]. Moreover, the mechanical stability of the imprinted texture, such as its adhesion to glass, should be taken into consideration in the case of module application. In order to circumvent this key limitation, a geometric light-coupling scheme that uses micrometer-scale hemispherical pit arrays (HPAs) is presented in this Letter. Compared to the pyramidal textures, one major advantage of this approach is that the hemispherical microstructures can be directly patterned on the glass substrates. The shape, size, and array pitch of the periodic structure can be easily controlled for the cell device application. Here, a design of HPAs is demonstrated to have both broadband and an omnidirectional antireflective effect, and we show a dramatic photocurrent enhancement of micromorph tandem cells by applying such microstructure arrays. HPAs can be fabricated on quartz glass substrates via a simple and low-cost wet etching method. The preparation flow includes three pattern transfers. First, a 100 nm chromium (Cr) was deposited on the quartz glass. Then, the glass was coated with a photoresist layer. Afterwards, the sample was exposed to ultraviolet light under a © 2015 Optical Society of America
April 1, 2015 / Vol. 40, No. 7 / OPTICS LETTERS
quartz glass mask with specific hole arrays. After developing, the patterns were transferred onto the photoresist layer. The Cr layer without the photoresist protection was first chemically etched in a de-chroming solution, and then the patterns were transferred onto the Cr layer. Finally, the quartz glass sample was dipped into a buffered hydrofluoric acid solution, and the HPA surface was obtained through isotropic etching. The period and depth of the hemispheric pit can be optimized by varying the mask size and the isotropic etching time in the buffered hydrofluoric acid solution. In this Letter, a mask with a pitch of 10 μm and a diameter of 5 μm was used, since textures with feature sizes of 10 μm are required in accordance with the spectral range of solar interest [2]. The time spent in the acid solutions was set at 5 min. To confirm the antireflection effect of the HPAs for practical applications in solar cells, micromorph tandem thin film silicon solar cells in a superstrate configuration were grown on two different substrates: a HPAs glass substrate and a bare glass substrate. Here, the bare glass served as a reference. The layers of the devices were deposited using a multi-chamber plasma-enhanced chemical vapor deposition system, and with following structure [see Fig. 1(d)]: glass (HPAs/bare)/ZnO (texture etched)/a-Si(p-i-n)/μc-Si (p-i-n)/ZnO:Al/Ag. The thicknesses of the a-Si top cell and the μc-Si bottom cell are 0.26 μm and 1.2 μm, respectively. In this cell configuration, the transparent ZnO film and the ZnO:Al/Ag layer are used as front and rear electrodes, respectively. As presented in Fig. 1(d), the electrodes are extracted with a conductive silver paste and connected by the wires for the purposes of testing. Figure 1(a) is an entity picture of the corresponding micromorph tandem solar cell (20 mm × 20 mm). The surface morphology of the micrometer-sized hemispherical pit textures, as shown in Fig. 1(b), was examined by means of a field emission scanning electron microscope. The reflectance of the cells was measured using a UV/VIS/NIR spectrophotometer equipped with an integrating sphere (PerkinElmer, Lambda 950). The external quantum efficiency (EQE) of the top cell (a-Si) and bottom cell (μc-Si) was measured under red and blue bias light illumination, respectively (Bentham PVE 300). The electrical
Fig. 1. (a) The photograph of a a-Si:H/μc-Si:H tandem solar cell incorporated into the HPAs glass substrate. (b) The SEM image of the fabricated HPAs. (c) The details of a hemisphere pit. The lines drawn in the graph are guides for the eyes. (d) Schematic representation of the micromorph tandem thin film silicon solar cells with HPAs. At the cell edges, a portion of Si material (the dashed yellow rectangle box, ∼1 mm in width) is used as a sacrificial layer to avoid a short circuit between the front and rear electrodes.
1291
properties of the current-voltage curves were characterized by a solar simulator under standard test conditions (AM 1.5 G, 100 mW∕cm2 at 25°C). Figure 2 shows the measured performance of the a-Si: H/μc-Si:H tandem solar cells with two different surface structures. From Fig. 2(a), it is obvious that the total reflectance from the cell device with HPAs (the red line) decreased dramatically compared to the device with the flat glass substrate (the dark line). Over the broadband wavelength, the reflectance reduced from about 6.5% to less than 1.5%. This superior reflection reduction phenomenon demonstrates that HPAs surface structure increased the optical absorption of the solar cells. The antireflection characteristics of the periodic air/glass interface can be explained by the following reasons. When the light strikes an interface, one part of the light is transmitted, and the other is reflected. For the flat air/glass interface, the proportion of the reflected light is about 5%. However, when the light is reflected from the HPAs surface, as opposed to the bare surface, the large-scale hemispherical pit allows the reflected light to have a second opportunity to interact with the surface of the glass and to be refracted into the absorbing layer of the device. The light experiences double rebounds in the hemispherical pit, decreasing its reflectance from R to R2 . Figure 3(a) shows how the incoming light interacts with the hemispherical surface structure. Considering the normal incidence case, there exists a critical angle, as shown in Fig. 3(a1), which makes a distinction between region S 1 (gray area) and S 2 (yellow area). When the incident
Fig. 2. Performance of a-Si:H/μc-Si:H solar cells with two different surface structures. (a) Total reflectance from the cell devices’ surface. (b) Quantum efficiency curves for the top and bottom cell components. (c) The current-voltage curves under the AM 1.5 spectrum.
1292
OPTICS LETTERS / Vol. 40, No. 7 / April 1, 2015
light hits the area of S 1 [see Fig. 3(a2)], the corresponding reflected light escapes directly from the cell devices’ surface. In contrast, the incident light interacting with the area of S 2 allows the corresponding reflected light to satisfy the condition of having a double rebound in the hemispherical pit [see Fig. 3(a3)]. Therefore, the proportion P of the reflected light that experienced double rebounds is P
AreaS 2 AreaS 1 1− : AreaS 1 S 2 AreaS 1 S 2
Since AreaS 1 πl2OA , AreaS 1 S 2 πl2OB lOB lOP , one can deduce that P 1−
πl2OA l2OA 1 − 1 − sin2 θ; πl2OB l2OP
(1) and
(2)
where lOA , lOB , and lOP are the lengths of OA, OB, and OP, respectively, and θ is the critical angle. According to the simple geometric relation, one can obtain the value of the critical angle, which is 24.29°. This means that, for the HPAs surface, the value of P is 0.83 and about 83% part of incoming light will hit a second hemispherical facet, thereby reducing the optical reflection. The performance enhancement also comes from the diffraction scattering, since the diffracted light with large angles can be well confined inside the absorbing layers. We investigated the diffraction characteristics of a glass (surface texturing)/ ZnO (1.4 μm) substrate via simulation by applying a rigorous coupled-wave analysis method. Figure 3(b) illustrates the diffuse transmission and haze of the glass/ZnO with different surface texturing. It is seen that the diffuse transmission is significantly enhanced as the hemisphere size increases, especially in the red and near infrared regions. When the array periods were designed to be 10 μm (the size we fabricated in our experiments), the
Fig. 3. (a) Schematic representation of how light interacts with the hemispherical pit texture under different incidents on case. (b) The simulated diffuse transmission and haze as a function of the wavelength for the glass/ZnO with different surface texturing.
hemisphere geometry exhibits superior diffuse transmission and haze performance to the typical pyramidal texture (10 μm in lateral size, 5 μm in depth, and a 54.7° inclined plane). This proves its potential advantages as an excellent light-trapping device. The absorption enhancement in the solar cell is further confirmed by the EQE curves, as shown in Fig. 2(b). For the top cell component, the EQE curves show a pronounced improvement in the spectral range around 480 nm in the case of the cell with the HPAs surface. However, in the wavelength region shorter than 400 nm, the increment is not as high as expected. The optical losses, in which the enhanced light comes from the antireflection, could be attributed to the type of p absorbing layer. For the bottom cell component, interestingly, we observed that the interference fringes became weak with the HPAs, which is due to the double rebounds. When accompanied by double rebounds, the interference between the reflected and incident light was broken. The significant conclusion we can infer from the EQE spectrum is that, compared to the bare glass solar cell, both the top and bottom component cells with HPAs have a higher current. For the tandem solar cells, the output J sc is determined by the lowest current between the two component cells. That is, to achieve a better solar cell performance, the currents of the top and bottom cells should be enhanced simultaneously. The value of J sc deduced from the EQE spectrum in Fig. 2(b) shows that, with the HPAs glass substrate, the current increases from 11.23 to 11.61 mA∕cm2 for the top cell, and from 10.98 to 11.49 mA∕cm2 for the bottom cell. The relative increases for the top and bottom cells are 3.38% and 4.64%, respectively. Figure 2(c) illustrates the current-voltage characteristics of the two sample cell devices. The value of J sc of the HPAs cell is 4.7% higher than that of the bare substrate. These results are in excellent agreement with the EQE calculations discussed above, which leads to a cell efficiency increase from 11.67% to 12.23%. In contrast, there is slight variation observed in the fill factor (FF) and open circuit voltage (V oc ). That is because, compared to the bare case, the method discussed in this Letter improved cell efficiency by decreasing reflectance at the air/glass interface, but invited no modifications in the silicon layers and the preparation process. To further investigate the angle-independent optical properties of the HPAs surface structure, the efficiency of the solar cells was measured under different angles of incident light. Figure 4 shows the efficiency of the two different cell devices as a function of the incident angles. For all the measured angles (0–85°), the efficiency of the
Fig. 4. Cell efficiency of the a-Si:H/μc-Si:H solar cells as a function of the incident angle.
April 1, 2015 / Vol. 40, No. 7 / OPTICS LETTERS
cell devices with HPAs remain higher than that of the bare one. Beside, its efficiency degrades remarkably at the angle of 60°, whereas the angle for the reference is about 45°. This observation demonstrates that the influence of the incident angle of light on the HPAs device is weaker than that on the bare device. The independent angle is significant for commercial applications, since the solar light is incident on the cell devices over a wide range of angles during the whole day. In conclusion, we have demonstrated a simple and novel periodic microstructured surface to enable broadband absorption enhancement in a-Si:H/μc-Si:H tandem solar cells. The HPAs surface structures were fabricated on quartz glass substrates via the wet chemical method, and incorporated into thin film solar cells. With this design, the cell devices displayed broadband reflection suppression from 400 to 1200 nm, resulting in a reflectance as low as 1.5%. The performance improvement in the cell was confirmed by the results measured from the EQE and current-voltage curves. Compared to the bare substrate, the cell efficiency of the HPAs glass substrate increased from 11.67% to 12.23%, without any deterioration in V oc or FF. The excellent cell performance is a result of the antireflection provided by the periodic interface. Another advantage of the HPAs surface structures for solar cell application is the angle-independent property, which is crucial considering the wide range of angle incidences contained in sunlight. This approach to enhancing absorption is not strictly limited to a-Si:H/μc-Si:H tandem solar cells, but provides an efficient light management strategy for a wide variety of photovoltaic devices. This research was supported by the State High-Tech Development Plan of China (863 Program, Grant No. 2011AA050518)
1293
References 1. A. Bessonov, Y. Cho, S.-J. Jung, E.-A. Park, E.-S. Hwang, J.-W. Lee, M. Shin, and S. Lee, Sol. Energy Mater. Sol. Cells 95, 2886 (2011). 2. J. Escarre, K. Soderstrom, M. Despeisse, S. Nicolay, C. Battaglia, G. Bugnon, L. Ding, F. Meillaud, F. Haug, and C. Ballif, Sol. Energy Mater. Sol. Cells 98, 185 (2012). 3. S. Pillai, K. R. Catchpole, T. Trupke, and M. A. Green, J. Appl. Phys. 101, 093105 (2007). 4. C. Das, A. Lambertz, J. Huepkes, W. Reetz, and F. Finger, Appl. Phys. Lett. 92, 053509 (2008). 5. C. Sun, P. Jiang, and B. Jiang, Appl. Phys. Lett. 92, 061112 (2008). 6. Y. Lee, D. S. Ruby, D. W. Peters, B. B. McKenzie, and J. W. P. Hsu, Nano Lett. 8, 1502 (2008). 7. S. Jang, Y. Song, J. Yu, C. I. Yeo, and Y. Lee, Opt. Lett. 36, 253 (2011). 8. Y. Song, S. Jang, J. Yu, and Y. Lee, Small 6, 984 (2010). 9. S. Jeong, S. Wang, and Y. Cui, J. Vac. Sci. Technol. 30, 060801 (2012). 10. Y. Song, J. Yu, and Y. Tak Lee, Opt. Lett. 35, 276 (2010). 11. T. Fujibayashi, T. Matsui, and M. Kondo, Appl. Phys. Lett. 88, 183508 (2006). 12. C. Battaglia, K. Söderström, J. Escarré, F. Haug, D. Dominé, P. Cuony, M. Boccard, G. Bugnon, C. Denizot, M. Despeisse, A. Feltrin, and C. Ballif, Appl. Phys. Lett. 96, 213504 (2010). 13. J. Zhu, Z. Yu, S. Fan, and Y. Cui, Mat. Sci. Eng. R 70, 330 (2010). 14. Q. Du, C. Kam, H. Demir, H. Yu, and X. Sun, Opt. Lett. 36, 1713 (2011). 15. C. Hsu, C. Battaglia, C. Pahud, Z. Ruan, F. Haug, S. Fan, C. Ballif, and Y. Cui, Adv. Energy Mater. 2, 628 (2012). 16. L. Zhou, X. Yu, and J. Zhu, Nano Lett. 14, 1093 (2014).
