**3. Material characterization of hybrid solar cells**

Relating to the configuration of hybrid SCs like HJ SCs or dye-sensitized SCs, various materials have been suggested by research groups. The BHJ devices were characterized by an interpenetrating network of donor and acceptor materials, providing a large interface area where photo-induced excitons could efficiently dissociate into separated electrons and holes. However, the interpenetrating network cannot be easily formed in the blended mixture. In addition, the organic materials are not good in carrier transport. Thus, the power conversion efficiency is still limited by the low dissociation probability of excitons and the inefficient hopping carrier transport (Huang et al., 2009, as cited in Sirringhaus et al.,1999; Shaw et al.,2008). Semiconductor nanostructures are used to be combined with the organic materials to provide not only a large interface area between organic and inorganic components for exciton dissociation but also fast electron transport in semiconductors. Therefore, many research groups combined organic materials with semiconductor nanostructures to overcome the drawbacks of the organic solar cells. Many inorganic nanowire (NW) had been experimented for this purpose, including CdTe, CdS, CdSe, ZnO, and TiO2 NWs (Huang et al., 2009, as cited in Kang & Kim, 2006). Totally, BHJ hybrid SCs itself have been demonstrated in various inorganic materials such as CdSe nanodots, nanorods and tetrapods, TiO2, ZnO, ZnS nanoparticles, CuInS2, CuInSe2, CuPc, CdS, SnS, CIS , PbSe or PbS nanocrystals , HgTe ncs, Si NWs, Si ncs (Chandrasekaran et al., 2010 as cited in Kwong et al.,2004, Arici et al., 2004 Greenham et al., 1996 Qi et al., 2005; Choudhury et al., 2005), etc. which act as acceptors and polymer materials acting as donors are P3HT, PPHT, P3OT, P3BT, P3MeT (Lin et al., 2006), MDMO-PPV, MEH-PPV, MOPPV, etc. (Chandrasekaran et al., 2010). However,CdTe, CdS,and CdSe materials are harmful to the environment, while ZnO and TiO2 have a bandgap higher than 3eV and so cannot effectively absorb the solar spectrum. To overcome this, SiNWs are suitable for this application because they are environmental friendly and have high absorption coefficient in the infrare dregion (Huang et al., 2009).

J.Haung et al. (2009) reported the fabrication of the SiNW/P3HT:PCBM blend hybrid SCs using the SiNW transfer technique.

Fig. 4. A schematic of the hybrid SC using SiNWs and P3HT:PCBM blend

Their investigation showed that after introducing the SiNWs, the Jsc increases from 7.17 to 11.61 mA/cm2 and η increases from 1.21% to 1.91% (Haung et al., 2009). This increase is due to this fact that the NWs act as a direct path for transport of charge without the presence of grain boundaries.( Movla et al., 2010b).

Relating to the configuration of hybrid SCs like HJ SCs or dye-sensitized SCs, various materials have been suggested by research groups. The BHJ devices were characterized by an interpenetrating network of donor and acceptor materials, providing a large interface area where photo-induced excitons could efficiently dissociate into separated electrons and holes. However, the interpenetrating network cannot be easily formed in the blended mixture. In addition, the organic materials are not good in carrier transport. Thus, the power conversion efficiency is still limited by the low dissociation probability of excitons and the inefficient hopping carrier transport (Huang et al., 2009, as cited in Sirringhaus et al.,1999; Shaw et al.,2008). Semiconductor nanostructures are used to be combined with the organic materials to provide not only a large interface area between organic and inorganic components for exciton dissociation but also fast electron transport in semiconductors. Therefore, many research groups combined organic materials with semiconductor nanostructures to overcome the drawbacks of the organic solar cells. Many inorganic nanowire (NW) had been experimented for this purpose, including CdTe, CdS, CdSe, ZnO, and TiO2 NWs (Huang et al., 2009, as cited in Kang & Kim, 2006). Totally, BHJ hybrid SCs itself have been demonstrated in various inorganic materials such as CdSe nanodots, nanorods and tetrapods, TiO2, ZnO, ZnS nanoparticles, CuInS2, CuInSe2, CuPc, CdS, SnS, CIS , PbSe or PbS nanocrystals , HgTe ncs, Si NWs, Si ncs (Chandrasekaran et al., 2010 as cited in Kwong et al.,2004, Arici et al., 2004 Greenham et al., 1996 Qi et al., 2005; Choudhury et al., 2005), etc. which act as acceptors and polymer materials acting as donors are P3HT, PPHT, P3OT, P3BT, P3MeT (Lin et al., 2006), MDMO-PPV, MEH-PPV, MOPPV, etc. (Chandrasekaran et al., 2010). However,CdTe, CdS,and CdSe materials are harmful to the environment, while ZnO and TiO2 have a bandgap higher than 3eV and so cannot effectively absorb the solar spectrum. To overcome this, SiNWs are suitable for this application because they are environmental friendly and have high absorption coefficient in

J.Haung et al. (2009) reported the fabrication of the SiNW/P3HT:PCBM blend hybrid SCs

Their investigation showed that after introducing the SiNWs, the Jsc increases from 7.17 to 11.61 mA/cm2 and η increases from 1.21% to 1.91% (Haung et al., 2009). This increase is due to this fact that the NWs act as a direct path for transport of charge without the presence of

Fig. 4. A schematic of the hybrid SC using SiNWs and P3HT:PCBM blend

**3. Material characterization of hybrid solar cells** 

the infrare dregion (Huang et al., 2009).

grain boundaries.( Movla et al., 2010b).

using the SiNW transfer technique.

Al

SiNW

ITO

P3HT:PCBM

PEDOT:PSS

Fig. 5. The current density–voltage characteristics for the SCs with and without the SiNWs under simulated AM1.5 illumination. Reprinted with permission from Solar Energy Materials & Solar Cells Vol.93, Huang, J. et al. Well-aligned single-crystalline silicon nanowire hybrid solar cells on glass, pp. 621–624 © 2009, Elsevier.

More precisely, the results clearly indicate that combination of the SiNWs and P3HT:PCBM blend is an attractive route to obtain high Jsc and efficiencies by improving the optical absorption, dissociation of excitons, and the electron transport. Silicon wafer is commercially available and cheap. SiNWs can be fabricated at low temperature from solution processing without any vacuum equipment or high-temperature processing. In addition, this transfer method for SiNWs is simple and fast. It is not a laborious way.This method is suitable for plastic SCs because it can be processed fast, is cheap and simple ( Haung et al., 2009).

Similar work was done by G. Kalita et al. (2009) for demonstrating hybrid SCs using Si NWs and polymer incorporating MWNTS. This fabricated device with the structure of Au/P3OT+O-MWNTS/n-Si NWs marked a conversion efficiency of 0.61% (Bredol et al., 2009). Another study was done by C. Y. Liu et al. (2009) about fabricating the hybrid SCs on blends of Si ncs and P3HT (Liu et al., 2009). Also, V. Svrcek et al. (2009), investigated the photoelectric property of BHJ SC based on Si-ncs and P3HT. They came into conclusion that I–V characteristic enhanced when BHJ was introduced into TiO2 nanotube (nt). The arrangement of Si-ncs/P3HT BHJ within ordered TiO2 nt perpendicular to the contact facilitated excition separation and charge transfer along nts (Chandrasekaran et al., 2010, as cited in Svrcek et al., 2009).

A new approach for hybrid metal-insulator-semiconductor (MIS) Si solar cells is adopted by the Institute of Fundamental Problems for High Technology, Ukrainian Academy of Sciences. In this technique, the porous silicon layers are created on both sides of single crystal wafers by chemical etching before an improved MIS cell preparation process. The porous Si exhibits unique properties. It works like a sunlight concentrator, light scattering diffuser and reemitter of sunlight as well as an electrical isolator in the multilayer Si structure. The most important advantage of using porous Si in SCs is its band gap which behaves as a direct band gap semiconductor with large quantum efficiency and may be adjusted for optimum sunlight absorption. Employing a specific surface modification, porous Si improves the PV efficiency in UV and NIR regions of solar spectra (Tuzun et al.,

Hybrid Solar Cells Based on Silicon 405

nanoparticles. A polymeric gel electrolyte is considered as a compromise between liquid electrolytes and hole conductors in quasi solid state dye-sensitized SCs (Günes & Sariciftci,2008, as cited in Nogueira et al.,2004;Murphy,1998; Megahed & Scosati,1995) . A mixture of NaI, ethylene carbonate, propylene carbonate and polyacrylonitrile was reported by Cao et al. [55]. Poly (vinylidenefluoride- co-hexafluoropropylene) (PVDF-HFP) used to solidify 3-methoxypropionitrile (MPN) was utilized by Wang et al. (Wang et al.,2004 ). The last efficiency for dye-sensitized reported by Sharp Corporation is about 10.4 which stands in lower rank in comparison with crystalline Si showing efficiency of approximately 25% (

Silicon is the leading material used in microelectronic technology and shows novel photoelectrochemical properties in electrolyte solutions ( Wang et al.,2010). Now and before Si-based cells especially crystalline Si has shown higher efficiencies. The last efficieny reported by UNSW PERL is 25% which exceeds other types of Si-based SCs (Green et al.,

Crystalline silicon (c-Si) is an extremely well suited material for terrestrial photovoltaics (PV). It is non-toxic and abundant (25% of the Earth's crust), has excellent electronic, chemical and mechanical properties, forms a simple monoelemental semiconductor that has an almost ideal bandgap (1.1 eV) for terrestrial PV, and gives long-term stable SCs and modules. Furthermore, it is the material of choice in the microelectronics industry, ensuring that a large range of processing equipment exists and is readily available. Given these properties of c-Si, it is not surprising that almost all (>90%) terrestrial PV modules sold today use Si wafer SCs. However, the fabrication of Si wafers is both material and energy intensive. Therefore, there is a need for a less material intensive c-Si technology calling a thin-film technology. Besides cost savings on the materials side, thin-film technologies offer the additional benefits of large-area processing (unit size about 1 m2) on a supporting material, enabling monolithic construction and cell interconnection. The recent c-Si thin-film PV approaches can broadly be classified as follows: (i) fabrication of thin, long stripes of c-Si material from thick single crystalline Si wafers (''ultrathin slicing''); (ii) growth of c-Si thin-

Recently photoelectrochemical (PEC) SCs based on 1D single crystalline semiconductor micro/nanostructures have attracted intense attention as they may rival the nanocrystalline dye-sensitized SCs (Wang et al., 2010, as cited in Law et al., 2005; Peng et al., 2008; Baxter & Aydil, 2005; Jiang et al., 2008; Mor et al., 2006; Hwang et al., 2009; Dalchiele et al., 2009; Goodey et al., 2007; Maiolo et al., 2007). Therefore, X.Wang et al. (2010) proposed single crystalline ordered silicon wire/Pt nanoparticle hybrids for solar energy harvesting. In this configuration, wafer-scale Si wire arrays are fabricated by the combination of ultraviolet lithography (UVL) and metal-assisted etching. This method emphasizes that the etching technology forms SiNWs non-contaminated by catalyst material while SiNWs by vaporliquid-solid (VLS) is generally contaminated by gold (Wang et al., 2010, as cited in Allen et al., 2008). It is found that PtNPs modified Si wire electrochemical PV cell generated significantly enhanced photocurrents and larger fill factors. The overall conversion efficiency of PtNPs modified Si wire PEC solar cell is up to 4.7%. Array of p-type Si wire modified with PtNPs also shows significant improvement for water splitting (Wang et al.,

Green et al., 2011).

2011, Zhao et al., 1998).

2010).

**3.1 Crystalline silicon thin film solar cells** 

films on native or foreign supporting materials (Aberle, 2006).

2006, as cited in Tiris et al., 2003). In this approach, due to high quality starting materials and rapid low-temperature (<800 0C) processing a high minority carrier life time is attainable; this, in turn, gives rise to a high photogenerated current collection. Therefore, the SCs with efficiencies above 15% have been obtained under AM1.5 condition (under 100 mW/cm2 illumination at 25 0C) (Tuzun et al., 2006).

Another study was reported by J. Ackermann et al. (2002) about the growth of quaterthiophene (4T), a linear conjugated oligomer of thiophene behaving as a p-type semiconductor, on n-doped GaAs and Si substrates to form hybrid HJ SCs. This study shows that in the case of Si as substrate there is almost defect-free high ordered films with grain sizes of several micrometers up to a film thickness of 250 mm (Ackermann et al., 2002). K.Yamamoto et al. (2001) investigated a-Si/poly-Si hybrid (stacked) SC paying attention to the stabilized efficiency, since a-Si has photo degradation, while a poly-Si is stable. This tandem cell exhibited a stabilized efficiency of 11.3%. Also, this research group prepared three-stacked cell of a-Si:H/poly-Si/poly-Si(triple), which will be less sensitive to degradation by using the thinner a-Si. This triple cell showed a stabilized efficiency of 12%( Yamamoto et al., 2001) . Their reasons for applying poly-Si are (i) high growth rate (ii) large area and uniform deposition at the same time and (iii) monolithic series interconnection. In addition, for enhancing the absorption, they suggested natural surface texture and a back reflector (See Fig .6.) (Yamamoto et al., 2001).

Fig. 6. Schematic view of thin film poly-Si solar cell with natural surface texture and enhanced absorption with back reflector structure. Reprinted with permission from Solar Energy Materials & Solar Cells Vol.93, Huang, J. et al. Well-aligned single-crystalline silicon nanowire hybrid solar cells on glass, pp. 621–624 © 2009, Elsevier.

Apart from BHJ SCs, solid state dye-sensitized SCs can experience various materials. CuI, CuBr, CuSCN, MgO (Tennakone et al.,2001) can be replacements for liquid crystals as inorganic p-type semiconductors. Also, 2,2',7,7'-tetranis(N,N-di-p-methoxyphenylamine)9,9'-spirobifluorene (OMETAD) can replace the liquid crystals as organic p-type semiconductor due to their low cost processability (Bach et al.,1995). Poly (3 alkylthiophenes) were used to replace the liquid electrolyte by Sicot et al. (Sicot et al., 1991) and Gebeyehu et al. (Gebeyenha et al.,2002a,2002b) as conjugated polymers although high molecular weight polymers cast from solution, do not penetrate into the pores of the

2006, as cited in Tiris et al., 2003). In this approach, due to high quality starting materials and rapid low-temperature (<800 0C) processing a high minority carrier life time is attainable; this, in turn, gives rise to a high photogenerated current collection. Therefore, the SCs with efficiencies above 15% have been obtained under AM1.5 condition (under 100

Another study was reported by J. Ackermann et al. (2002) about the growth of quaterthiophene (4T), a linear conjugated oligomer of thiophene behaving as a p-type semiconductor, on n-doped GaAs and Si substrates to form hybrid HJ SCs. This study shows that in the case of Si as substrate there is almost defect-free high ordered films with grain sizes of several micrometers up to a film thickness of 250 mm (Ackermann et al., 2002). K.Yamamoto et al. (2001) investigated a-Si/poly-Si hybrid (stacked) SC paying attention to the stabilized efficiency, since a-Si has photo degradation, while a poly-Si is stable. This tandem cell exhibited a stabilized efficiency of 11.3%. Also, this research group prepared three-stacked cell of a-Si:H/poly-Si/poly-Si(triple), which will be less sensitive to degradation by using the thinner a-Si. This triple cell showed a stabilized efficiency of 12%( Yamamoto et al., 2001) . Their reasons for applying poly-Si are (i) high growth rate (ii) large area and uniform deposition at the same time and (iii) monolithic series interconnection. In addition, for enhancing the absorption, they suggested natural surface texture and a back

Fig. 6. Schematic view of thin film poly-Si solar cell with natural surface texture and enhanced absorption with back reflector structure. Reprinted with permission from Solar Energy Materials & Solar Cells Vol.93, Huang, J. et al. Well-aligned single-crystalline silicon

Apart from BHJ SCs, solid state dye-sensitized SCs can experience various materials. CuI, CuBr, CuSCN, MgO (Tennakone et al.,2001) can be replacements for liquid crystals as inorganic p-type semiconductors. Also, 2,2',7,7'-tetranis(N,N-di-p-methoxyphenylamine)9,9'-spirobifluorene (OMETAD) can replace the liquid crystals as organic p-type semiconductor due to their low cost processability (Bach et al.,1995). Poly (3 alkylthiophenes) were used to replace the liquid electrolyte by Sicot et al. (Sicot et al., 1991) and Gebeyehu et al. (Gebeyenha et al.,2002a,2002b) as conjugated polymers although high molecular weight polymers cast from solution, do not penetrate into the pores of the

nanowire hybrid solar cells on glass, pp. 621–624 © 2009, Elsevier.

mW/cm2 illumination at 25 0C) (Tuzun et al., 2006).

reflector (See Fig .6.) (Yamamoto et al., 2001).

nanoparticles. A polymeric gel electrolyte is considered as a compromise between liquid electrolytes and hole conductors in quasi solid state dye-sensitized SCs (Günes & Sariciftci,2008, as cited in Nogueira et al.,2004;Murphy,1998; Megahed & Scosati,1995) . A mixture of NaI, ethylene carbonate, propylene carbonate and polyacrylonitrile was reported by Cao et al. [55]. Poly (vinylidenefluoride- co-hexafluoropropylene) (PVDF-HFP) used to solidify 3-methoxypropionitrile (MPN) was utilized by Wang et al. (Wang et al.,2004 ). The last efficiency for dye-sensitized reported by Sharp Corporation is about 10.4 which stands in lower rank in comparison with crystalline Si showing efficiency of approximately 25% ( Green et al., 2011).

### **3.1 Crystalline silicon thin film solar cells**

Silicon is the leading material used in microelectronic technology and shows novel photoelectrochemical properties in electrolyte solutions ( Wang et al.,2010). Now and before Si-based cells especially crystalline Si has shown higher efficiencies. The last efficieny reported by UNSW PERL is 25% which exceeds other types of Si-based SCs (Green et al., 2011, Zhao et al., 1998).

Crystalline silicon (c-Si) is an extremely well suited material for terrestrial photovoltaics (PV). It is non-toxic and abundant (25% of the Earth's crust), has excellent electronic, chemical and mechanical properties, forms a simple monoelemental semiconductor that has an almost ideal bandgap (1.1 eV) for terrestrial PV, and gives long-term stable SCs and modules. Furthermore, it is the material of choice in the microelectronics industry, ensuring that a large range of processing equipment exists and is readily available. Given these properties of c-Si, it is not surprising that almost all (>90%) terrestrial PV modules sold today use Si wafer SCs. However, the fabrication of Si wafers is both material and energy intensive. Therefore, there is a need for a less material intensive c-Si technology calling a thin-film technology. Besides cost savings on the materials side, thin-film technologies offer the additional benefits of large-area processing (unit size about 1 m2) on a supporting material, enabling monolithic construction and cell interconnection. The recent c-Si thin-film PV approaches can broadly be classified as follows: (i) fabrication of thin, long stripes of c-Si material from thick single crystalline Si wafers (''ultrathin slicing''); (ii) growth of c-Si thinfilms on native or foreign supporting materials (Aberle, 2006).

Recently photoelectrochemical (PEC) SCs based on 1D single crystalline semiconductor micro/nanostructures have attracted intense attention as they may rival the nanocrystalline dye-sensitized SCs (Wang et al., 2010, as cited in Law et al., 2005; Peng et al., 2008; Baxter & Aydil, 2005; Jiang et al., 2008; Mor et al., 2006; Hwang et al., 2009; Dalchiele et al., 2009; Goodey et al., 2007; Maiolo et al., 2007). Therefore, X.Wang et al. (2010) proposed single crystalline ordered silicon wire/Pt nanoparticle hybrids for solar energy harvesting. In this configuration, wafer-scale Si wire arrays are fabricated by the combination of ultraviolet lithography (UVL) and metal-assisted etching. This method emphasizes that the etching technology forms SiNWs non-contaminated by catalyst material while SiNWs by vaporliquid-solid (VLS) is generally contaminated by gold (Wang et al., 2010, as cited in Allen et al., 2008). It is found that PtNPs modified Si wire electrochemical PV cell generated significantly enhanced photocurrents and larger fill factors. The overall conversion efficiency of PtNPs modified Si wire PEC solar cell is up to 4.7%. Array of p-type Si wire modified with PtNPs also shows significant improvement for water splitting (Wang et al., 2010).

Hybrid Solar Cells Based on Silicon 407

cells with different light absorption characteristics are stacked together. This approach allows better characteristics to be obtained with existing materials and processes. The advantages of using a layered structure include the following: (1) it is possible to receive light by partitioning it over a wider spectral region, thereby using the light more effectively; (2) it is possible to obtain a higher open-circuit voltage; and (3) it is possible to suppress to some extent the rate of reduction in cell performances caused by photo-degradation phenomena that are observed when using a-silicon based materials. Therefore, they have engaged in thin film amorphous and microcrystalline (a-Si/μc-Si) stacked solar cell

The advantage of a high Jsc for our μc-Si Sc as mentioned before was applied to the stacked cell with the combination of a-Si cell to gain stabilized efficiency as the study done by K.Yamamoto et al. (2001) since a-Si has a photo-degradation while a μc-Si cell is stable. They have also prepared three stacked cell of a-Si:H/μc-Si/c-Si (triple), which will be less sensitive to degradation by using the thinner a-Si and they have investigated the stability of a-Si:H/μc-Si/μc-Si (triple) cell, too (Yamamoto et al., 2001, 2004). Some other three stacked Si-based SCs can be named such as a-Si/a-SiGe/a-SiGe(tandem) and a-Si/nc-Si/nc-Si (tandem) SCs. The last efficiency reported for a-Si/a-SiGe/a-SiGe(tandem) is about 10.4%

As a next generation of further high efficiency of SC, the new stacked thin film Si SC is proposed where the transparent inter-layer was inserted between a-Si and μc-Si layer to enhance a partial reflection of light back into the a-Si top cell (see fig.8.). This structure is called as internal light trapping enabling the increase of current of top cell without increasing the thickness of top cell, which leads less photo-degradation of stacked cell

Fig. 8. Schematic view of a-Si/poly-Si (μc-Si) stacked cells with an interlayer. Reprinted with permission from Solar Energy Vol. 77, Kenji Yamamoto et al., A high efficiency thin film

silicon solar cell and module, pp. 939–949 © 2004, Elsevier.

and for a-Si/nc-Si/nc-Si (tandem) is approximately 12.5% (Green et al., 2011).

(Yamamoto et al., 2004).

(Yamamoto et al., 2004).

### **3.2 Amorphous ( protocrystalline) and microcrystalline silicon solar cells**

The use of thin-film silicon for SCs is one of the most promising approaches to realize both high performance and low cost due to its low material cost, ease of manufacturing and high efficiency. Microcrystalline silicon (μc) SCs as a family of thin film SCs formed by plasma CVD at low temperature are assumed to have a shorter carrier lifetime than single-crystal cells, and it is common to employ a p–i–n structure including an internal electric field in the same way as an amorphous SC. These cells can be divided into p–i– n and n–i–p types according to the film deposition order, although the window layer of the SC is the p-type layer in both cases. A large difference is that the underlying layer of a p–i–n cell is the transparent p-type electrode, whereas the underlying layer of an n–i–p cell is the n-type back electrode. Light-trapping techniques are a way of increasing the performance of mc-SCs. This is a core technique for cells made from μc-silicon because—unlike a-silicon—it is essentially an indirect absorber with a low absorption coefficient. That is, the thickness of the Si film that forms the active layer in a mc-silicon SC is just a few μm, so it is not able to absorb enough incident light compared with SCs using ordinary crystalline substrates. As a result, it is difficult to obtain a high photoelectric current. Light trapping technology provides a means of extending the optical path of the incident light inside the SC by causing multiple reflections, thereby improving the light absorption in the active layer (Yamamoto et al., 2004). Light trapping in this method of categorizing the SCs according to p-i-n or n-i-p types, can be achieved in two ways: (1) by introducing a highly reflective layer at the back surface to reflect the incident light without absorption loss, and (2) by introducing a textured structure at the back surface of the thin-film Si SC (see Fig.7.) (Komatsu et al.,2002; Yamamoto et al.,2004).

Fig. 7. Cross-sections through light-trapping μc-silicon SC devices: (a) first generation (flat back reflector); (b) second generation (textured back reflector, thinner polycrystalline silicon layer). Reprinted with permission from Solar Energy Vol. 77, Kenji Yamamoto et al., A high efficiency thin film silicon solar cell and module, pp. 939–949 © 2004, Elsevier.

Although the μc-silicon cells formed at low temperature have a potential for high efficiently, their efficiency in single-cell structures is currently only about 10%, which is much lower than that of bulk polycrystalline cells. In order to achieve high efficiencies, Yamamoto et al.(2004) investigated the use of two-and three-stacked (hybrid) structures in which multiple

The use of thin-film silicon for SCs is one of the most promising approaches to realize both high performance and low cost due to its low material cost, ease of manufacturing and high efficiency. Microcrystalline silicon (μc) SCs as a family of thin film SCs formed by plasma CVD at low temperature are assumed to have a shorter carrier lifetime than single-crystal cells, and it is common to employ a p–i–n structure including an internal electric field in the same way as an amorphous SC. These cells can be divided into p–i– n and n–i–p types according to the film deposition order, although the window layer of the SC is the p-type layer in both cases. A large difference is that the underlying layer of a p–i–n cell is the transparent p-type electrode, whereas the underlying layer of an n–i–p cell is the n-type back electrode. Light-trapping techniques are a way of increasing the performance of mc-SCs. This is a core technique for cells made from μc-silicon because—unlike a-silicon—it is essentially an indirect absorber with a low absorption coefficient. That is, the thickness of the Si film that forms the active layer in a mc-silicon SC is just a few μm, so it is not able to absorb enough incident light compared with SCs using ordinary crystalline substrates. As a result, it is difficult to obtain a high photoelectric current. Light trapping technology provides a means of extending the optical path of the incident light inside the SC by causing multiple reflections, thereby improving the light absorption in the active layer (Yamamoto et al., 2004). Light trapping in this method of categorizing the SCs according to p-i-n or n-i-p types, can be achieved in two ways: (1) by introducing a highly reflective layer at the back surface to reflect the incident light without absorption loss, and (2) by introducing a textured structure at the back surface of the thin-film Si SC (see Fig.7.) (Komatsu et al.,2002;

Fig. 7. Cross-sections through light-trapping μc-silicon SC devices: (a) first generation (flat back reflector); (b) second generation (textured back reflector, thinner polycrystalline silicon layer). Reprinted with permission from Solar Energy Vol. 77, Kenji Yamamoto et al., A high

Although the μc-silicon cells formed at low temperature have a potential for high efficiently, their efficiency in single-cell structures is currently only about 10%, which is much lower than that of bulk polycrystalline cells. In order to achieve high efficiencies, Yamamoto et al.(2004) investigated the use of two-and three-stacked (hybrid) structures in which multiple

efficiency thin film silicon solar cell and module, pp. 939–949 © 2004, Elsevier.

**3.2 Amorphous ( protocrystalline) and microcrystalline silicon solar cells** 

Yamamoto et al.,2004).

cells with different light absorption characteristics are stacked together. This approach allows better characteristics to be obtained with existing materials and processes. The advantages of using a layered structure include the following: (1) it is possible to receive light by partitioning it over a wider spectral region, thereby using the light more effectively; (2) it is possible to obtain a higher open-circuit voltage; and (3) it is possible to suppress to some extent the rate of reduction in cell performances caused by photo-degradation phenomena that are observed when using a-silicon based materials. Therefore, they have engaged in thin film amorphous and microcrystalline (a-Si/μc-Si) stacked solar cell (Yamamoto et al., 2004).

The advantage of a high Jsc for our μc-Si Sc as mentioned before was applied to the stacked cell with the combination of a-Si cell to gain stabilized efficiency as the study done by K.Yamamoto et al. (2001) since a-Si has a photo-degradation while a μc-Si cell is stable. They have also prepared three stacked cell of a-Si:H/μc-Si/c-Si (triple), which will be less sensitive to degradation by using the thinner a-Si and they have investigated the stability of a-Si:H/μc-Si/μc-Si (triple) cell, too (Yamamoto et al., 2001, 2004). Some other three stacked Si-based SCs can be named such as a-Si/a-SiGe/a-SiGe(tandem) and a-Si/nc-Si/nc-Si (tandem) SCs. The last efficiency reported for a-Si/a-SiGe/a-SiGe(tandem) is about 10.4% and for a-Si/nc-Si/nc-Si (tandem) is approximately 12.5% (Green et al., 2011).

As a next generation of further high efficiency of SC, the new stacked thin film Si SC is proposed where the transparent inter-layer was inserted between a-Si and μc-Si layer to enhance a partial reflection of light back into the a-Si top cell (see fig.8.). This structure is called as internal light trapping enabling the increase of current of top cell without increasing the thickness of top cell, which leads less photo-degradation of stacked cell (Yamamoto et al., 2004).

Fig. 8. Schematic view of a-Si/poly-Si (μc-Si) stacked cells with an interlayer. Reprinted with permission from Solar Energy Vol. 77, Kenji Yamamoto et al., A high efficiency thin film silicon solar cell and module, pp. 939–949 © 2004, Elsevier.

Hybrid Solar Cells Based on Silicon 409

Fig. 10. KW system at the top of the building Osaka (HYBRID modules). Note that modules

Fig. 11. External quantum efficiency of the top cell in an n–i–p/n–i–p micromorph on flexible substrate (i-layer thickness:200nm). Reprinted with permission from Solar Energy Vol. 95, Meillaud et al., Realization of high efficiency micromorph tandem silicon solar cells

The last efficiency reported by Oerlikon Solar Lab, Neuchatel for a-Si/μc-Si (thin film cell) is about 11.9% (Green et al., 2011, as cited in Bailat et al., 2010). More precisely talk about

The tandem junction cell is a high-performance silicon solar cell, which is best suited for terrestrial solar power systems. The most distinctive design feature of this device is the use

on glass and plastic substrates: Issues and potential, pp. 127-130 © 2010, Elsevier.

are installed at low angle (5 degree off from horizontal).

tandem cells will be done in following sections.

**3.3 Tandem cell** 

By introduction of this interlayer, a partial reflection of light back into the a-Si top cell can be achieved. The reflection effect results from the difference in index of refraction between the interlayer and the surrounding silicon layers. If *n* is the refractive index and *d* is the thickness of inter-layer, the product of Δ*n*×*d* determine the ability of partial reflection. Namely, the light trapping is occurred between the front and back electrode without interlayer, while with inter-layer, it is also occurred between inter-layer and back electrode. This could reduce the absorption loss of TCO and a-Si:H (see Fig.9.) (Yamamoto et al., 2004).

Fig. 9. Spectral-response of the cell with and without interlayer. Bold and normal line shows the spectral-response of the cell with and without inter-layer. Reprinted with permission from Solar Energy Vol. 77, Kenji Yamamoto et al., A high efficiency thin film silicon solar cell and module, pp. 939–949 © 2004, Elsevier.

Some of the advantages of thin film SCs are being characterized to low temperature coefficient, the design flexibility with a variety of voltage and cost potential. Therefore, the thin film Si SCs can be used for the PV systems on the roof of private houses as seen in Fig.10. (Yamamoto et al., 2004).

Another similar work was done by F. Meillaud et al.(2010). They investigated the high efficiency (amorphous/microcrystalline) "micromorph" tandem silicon SCs on glass and plastic substrates. High conversion efficiency for micromorph tandem SCs as mentioned before requires both a dedicated light management, to keep the absorber layers as thin as possible, and optimized growth conditions of the μc-silicon(μc-Si:H) material. Efficient light trapping is achieved in their work by use of textured front and back contacts as well as by implementing an intermediate reflecting layer (IRL)between the individual cells of the tandem.The latest developments of IRLs at IMT Neuchˆatel are: SiOx based for micromorphs on glass and ZnO based IRLs for micromorphs on flexible substrates successfully incorporated in micromorph tandem cells leading to high,matched, current above 13.8mA/cm2 for p–i–n tandems. In n–i–p configuration, asymmetric intermediate reflectors (AIRs) were employed to achieve currents of up to 12.5mA/cm2 (see fig.11.) (Meillaud et al., 2011).

By introduction of this interlayer, a partial reflection of light back into the a-Si top cell can be achieved. The reflection effect results from the difference in index of refraction between the interlayer and the surrounding silicon layers. If *n* is the refractive index and *d* is the thickness of inter-layer, the product of Δ*n*×*d* determine the ability of partial reflection. Namely, the light trapping is occurred between the front and back electrode without interlayer, while with inter-layer, it is also occurred between inter-layer and back electrode. This could reduce the absorption loss of TCO and a-Si:H (see Fig.9.) (Yamamoto et al., 2004).

Fig. 9. Spectral-response of the cell with and without interlayer. Bold and normal line shows the spectral-response of the cell with and without inter-layer. Reprinted with permission from Solar Energy Vol. 77, Kenji Yamamoto et al., A high efficiency thin film silicon solar

Some of the advantages of thin film SCs are being characterized to low temperature coefficient, the design flexibility with a variety of voltage and cost potential. Therefore, the thin film Si SCs can be used for the PV systems on the roof of private houses as seen in

Another similar work was done by F. Meillaud et al.(2010). They investigated the high efficiency (amorphous/microcrystalline) "micromorph" tandem silicon SCs on glass and plastic substrates. High conversion efficiency for micromorph tandem SCs as mentioned before requires both a dedicated light management, to keep the absorber layers as thin as possible, and optimized growth conditions of the μc-silicon(μc-Si:H) material. Efficient light trapping is achieved in their work by use of textured front and back contacts as well as by implementing an intermediate reflecting layer (IRL)between the individual cells of the tandem.The latest developments of IRLs at IMT Neuchˆatel are: SiOx based for micromorphs on glass and ZnO based IRLs for micromorphs on flexible substrates successfully incorporated in micromorph tandem cells leading to high,matched, current above 13.8mA/cm2 for p–i–n tandems. In n–i–p configuration, asymmetric intermediate reflectors (AIRs) were employed to achieve currents of up to 12.5mA/cm2 (see fig.11.)

cell and module, pp. 939–949 © 2004, Elsevier.

Fig.10. (Yamamoto et al., 2004).

(Meillaud et al., 2011).

Fig. 10. KW system at the top of the building Osaka (HYBRID modules). Note that modules are installed at low angle (5 degree off from horizontal).

Fig. 11. External quantum efficiency of the top cell in an n–i–p/n–i–p micromorph on flexible substrate (i-layer thickness:200nm). Reprinted with permission from Solar Energy Vol. 95, Meillaud et al., Realization of high efficiency micromorph tandem silicon solar cells on glass and plastic substrates: Issues and potential, pp. 127-130 © 2010, Elsevier.

The last efficiency reported by Oerlikon Solar Lab, Neuchatel for a-Si/μc-Si (thin film cell) is about 11.9% (Green et al., 2011, as cited in Bailat et al., 2010). More precisely talk about tandem cells will be done in following sections.

### **3.3 Tandem cell**

The tandem junction cell is a high-performance silicon solar cell, which is best suited for terrestrial solar power systems. The most distinctive design feature of this device is the use

Hybrid Solar Cells Based on Silicon 411

µc-Si: H layers. In 2002, Meier published a micromorph tandem cell with efficiency of 10.8% in which the bottom cell was deposited at a rate of Rd =0.5 nm/s with the thickness of 2 µm (Meier et al.,2002). At least in this type of devices the highest efficiencies reported to date are 15.4% (tandem cell consisting of microcrystalline silicon cell and amorphous silicon cell)

Further development and optimization of a-Si: H/µc-Si: H tandems will remain very important because it is expected that in the near future, its market share can be considerable. For example, in the European Roadmap for PV R&D, it is predicted that in 2020, the European market share for thin film silicon (most probably a-Si: H/µc-Si: H tandems) will be 30%. This shows the importance of thin film multibandgap cells as second-generation

Although conventional SCs based on inorganic materials specially Si exhibit high efficiency, very expensive materials and energy intensive processing techniques are required. In comparison with the conventional scheme, the hybrid Si-based SC system has advantages such as; (1) Higher charging current and longer timescale, which make the hybrid system have improved performances and be able to full-charge a storage battery with larger capacity during a daytime so as to power the load for a longer time; (2) much more cost effective, which makes the cost for the hybrid PV system reduced by at least 15% (Wu et al., 2005). Therefore, hybrid SCs can be suitable alternative for conventional SCs. Among hybrid SCs which can be divided into two main groups including HJ hybrid SCs and dye-sensitized hybrid SCs, HJ hybrid SCs based on Si demonstrate the highest efficiency. Thus, the combination of a-Si/μc-Si has been investigated. These configurations of SCs can compensate the imperfection of each other. For example, a-Si has a photo-degradation while a μc-Si cell is stable so the combination is well stabilized. Furthermore, applying textured structures for front and back contacts and implementing an IRL between the individual cells of the tandem will be beneficial to enhancement of the efficiencies in these types of hybrid SCs. Due to recent studies; a-Si/μc-Si (thin film cell) has an efficiency of about 11.9%. Another study is done over three stacked cell of a-Si:H/μc-Si/c-Si (triple), which will be less sensitive to degradation by using the thinner a-Si. The last efficiency reported for a-Si/a-SiGe/a-SiGe(tandem) is about 10.4% and for a-Si/nc-Si/nc-Si (tandem) is approximately

Furthermore, Si based SC systems are being characterized to low temperature coefficient, the design flexibility with a variety of voltage and cost potential, so it can be utilized in large scale. In near future, it will be feasible to see roofs of many private houses constructed by thin film Si solar tiles. Although hybrid SCs are suitable replacements for conventional SCs, these kinds of SCs based on inorganic semiconductor nanoparticles are dependent on the synthesis routes and the reproducibility of such nanoparticle synthesis routes. The surfactant which prevents the particles from further growth is, on the other hand, an insulating layer which blocks the electrical transport between nanoparticles for hybrid SCs son such surfactants should be tailored considering the device requirements. Therefore, there is an increased demand for more studies in the field of hybrid SCs to find solutions to

(Yan et al., 2010).

solar cells.

12.5%.

overcome these weak points.

**4. Conclusions and outlook** 

of only back contacts to eliminate the metal shadowing effects because of lower conversion efficiency and steady-state bias requirement. Here we discuss tandem devices consisting of the a-si:H with other forms of silicon:
