Open access peer-reviewed chapter - ONLINE FIRST

Review on Metallization in Crystalline Silicon Solar Cells

By Nagarajan Balaji, Mehul C. Raval and S. Saravanan

Submitted: October 5th 2018Reviewed: January 30th 2019Published: July 1st 2019

DOI: 10.5772/intechopen.84820

Downloaded: 85

Abstract

Solar cell market is led by silicon photovoltaics and holds around 92% of the total market. Silicon solar cell fabrication process involves several critical steps which affects cell efficiency to large extent. This includes surface texturization, diffusion, antireflective coatings, and contact metallization. Among the critical processes, metallization is more significant. By optimizing contact metallization, electrical and optical losses of the solar cells can be reduced or controlled. Conventional and advanced silicon solar cell processes are discussed briefly. Subsequently, different metallization technologies used for front contacts in conventional silicon solar cells such as screen printing and nickel/copper plating are reviewed in detail. Rear metallization is important to improve efficiency in passivated emitter rear contact cells and interdigitated back contact cells. Current models on local Al contact formation in passivated emitter rear contact (PERC) cells are reviewed, and the influence of process parameters on the formation of local Al contacts is discussed. Also, the contact mechanism and the influence of metal contacts in interdigitated back contact (IBC) cells are reviewed briefly. The research highlights on metallization of conventional screen printed solar cells are compared with PERC and IBC cells.

Keywords

  • silicon solar cells
  • process flows
  • metallization
  • passivated emitter rear contact cells
  • interdigitated back contact cells

1. Introduction

The photovoltaic industry plays a critical part in the global energy scenario [1] to compete with the other renewable and conventional energy sources. Crystalline silicon (c-Si) wafer-based technologies [2, 3] dominate the photovoltaic market for terrestrial application due to its high efficiency, stability, and benefits arising out of microelectronic industry. Due to high production cost (i.e., high $/watt), researchers are continuously putting their efforts to improve low cost Si solar cell technology. Silicon solar cell fabrication process involves various vital steps [4, 5] which includes texturing [6], n+ and p+ diffusion [7, 8], antireflection coatings [9], and contact metallization [7, 8, 9, 10, 11, 12, 13, 14, 15]. Electrical parameters of the solar cell, namely open circuit voltage (Voc), short circuit current (Isc), and fill factor (FF) vary with processing conditions. Though the conventional Si processing technology is mature, it is important to modify fabrication process and device structure to improve electrical performance. Approaches such as nickel/copper metallization in conventional solar cell structure [16, 17, 18], passivated emitter rear contact (PERC) cells [19] and interdigitated back contacts (IBC) cells [20], etc. are being used in lab scale and production. In this chapter, contact mechanism in conventional structure and novel structures is reviewed.

2. Process flow and current status

Conventional silicon solar cell process and its current status in PV industry are discussed in detail. Subsequently, the process steps of advanced process techniques such as Ni/Cu plating-based silicon solar cell, PERC, and IBC are also discussed.

2.1 Conventional Si solar cell

Currently, most of the PV industries use boron-doped p-type wafers as the starting material for c-Si solar cell fabrication. The schematic diagram of conventional fabrication process is shown in Figure 1. As reported in [2, 21], every processing step contributes to losses in conventional screen printing solar cell. Screen printing metallization is cost-competitive and robust technology used in production. Screen printing technology has attracted considerable attention due to significant improvement in printing medium and simplicity of the process. Also, this technology increases the throughput and decreases the production cost. For metallization, several alternatives to screen printing are available to improve cell efficiency [22, 23]. However, the existing screen printing technology is the matured and cost effective technology [24, 25] compared to recently developed technologies such as PERC, IBC, and HIT. Hence, around 85% of Si solar cells are manufactured using screen printing of thick film pastes. In a typical solar cell process, screen printing has the potential to improve efficiency and lower the cost, since metallization pastes are continuously evolving and new generation of pastes are available.

Figure 1.

Si solar cell process flow.

In addition to the new generation pastes, the right choice of front grid design and screen pattern results in better efficiency with reduced cost. The new generation paste provided a better aspect ratio (the ratio of line height to line width). The improvement in the aspect ratio improves the current carrying capacity of the contacts, as the shadow loss is decreased as well as the series resistance also decreases. In addition to the paste rheology, enhancement of the aspect ratio relies on choosing the right screen parameters such as mesh count, wire dimension, and emulsion thickness. Along with the paste material, optimized screen parameters are also the important factors for making the front contact with high aspect ratio and reduced shadow loss, which are desired for getting high efficiency solar cells. The silicon solar cell researchers or industries have achieved a maximum efficiency of 19% on multicrystalline silicon and around 20% on mono crystalline silicon-based solar cells by using the conventional process as shown in Figure 1 and are still working to enhance the efficiency using advanced materials.

Screen printing-based metallization technology occupies the significant role in solar cell manufacturing due to high throughput in cell production with better efficiencies. Though it is a mature technology, the finger aspects of the cells were limited by screen specifications and paste rheology.

2.2 Ni/Cu metallization-based solar cell

The limitations of screen printing and chance of considering alternate materials for front contact led many researchers [26] to look for the Ni/Cu metallization-based cell process. It is observed that it is important to optimize the Ni/Cu metallization to compete with screen printing technology in terms of reliability, cost competitiveness, and high throughput production. Ni plating on solar cells started in 1959, and the process has been developed in subsequent years and came into the present process flow in early 1980s. Interdigitated back contact solar cells by Sun Power Corporation integrated metallization scheme of patterned Al followed by plated Ni▬Cu▬Ag which was further annealed to realize the contact [27] and its world record efficiency is 24% [28].

Table 1 shows the cost and properties comparison of copper (Cu), silver (Ag), and nickel (Ni). It can be seen that the resistivity of copper is only more by 3.7% as compared to Ag, while the cost being less by around a 100 times, which process to be an important factor for cost reduction. Moreover, Cu is widely used as interconnects in ultra-large-scale integrated circuits owing to its low resistivity and good resistance to electro migration and has a proven track record in the microelectronics industry. Hence, Cu is a possible choice for metallization of solar cells. The main drawback of Cu is its high mobility and being a highly reactive recombination center in silicon. This necessitates a diffusion barrier like Ni to prevent its diffusion in Si. The nickel silicide formed at the interface reduces the contact resistance, which will ensure minimum power loss due to series resistance (RS) in a solar cell. Many groups have demonstrated cells based on Ni▬Cu front side metallization with improved fill factor (FF) and higher efficiency (ƞ) compared to Ag-based cells. The laser grooved buried contact (LGBC) technology utilizes Ni-Cu-based front-side metallization and has been successfully commercialized by BP Solar. The process flow of silicon solar cells with Ni/Cu front contact is shown in Figure 2.

ParametersAgCuNi
Conductivity (106 S/m)61.459.113.9
Density (gm/cm3)10.58.98.9
Typical cost ($/kg)431.04.514.2

Table 1.

Cost and properties comparison of copper (Cu), silver (Ag), and nickel (Ni).

Figure 2.

Ni-Cu process flow.

One of the crucial steps in Ni/Cu metallization is opening of ARC to make selective contact with an emitter. Literatures reported for patterning ARC and subsequent metal deposition to make the front contact; however, it is important to choose the process which can compete with screen printing technology both in cost and performance. It has been found that one such a technique is Ni/Cu metallization which can be commercialized with few additional process steps. But in Ni/Cu metallization, it is important to standardize and optimize few critical parameters such as Ni thickness, annealing temperature for silicide formation, and Ni/Cu deposition parameters. The 80-μm thick stencil printed grid lines were thickened by electroplating of Ni▬Cu▬Sn stack with a commercial plating tool, improving the efficiency of the solar cells by 0.4% abs [29]. The platform had single side wafer processing and hence no chemical attack on the back side Al. Complete solar cell metallization based on electrochemical deposition of Ni and Cu has also been demonstrated [30]. Pulsed plating was used as compared to direct plating in the work to ensure homogeneous and well adhered contacts. LIP-based thickening of screen-printed contacts was first reported by Mette et al. [31]. An absolute ƞ gain of 0.4% was obtained for large area cells based on standard production process, while an improvement of more than 1% absolute was possible for fine line 70-μm printed contacts. For an optimized grid design, an absolute ƞ gain of 0.7% was achieved for large area cells as shown in Table 2. With an optimized SiNx:H, ƞ of 19.2% has been demonstrated on industrial grade Si with Ni and Cu layers deposited by LIP on LDSE cells [32]. pFF nearing 80% were possible with Ni▬Cu▬Sn-based contacts deposited by LIP, with no observable degradation in the pFF after a stress test at 200°C for 1000 hours.

Result typeProcess differenceSubstrate typeVOC (mV)JSC (mA/cm2)FF (%)η (%)
BestBefore LIPmc-Si612.134.674.915.9
BestAfter LIPmc-Si613.434.179.216.6
AverageBefore LIPmc-Si610.0 ± 234.5 ± 0.174.3 ± 0.615.6 ± 0.2
AverageAfter LIPmc-Si611.0 ± 234.1 ± 0.177.9 ± 0.916.3 ± 0.2
BestNi▬Ni▬Cu▬SnCz619.035.772.716.4
BestNi▬Cu▬SnCz624.035.469.515.4
BestNi▬Cu▬SnCz623.037.374.217.2
BestBackground plating presentCz639.038.774.218.3
BestNo background platingCz634.338.778.319.2
BestCz638.3 ± 238.4 ± 0.778.8 ± 0.519.3 ± 0.4

Table 2.

Solar cell performance data for front contacts with electroplated metal layer(s).

After plating the conducting Cu layer, an Ag or Sn capping layer is deposited to protect the Cu conducting lines from oxidation and facilitate soldering of the interconnection tabs. Another important function of the capping layer is to prevent interaction of Cu with the EVA encapsulant. An impulse voltage test failure indicated by development of discharge sites was observed due to the presence of Cu particles at the Al/EVA interface on the back side [33]. Though the Ni▬Cu contact-based LGBC cells were commercialized, low throughput rates and increased processing costs led to screen printed contacts becoming the standard for solar cell metallization [34]. Other major concern is regarding the chemical waste due to the metal baths which can lead to serious environment contamination issues [35]. Steady research and advances in plating techniques have enabled transition of solar cell with Ni▬Cu-based metallization from labs to commercial scale production. Economic factors play vital role when considering an alternative technology with the introduction of new equipment in the fabrication line. As per the ITRPV roadmap, direct plating and plating on the seed layer are expected to have a share of around 15% in 2028 for the front-side metallization [36].

The Pluto series from Suntech Power is based on Ni▬Cu metallization with stabilized efficiencies of 19.0% on large area mono-Si solar cells. There was an improvement of over 6% as compared to the screen printed contacts due to reduced shading and improvement in VOC. IMEC has demonstrated conversion efficiencies of 20.3% on large area i-PERL cells with plated contacts. Using a PERC structure, Schott Solar along with Schmid Group demonstrated 20.9% efficient 6″ cells. Schott Solar has also demonstrated a median ƞ of 20.8% with a best ƞ of 21.3% on an industrial production line with electroplated contacts. Rena has recently demonstrated solar cells based on PERC technology reaching 20.8% with Ni▬Cu metallization. Modules made with these cells successfully passed IEC 61215 test three times and adhesion of >1 N/mm. The technology can lead to a reduction in the cell production cost by $ 0.06.

Ni▬Cu metallization yields the better efficiency compared to the conventional screen printing solar cells; however, due to the low throughput rates and increased processing costs, standard solar cell metallization dominates the PV industry. Also the chemical wastes of the metal baths in Ni▬Cu metallization lead to environment disputes.

2.3 Passivated emitter rear contact solar cell

The conventional screen-printed Al-BSF cells suffer from the optical losses (front reflectance), transmission losses, and the recombination (rear side). The major limitation arises from the rear surface recombination which is due to the low solubility of Al in Si (doping concentration 7 × 1018 cm3) during the very short firing process employed for alloying of screen printed Al paste. Though the boron co-doping with Al-BSF improves the doping concentration of Al-BSF [37, 38] owing to the higher solubility of boron, only 65% of the internally reflected longer wavelength light reaches the rear side, and hence the rear surface recombination is still being high [39]. One way to overcome the drawback of Al-BSF is the introduction of a dielectric rear side passivation with local contact points, which improves the optical properties with less surface recombination. One such cell architecture is the passivated emitter and rear cell [40]. With this structure, low rear surface recombination velocity of 60–200 cm−1 and internal reflectance over 95% have been realized so far. The dielectric passivation layer is locally opened for contact formation [40, 41, 42, 43, 44, 45, 46, 47] by laser [48, 49, 50] or by printing etching pastes [51]. About 1% of the total rear surface is covered by the local point contacts. The local point contact is realized by photolithography in the laboratory level, and in mass production, the contacts are formed either by screen printed Al [38, 46, 47, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60] or by physical vapor deposition (PVD) of Al [40, 47]. The process flow for the PERC cell is shown in Figure 3. The key challenge in the local aluminum contact formation is that the Al should not be penetrated into the dielectric passivation layer [45]. The local Al-BSFs produced during the alloying process create voids below the Al contacts. These voids result in incomplete BSF formation and hence the rear surface recombination and contact resistances are increased [38, 46, 47, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60]. An effective local Al-BSF is formed [61, 62] in the laser-fired contact process [63]. In this process, the deposition of the passivation layer followed by rear metallization (screen printed or evaporated contact) is carried out and finally with laser local contacts were formed. As a novel route, the rear Al electrode is formed by using commercial Al foil, thus complicated equipment such as evaporation or screen printing systems are avoided [64].

Figure 3.

PERC process flow.

The best commercial PERC cell with 20–21% (mc-Si) and 21–22% (mono c-Si) has been achieved in the commercial scale [36]. A detailed investigation on the various factors involved in the formation of local Al contact formation and the influences of process steps have been studied by various authors and are described in the following sections. Meemongkolkiat et al. [52] observed that voids are created beneath the local Al contacts during the alloying process. Rauer et al. [38] avoided these voids by adding Si powder to the Al paste. The various factors that influence the contact formation are: (a) dielectric opening method, (b) rear-side contact geometry, (c) the amount of Al in the metallization paste, and (d) firing process.

2.3.1 Influence of process parameters on local-BSF formation

To form a high-quality localized contact, a deep Al-BSF is required for Al▬Si contact interface to minimize the rear surface combination along with shunt free rear surface passivation. Urrejola [58] reported a shallow BSF or the presence of Kirkendall voids at the Al▬Si interface as shown in Figure 4. Figure 4(a) shows the uniform BSF with a thickness of (4 μm). Void formation is due to sub-optimal conditions (Figure 4(b)) or inadequate BSF depth (Figure 4(c and d)). These voids reduce the FF as well as act as high recombination centers which affects Jsc and Voc. Though the electrical contact is not affected with inadequate BSF depth, a very high contact recombination is expected. To minimize Rs with low contact resistivity, a very narrow local opening is required, hence high recombination beneath the metal contact is reduced. Urejola [56] obtained a lowest contact resistivity of 8 mΩ cm2 for a shallow dielectric opening which lead to the FF loss minimization. However, narrow Al▬Si alloy formation increases the dielectric passivated area under the contact, thus reducing Jsc and Voc. The influence of contact size and finger spacing was investigated by Urejola [58]. The decrease in contact spacing reduced the overlap of Al on each side of the local opening leading to high quality BSF, thus lowering the presence of voids. For a contact spacing of 100 μm, the BSF thickness around 6–7 μm with less void (8%) was obtained. Similarly, Rauer et al. [38] concluded that the thickness of the local BSF depends on the contact spacing and obtained a BSF thickness up to 4 μm for a contact spacing of 400 μm. Further to increase the local BSF thickness and to avoid the void formation, the authors added more Si powder to the Al paste. This prevents the contact penetration into the Si bulk with enhanced Al-BSF thickness. Moreover, this increase in Si powder diminishes the emitter saturation current density (J0e).

Figure 4.

The Al-Si interface showing (a) a well formed contact with deep BSF and no void formation, (b) a Kirkendall void, and (c) & (d) contact with shallow BSF [58].

2.3.2 Impact of dielectric opening method

A detailed investigation on the influence of formation of BSF using industrially screen printable local BSF Al paste and laser processing for removing the dielectric barrier was carried by Fang et al. and Bahr et al. [51, 53]. The laser ablation was carried out using nanosecond [wavelength (λ) = 1064 nm, pulse duration = 300ns, and pulse energy = <1.6 mJ] and femtosecond [wavelength (λ) = 1025 nm, pulse duration = 300fs and pulse energy = <36 μJ] laser. The ns laser has a strong influence on the local removing of the dielectric stack AlOx/SiNx:H, whereas fs showed a moderate influence. The strong and moderate influence is attributed to the interaction time between the laser pulse and the silicon substrate. With ns laser, few ten micrometer etching depth is achieved depending on the laser power, whereas in fs laser with very short interaction, few-micrometer depth is obtained. With screen printable local BSF Al etching paste, the passivation stack with 105 nm thick is etched off. After firing in an IR furnace with optimum belt speed, the local BSF formed with etching paste was more thicker (around 5 μm) and homogeneous with less voids. In the case of ns and fs, laser showed more voids with inhomogeneous thinner BSF (1–2 μm) due to the increased surface roughness.

2.3.3 Impact of contact resistivity

In recent days, the aluminum pastes are improved in such a way that even for thin laser contact opening (LCO), very low surface recombination is achievable. In future, decrease in the fraction of metallized area at the rear might is expected, hence Rs plays a vital role in the contact resistance of the Al▬Si interface which is given by

Rc,rear=ρc/frearE1

where frear is the rear metallization fraction and ρc is the specific contact resistivity. However, ρc is independent of the contact size [65]. Similarly, Rohatgi et al. [66], on 2.3 Ω-cm wafers, obtained a ρc = 10 mΩ cm2. Urrejola et al. [56] carried out the contact measurements with a PERC structure. The Al paste is printed on the top of the locally opened dielectric, and the transmission line model revealed the dependence of the ρc on the contact area. They obtained the ρc of 9–17 mΩ cm2 for the dielectric opening width of 80–170 μm. Gatz et al. [67], to determine ρc, varied the rear contact pitch of PERC solar cells and obtained a ρc of 40–55 mΩ cm2. The contribution of the bulk to the series resistance Rb is acquired either by calculation or numerical simulation. Kranz et al. [68] processed PERC-like TLM samples and measured the ρc of 3 mΩ cm2, whereas the fit to the solar cell data resulted in ρc of 0.2–2 mΩ cm2 and is shown in Figure 5.

Figure 5.

Rs-Rb vs. inverse metallization fraction 1/f. Reproduced with permission from [68].

2.3.4 PVD metallization

In most of the high efficiency solar cell concepts, the metallization is carried out using three different physical vapor deposition (PVD) techniques: sputtering, electron gun, and thermal evaporation. During the deposition of aluminum layer (2 μm), the substrate temperature increases to ~350°C, which mainly arises from the recrystallization heat of the aluminum. Comparing with the screen printing process, the mechanical and thermal impact on the wafer is substantially reduced. After the deposition of PVD aluminum layers, the contacts can be formed using laser pulses with different laser parameters which results in a much shallower profile. Hoffmann et al., on a 0.5 Ω cm p-type silicon, demonstrated a solar cell efficiencies up to 21.7% [69]. Reinwand et.al. [70] investigated PERC cells with sputtered aluminum on the rear side and a Ti▬Ag (50/100 nm) seed layer on top prior to the silver plating. With the optimized annealing temperature, the highest efficiency η = 21.1 and 19.4% for FZ and CZ wafers, respectively, was determined with the lowest contact resistivity ρc = 0.36 mΩ cm2.

2.3.5 Foil metallization

In 2007, researchers from F-ISE introduced the laser-based foil metallization technology called “FolMet.” With this technology, the conventional aluminum foil is attached to the silicon wafer [71], and thus the laser fired contact process forms both the electrical contact at the rear side of PERC cell as well as the mechanical contact by locally melting the aluminum through the passivation layer into the bulk silicon [72]. The key advantages of this process is its enhanced internal optical properties obtained due to the air gap between foil and passivation layer [73], cost reduction potential by decreasing the capping layer thickness, and ease of cell production process [74].

Figure 6 shows the internal reflection R at the rear side, after foil attachment and laser fired contacts. Nekarda et al., [73] by using the thick passivation layer optimized for the screen printed Al-paste, obtained an efficiency of 20.5%. In order to further reduce the cost, Graf et al. [74] adapted the rear side passivation layer with thinner capping layer and demonstrated an efficiency of 21.3% with a high Jsc due to the improved internal reflectance. Moreover, a low series resistance of 9 mΩ cm2 of Al foil improved the FF to 80%. Pros and cons of various metallization schemes such as screen printing (SP), physical vapor deposition (PVD), and foil are tabulated in Table 3.

Figure 6.

Measurement of the internal reflection R at the rear side, after foil attachment and laser fired contacts dependent on the SiNx capping layer thickness. The reflection varies by < 0.5% in the IR regime [74].

IssuesPVDScreen printFoil
Efficiency-potentialHighHighHigh
FlexibilityMediumLowHigh
MaturityHighHighLow
Process temperature300 °C800–900°C
CostMediumHighLow

Table 3.

Advantages and disadvantages of different printing mechanisms [75].

2.3.6 Metal wrap through PERC

MWT cell (Figure 7) is similar to the conventional solar cell design, and the external front contact busbars for interconnection are located at the rear side which increases Jsc due to the reduced shading loss. Lohmüller et al. [76], from FhG-ISE, combined the MWT concept (Jsc improvement) and passivated emitter rear contact (PERC) concept (reduced rear SRV) and reported a conversion efficiency of 18.7% with Jsc of 39.9 mA/cm2, Voc of 638 mV, and FF of 80.9% on a boron-doped p-type Cz grown silicon. The higher FF is due to the successful implementation of seed and plate technology [76]. Thaidigsmann from the same group introduced a simplified MWT-PERC cell called HIP-MWT (high performance metal wrap) to improve the efficiency by reducing the process complexity. In HIP-MWT structure, the formation of rear emitter is neglected hence no need of structuring steps. On a p-type FZ wafer with 0.5 Ω cm, a substrate thickness of 160 μm, on a cell area of 149 cm2 resulted in an efficiency of 20.1% with Jsc of 39.1 mA/cm2, Voc of 659 mV, and FF of 77.8%, was obtained. The HIP-MWT cell demonstrated an efficiency of 19.6% with Jsc of 40.2 mA/cm2, Voc of 649 mV, and FF of 75.1% on Cz grown wafer with 2.6 Ω cm for the same substrate thickness and cell area [77].

Figure 7.

Process flow of MWT PERC solar cells.

Passivated emitter rear contact solar cell with dielectric layer at the rear side and locally rear aluminum contacts reduces the recombination losses which increases the open circuit voltage. Also the rear dielectric layer increases the internal reflection and thus increases the current of the solar cell. Though the performance of PERC cell is better, the efficiency of PERC cell decreases after light-induced degradation which is around 0.5–1.0% absolute.

2.4 Interdigitated back contact solar cell

In IBC solar cells, optical shading loss is eliminated as both polarities of the metal contact are placed on the rear surface. In addition, the resistive power loss is reduced largely as the rear surface furnishes an opportunity for best design of metal contact formation. The other key advantages of IBC cell are (a) module manufacturing cost is reduced as the interconnection between the cells is simplified and (b) higher cell packing density increases the module power. The process flow of the IBC cell is shown in Figure 8. The major challenges present in the metallization of IBC cell includes: (i) shunting between the two polarities of metal contacts must be prevented and (ii) the metal conductors must be thick enough to ascertain the low resistive power loss. To isolate both the contacts, different cell-based metallization techniques can be used. One such method is patterning metal seed layer [78, 79], with electroplating to reduce resistance [80]. However, this plating up process needs electrical contact to the seed metal lines, which may lead to be problem with thinner wafers.

Figure 8.

IBC process flow.

IBC solar cells with a record high efficiency of 25.6% were obtained by Sanyo/Panasonic [81], and the pioneer SunPower Corporation achieved 25% efficiency [82]. For IBC cells, the front surface field (FSF) reduces the surface recombination at the front as it acts as an electrical field which pushes back the minority carriers at the front surface [83]. The high expensive photolithography process is replaced with laser processing or screen printing which leads to a significant reduction in position accuracy which increases the pitch. This makes the majority carriers to travel from vertical to lateral direction. Depending on the pitch and base resistivity, series resistance over 90% contributing to the lateral majority carrier transport reduces the cell efficiency. Moreover the lateral majority carrier’s current transport as well as the front surface passivation has been enhanced by FSF and finally the series resistance also significantly reduced to 0.1 and 1.3 Ω cm2 for the base resistivity of 1 and 8 Ω cm, respectively, for the pitch of 3.5 mm [84].

The rear metallization of IBC cells is usually done with silver (Ag) and aluminum (Al) pastes [85], and Si/Ti/Pd/Ag or Si/Al/Ti/Pd/Ag metal stack and Al-deposited by PVD form a good ohmic contact with both n- and p-type silicon [86]. In Si/Ti/Pd/Ag or Si/Al/Ti/Pd/Ag metal stack, the Ag layer is used as a conductive layer because of its low resistivity. To avoid the reaction between Ti and Ag, the Pd layer is deposited between Ti and Ag layer. The work function of Ti or Al makes it suitable to contact with low contact resistivity [87] on both p- and n-doped regions. As Ti is a poor reflector, the Si/Al/Ti/Pd/Ag structure is adopted owing to the higher reflectance obtained with an Al layer which increases the light trapping. Couderca et al. [20] used Si/Ti/Ag stack with a thin Ti layer which has a low contact resistivity on both n- and p- doped regions. For n-doped surfaces, the specific resistivity is very low, and hence, the resistive losses are negligible. As the p-doped surfaces are lowly doped, the specific resistivity is higher though it stays under the crucial limit of 0.01 Ω cm2. As a low-cost approach, Chena et al. [88] applied the aluminum electrodes deposited by e-beam evaporation using Al contact for high performance IBC cells. The specific contact resistivity of the Al contact cell resulted in 0.7 and 0.05 mΩ cm2 on an n-doped and p-doped surface, respectively, and the final Al-contacted IBC cell resulted in an efficiency of 22.72%.

Recently, carrier selective contacts [TOPcon] using tunnel oxide and amorphous (a-Si) layer resulted in Voc values around 720 mV and contact resistivities less than 10 mΩ cm2 [89]. With poly-Si/SiOx approach, similar values for their passivated contact have been achieved by various researchers [90, 91, 92]. Young et al. used the similar contact for the IBC solar cell patterned with ion implantation. The metallization layer consisting of thin Ti/Pd adhesion layer with 1-μm thick Ag layer and a Pd capping layer resulted in the contact resistivities less than 0.1 mΩ cm2 [93].

Electrical shading loss plays a detrimental role as it reduces the collection efficiency of the minority carriers over the BSF regions [94], which compromises Jsc. By decreasing the width of the BSF region, this detrimental effect can be resolved. In the active cell area, by decreasing the finger pitch and BSF finger width, the electrical shading loss is minimized. Nevertheless, the base busbar still enforces electrical shading. In addition, as the majority carriers generated over the emitter busbar have to traverse over the entire wafer area, result in transport losses. The electrical shading loss and the transport loss contribute to the resistive losses and FF losses in the busbar metallization. Hence for further efficiency improvement, research group from ISFH developed a busbar less metallization which omits the busbar and eliminates the resistive losses in metallization with aluminum-based mechanical and electrical laser interconnection (AMELI) process for contacting aluminum-metalized IBC cells [95] and obtained a conversion efficiency of 22.1% with a Voc of 683 mV, Jsc of 41.4 mA/cm2, and FF of 78.1% on a cell area of 132 cm2. AMELI process interconnects the solar cell with highly flexible interconnection geometry performed by a laser as structuring of the metallization. In addition, this AMELI process can interconnect that are as wide as the whole cell edge with a lower electrical resistance between the cells [96, 97]. Figure 9 depicts the AMELI interconnection scheme for busbar-free solar cells. Woehl et al. introduced a point-shaped metalized IBC cells interconnected to a printed circuit board. The presence of only point-shaped metal contacts, increases the Voc as the recombination area, is significantly reduced [98].

Figure 9.

Process flow of AMELI interconnection process.

The main advantage of interdigitated back contact solar cells over other type of solar cells is zero shadow loss due to the absence of complete front contact. Although IBC is the high efficiency single junction cells among all other type of silicon solar cells, the carrier collection efficiency in front of the back surface field is low.

3. Conclusions

Process flows of conventional silicon solar cell, Ni/Cu plating for silicon solar cells, passivated emitter rear contact solar cells, and interdigitated back contact solar cells were discussed. Influences of process parameters in electrical parameters were analyzed. Though the contact formed by lithography, sputtering, etc. is reliable and resulting in good energy conversion efficiency, however, it is expensive due to the vacuum evaporation and single wafer type process. In this context, screen printed contacts are consistent in reliable and providing the best approach in production industry. Screen printing-based metallization is one of the key and crucial processes in silicon solar cell fabrication process which was discussed by interpreting the paste rheology, screen, and printing parameters. The screen printing paste used for contacting the solar cells is the other expensive element after the silicon wafer, and thus it is important to find an alternate technique for silver paste-based printing mechanism. Researchers arrived at a Ni/Cu plating technique for contact mechanism, and the technique has proved its capability in manufacturing industry as well. However, advanced structures such as PERC and IBC are using either screen printing or evaporation technique for making contacts. By seeing current scenario of metallization in different types of solar cells, it has been concluded that screen printing will continue to be an important and reliable metallization technique. The current efficiency status of different silicon solar cell technologies is depicted in Table 4.

S/NTechnologyEfficiency in %
1Al BSF multi Si cell19.0
2Al BSF mono Si cell20.0
3Ni/Cu plating Si cell21.3
4PERC multi Si cell22.0
5PERC mono Si cell25.0
6MWT PERC Si cell19.6
7IBC Si cell26.7

Table 4.

Current efficiency trend of different technology solar cells [36, 77, 99].

Acknowledgments

SS thanks the Management, RenewSys India Pvt. Ltd. and Faculties, National Centre for Photovoltaic Research and education, Indian Institute of Bombay, Mumbai for their support.

Disclaimer

The technical content in this article is the literature review of authors and do not necessarily reflect the official policy or position of RenewSys India Pvt. Ltd. Further, the authors hereby undertake to protect the proprietary technical and confidential information of RenewSys India Pvt. Ltd, including but not limited to the process parameters, recipes, formulations, sequences, methods, data, information etc. if any, referred in the above article, and have taken sufficient and appropriate measures and precaution to avoid the same from being inferred directly or indirectly in any manner whatsoever.

Definitions

Voc

open circuit voltage

Isc

short circuit current

Jsc

short circuit current density

J0e

emitter saturation current density

FF

fill factor

Rs

series resistance

η

efficiency

ARC

anti-reflective coating

SP

screen printing

FSF

front surface field

AlBSF

aluminum back surface field

PERC

passivated emitter rear contact

IBC

interdigitated back contact

HIT

heterojunction with intrinsic thin layer

LGBC

laser grooved buried contacts

LIP

light induced plating

LDSE

laser-doped selective emitter

EVA

ethylene vinyl acetate

ITRPV

international technology roadmap for photovoltaic

i-PERL

passivated emitter with rear locally diffused

PVD

physical vapor deposition

LCO

laser contact opening

MWT

metal wrap through

HIP MWT

high performance metal wrap through

CZ

czochralski

FZ

float zone

SRV

surface recombination velocity

AMELI

aluminum based mechanical and electrical laser interconnection

Download

chapter PDF

© 2019 The Author(s). Licensee IntechOpen. This chapter is distributed under the terms of the Creative Commons Attribution 3.0 License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

How to cite and reference

Link to this chapter Copy to clipboard

Cite this chapter Copy to clipboard

Nagarajan Balaji, Mehul C. Raval and S. Saravanan (July 1st 2019). Review on Metallization in Crystalline Silicon Solar Cells [Online First], IntechOpen, DOI: 10.5772/intechopen.84820. Available from:

chapter statistics

85total chapter downloads

More statistics for editors and authors

Login to your personal dashboard for more detailed statistics on your publications.

Access personal reporting

We are IntechOpen, the world's leading publisher of Open Access books. Built by scientists, for scientists. Our readership spans scientists, professors, researchers, librarians, and students, as well as business professionals. We share our knowledge and peer-reveiwed research papers with libraries, scientific and engineering societies, and also work with corporate R&D departments and government entities.

More about us