Scholarly article on topic 'Novel front and back side screen designs for improved cost/performance ratio'

Novel front and back side screen designs for improved cost/performance ratio Academic research paper on "Materials engineering"

CC BY-NC-ND
0
0
Share paper
Academic journal
Energy Procedia
OECD Field of science
Keywords
{"Heraeus paste" / "cost performance ratio" / "silicon solar cells" / "reduced silver consumption" / "low cost" / "screen printing" / photovoltaic}

Abstract of research paper on Materials engineering, author of scientific article — E. Kurtz, W. Zhang

Abstract As the PV industry strives toward grid parity, the driving factor for cell manufacturers is the cost-to-performance ratio. While the authors of this paper are in the business of constantly improving metallization pastes, we also believe it is to our benefit to help improve the product of our customers by any means available to us. There have been many ideas discussed to reduce the cost/performance ratio of crystalline Silicon cells, however with this paper we will show that this ratio can be improved by reducing the amount of silver paste that is deposited on a Silicon wafer via newly conceived screen designs. New front side designs show up to a 25% reduction in paste consumption while maintaining electrical and adhesive integrity using Heraeus SOL-9235HL front contact paste. New back side designs showed up to a 40% reduction in paste consumption while maintaining electrical and adhesive integrity using Heraeus SOL-215H back contact paste.

Academic research paper on topic "Novel front and back side screen designs for improved cost/performance ratio"

Available online at www.sciencedirect.com

ScienceDirect

Procedía

Energy

ELSEVIER

Energy Procedía 8a (2011) 620-624

SiliconPV: 17-20 April 2011, Freiburg, Germany

Novel front and back side screen designs for improved

cost/performance ratio

E. Kurtz, W. Zhang

Heraeus Materials Technology LLC, Photovoltaics B. U. 24 Union Hill Rd, W. Conshohocken, PA 19428 USA Tel: 001.610.825.6050, Fax: 001.610.825.7061 Email: eric.kurtz@heraeus.com

Abstract

As the PV industry strives toward grid parity, the driving factor for cell manufacturers is the cost-to-performance ratio. While the authors of this paper are in the business of constantly improving metallization pastes, we also believe it is to our benefit to help improve the product of our customers by any means available to us. There have been many ideas discussed to reduce the cost/performance ratio of crystalline Silicon cells, however with this paper we will show that this ratio can be improved by reducing the amount of silver paste that is deposited on a Silicon wafer via newly conceived screen designs. New front side designs show up to a 25% reduction in paste consumption while maintaining electrical and adhesive integrity using Heraeus SOL-9235HL front contact paste. New back side designs showed up to a 40% reduction in paste consumption while maintaining electrical and adhesive integrity using Heraeus SOL-215H back contact paste.

© 2011 Published by Elsevier Ltd. Selection and/or peer-review under responsibility of SiliconPV 2011.

Keywords: Heraeus paste, cost performance ratio; silicon solar cells; reduced silver consumption; low cost; screen printing; photovoltaic

1. Introduction

The increasing cost of silver is a major material cost in solar cell manufacturing. A step that can be taken to improve the cost/performance ratio for a manufacturer would be to print less silver paste on solar cells, all while maintaining electrical and adhesive performance. Here we focus on reducing the amount of paste that needs to be consumed, all while striving to maintain electrical and adhesive integrity of the completed solar cell.

Several different novel designs have been conceived for both the front and back side bus bar designs. While many of these new designs result in exposed silicon where there is none in standard designs, we compare electrical results only after applying standard solder connectivity wires (described later) to the cell to mimic, as best we can, real manufacturing conditions in a laboratory setting. Comparison of

1876-6102 © 2011 Published by Elsevier Ltd. Selection and/or peer-review under responsibility of SiliconPV 2011. doi:10.1016/j.egypro.2011.06.192

adhesion values should be considered relative, as our previous research [1] has shown that the soldering or adhesion evaluation process can vary greatly from one method to the next.

All experiments discussed in this paper were conducted with 156mm single crystal 55Q/sq p-type silicon wafers, and all processed in the same manner unless mentioned otherwise.

2. Front Contact Design

A three bus bar front side pattern was chosen to magnify any flaws in the experimental design. A standard screen design was chosen with 1.6mm wide bus bars, and 80^m finger line openings. The experimental screen modified the control bus bar by instead filling the bus bar area with a regular block pattern, where the widest areas are 1.6mm, and the thin areas are 590^m, shown in Fig. 1 and Fig. 2 at high magnification. The control pattern was of similar design, except for having the full 1.6mm wide bus bars. The difference in bus bar designs results in a 25% reduction in printable area over the bus bars.

Fig 1. CAD drawing of the experimental front side screen pattern, low zoom.

Illlllllllllllllllllllllll

iiiiiiiiiiiiiiiiiiiimnii

Fig 2. CAD drawing at High zoom of experimental bus bar, 1.6mm at its widest, 590^m at its thinnest

In averaging 5 cells per group, a 25% average reduction in print weight was achieved using an industry standard front side paste, Heraeus SOL-9235HL, while making sure to keep the average finger line width constant between the groups. The cells were then hand soldered with standard solder wire, and tested for

electrical results. The reason for applying solder wires to the cells is that there are areas of exposed Si on cells in the experimental group which are not there on the control cells. The finished cells will need to be strung together for module implementation, so we do this process here to mimic the final product. When we compare the control and experimental groups' electrical data, Table 1 shows comparable electrical results with a 25% reduction in paste weight.

Table 1. Electrical results comparing front side control and experimental cells. Each group is average of 5 cells

Wet Print Weight Eta (%) FF Isc (A) Voc (V) Pmax (W)

Control 0.230g 17.7 78.6 8.593 .6243 4.22

Experimental 0.171g 17.7 78.6 8.628 .6247 4.24

To evaluate adhesion of the samples, they were hand soldered with a solder wire from a widely used supplier, and pulled at 180° while a force gauge and software package recorded the force data. The parameters of the solder wire in this case were as follows: 0.13x2.0mm Cu wire coated with 62/36/2 (Sn/Pb/Ag) at 18-22^m. Four "pulls" per group were done and their average was taken. The control group showed an average adhesion force of 8.3N, and the experimental showed an average force of 6.3N, a 25% reduction yet still at a high average. Our previous research has shown that adhesion force is dependent on the amount of silver that is able to be soldered [1]. With this in mind, Fig. 3 shows that the experimental pattern provides localized areas of adhesion equal to the control in areas where the experimental bus bars are the widest.

o H-1-1-1-1-1-1-1

0 20 40 60 SO 100 120 140

Pull distance (mm) —>

Fig 3. Adhesion Force (grams force) vs. distance (x) chart shows localized high adhesion on experimental design. This plot presents data from single pulls, as opposed to the average taken of the four.

3. Back Contact Design

Again, a three bus bar pattern was chosen to magnify any flaws in the experimental design. A standard screen design was chosen with 153 x 4.0mm bus bars. The aluminum back surface field (Al BSF) pattern will overlap each bus bar by 0.75mm on each side, or 1.5mm coverage per bus bar to ensure good electrical contacts. Early attempts to design a "low weight" bus bar highlighted the need for the following proposed design: the experimental screen here modified the control bus bar by instead filling the bus bar area with discrete lines at 45° angles, as can be seen in Fig 4. The reason for angling the lines is to allow for some paste spreading while keeping an ideally consistent adhesion profile. Additionally, two experimental designs were made: one with 320^m openings (referred to as "320^m), and one with

120^m openings (referred to 120^m), a 20% and 40% area reduction, respectively. Using the Heraeus back contact paste SOL-215H, and as can be seen in Figure 5, the very fine line openings in the "120 ^m" screen allow the lines to bleed together to form one continuous paste layer. Table 2 shows comparable electrical results with both experimental screens.

Fig 4. (a) Control pattern, (b) "120um" opening, (c) "320 um " opening. Actual width of the bus bar is the same between all 3 designs. Experimental screen patterns enlarged to show design.

Fig 5. 50x zoom, 120um (top) and 320um (bottom) designs after print/fire

Table 2. Electrical results comparing back side control and experimental cells. Each group is average of 5 cells

Eta (%) FF Isc (A) Voc (V) Pmax (W)

Control 17.5% 77.8 8.592 .6236 4.17

Tester 1 120 ^m 17.5% 77.9 8.597 .6239 4.18

320^m 17.5% 78.2 8.595 .6237 4.19

Adhesion testing was performed similarly to the front side samples. The control gave an average adhesion of 5.8N, the "120^m" screen gave 3.1N, and the "320^m" screen gave 4.4N, a drop of 4 6% and 24% respectively which is similar to the paste weight reduction.

4. Other conceived designs

With the results of the discussed front and back side trials in mind, more designs can be conceived to obtain any particular electrical or adhesive goal. For example, if higher adhesion forces are needed for the proposed front side design, one could design the wide bus bar blocks to be longer as necessary. For the discussed back side designs, one could make those openings wider to allow for higher adhesion forces. Of course, the cost for higher adhesion forces is the need to print more paste. Rather than design all of these ideas and test them one by one, we choose here to demonstrate that the concept of reducing

the amount of paste used all while maintaining electrical and adhesive integrity is very possible and can be achieved through many different screen designs.

5. Conclusions

While we are constantly helping our customers achieve higher cost-performance ratios by developing and supplying metallization pastes that allow for higher efficiency and wider processing windows, we have also demonstrated that it is possible to significantly reduce the amount of paste used on silicon solar cells and maintain equal electrical and adhesive results. All samples were tested electrically after connectivity wires were attached to mimic a final production solar cell. A 25% front side weight reduction was achieved with comparable electrical and adhesive results using Heraeus SOL-9235HL front contact paste. Up to a 40% back side weight reduction was achieved with equal electrical results using Heraeus SOL-215H back contact paste, and the concept of being able to control the adhesion by manipulating the amount of silver deposited was discussed. The proposed designs can be modified in many ways to achieve a particular electrical or adhesive goal, however the achievement of this goal may come at the cost of needing to use more silver paste.

References

[1] J. Moyer, W. Zhang, E. Kurtz, R. Tavares, D. Buzby, S. Kleinbach. The role of silver contact paste on reliable connectivity systems. 25th European Photovoltaic Solar Energy Conference and Exhibition; 2010.