Scholarly article on topic 'Technological and economic assessment of two-steps printing processes in a mc-Si solar cells production environment'

Technological and economic assessment of two-steps printing processes in a mc-Si solar cells production environment Academic research paper on "Materials engineering"

CC BY-NC-ND
0
0
Share paper
Academic journal
Energy Procedia
OECD Field of science
Keywords
{"Silicon solar cells" / metallization / "screen printing" / "two-step printing" / "double printing" / "print on print" / "dual print" / "production cost reduction"}

Abstract of research paper on Materials engineering, author of scientific article — M. Pesce, M. Maugeri, M. Marsili, M. Zarcone, D. Tonini, et al.

Abstract The possibility of obtaining economic benefits out of two-step printing processes is still a matter of debate for the industry. This is due to both the evolution of the single printing process and to the superimposition of many technological and economic variables. In the present work we analyze production cost and cell performance benefits achieved in a mc-Si solar cells production environment. Three different screen technologies are compared; we find that for all of them the finger aspect ratio is improved with respect to the single print baseline. An efficiency gain of approximately 0.1 absolute is also found. The overall performance of the double printing technology in terms of €/Wp is evaluated taking into account the efficiency gain, the screen lifetime and cost, the Ag consumption and the additional costs due to labor and equipment depreciation. The three screen technologies offer savings ranging from 0.4% to 1.4% of the total production costs.

Academic research paper on topic "Technological and economic assessment of two-steps printing processes in a mc-Si solar cells production environment"

Available online at www.sciencedirect.com

Procedía

Energy

ELSEVIER

Energy Procedía 21 (2012) 24 - 31

3rd Workshop on Metallization for Crystalline Silicon Solar Cells, 25 - 26 October 2011, Charleroi, Belgium

Technological and economic assessment of two-steps printing processes in a mc-Si solar cells production environment

M. Pescea, M. Maugeria, M. Marsilia, M. Zarconea, D. Toninib, C. Bottossob,

M. Galiazzob, and A. Tomasia*

The possibility of obtaining economic benefits out of two-step printing processes is still a matter of debate for the industry. This is due to both the evolution of the single printing process and to the superimposition of many technological and economic variables. In the present work we analyze production cost and cell performance benefits achieved in a mc-Si solar cells production environment. Three different screen technologies are compared; we find that for all of them the finger aspect ratio is improved with respect to the single print baseline. An efficiency gain of approximately 0.1 absolute is also found. The overall performance of the double printing technology in terms of €/Wp is evaluated taking into account the efficiency gain, the screen lifetime and cost, the Ag consumption and the additional costs due to labor and equipment depreciation. The three screen technologies offer savings ranging from 0.4 % to 1.4 % of the total production costs.

© 2012 Published by Elsevier Ltd. Selection and/or peer review under responsibility of Guy Beaucarne

Keywords:Silicon solar cells; metallization, screen printing, two-step printing; double printing, print on print, dual print, production cost reduction.

* Corresponding author. Tel.: +39-0429-719244; fax: +39-0429-719015. E-mail address: andrea.tomasi82@gmail.com

aXgroup SPA, Via dell'Artgianato ZI, Vanzo di S. Pietro Viminario (PD), 35020, Italy bApplied Materials Italia, Via Postumia Ovest 244 Olmi di S. Biagio Di Callalta (TV), 31050 Italy

Abstract

1876-6102 © 2012 Published by Elsevier Ltd. Selection and/or peer review under responsibility of Guy Beaucarne doi:10.1016/j.egypro.2012.05.004

1. Introduction

The pressure towards lower production costs, due to the high competitive scenario of the PV industry, demands at the same time for efficiency gain and material cost reduction. During the last twelve months, the metallization material consumption became very important in regard to production cost due to the reduction of Si wafer prices (- 50 %) and the increase of Ag value (+ 30 %). The relative incidence of the metallization cost on the total cell production cost increased from below 10 % to a value around 18 %. High aspect ratio (AR) fingers are strongly desirable since they can provide silver paste consumption reductions and efficiency gains. The fingers AR is limited by technology issues concerning screens, paste thixotropic properties, and the printing process itself. While improvements on all these fields yields higher finger AR already with the standard single print (SP) technology, printing two grids, one on top of the other (the so called double printing (DP) process), is a possible way to achieve this goal faster. Within the DP process, efficiency gain ranging from 0 to 0.5 % and paste consumption reduction up to 30 % have been reported [1-4]. The effectiveness of such benefits at the €/Wp level is still not clarified. Difficulties can arise when two steps processes, as a whole, are analyzed in terms of cost and process constraints and requirements.

In this paper we analyze different sets of data collected at our mc-Si solar cell production facility. We implemented the DP process using three different screen technologies. Data concerning morphology of the fingers and electrical parameters of sorted production groups are reported in comparison to standard SP reference groups following the evolution of the printing during the entire 'life' of the screens. In this way we are able to probe, at the industrial scale, the potentiality of the DP processes and to assess their cost effectiveness.

2. Double printing

2.1. Experimental details

156x156 mm2 mc-Si production wafers have been used throughout the present work. All the wafers, with a 70 Ohm/square emitter, were processed in one of Xgroup production lines up to the PECVD SiN anti-reflection coating (ARC) deposition process. The production flow was then split, and half of the wafers underwent a DP process, while the remaining half underwent the standard production SP process on the same printing.

The reference SP process employed the standard stainless steel mesh screen technology while for DP applications the different screen technologies tested were V-mesh (DP VS), hybrid stencil mounted on polyester fabric (DP HS) and stainless steel mesh mounted on polyester fabric (DP CS). The number of finger and bus bars of the screen layouts was the same for all the screen technologies under investigation.

The cells were characterized in terms of grid morphology and electrical parameters. The electrical parameters of each cell have been measured with the standard in-line I-V testing equipment, whereas the finger morphology has been studied for selected wafers at increasing print number using the LEICA DCM-3D microscope. The finger width, height and section area (FSA), as shown in Fig. 1, were measured in three points (at wafer center and edges) thus bringing out any print misalignment (an uneven distribution of the finger width is a clear evidence of a non-optimal matching of first and second print).

Fig. 1: 3D profile acquisition locations (left). The profiles were acquired for 600 ^m each. The finger width (w), height (h), and section area (FSA) were taken on 30 sections at 20 |im distance one from the other, and then averaged.

2.2. Results and discussion

The evolution with print number of the average finger width is reported in Fig. 2a. The error bars indicate the standard deviation of the three measures acquired at point 1,2 and 3. From a finger width of around 90 |jm with SP process (80 ^ opening) we went down to 80 |jm with VS (60 |jm opening) and to 65 |jm with HS and CS (65 and 60 ^m opening). The standard deviations are much bigger for VS and are indicating a larger deformation for this screen type. The spreading, i.e. the difference between the final finger width and screen opening, is close to 15 |jm for the reference SP process. We find that it is reduced to 5 |jm and 7 |jm for the HS and CS technologies, while for VS we obtain a significantly higher spreading (around 20 |jm), an artifact of the second print misalignment. For all the screen technologies we find that the finger width increases with print number, presumably indicating an emulsion wearing.

Screen Technology SP REF DP VS DP HS DP CP

Avg. Deposition [mg] 205 200 183 164

Variation with respect to SP baseline [%] - -2 -10 -20

Estimated lifetime [print#] 12k 24k 28k 29k

Table 1. Average Ag deposition and screen lifetime.

From the different runs in production we estimated a screen average lifetime and evaluated the silver consumption evolution for increasing print number. In Tab. 1 the average silver consumption per cell at the estimated screen lifetime is given. The high silver consumption reported for the VS technology may be related to a loss of sealing effect caused by the misalignment of the two prints due to the enhanced screen deformation. In Fig. 2b we show the evolution of the Ag consumption for the different screen technologies. The CS technology shows the strongest variation of Ag deposition with print number.

In Fig. 3 we report the evolution of the main finger parameters. The height values range between 21 |jm and 27 |jm. The standard deviation of the finger heights along each acquired profile, and averaged over the 3 spots (see Fig. 1) is reported in Fig 3b. The good performance of the VS technology, that allows a low finger roughness, is clearly visible. The average finger section area (FSA) is shown in Fig. 3c. The VS technology produces FSA comparable to the SP baseline process (around 1200 |im2), whereas

10000 15000 20000 25000 30000 Print §

-SPREF -VS -HS -CS

- 0.130 0.120

10000 15000 20000 25000 30000 Print #

Fig. 2: Finger width (a) and paste consumption (b) evolution vs. print number. Black squares correspond to the SP reference process, blue circles to DP VS, red triangles to DP HS and green diamonds to DP CS.

the CS and HS FSAs are much lower. Finally concerning the finger aspect ratio (AR) we see (Fig 3d) that two different levels are achieved, one for the SP process between 0.2 and 0.3 and one for the DP processes between 0.3 and 0.4.

The average efficiency, Isc, Voc and FF for the different trials are reported in Tab. 2. In all cases slightly higher efficiencies were achieved for the DP process. The higher efficiencies are obtained thanks to an improved Isc, but the benefits are limited by FF losses. Comparing Tab. 2 and Fig. 2 we see that the Isc gain well correlates with the finger width decrease (due to reduced shadowing) at the same time the FF loss can be linked to the FSA (Fig. 3c).

Fig. 3: Finger parameter evolution vs. print number. (a) Average finger height; (b) average standard deviation along the 3D profile of the finger height; (c) average finger section area; (d) average aspect ratio. Black squares correspond to the SP refer.

Screen Technology Voc [V] Isc [A] Efficiency [%] FF [%] Width [pm] FSA [^m2]

SP reference 0.612 8.37 16.31 77.5 94 1210

DP VS 0.613 8.41 16.38 77.3 81 1180

SP reference 0.613 8.38 16.48 78.0

DP HS 0.616 8.48 16.56 77.2 70 810

SP reference 0.612 8.41 16.53 78.2

DP CS 0.614 8.53 16.61 77.2 67 830

Table 2. Average electric parameters from the production experiment.

3. Cost analysis

In the economic evaluation of the DP technologies the benefits resulting from the efficiency gain and the reduced paste consumption must be compared with the economic loss due to an increased screen, equipment depreciation, and labor costs. Of course, as the scenario of raw material prices is rapidly varying, the relative weight of each term varies too.

We define the break-even point for each screen technology as the print number in which DP screen additional cost equals the economic gain due to paste consumption reduction and efficiency increase, i.e. the minimum screen lifetime after which the DP processes become cost-effective. In Fig. 4 the way in which the break-even point is determined for the case of the CS technology is shown. The average DP economic performance per cell as a function of screen lifetime (given as a print number) is the sum of the additional DP cost and of the cost saving at each print number under consideration as screen lifetime. To determine the cost savings the evolution of the paste consumption (Fig. 2b) was fully taken into account while the efficiency gain as been kept constant at its average value over the entire screen lifetime. The break-even point for the different screen technologies are shown in Tab. 3. Mesh screens or stencils mounted on polyester fabric can provide lower break-even points although they are more expensive screen technologies. For all the screens under consideration the break-even point is below the estimated screen lifetime (see Tab. 1).

Screen Technology DP VS DP HS DP CP

Break-even point (Print #) 19k 12k 13k

Table 3. Break-even point in terms of screen lifetime for each screen technology.

The contribution of the metallization step to the final cost per Wp is reported in Tab. 4. We assumed an additional 3 % of labor cost for the DP process and computed the equipment depreciation from the additional equipment cost considering a 5 years amortization schedule. We see that with the DP process the gain in cost per Wp ranges between 0.4 % and 1.4 % of the final cost. The HS and VS technologies provide a gain below 1 % with different saving distributions: in comparison to the SP process, the HS-DP has higher screen costs but quite lower metallization costs, whereas the screen cost increase is not so pronounced in the case of DP VS but so is the paste consumption reduction. Concerning the DP CS, the very low Ag paste consumption is the main responsible for the highest savings (1.4 % of the final cost).

Screen Technology AEff. abs [%] AAg [%] Rel. single screen lifetime Rel. single screen cost Front screen cost [%] Metallization cost [%] A Cell production cost [%]

SP reference 0 0 1.0 1.0 0.4 17.7 0.0

DP VS 0.07 -2 2.4 1.1 0.4 16.9 -0.4

DP HS 0.08 -10 2.8 2.3 0.7 16.6 -0.7

DO CS 0.08 -20 2.9 3.5 1.1 15.6 -1.4

Table 4. DP cost impact on final €/Wp production cost.

4. Conclusions

We implemented the DP process on one of our production lines, comparing three different screen technologies (V-mesh, hybrid stencil mounted on polyester fabric, and stainless steel mesh mounted on polyester fabric). We found that within DP it is possible to obtain lower finger width and higher AR also

Print»

Fig. 4: Break-even point determination for the DP CS technology. The break-even point is the print number in which DP additional cost and benefits are equal.

in production. We found that mesh screens or stencils mounted on polyester fabric can provide higher performances in term of screen deformation and resulting finger widths. This leads to different silver paste consumptions for the different technologies. Higher paste consumption may be addressed, in the case of the V-mesh, to a loss of sealing effect for the side fingers. This is due to the second print misalignment at the cell borders provided by the VS deformation. For the DP HS and DP CS processes the paste consumption is reduced by 10 % and 20 % respectively.

The DP processes allowed in a production environment for a 0.1 % absolute efficiency gain, without front grid optimization according to resulting finger widths.

In the final evaluation of the DP performance in terms of €/Wp the benefits of the reduced paste consumptions and of the improved efficiency were taken into account together with the higher screen, labor and equipment depreciation costs. We found that the different technologies have different breakeven points in terms of screen lifetime. Lower break-even points were found for the more expensive HS and CS technologies due to the strong reduction in Ag paste consumption. All the break-even points are below the empirically estimated screen lifetime. Overall €/Wp cell production cost reductions ranging from 0.4 % to 1.4 % were shown to be feasible in production.

References

[1] E. Kossen, B. Heurtault, and A. F. Stassen, "Comparison of Two Step Printing Methods for Front Side Metallization", Conference Proceedings, 25th EUPVSEC, Valencia (2010).

[2] T. Falcon and A. Hobby "Development of a 'Print On Print' Process For High Aspect Ratio Frontside Conductors", 2nd Workshop on Metallization of Crystalline Si Solar Cells, Konstanz (2010).

[3] M. Galiazzo et al. "Reliable double printing of Ag contacts for c-Si cell manufactoring", 2nd Workshop on Metallization of Crystalline Si Solar Cells, Konstanz (2010).

[4] M. Galliazzo et al. " Double printing with esatto technology" Applied Baccini Innovation Summit, Treviso (2010).

[5] M.A.Green "Solar cells, operating principles technology and system application" Englewood Cliffs, N.J. (USA), Prentice Hall (1981).

[6] P. Magnone et al. "Understanding the Impact of Double Screen-Printing on Silicon Solar Cells by 2-D Numerical Simulation", Conference Proceedings, 37th IEEE-PVSC, Seattle (2011).