Pulsed laser ablation is increasingly being applied to locally open the rear dielectric layer of advanced silicon wafer solar cell structures, such as aluminum local back surface field solar cells. We report that the laser ablation process on the rear surface of the solar cell at a relatively low laser fluence can cause undesirable spallation at the front surface which is textured with random upright pyramids. This phenomenon is attributed to the enhancement of the surface spallation effect by up to 3 times due to the confinement of the pressure waves at the tips of these random pyramids. Laser ablation at different laser focus positions and laser fluences is carried out to achieve optimized laser processing of the solar cells.
© 2012 OSA
Photovoltaic (PV) electricity is sustainable, abundant, affordable and clean. In the past decade the PV industry grew rapidly, predominantly with the help of government incentives. To further reduce the cost per kWh of PV electricity, laser processes have drawn intense interest in the past years for various potential applications in the manufacturing of silicon wafer based solar cells. The fabrication of higher-efficiency silicon solar cell structures, such as all-back-contact cells, laser-doped selective emitters, aluminum local back surface field (Al-LBSF) cells and emitter and metal wrap through cells, typically involve one or more laser processing steps [1–4]. These new solar cell concepts would probably not be economically feasible without the use of lasers. The Al-LBSF solar cell is one of the potential candidates to be commercialized in the near term because it allows a significant increase in the solar cell efficiency by the addition of only a few steps to the standard silicon wafer solar cell processing sequence. The rear surface of an Al-LBSF solar cell features a planar Si surface with a dielectric film covered by an Al rear electrode and local contact openings . Compared to the standard Si wafer cells, this rear surface structure has a greatly reduced surface recombination rate and increased reflection for near-bandgap photons. The front surface of Al-LBSF solar cells features a random pyramid texture coated with an anti-reflection coating (ARC) and a homogeneous or selective emitter. The local contact openings at the rear of Al-LBSF solar cells are typically formed by pulsed laser ablation, which can induce damage to the underlying silicon, potentially adversely affecting the efficiency of the final solar cell. Therefore extensive studies were carried out to minimize laser induced damage [5, 6].
In this work, it is shown that laser ablation at the rear surface of the solar cell using picosecond laser pulses can result in damage to the front surface of the solar cell, although this surface is not directly exposed to laser radiation. The damage occurs at the tips of the upright random pyramids on the front surface of the cell. This experimental observation can be explained by the fact that the shock wave induced by the picosecond laser pulses at the rear surface of the solar cell propagates through the silicon wafer and is enhanced significantly by the pyramids at the opposite surface (i.e., the front surface) of the solar cell. This enhanced shock wave causes fracture and spallation at the tips of the pyramids. This phenomenon is related to the “rear surface spallation” effect reported in the literature , which was previously only reported for planar surfaces. To our knowledge, this is the first time that “enhanced rear surface spallation” has been observed during laser ablation. Experiment and numerical calculation are conducted to investigate the role of the laser induced pressure wave and the textured pyramids. The effect of the laser focus position and laser fluence on the degree of surface spallation is also studied.
2. Sample fabrication and characterization
Al-LBSF solar cells were fabricated on 156 mm × 156 mm pseudosquare p-type Cz mono-Si wafers with a bulk resistivity of 1-3 Ωcm. KOH-based chemicals were used to remove the saw damage of the raw wafers. After a standard RCA cleaning with a final HF dip and DI water rinse, a silicon nitride masking layer was deposited onto the wafer’s rear surface, followed by single-sided surface texturing and POCl3 diffusion (TS 81004, Tempress) at the front surface. Subsequently, 10% HF was used to remove the phosphosilicate glass (PSG) and SiNx masking layer. A 75 nm thick SiNx antireflection coating (ARC) was then deposited onto the front surface of the wafers, followed by the deposition of 40 nm of aluminum oxide (AlOx) capped with 100 nm of SiNx onto the rear surface. All dielectric films were deposited by plasma-enhanced chemical vapor deposition (PECVD, SiNA-XS, Roth & Rau). Subsequently the rear dielectric stack was locally opened by picosecond (ps) laser ablation using different laser fluences and laser focus positions. The laser beam was focused either slightly above, slightly below or exactly at the sample surface. The ps solid-state laser (Nd:YAG) operated at a wavelength of 532 nm and with a pulse duration of ~10 ps. Some of these laser processed samples were used as test samples to study the enhanced rear surface spallation effect, while the remaining samples were completed into Al-LBSF solar cells, involving Ag and Al screen printing at the front and rear surfaces followed by an industrial co-firing process (Ultraflex, Despatch).
Figure 1(a) is a schematic drawing to illustrate the important effects during the laser ablation process, such as the propagation and focusing of the pressure wave. The laser light was focused by a lens onto the rear surface of the solar cell, to locally open the rear AlOx/SiNx dielectric stack. A laser fluence of 8.5 J/cm2 was used. The thickness of the Si wafer was 200 ± 20 μm, while the size of the pyramids was distributed randomly in the range from 4 to 7 μm. Scanning electron microscopy (SEM, Auriga, Carl Zeiss) was used to characterize the rear surface spallation effect occurring at the front surface of the solar cell wafer. Figure 1(b) shows a cross-sectional SEM image taken in the secondary electron (SE) mode. It can clearly be seen that some of the pyramid tips at the front surface of the solar cell wafer were missing after the laser processing of the rear surface of the solar cell. Figure 1(b) also shows that the topological changes are limited to the tips of the pyramids. It is thus argued that the pyramids confine the laser induced shock/pressure wave and thereby increase the local pressure to a level that exceeds the spall strength of crystalline silicon, thus resulting in spallation at the pyramid tips. The planar view SEM image of the solar cell’s front surface was taken in the SE mode and is shown in Fig. 2(a) . Spallation of the SiNx ARC coating as well as the silicon at the tips of the textured pyramids can be clearly observed. Figure 2(b) shows another SEM image that was taken at the same position but now using the back-scattering electron (BSE) detector. In the BSE mode, differences in intensity indicate differences in chemical composition. Clearly the spalled areas have a very different chemical composition compared to the unaffected front surface regions, and this can mainly be related to the removal of the SiNx film at these areas. Hence, the BSE mode is more sensitive for the detection of the rear surface spallation effect. In the remaining of the paper the degree of the rear surface spallation effect was evaluated qualitatively using BSE SEM images and is calculated relative to the laser irradiated area at the rear surface of the solar cell wafer. The relative spallation area (SA) was defined as the ratio of color change area in the BSE mode images to the laser irradiated area on the cell’s rear surface. In the following Section the influence of laser fluence and the laser focus position on the rear surface spallation effect will be investigated.
3. Results and discussion
It is well known that short-pulse high-power laser-matter interaction can induce high-pressure high-temperature shock waves . Laser ablation is a process of explosive material removal which generates a strong recoil pressure on the substrate. Shock waves are generated due to the rapid rise in pressure at the material surface caused by the evaporation of material  and/or adiabatic plasma expansion  during the laser ablation process. These shock waves decay into stress waves and acoustic waves during their propagation inside the material . Theoretical studies confirmed the correlation between shock-wave pressure and laser parameters, such as laser intensity, wavelength and pulse duration . Laser induced pressures up to a few gigapascals (GPa) were demonstrated on confined surfaces  which were covered by an overlayer that was transparent to the laser beam. The pressure induced by the laser ablation on the confined surfaces can be 5 to 10 times higher than that on unconfined surfaces . Under these confined conditions, the material’s vapor and plasma expansion are constrained in a small volume by the overlayer, leading to a significant increase in the peak pressure. Laser induced stresses are maximal under the “stress confinement” conditions [7, 15] when the heating pulse is much shorter than the characteristic time of the materials’ mechanical relaxation τs. The formula for this criterion is 16] and speed of sound inside silicon is about 8433 m/s. τs can thus be estimated as about 150 ps, which is much longer than the laser pulse duration used in our experiment. Therefore, the stress confinement condition was valid in the ps laser dielectric ablation experiments presented in this paper. The maximum pressure Pmax developed on the confined surface can be calculated as 7, 17]. Stress waves are also attenuated during their propagation inside the material due to the hydrodynamic decay effect. The attenuation of the laser induced stress in monocrystalline silicon was estimated to be 0.155% per μm .
In our experiment, it was observed that the laser focus position strongly affects the degree of the rear surface spallation effect. The laser beam was focused at either 3 mm above, 3 mm below or exactly at the sample’s rear surface and the area was laser ablated three times to show this focus position effect more dramatically. For the remaining of the experiments, only single ablation was applied. Figures 2(b-d) show that the SA was 10%, 5% and 70%, respectively for the various laser focus positions using a laser pulse energy of ~0.1 mJ. The strongest spallation effect was observed when the laser beam was focused at 3 mm below the sample’s rear surface, resulting in a SA of ~70% [Fig. 2(d)] while the weakest spallation effect was found when the laser was focused at 3 mm above the sample’s rear surface, resulting in a SA of ~5% [Fig. 2(c)]. When the laser is focused at 3 mm above or below the sample’s surface, the laser fluence is only 3.0 J/cm2 which is much lower than the laser fluence of 8.5 J/cm2 at the in-focus position. This is due to a larger laser spot size when the laser is focused above or below the sample surface and thus the laser induced pressure is lower, resulting in a weaker spallation effect. Although the laser fluence at the rear surface is identical when the laser is focused at 3 mm below or above the sample’s rear surface, the rear surface spallation effect occurring at the solar cell’s front surface is much more pronounced when the laser is focused at 3 mm below the sample’s rear surface. The reason behind this fact is not yet understood and further research is ongoing.
Next, the impact of the laser fluence on the rear surface spallation effect was investigated with the laser being focused at the sample’s rear surface. Figure 3 shows the variation of the SA with the laser fluence. The large variation of the error bar is attributed to different sizes of the randomly located upright pyramids. The fitted curve shows the trend that the SA increases with the laser fluence, which is expected as it is well known that the laser fluence increases with the strength of the shock wave. Although the laser fluence used in our experiment is not sufficient to cause any spallation effect on a flat silicon wafer front surface, the local pressure at the pyramid tips is enhanced and consequently high enough to induce spallation. No spallation was observed when the laser fluence was below 2.4 J/cm2. When the laser fluence increases to 3.8 J/cm2, surface spallation becomes detectable with a SA of 0.07% as shown in Fig. 3. By numerical fitting, the threshold laser fluence for the enhanced rear surface spallation effect was estimated to be ~2.9 J/cm2 for this specific sample. The laser ablation induced pressure at this threshold fluence can be estimated as 0.62 GPa at the rear surface of the solar cell by using Eq. (2). This pressure is attenuated to about 0.43 GPa after propagating from the rear surface to the front surface of the ~200 μm thick silicon solar cell sample. Since the spall strength of silicon is about 1.4 GPa , the rear surface spallation effect is thus enhanced by roughly a factor of 3 by the upright pyramids on the front surface of the solar cell. The rear surface spallation effect was not observed during ns laser ablation due to the much lower laser induced pressure which is attributed to the much longer pulse duration according to Eq. (2) and the fact that stress confinement condition was not satisfied.
Two types of Al-LBSF solar cells were fabricated to compare the device performance with and without the enhanced rear surface spallation effect: Type-A cells were processed at a laser fluence of 2.3 J/cm2 while type-B cells were processed at a slightly higher fluence of 3.0 J/cm2. For type-A cells no spallation on the front surface of the solar cell is expected, as the laser fluence is below the spallation threshold. Type-B solar cells, however, were processed at a laser fluence above the spallation threshold. Table 1 shows the measured one-sun current-voltage (I-V) curve of the type-A and type-B solar cells. Type-B cells have extremely low open-circuit voltage (Voc) and fill factor (FF). The low Voc and FF can fully be attributed to the enhanced rear surface spallation effect. The spallation effect removes the emitter (i.e., the p-n junction) at most affected areas thereby increasing the surface recombination and locally causing shunting as the front metal grid is in direct contact with the base of the solar cell. Even when the emitter is not fully removed, the spallated area will have an increased surface recombination, hence, this explains the significantly lower Voc and FF of the solar cell. Clearly, the enhanced rear surface spallation effect resulting from the upright pyramids is extremely detrimental to the solar cell’s electrical performance. This is also of relevance for most of the new solar cell structures that involve laser processing. As the PV industry moves towards thinner wafers and wider applications of laser processing, the pyramid-enhanced rear surface spallation effect could become increasingly important in the coming years.
Laser ablation of rear surface dielectrics by a ps laser was studied in detail for application in industrial Al-LBSF silicon wafer solar cells with a pyramid textured front surface. It was found that certain laser processing conditions can result in an undesired and very detrimental rear surface spallation effect at the textured front surface, despite the use of laser conditions that are well below the spallation threshold for planar surfaces. Confinement of the pressure waves at the upright pyramids of the front surface of the solar cell wafer was the root cause for this enhanced rear surface spallation effect. The threshold laser fluence for inducing enhanced spallation was estimated to be ~2.9 J/cm2 for a solar cell sample at a thickness of ~200 μm, a planar rear and a pyramid textured front surface. This is about 3 times lower compared to the spallation threshold for planar surfaces. At a higher laser fluence, the spallation area increases with the laser ablation induced pressure. The laser focus position also strongly affects the degree of the laser induced spallation. The enhanced rear surface spallation effect is extremely detrimental for the solar cell performance and hence must be avoided during the laser processing in solar cell manufacturing.
SERIS is sponsored by the National University of Singapore and Singapore’s National Research Foundation (NRF) through the Singapore Economic Development Board. This work was sponsored by NRF grant NRF2009EWT-CERP001-056.
References and links
1. J. Knobloch, A. Aberle, and B. Voss, “Cost effective processes for silicon solar cells with high performance,” in Proceedings of 9th EU PVSEC, Freiburg, Germany (1989), pp.777–780.
2. T. Roder, P. Grabitz, S. Eisele, C. Wagner, J. R. Kohler, and J. H. Werner, “0.4% absolute efficiency gain of industrial solar cells by laser doped selective emitter,” in Proceedings of 34th IEEE Photovoltaic Specialists Conference (PVSC), PA, USA, (2009), pp. 871–873.
3. T. Fellmeth, M. Menkoe, F. Clement, D. Biro, and R. Preu, “Highly efficient industrially feasible metal wrap through (MWT) silicon solar cells,” Sol. Energy Mater. Sol. Cells 94(12), 1996–2001 (2010). [CrossRef]
4. E. V. Kerschaver and G. Beaucarne, “Back-contact solar cells: a review,” Prog. Photovolt. Res. Appl. 14(2), 107–123 (2006). [CrossRef]
5. P. Engelhart, S. Hermann, T. Neubert, H. Plagwitz, R. Grischke, R. Meyer, U. Klug, A. Schoonderbeek, U. Stute, and R. Brendel, “Laser ablation of SiO2 for locally contacted Si solar cells with ultra-short pulses,” Prog. Photovolt. Res. Appl. 15(6), 521–527 (2007). [CrossRef]
6. S. Baumann, D. Kray, K. Mayer, A. Eyer, and G. P. Willeke, “Comparative study of laser induced damage in silicon wafers,” in Proceedings of 4th IEEE World Conference on the Photovoltaic Energy Conversion, Hawaii, USA, (2006), pp. 1142–1145.
8. X. Z. Zeng, X. L. Mao, S. S. Mao, S. B. Wen, R. Greif, and R. E. Russo, “Laser-induced shockwave propagation from ablation in a cavity,” Appl. Phys. Lett. 88(6), 061502 (2006). [CrossRef]
9. Y. F. Lu, M. H. Hong, S. J. Chua, B. S. Teo, and T. S. Low, “Audible acoustic wave emission in excimer laser interaction with materials,” J. Appl. Phys. 79(5), 2186–2191 (1996). [CrossRef]
10. S. Zhu, Y. F. Lu, M. H. Hong, and X. Y. Chen, “Laser ablation of solid substrates in water and ambient air,” J. Appl. Phys. 89(4), 2400–2403 (2001). [CrossRef]
11. S. Zhu, Y. F. Lu, and M. H. Hong, “Laser ablation of solid substrates in a water-confined environment,” Appl. Phys. Lett. 79(9), 1396–1398 (2001). [CrossRef]
12. C. R. Phipps, T. P. Turner, R. F. Harrison, G. W. York, W. Z. Osborne, G. K. Anderson, X. F. Corlis, L. C. Haynes, H. S. Steele, K. C. Spicochi, and T. R. King, “Impulse coupling to targets in vacuum by KrF, HF, and CO2 single pulse lasers,” J. Appl. Phys. 64(3), 1083–1096 (1988). [CrossRef]
13. N. C. Anderholm, “Laser generated stress waves,” Appl. Phys. Lett. 16(3), 113–115 (1970). [CrossRef]
14. M. Boustie, L. Berthe, T. Resseguier, and M. Arrigoni, “Laser shock waves: fundamentals and applications,” in Proceedings of 1st International Symposium on Laser Ultrasonics: Science, Technology and Applications. Montreal, Canada, (2008), pp. 32–40.
15. L. V. Zhigilei, Z. B. Lin, and D. S. Ivanov, “Atomistic modeling of short pulse laser ablation of metals: connections between melting, spallation, and phase explosion,” J. Phys. Chem. C 113(27), 11892–11906 (2009). [CrossRef]
16. M. A. Green and M. J. Keevers, “Optical properties of intrinsic silicon at 300 K,” Prog. Photovolt. Res. Appl. 3(3), 189–192 (1995). [CrossRef]
17. A. W. Webb, E. F. Skelton, D. J. Nagel, and S. B. Qadri, “Effects of laser-driven shocks on silicon single crystals,” J. Appl. Phys. 61(3), 1155–1161 (1987). [CrossRef]
18. J. Ren, S. S. Orlov, and L. Hesselink, “Rear surface spallation on single-crystal silicon in nanosecond laser micromachining,” J. Appl. Phys. 97(10), 104304 (2005). [CrossRef]