- Email: [email protected]
Efficiency and stability enhancement of laser-crystallized polycrystalline silicon thin-film solar cells by laser firing of the absorber contacts
Solar Energy Materials & Solar Cells 120 (2014) 521–525
Contents lists available at ScienceDirect
Solar Energy Materials & Solar Cells journal homepage: www.elsevier.com/locate/solmat
Letter
Efficiency and stability enhancement of laser-crystallized polycrystalline silicon thin-film solar cells by laser firing of the absorber contacts M. Weizman a,n, H. Rhein b, J. Dore c, S. Gall b, C. Klimm b, G. Andrä d, C. Schultz a, F. Fink a, B. Rau b, R. Schlatmann a,b a
University of Applied Sciences Berlin (HTW), Schwarzschildstr. 3, Berlin 12489, Germany Helmholtz-Zentrum Berlin für Materialien und Energie (HZB), Germany University of New South Wales (UNSW), Sydney, Australia d Institute of Photonic Technology (IPHT), Jena, Germany b c
art ic l e i nf o
a b s t r a c t
Article history: Received 28 July 2013 Received in revised form 21 September 2013 Accepted 23 September 2013 Available online 14 October 2013
Polycrystalline silicon thin-film solar cells produced by continuous-wave diode-laser crystallization at the University of New South Wales were recently reported to have reached a conversion efficiency above 10%. One drawback of these cells, however, was that they exhibited efficiency degradation within several hours after the cell fabrication was completed. In this work we show that by applying laser firing to the rear point contacts of the solar cells, it is possible to stabilize and even to enhance the performance of these devices. Our investigation indicates that it is the poor quality of the contact between the aluminum and the silicon absorber that causes the cell degradation and offers an elegant and industrial-compatible process to improve the cell performance. This is the first time that the laser firing process, initially developed for alloying an aluminum layer through a dielectric layer on crystalline silicon wafer solar cells, is being applied to polycrystalline silicon thin-film solar cells. & 2013 Elsevier B.V. All rights reserved.
Keywords: Polycrystalline silicon Thin-film Solar cells Crystallization Laser firing
1. Introduction The fabrication of thin-film polycrystalline silicon (poly-Si) solar cells by liquid phase crystallization is an emerging technology that shows promise for combining the high efficiencies of wafer silicon solar cells with the low costs of thin-film production. In the last few years it has been demonstrated that high-quality silicon absorber layers can be produced on a glass substrate by a single crystallization pass using either an electron-beam line [1] or a continuous-wave diode-laser line [2–4]. These layers have centimeter-long grains in the scan direction and a much lower defect concentration compared to solid-phase-crystallized layers, which were developed earlier for solar cell applications [5–7]. Solar cells prepared by e-beam and laser crystallization have now reached an open circuit voltage Voc of around 580 mV [4,8] whereas the values reported for solid-phase-crystallized solar cells never increased significantly above 500 mV [5,6]. The solar cell efficiency record for polycrystalline solar cells prepared at UNSW has been rapidly improved over the last 2 years [2–4]. The best initial efficiency value reported so far is 11.7% and
n
Corresponding author. Tel.: þ 49 30 8062 18147. E-mail address: [email protected] (M. Weizman).
0927-0248/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.solmat.2013.09.033
the best stabilized value is 10.4% [4], which matches roughly the efficiency record for solid-phase crystallization [6] even without an optimized light trapping structure. One of the problems that was reported with these solar cells, however, is that they exhibit an efficiency degradation within several hours after the metallization of the rear side of the cell [2]. In this work, we show that the inadequate contact between the aluminum (Al) and the silicon absorber is most likely the cause for this degradation and that by applying laser firing to the absorber point contacts, the cell performance can be stabilized and even enhanced. Laser firing has been utilized since 2002 for creating point contacts by alloying the aluminum back contact with the crystalline silicon wafer through a dielectric layer [9]. Showing that laser firing can be applied also for polycrystalline silicon thin-film solar cells on glass opens new possibilities for the development of the contact system for this type of device. 2. Experimental details The poly-Si solar cells investigated in this study were prepared at UNSW according to the process described in Ref. [3]. The substrate used for the cells was 3.3 mm thick borosilicate float glass that was coated with a 200 nm SiOx intermediate layer. The
522
M. Weizman et al. / Solar Energy Materials & Solar Cells 120 (2014) 521–525
silicon absorber layer was deposited on the intermediate layers by electron-beam evaporation at 650 1C, was doped with boron (ptype), and had a thickness of about 10 mm. Then, the absorber layer was crystallized with a LIMO continuous-wave (cw) diode-laser line operating at a wavelength of 808 nm. The laser crystallization was performed at an energy fluence density of about 3 J/mm2, a substrate preheating temperature of about 600 1C, and a scan velocity of about 10 mm/s. The laser line had a top hat intensity profile along the long axis and a Gaussian profile along the short axis (scan direction). The full width at half maximum (FWHM) of the long axis was 12 mm and of the short axis was 0.17 mm. The ntype emitter layer was formed by spinning a phosphorous source onto the surface and diffusing it into the absorber by rapid thermal annealing. The final step before starting the contacting procedure was a hydrogen plasma passivation treatment at 600 1C applied for about 10 min. After finishing preparation of the layer stack, a rear contact scheme based on the process developed by CSG Solar [5] was utilized to form the emitter and absorber contacts. The cell size was defined by isolation laser scribes and had the dimensions of 0.6 cm 1.7 cm. A white paint resist was then coated on the emitter and holes were produced in this layer by inkjet printing. The holes were etched down to the emitter and to the absorber. The side walls of the absorber contact holes were isolated by reflowing the resist. A 100 nm Al layer was deposited on the surface in order to form the n-type (Dimples) and p-type (Craters) contacts. Cutting the Al in order to separate the emitter and the absorber contact regions was performed by a laser. Finally, annealing at 135 1C was performed for an hour to improve the metal contacts. At this stage, the cells were delivered to HZB for the subsequent laser firing treatment. The firing was performed by a neodymiumdoped vanadate solid-state laser from Rofin operating at a wavelength of 532 nm and a pulse frequency of 20 kHz. The pulse duration of this laser was about 16 ns and it had a Gaussian intensity profile.
Al
Poly-Si
8 µm
Fig. 1. Scanning electron microscopy images of 100 nm aluminum on a polycrystalline silicon absorber that was fired with a single pulse of a nanosecond laser, operating at a wavelength of 532 nm, using a laser fluence of a) 1.46 J/cm2, b) 1.85 J/ cm2, and c) 2.7 J/cm2. The dashed line marks the area of the laser treatment.
3. Results Section 3.1 presents the effect of firing with a single laser shot whereas Section 3.2 deals with laser firing performed by multiple irradiations on the same spot. The quality of the laser-fired contacts is evaluated in Section 3.3 using the transfer length method (TLM) and finally in Section 3.4 the effect of the laser firing on the solar cells is presented.
Fig. 2. Scanning electron microscopy (SEM) and energy dispersive X-ray (EDX) investigation done on a laser-fired spot that was broken in the middle. The laser firing was done with about 60 laser shots and a laser fluence of about 1.5 J/cm2. a) SEM of the sample tilted by 301. b) SEM of the sample tilted by 01. c) EDX crosssection map of Al. d) EDX cross-section map of Si. (For interpretation of the references to color in this figure, the reader is referred to the web version of this article.)
3.1. Aluminum ablation threshold
3.2. Laser firing with multiple irradiations
Since a very thin Al layer of about 100 nm was used for the back contact of the poly-Si cells, a crucial experiment was to test at what laser fluence this layer ablates. Fig. 1 shows scanning electron microscopy (SEM) images of the Al regions that were exposed to a single laser shot with different laser fluences. In part (a) of the figure, the laser fluence was set at 1.46 J/cm2, causing some small regions in the laser spot to ablate. As the laser fluence increases in part (b) to 1.85 J/cm2, the sample surface becomes rough but according to the broken sample edge the Al layer still covers the poly-Si layer completely. In part (c) where the laser fluence was 2.7 J/cm2, it can be seen that the Al piles up at the rim of the irradiated spot and is absent at the center. Our investigation revealed that the appearance of an elevated Al ring at the circumference of the laser spot indicated that ablation occurred at the center of the spot. The critical laser fluence for ablation was found to be about 2 J/cm2.
Applying multiple laser shots on the same spot is a method to enlarge the parameter window in which a sufficient amount of energy can be deposited into the Al layer without causing ablation. Fig. 2 shows two SEM images and two energy dispersive X-ray (EDX) maps from a spot that was irradiated with about 60 laser shots and then broken in the middle for investigation. Parts (a) and (b) show the SEM image of the laser-fired spot with a 301 surface tilting and with no tilting, respectively. In both images it is possible to see that a fine-grained layer of about 400 nm is formed on the sample surface. Parts (c) and (d) show the EDX map of the elements aluminum and silicon, respectively. The black regions stand for no signal and the brightness of the red (Al) or the green (Si) color corresponds to the concentration of the element. The EDX results show that the laser firing causes Al and Si to intermix and to be distributed over the entire region of the fine-grained layer seen in the SEM images of parts (a) and (b).
M. Weizman et al. / Solar Energy Materials & Solar Cells 120 (2014) 521–525
The appearance of such a fine-grained Si–Al alloy is not observed after a single laser shot, indicating that this effect builds up and evolves with an increasing number of laser shots.
assure that w can be taken as the length of the Crater line in spite of the discontinuity caused by the regions between the point contacts, the resistivity of the Si was determined by four-probe conductivity measurements and was found not to differ significantly from the results obtained below for the test sample. The resistance R measured on the test structure in Fig. 3 is linear with respect to d, thus allowing to derive according to Eq. (2) a bulk resistivity for the Si absorber layer of ρb ¼ 0.04 Ω cm. This resistivity value of the test sample is somewhat lower than usually chosen for the absorber in the solar cell but in this low-doping region such a deviation has a rather small effect on the specific contact resistance of less than one order of magnitude [11]. Extrapolating the linear fit to d ¼0 gives a contact resistance of Rc ¼1.38 Ω. The specific contact resistance ρc was estimated as
3.3. Contact resistance of the laser-fired spots To evaluate the quality of the laser-fired contacts, a test structure according to the transfer length method (TLM) [10] was produced. Besides the laser crystallization that was performed at IPHT Jena by LIMO laser irradiation as described in Section 2, all other process steps for this sample were done at HZB. As can be seen in the lower right inset of Fig. 3, the test structure consisted of Crater lines that were positioned at different distances d from each other. Since the distance between the lines is much larger than the distance between the point contacts, the major part of the current flows perpendicular to the lines as in the traditional TLM structure. In addition, the current collection is similar to the case of contacts placed on a planar silicon surface because the Crater side walls are isolated by the resist and the silicon–aluminum contact is only at the bottom of the Craters. The resistance between two Crater lines of this structure is given by R ¼ 2Rc þ Rb
ρc ¼ Rc A;
ρ b
tw
ð1Þ
ð2Þ
d;
ð3Þ
where A is the area of the point contact multiplied by the 41 points along the line. The lower limit for the area of the point contact was the visible laser-modified spot area and the upper limit was the entire area of the etched Crater. We believe that Eq. (3) is valid for our case since the transfer length, which characterizes the distance from the contact edge in which the current crowds, is larger than the diameter of our point contacts. This suggests that we have a rather homogenous current distribution underneath our contacts. The lower and upper-limit values derived for ρc after the laser firing were 4 10 4 Ω cm2 and 5 10 3 Ω cm2, respectively, which are several orders of magnitude better than before the laser treatment and are getting close to optimized values reported for laser firing of c-Si wafer of 5 10 5–5 10 4 Ω cm2 [12,13]. It should be noted that before the laser firing, the resistance R showed no dependency on the distance between contact lines d and exhibited a wide spread in the data values (see the upper left inset in Fig. 3). For the sake of clarity, we summarize the relevant parameters of this experiment in Table 1.
with Rb ¼
523
where R is the total measured resistance, Rc is the contact resistance of one Crater line, Rb is the resistance of the Si bulk between the two Crater lines, ρb is the bulk resistivity, t is the thickness of the Si layer, and w is the length of the Crater line. To
3.4. Laser firing on solar cells After deriving suitable laser firing parameters, the process was applied to the laser-crystallized poly-Si solar cells supplied by UNSW. Fig. 4 shows a schematic illustration of the poly-Si cell with the point contacts on the rear side. The red arrow pointing to the middle of a Crater shows where the laser firing was performed. Fig. 5 shows optical microscopy images of a solar cell after applying laser firing in the middle of the Craters with about 60 laser shots and a laser fluence of about 1.5 J/cm2. The magnification image on the right side shows a laser-fired spot within the Crater that appears darker due to the roughening of the Al surface in this region. The poly-Si solar cells were measured with an AAA dual-source sun simulator before and after applying the laser firing. Fig. 6 shows the J–V curves of the cell that had the largest improvement in performance after the laser firing. The cell parameters before
Fig. 3. Resistance R measured between lines of point contacts produced on a lasercrystallized test sample as function of the distance between the lines d according to the transfer length method (TLM). The sample structure is shown in the lower right inset. The contact resistance Rc between the Al and Si absorber before and after the laser firing is shown in the upper left inset.
Table 1 Summary of the parameters derived from the transfer length method (TLM) measurements. Sample structure
t (lm)
No. of point contacts per line
Crater diameter (lm)
Diameter of fired point
No. of shots per point contact
Laser fluence (J/cm2)
ρb (Ω cm)
Al/resist/polySi/SiOx/glass
6.6
41
110
29
60
1.5
0.04
Rc (Ω)
Before firing 8–215 2 103–5.5 104
1.38 t is the thickness of the silicon absorber layer,
ρc (Ω cm2)
After firing 4 10 4–5 10 3
ρb is the bulk silicon resistivity, Rc is the total contact resistance, and ρc is the specific contact resistance.
524
M. Weizman et al. / Solar Energy Materials & Solar Cells 120 (2014) 521–525
6.8
Laser firing Absorber contact line
Dimple
Crater
Resist n+ Si
p-Si absorber ~10 µm
6.6
Efficiency (%)
Emitter contact line
Cell 1
6.4
Cell 2
6.2 6.0 5.8 Annealing 135°C, 1h
Intermediate layer
Laser firing
5.6
Initial
Glass
28 days after 0 annealing
15
30
45
60
Time after firing (days)
Fig. 7. Efficiency stability of two solar cells after an hour annealing at 135 1C and after laser firing of the Crater point contacts. Fig. 4. Schematic structure of the laser-crystallized polycrystalline silicon solar cell. Laser firing was performed at the absorber point contacts (Craters) as illustrated by the red arrow. (For interpretation of the references to color in this figure the reader is referred to the web version of this article.)
37 µm
Dimple
Crater
Area fired with laser
Fig. 5. Optical microscopy images of the solar cell contact system. Left: alternating absorber and emitter contact lines on which Crater and Dimple points were placed, respectively. Right: higher magnification view of a Crater point contact that was fired at the middle region with about 60 ns-laser pulses at a wavelength of 532 nm.
time after applying the furnace annealing and the laser firing to the solar cell presented in Fig. 6 (cell 1) and to another solar cell (cell 2). After the furnace annealing and laser treatment, the cells were stored under dark-room conditions. The results show that annealing at 135 1C creates an increase in efficiency that diminishes after 28 days, whereas the laser firing creates a much larger increase in efficiency that is maintained for 69 days. The magnitude of the degradation after the furnace annealing treatment was reported by Dore et al. to decrease with repeated annealing cycles [4]. In our case, since the furnace annealing step presented in Fig. 7 was already the second annealing cycle after completing fabrication, the degradation magnitude is expected to be somewhat lower than in the first cycle. It should be noted that we performed additional annealing for cells 1 and 2 after the laser firing and found that they still exhibit a temporary efficiency increase of about 0.1%, which suggests that there might be some interaction between the annealing history of the sample and the laser firing. Further investigation is needed to test the effect of laser firing after a varying number of furnace annealing cycles. Altogether, our results clearly show that the laser firing enhances the cell efficiency beyond the level reached by furnace annealing and contributes to a stable operation of the cells at a high efficiency level.
4. Conclusions
Fig. 6. J–V curves before and after laser firing of the Crater point contacts.
and after the firing are given in the inset of the figure. The improvement of the cell efficiency η in this case was about 0.7% from 5.9% to 6.6%. The most significant enhancement is in the fill factor FF, particularly due to the reduction in the series resistance Rs from 5.6 Ω cm2 to 4.5 Ω cm2. Finally, we investigated the most crucial issue, the solar cell stability over time. Before applying the laser firing treatment, annealing at about 135 1C for about an hour was carried out. This was the annealing procedure used in Refs. [2] and [4] to recover cell performance after degradation. Fig. 7 shows the efficiency behavior over
The results presented here offer a new insight into the degradation phenomenon observed recently for laser-crystallized polycrystalline thin-film solar cells. Our finding that applying laser firing locally to the absorber point contacts stabilizes the cell performance allows us to rule out metastability phenomena related to the Si bulk material and to assign the degradation to the quality of the aluminum–silicon interface at the contact. We demonstrate here an elegant laser firing procedure which is able to stabilize and even enhance the efficiency of the solar cells by up to 0.7% absolute. The firing treatment is done with multiple irradiations on a 100 nm aluminum layer, yielding a fine-grained silicon–aluminum alloy contact with a specific contact resistance lower than 10 3 Ω cm2 and electrical characteristics that are more stable than those of the untreated weakly-doped silicon to aluminum contact. The mechanism behind the enhanced stability seems to be related to the reduction of the contact resistance but other possibilities such as the suppression of minority charge carrier recombination near the contact surface cannot be excluded yet. Further improvement of the aluminum–silicon contact for cell application is expected by adding a suitable passivation layer between silicon and aluminum. These results are an exciting step
M. Weizman et al. / Solar Energy Materials & Solar Cells 120 (2014) 521–525
on the way to a polycrystalline silicon solar cell based on liquidphase crystallization with a stable conversion efficiency above 12%.
Acknowledgments The authors gratefully acknowledge V. Juzumas for electrical measurements and M. Schüle for assistance with the laser scribing tool. The work done at UNSW has been supported by the Australian Government through the Australian Renewable Energy Agency (ARENA). The Australian Government, through ARENA, is supporting Australian research and development in solar photovoltaic and solar thermal technologies to help solar power become cost competitive with other energy sources.
[4]
[5]
[6] [7]
[8]
[9]
References [10] [1] D. Amkreutz, J. Müller, M. Schmidt, T. Hänel, T.F. Schulze, Electron-beam crystallized large grained silicon solar cell on glass substrate, Progress in Photovoltaics: Research and Applications 19 (2011) 937–945. [2] J. Dore, R. Evans, U. Schubert, B.D. Eggleston, D. Ong, K. Kim, J. Huang, O. Kunz, M. Keevers, R. Egan, S. Varlamov, M.A. Green, Thin-film polycrystalline silicon solar cells formed by diode laser crystallization, Progress in Photovoltaics: Research and Applications 21 (2013) 1377–1383, http://dx.doi.org/10.1002/ pip.2282. [3] J. Dore, S. Varlamov, R. Evans, B. Eggleston, D. Ong, O. Kunz, J. Huang, U. Schubert, K.H. Kim, R. Egan, M. Green, Performance potential of low-defect density silicon
[11] [12]
[13]
525
thin-film solar cells obtained by electron beam evaporation, EPJ Photovoltaics 4 (2013) 40301. J. Dore, D. Ong, S. Varlamov, R. Egan, M.A. Green, Progress in laser-crystallized thin-film polycrystalline silicon solar cells: intermediate layers, light Trapping, and metallization, IEEE Journal of Photovoltaics, http://dx.doi.org/10.1109/ JPHOTOV.2013.2280016. M. Green, P. Basore, N. Chang, D. Clugston, R. Egan, R. Evans, D. Hogg, S. Jarnason, M. Keevers, P. Lasswell, Crystalline silicon on glass (CSG) thinfilm solar cell modules, Solar Energy 77 (2004) 857–863. M. Keevers, T. Young, U. Schubert, M. Green, 10% efficient CSG minimodules, in: Proceedings of the 22nd EUPVSEC, Milan, Italy, 2007, pp. 1783–1790. R. Egan, M. Keevers, U. Schubert, T. Young, R. Evans, S. Partlin, M. Wolf, J. Schneider, D. Hogg, B. Eggleston, M. Green, F. Falk, A. Gawlik, G. Andrä, M. Werner, C. Hagendorf, P. Dogan, T. Sontheimer, S. Gall, CSG minimodules using electron-beam evaporated silicon, in: Proceedings of the 24th EUPVSEC, Hamburg, Germany, 2009, pp. 2279–2285. J. Haschke, L. Jogschies, D. Amkreutz, L. Korte, B. Rech, Polycrystalline silicon heterojunction thin-film solar cells on glass exhibiting 582 mV open-circuit voltage, Solar Energy Materials and Solar Cells 115 (2013) 7–10. E. Schneiderloechner, R. Preu, R. Luedemann, S.W. Glunz, Laser-fired rear contacts for crystalline silicon solar cells, Progress in Photovoltaics: Research and Applications 10 (2002) 29–34. W. Shockley, Research and investigation of inverse epitaxial uhf power transistors, Report No AT TDR 64-207, 1964. D.K. Schroder, D.L. Meier, Solar cell contact resistance—a review, IEEE Transactions on Electron Devices 31 (1984) 637–647. I. Sánchez-Aniorte, R. Barrio, A. Casado, M. Morales, J. Cárabe, J.J. Gandía, C. Molpeceres, Optimization of laser-firing processes for siliconheterojunction solar-cell back contacts, Applied Surface Sciences 258 (2012) 9443–9446. P. Ortega, A. Orpella, I. Martı, M. Colina, G. Lopez, C. Voz, M.I. Sanchez, C. Molpeceres, R. Alcubilla, Laser-fired contact optimization in c-Si solar cells, Progress in Photovoltaics: Research and Applications 20 (2012) 173–180.
Contents lists available at ScienceDirect
Solar Energy Materials & Solar Cells journal homepage: www.elsevier.com/locate/solmat
Letter
Efficiency and stability enhancement of laser-crystallized polycrystalline silicon thin-film solar cells by laser firing of the absorber contacts M. Weizman a,n, H. Rhein b, J. Dore c, S. Gall b, C. Klimm b, G. Andrä d, C. Schultz a, F. Fink a, B. Rau b, R. Schlatmann a,b a
University of Applied Sciences Berlin (HTW), Schwarzschildstr. 3, Berlin 12489, Germany Helmholtz-Zentrum Berlin für Materialien und Energie (HZB), Germany University of New South Wales (UNSW), Sydney, Australia d Institute of Photonic Technology (IPHT), Jena, Germany b c
art ic l e i nf o
a b s t r a c t
Article history: Received 28 July 2013 Received in revised form 21 September 2013 Accepted 23 September 2013 Available online 14 October 2013
Polycrystalline silicon thin-film solar cells produced by continuous-wave diode-laser crystallization at the University of New South Wales were recently reported to have reached a conversion efficiency above 10%. One drawback of these cells, however, was that they exhibited efficiency degradation within several hours after the cell fabrication was completed. In this work we show that by applying laser firing to the rear point contacts of the solar cells, it is possible to stabilize and even to enhance the performance of these devices. Our investigation indicates that it is the poor quality of the contact between the aluminum and the silicon absorber that causes the cell degradation and offers an elegant and industrial-compatible process to improve the cell performance. This is the first time that the laser firing process, initially developed for alloying an aluminum layer through a dielectric layer on crystalline silicon wafer solar cells, is being applied to polycrystalline silicon thin-film solar cells. & 2013 Elsevier B.V. All rights reserved.
Keywords: Polycrystalline silicon Thin-film Solar cells Crystallization Laser firing
1. Introduction The fabrication of thin-film polycrystalline silicon (poly-Si) solar cells by liquid phase crystallization is an emerging technology that shows promise for combining the high efficiencies of wafer silicon solar cells with the low costs of thin-film production. In the last few years it has been demonstrated that high-quality silicon absorber layers can be produced on a glass substrate by a single crystallization pass using either an electron-beam line [1] or a continuous-wave diode-laser line [2–4]. These layers have centimeter-long grains in the scan direction and a much lower defect concentration compared to solid-phase-crystallized layers, which were developed earlier for solar cell applications [5–7]. Solar cells prepared by e-beam and laser crystallization have now reached an open circuit voltage Voc of around 580 mV [4,8] whereas the values reported for solid-phase-crystallized solar cells never increased significantly above 500 mV [5,6]. The solar cell efficiency record for polycrystalline solar cells prepared at UNSW has been rapidly improved over the last 2 years [2–4]. The best initial efficiency value reported so far is 11.7% and
n
Corresponding author. Tel.: þ 49 30 8062 18147. E-mail address: [email protected] (M. Weizman).
0927-0248/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.solmat.2013.09.033
the best stabilized value is 10.4% [4], which matches roughly the efficiency record for solid-phase crystallization [6] even without an optimized light trapping structure. One of the problems that was reported with these solar cells, however, is that they exhibit an efficiency degradation within several hours after the metallization of the rear side of the cell [2]. In this work, we show that the inadequate contact between the aluminum (Al) and the silicon absorber is most likely the cause for this degradation and that by applying laser firing to the absorber point contacts, the cell performance can be stabilized and even enhanced. Laser firing has been utilized since 2002 for creating point contacts by alloying the aluminum back contact with the crystalline silicon wafer through a dielectric layer [9]. Showing that laser firing can be applied also for polycrystalline silicon thin-film solar cells on glass opens new possibilities for the development of the contact system for this type of device. 2. Experimental details The poly-Si solar cells investigated in this study were prepared at UNSW according to the process described in Ref. [3]. The substrate used for the cells was 3.3 mm thick borosilicate float glass that was coated with a 200 nm SiOx intermediate layer. The
522
M. Weizman et al. / Solar Energy Materials & Solar Cells 120 (2014) 521–525
silicon absorber layer was deposited on the intermediate layers by electron-beam evaporation at 650 1C, was doped with boron (ptype), and had a thickness of about 10 mm. Then, the absorber layer was crystallized with a LIMO continuous-wave (cw) diode-laser line operating at a wavelength of 808 nm. The laser crystallization was performed at an energy fluence density of about 3 J/mm2, a substrate preheating temperature of about 600 1C, and a scan velocity of about 10 mm/s. The laser line had a top hat intensity profile along the long axis and a Gaussian profile along the short axis (scan direction). The full width at half maximum (FWHM) of the long axis was 12 mm and of the short axis was 0.17 mm. The ntype emitter layer was formed by spinning a phosphorous source onto the surface and diffusing it into the absorber by rapid thermal annealing. The final step before starting the contacting procedure was a hydrogen plasma passivation treatment at 600 1C applied for about 10 min. After finishing preparation of the layer stack, a rear contact scheme based on the process developed by CSG Solar [5] was utilized to form the emitter and absorber contacts. The cell size was defined by isolation laser scribes and had the dimensions of 0.6 cm 1.7 cm. A white paint resist was then coated on the emitter and holes were produced in this layer by inkjet printing. The holes were etched down to the emitter and to the absorber. The side walls of the absorber contact holes were isolated by reflowing the resist. A 100 nm Al layer was deposited on the surface in order to form the n-type (Dimples) and p-type (Craters) contacts. Cutting the Al in order to separate the emitter and the absorber contact regions was performed by a laser. Finally, annealing at 135 1C was performed for an hour to improve the metal contacts. At this stage, the cells were delivered to HZB for the subsequent laser firing treatment. The firing was performed by a neodymiumdoped vanadate solid-state laser from Rofin operating at a wavelength of 532 nm and a pulse frequency of 20 kHz. The pulse duration of this laser was about 16 ns and it had a Gaussian intensity profile.
Al
Poly-Si
8 µm
Fig. 1. Scanning electron microscopy images of 100 nm aluminum on a polycrystalline silicon absorber that was fired with a single pulse of a nanosecond laser, operating at a wavelength of 532 nm, using a laser fluence of a) 1.46 J/cm2, b) 1.85 J/ cm2, and c) 2.7 J/cm2. The dashed line marks the area of the laser treatment.
3. Results Section 3.1 presents the effect of firing with a single laser shot whereas Section 3.2 deals with laser firing performed by multiple irradiations on the same spot. The quality of the laser-fired contacts is evaluated in Section 3.3 using the transfer length method (TLM) and finally in Section 3.4 the effect of the laser firing on the solar cells is presented.
Fig. 2. Scanning electron microscopy (SEM) and energy dispersive X-ray (EDX) investigation done on a laser-fired spot that was broken in the middle. The laser firing was done with about 60 laser shots and a laser fluence of about 1.5 J/cm2. a) SEM of the sample tilted by 301. b) SEM of the sample tilted by 01. c) EDX crosssection map of Al. d) EDX cross-section map of Si. (For interpretation of the references to color in this figure, the reader is referred to the web version of this article.)
3.1. Aluminum ablation threshold
3.2. Laser firing with multiple irradiations
Since a very thin Al layer of about 100 nm was used for the back contact of the poly-Si cells, a crucial experiment was to test at what laser fluence this layer ablates. Fig. 1 shows scanning electron microscopy (SEM) images of the Al regions that were exposed to a single laser shot with different laser fluences. In part (a) of the figure, the laser fluence was set at 1.46 J/cm2, causing some small regions in the laser spot to ablate. As the laser fluence increases in part (b) to 1.85 J/cm2, the sample surface becomes rough but according to the broken sample edge the Al layer still covers the poly-Si layer completely. In part (c) where the laser fluence was 2.7 J/cm2, it can be seen that the Al piles up at the rim of the irradiated spot and is absent at the center. Our investigation revealed that the appearance of an elevated Al ring at the circumference of the laser spot indicated that ablation occurred at the center of the spot. The critical laser fluence for ablation was found to be about 2 J/cm2.
Applying multiple laser shots on the same spot is a method to enlarge the parameter window in which a sufficient amount of energy can be deposited into the Al layer without causing ablation. Fig. 2 shows two SEM images and two energy dispersive X-ray (EDX) maps from a spot that was irradiated with about 60 laser shots and then broken in the middle for investigation. Parts (a) and (b) show the SEM image of the laser-fired spot with a 301 surface tilting and with no tilting, respectively. In both images it is possible to see that a fine-grained layer of about 400 nm is formed on the sample surface. Parts (c) and (d) show the EDX map of the elements aluminum and silicon, respectively. The black regions stand for no signal and the brightness of the red (Al) or the green (Si) color corresponds to the concentration of the element. The EDX results show that the laser firing causes Al and Si to intermix and to be distributed over the entire region of the fine-grained layer seen in the SEM images of parts (a) and (b).
M. Weizman et al. / Solar Energy Materials & Solar Cells 120 (2014) 521–525
The appearance of such a fine-grained Si–Al alloy is not observed after a single laser shot, indicating that this effect builds up and evolves with an increasing number of laser shots.
assure that w can be taken as the length of the Crater line in spite of the discontinuity caused by the regions between the point contacts, the resistivity of the Si was determined by four-probe conductivity measurements and was found not to differ significantly from the results obtained below for the test sample. The resistance R measured on the test structure in Fig. 3 is linear with respect to d, thus allowing to derive according to Eq. (2) a bulk resistivity for the Si absorber layer of ρb ¼ 0.04 Ω cm. This resistivity value of the test sample is somewhat lower than usually chosen for the absorber in the solar cell but in this low-doping region such a deviation has a rather small effect on the specific contact resistance of less than one order of magnitude [11]. Extrapolating the linear fit to d ¼0 gives a contact resistance of Rc ¼1.38 Ω. The specific contact resistance ρc was estimated as
3.3. Contact resistance of the laser-fired spots To evaluate the quality of the laser-fired contacts, a test structure according to the transfer length method (TLM) [10] was produced. Besides the laser crystallization that was performed at IPHT Jena by LIMO laser irradiation as described in Section 2, all other process steps for this sample were done at HZB. As can be seen in the lower right inset of Fig. 3, the test structure consisted of Crater lines that were positioned at different distances d from each other. Since the distance between the lines is much larger than the distance between the point contacts, the major part of the current flows perpendicular to the lines as in the traditional TLM structure. In addition, the current collection is similar to the case of contacts placed on a planar silicon surface because the Crater side walls are isolated by the resist and the silicon–aluminum contact is only at the bottom of the Craters. The resistance between two Crater lines of this structure is given by R ¼ 2Rc þ Rb
ρc ¼ Rc A;
ρ b
tw
ð1Þ
ð2Þ
d;
ð3Þ
where A is the area of the point contact multiplied by the 41 points along the line. The lower limit for the area of the point contact was the visible laser-modified spot area and the upper limit was the entire area of the etched Crater. We believe that Eq. (3) is valid for our case since the transfer length, which characterizes the distance from the contact edge in which the current crowds, is larger than the diameter of our point contacts. This suggests that we have a rather homogenous current distribution underneath our contacts. The lower and upper-limit values derived for ρc after the laser firing were 4 10 4 Ω cm2 and 5 10 3 Ω cm2, respectively, which are several orders of magnitude better than before the laser treatment and are getting close to optimized values reported for laser firing of c-Si wafer of 5 10 5–5 10 4 Ω cm2 [12,13]. It should be noted that before the laser firing, the resistance R showed no dependency on the distance between contact lines d and exhibited a wide spread in the data values (see the upper left inset in Fig. 3). For the sake of clarity, we summarize the relevant parameters of this experiment in Table 1.
with Rb ¼
523
where R is the total measured resistance, Rc is the contact resistance of one Crater line, Rb is the resistance of the Si bulk between the two Crater lines, ρb is the bulk resistivity, t is the thickness of the Si layer, and w is the length of the Crater line. To
3.4. Laser firing on solar cells After deriving suitable laser firing parameters, the process was applied to the laser-crystallized poly-Si solar cells supplied by UNSW. Fig. 4 shows a schematic illustration of the poly-Si cell with the point contacts on the rear side. The red arrow pointing to the middle of a Crater shows where the laser firing was performed. Fig. 5 shows optical microscopy images of a solar cell after applying laser firing in the middle of the Craters with about 60 laser shots and a laser fluence of about 1.5 J/cm2. The magnification image on the right side shows a laser-fired spot within the Crater that appears darker due to the roughening of the Al surface in this region. The poly-Si solar cells were measured with an AAA dual-source sun simulator before and after applying the laser firing. Fig. 6 shows the J–V curves of the cell that had the largest improvement in performance after the laser firing. The cell parameters before
Fig. 3. Resistance R measured between lines of point contacts produced on a lasercrystallized test sample as function of the distance between the lines d according to the transfer length method (TLM). The sample structure is shown in the lower right inset. The contact resistance Rc between the Al and Si absorber before and after the laser firing is shown in the upper left inset.
Table 1 Summary of the parameters derived from the transfer length method (TLM) measurements. Sample structure
t (lm)
No. of point contacts per line
Crater diameter (lm)
Diameter of fired point
No. of shots per point contact
Laser fluence (J/cm2)
ρb (Ω cm)
Al/resist/polySi/SiOx/glass
6.6
41
110
29
60
1.5
0.04
Rc (Ω)
Before firing 8–215 2 103–5.5 104
1.38 t is the thickness of the silicon absorber layer,
ρc (Ω cm2)
After firing 4 10 4–5 10 3
ρb is the bulk silicon resistivity, Rc is the total contact resistance, and ρc is the specific contact resistance.
524
M. Weizman et al. / Solar Energy Materials & Solar Cells 120 (2014) 521–525
6.8
Laser firing Absorber contact line
Dimple
Crater
Resist n+ Si
p-Si absorber ~10 µm
6.6
Efficiency (%)
Emitter contact line
Cell 1
6.4
Cell 2
6.2 6.0 5.8 Annealing 135°C, 1h
Intermediate layer
Laser firing
5.6
Initial
Glass
28 days after 0 annealing
15
30
45
60
Time after firing (days)
Fig. 7. Efficiency stability of two solar cells after an hour annealing at 135 1C and after laser firing of the Crater point contacts. Fig. 4. Schematic structure of the laser-crystallized polycrystalline silicon solar cell. Laser firing was performed at the absorber point contacts (Craters) as illustrated by the red arrow. (For interpretation of the references to color in this figure the reader is referred to the web version of this article.)
37 µm
Dimple
Crater
Area fired with laser
Fig. 5. Optical microscopy images of the solar cell contact system. Left: alternating absorber and emitter contact lines on which Crater and Dimple points were placed, respectively. Right: higher magnification view of a Crater point contact that was fired at the middle region with about 60 ns-laser pulses at a wavelength of 532 nm.
time after applying the furnace annealing and the laser firing to the solar cell presented in Fig. 6 (cell 1) and to another solar cell (cell 2). After the furnace annealing and laser treatment, the cells were stored under dark-room conditions. The results show that annealing at 135 1C creates an increase in efficiency that diminishes after 28 days, whereas the laser firing creates a much larger increase in efficiency that is maintained for 69 days. The magnitude of the degradation after the furnace annealing treatment was reported by Dore et al. to decrease with repeated annealing cycles [4]. In our case, since the furnace annealing step presented in Fig. 7 was already the second annealing cycle after completing fabrication, the degradation magnitude is expected to be somewhat lower than in the first cycle. It should be noted that we performed additional annealing for cells 1 and 2 after the laser firing and found that they still exhibit a temporary efficiency increase of about 0.1%, which suggests that there might be some interaction between the annealing history of the sample and the laser firing. Further investigation is needed to test the effect of laser firing after a varying number of furnace annealing cycles. Altogether, our results clearly show that the laser firing enhances the cell efficiency beyond the level reached by furnace annealing and contributes to a stable operation of the cells at a high efficiency level.
4. Conclusions
Fig. 6. J–V curves before and after laser firing of the Crater point contacts.
and after the firing are given in the inset of the figure. The improvement of the cell efficiency η in this case was about 0.7% from 5.9% to 6.6%. The most significant enhancement is in the fill factor FF, particularly due to the reduction in the series resistance Rs from 5.6 Ω cm2 to 4.5 Ω cm2. Finally, we investigated the most crucial issue, the solar cell stability over time. Before applying the laser firing treatment, annealing at about 135 1C for about an hour was carried out. This was the annealing procedure used in Refs. [2] and [4] to recover cell performance after degradation. Fig. 7 shows the efficiency behavior over
The results presented here offer a new insight into the degradation phenomenon observed recently for laser-crystallized polycrystalline thin-film solar cells. Our finding that applying laser firing locally to the absorber point contacts stabilizes the cell performance allows us to rule out metastability phenomena related to the Si bulk material and to assign the degradation to the quality of the aluminum–silicon interface at the contact. We demonstrate here an elegant laser firing procedure which is able to stabilize and even enhance the efficiency of the solar cells by up to 0.7% absolute. The firing treatment is done with multiple irradiations on a 100 nm aluminum layer, yielding a fine-grained silicon–aluminum alloy contact with a specific contact resistance lower than 10 3 Ω cm2 and electrical characteristics that are more stable than those of the untreated weakly-doped silicon to aluminum contact. The mechanism behind the enhanced stability seems to be related to the reduction of the contact resistance but other possibilities such as the suppression of minority charge carrier recombination near the contact surface cannot be excluded yet. Further improvement of the aluminum–silicon contact for cell application is expected by adding a suitable passivation layer between silicon and aluminum. These results are an exciting step
M. Weizman et al. / Solar Energy Materials & Solar Cells 120 (2014) 521–525
on the way to a polycrystalline silicon solar cell based on liquidphase crystallization with a stable conversion efficiency above 12%.
Acknowledgments The authors gratefully acknowledge V. Juzumas for electrical measurements and M. Schüle for assistance with the laser scribing tool. The work done at UNSW has been supported by the Australian Government through the Australian Renewable Energy Agency (ARENA). The Australian Government, through ARENA, is supporting Australian research and development in solar photovoltaic and solar thermal technologies to help solar power become cost competitive with other energy sources.
[4]
[5]
[6] [7]
[8]
[9]
References [10] [1] D. Amkreutz, J. Müller, M. Schmidt, T. Hänel, T.F. Schulze, Electron-beam crystallized large grained silicon solar cell on glass substrate, Progress in Photovoltaics: Research and Applications 19 (2011) 937–945. [2] J. Dore, R. Evans, U. Schubert, B.D. Eggleston, D. Ong, K. Kim, J. Huang, O. Kunz, M. Keevers, R. Egan, S. Varlamov, M.A. Green, Thin-film polycrystalline silicon solar cells formed by diode laser crystallization, Progress in Photovoltaics: Research and Applications 21 (2013) 1377–1383, http://dx.doi.org/10.1002/ pip.2282. [3] J. Dore, S. Varlamov, R. Evans, B. Eggleston, D. Ong, O. Kunz, J. Huang, U. Schubert, K.H. Kim, R. Egan, M. Green, Performance potential of low-defect density silicon
[11] [12]
[13]
525
thin-film solar cells obtained by electron beam evaporation, EPJ Photovoltaics 4 (2013) 40301. J. Dore, D. Ong, S. Varlamov, R. Egan, M.A. Green, Progress in laser-crystallized thin-film polycrystalline silicon solar cells: intermediate layers, light Trapping, and metallization, IEEE Journal of Photovoltaics, http://dx.doi.org/10.1109/ JPHOTOV.2013.2280016. M. Green, P. Basore, N. Chang, D. Clugston, R. Egan, R. Evans, D. Hogg, S. Jarnason, M. Keevers, P. Lasswell, Crystalline silicon on glass (CSG) thinfilm solar cell modules, Solar Energy 77 (2004) 857–863. M. Keevers, T. Young, U. Schubert, M. Green, 10% efficient CSG minimodules, in: Proceedings of the 22nd EUPVSEC, Milan, Italy, 2007, pp. 1783–1790. R. Egan, M. Keevers, U. Schubert, T. Young, R. Evans, S. Partlin, M. Wolf, J. Schneider, D. Hogg, B. Eggleston, M. Green, F. Falk, A. Gawlik, G. Andrä, M. Werner, C. Hagendorf, P. Dogan, T. Sontheimer, S. Gall, CSG minimodules using electron-beam evaporated silicon, in: Proceedings of the 24th EUPVSEC, Hamburg, Germany, 2009, pp. 2279–2285. J. Haschke, L. Jogschies, D. Amkreutz, L. Korte, B. Rech, Polycrystalline silicon heterojunction thin-film solar cells on glass exhibiting 582 mV open-circuit voltage, Solar Energy Materials and Solar Cells 115 (2013) 7–10. E. Schneiderloechner, R. Preu, R. Luedemann, S.W. Glunz, Laser-fired rear contacts for crystalline silicon solar cells, Progress in Photovoltaics: Research and Applications 10 (2002) 29–34. W. Shockley, Research and investigation of inverse epitaxial uhf power transistors, Report No AT TDR 64-207, 1964. D.K. Schroder, D.L. Meier, Solar cell contact resistance—a review, IEEE Transactions on Electron Devices 31 (1984) 637–647. I. Sánchez-Aniorte, R. Barrio, A. Casado, M. Morales, J. Cárabe, J.J. Gandía, C. Molpeceres, Optimization of laser-firing processes for siliconheterojunction solar-cell back contacts, Applied Surface Sciences 258 (2012) 9443–9446. P. Ortega, A. Orpella, I. Martı, M. Colina, G. Lopez, C. Voz, M.I. Sanchez, C. Molpeceres, R. Alcubilla, Laser-fired contact optimization in c-Si solar cells, Progress in Photovoltaics: Research and Applications 20 (2012) 173–180.