- Email: [email protected]
Back-contact back-junction silicon solar cells under UV illumination
Solar Energy Materials & Solar Cells 94 (2010) 1734–1740
Contents lists available at ScienceDirect
Solar Energy Materials & Solar Cells journal homepage: www.elsevier.com/locate/solmat
Back-contact back-junction silicon solar cells under UV illumination F. Granek n, C. Reichel Fraunhofer Institute for Solar Energy Systems, Heidenhofstr. 2, D-79110 Freiburg, Germany
a r t i c l e in f o
a b s t r a c t
Article history: Received 8 December 2009 Received in revised form 4 May 2010 Accepted 14 May 2010
The performance of n-type Si back-contact back-junction (BC-BJ) solar cells under illumination with high energy ultraviolet (UV) photons was investigated. The impact of the phosphorus doped front surface field (FSF) layer on the stability of the front surface passivation under UV illumination was investigated. Lifetime samples and solar cells without the front surface field showed a significant performance reduction when exposed to ultraviolet light. The surface saturation current density (J0e) increased from 48 to 446 fA/cm2 after the UV exposure. At the same time the efficiency of the BC-BJ solar cells without the FSF diffusion reduced from 19.8% to 14.3%. In contrast to the lifetime samples and solar cells without the FSF diffusion, the tested n + nn + structures and the BC-BJ solar cells with applied FSF diffusion profiles were significantly more stable under UV exposure, i.e. J0e increased only by a factor of 25% and the efficiency of these cells decreased only 0.3%abs by the UV illumination. Finally it was shown that the performance of the UV-degraded solar cells without FSF could be improved during a forming gas anneal (FGA). Due to application of FGA the efficiency almost fully recovered from 14.3% to 19.6%. & 2010 Elsevier B.V. All rights reserved.
Keywords: c-Si Back-contact UV-stability Front surface field
1. Introduction Back-contact back-junction (BC-BJ) silicon solar cells represent an attractive high-efficiency cell structure. In mass production, BC-BJ solar cells achieve average efficiencies of 22.4% as presented by De Ceuster et al. [1]. For comparison, standard silicon solar cells currently have efficiency in the range of 16–18%. Recently a new record efficiency of 23.4% for a large area (149 cm2) BC-BJ solar cell was announced by Swanson [2]. Due to the fact that the collecting p–n junction is placed on the rear surface of BC-BJ solar cells, and that most of the photogeneration takes place close to the front side, the requirements on the front surface passivation quality are very high. Thus, a low front surface recombination rate is one of the critical factors influencing the efficiency of the back-junction solar cells. The front surface passivation scheme needs not only to be of very high quality. It is also essential that the applied front surface passivation scheme is stable under solar cell operating conditions. Especially the stability of the surface passivation under the high energy ultraviolet (UV) part of the solar spectrum is of main importance in order to maintain high device performance during the long-term field operation of the photovoltaic modules. In our previous work [3] the front surface passivation scheme using a phosphorus doped front surface field (FSF) and a stack system of a thermal oxide and a PECVD silicon nitride for the
n
Corresponding author. Tel.: + 49 761 4588 5355; fax: + 49 761 4588 9250. E-mail address: fi[email protected] (F. Granek).
0927-0248/$ - see front matter & 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.solmat.2010.05.038
BC-BJ solar cells was developed and studied. Also other positive effects of the FSF on the performance of the BC-BJ solar cells were investigated, such as improvement of the low-illumination performance [4] and improvement of the lateral transport of the majority carriers [5]. In the present work the influence of the exposure of the front surfaces of the BC-BJ solar cells to UV light is investigated. Additionally the influence of the FSF diffusion on the UV stability of lifetime samples and on the performance of the BC-BJ solar cells is presented.
2. Influence of the UV light on the front surface passivation of BC-BJ solar cells The impact of the UV illumination of the oxide passivated silicon surfaces was investigated by many authors in the microelectronic field [6,7]. In the field of the silicon solar cells the work of Gruenbaum was pioneering. Already in 1988 Gruenbaum et al. [8,9] reported that the efficiency of some of the point-contact concentrator solar cells, developed at the Stanford University by Sinton et al. [10], decreased after exposure to concentrated sunlight. The decrease in solar cell performance was caused by an increase in the front surface recombination velocity. The studies of Gruenbaum et al. showed that the ultraviolet component of the incident light spectrum caused damage on the front surface.
F. Granek, C. Reichel / Solar Energy Materials & Solar Cells 94 (2010) 1734–1740
Gruenbaum et al. [11] performed the UV exposure and photoinjection experiments. It was discovered that the UV light of energy greater than 3.1 eV causes an increase in the surface recombination velocity (S0,front) and increases interface state densities of the surfaces passivated with oxide. The energy of 3.1 eV corresponds to light with wavelength shorter than 400 nm. In the terrestrial solar spectrum there is a significant amount of photons with such energy [12]. The absorption of the UV photons with wavelength shorter than 400 nm could inject electrons from the conduction band in silicon into the conduction band of silicon oxide. This photoinjection would then create defects at the Si/SiO2 interface. However, Gruenbaum et al. [13] and Ruby and Schubert [14] showed that not all solar cell structures are prone to degradation under UV light. The formation of the diffused phosphorus region on the front side creates a high field region. This field reduces the concentration of the minority charge carriers close to the recombination centres at the front silicon surface. Even if S0,front increases substantially, the effective surface recombination velocity (Seff) may increase minimally, leading to just a small reduction in efficiency of the solar cell.
3. Experimental Symmetrical lifetime test structures, as well as high-efficiency BC-BJ solar cells were processed and analyzed in order to investigate the impact of the high energy UV photons on front surface saturation current density (J0e) and on the solar cell performance.
1735
temperature drive-in oxidation process at the temperature of 1050 1C. The high oxidation temperature caused redistribution of the phosphorus atoms and transition from the error-function dopant profile (profile: FSF-shallow) to Gaussian profile (profile: FSF-deep). The silicon oxide layer was then etched back in HF solution. Sheet resistance (shown in Fig. 2) of the investigated front surface field phosphorus profiles was calculated by integrating the dopant profiles using the mobility model of Masetti et al. [15]. Next, all samples were passivated with a thin (10 nm) thermal SiO2 layer and an antireflection-SiNx coating (thickness of 60 nm). Finally, the samples were annealed in a forming gas atmosphere at a temperature of 425 1C for 15 min. 3.2. Back-contact back-junction Si solar cells The schematic cross-section of the BC–BJ solar cell analyzed in this work is shown in Fig. 3. The cells were also fabricated from n-type FZ–Si wafers with the resistivity of 1 O cm and a thickness of around 160 mm. The front cell side has the same structure as both surfaces of the lifetime structures. Thus, the front side of the BC–BJ solar cells is textured with random pyramids. For selected samples, the front surface was additionally passivated with phosphorus n + front surface field diffusion. The same phosphorus diffusion profiles (FSF-deep and FSF-shallow) as discussed in Section 3.1 were applied to the solar cells. The front surface was then passivated
1020 ρ
sheet
= 150 Ω/sq
FSF-shallow ρ
sheet
= 350 Ω/sq
Symmetrical n + nn + test structures for the minority carrier lifetime measurements were processed on n-type FZ–Si wafers with a thickness of 250 mm and specific resistivity of 1 O cm. Due to very high bulk lifetime of the minority carriers in the applied ntype FZ–Si wafers, the effective minority carrier lifetime is almost entirely limited by the surface recombination. In this way an accurate determination of the surface recombination is possible. The lifetime structures shown in Fig. 1 were processed on textured wafers (random pyramids texture). For selected samples, directly after the texturization process, both surfaces were additionally passivated by the application of the phosphorus doped front surface field. FSF doping profiles were formed by diffusion from a liquid POCl3 source in a tube furnace. Two different FSF doping profiles investigated in this study are shown in Fig. 2. Both profiles were formed at the same diffusion temperature. After diffusion the phosphorus glass was fully etched back in a buffered HF solution. The deep diffused (‘‘FSF-deep’’ in Fig. 2) profile was formed by an additional high
Phosphorus concentration ND [cm-3]
FSF-deep 3.1. Lifetime structures for determination of the surface saturation current density
1019
1018
1017
0.0
0.2
0.6
0.8
1.0
1.2
1.4
Depth [μm] Fig. 2. Secondary ion mass spectroscopy (SIMS) profiles of the studied phosphorus dopant profiles measured after all high temperature processing steps. Sheet resistance of both diffusion profiles is shown in the graph.
AR SiNX
AR SiNX
0.4
SiO 2
SiO2 n+ FSF (optional) n+ BSF
n++ FSF FSF(optional) (optional) n
n-Si
p+ emitter
n-Si n-Si metal fingers pitch
Fig. 1. Textured symmetrical structures used for the lifetime measurements and for the determination of the surface recombination current density.
SiO 2
Fig. 3. Schematic cross-section of the n-type high-efficiency back-contact backjunction silicon solar cell investigated in this work. Note that sketch is not to scale.
1736
F. Granek, C. Reichel / Solar Energy Materials & Solar Cells 94 (2010) 1734–1740
with a thin (10 nm) thermally grown silicon dioxide layer. PECVD SiNx antireflection coating layer (60 nm) was deposited on top of the oxide layer. The rear side was patterned using industrially feasible technology (e.g. screen-printing and laser ablation). These technologies lead to a pitch of the finished solar cells of 1.8 mm. Both emitter p++ and back surface field n++ line diffusions are separated by an undiffused gap, the rear structure forms an interdigitated grid. The rear cell surface is passivated with thermally grown silicon oxide. Metal fingers are contacted to the diffused regions through local openings in the thermal oxide layer. The size of the active cell area is 2 2 cm2. Solar cells were cut from the host wafer with a laser, where the distance from the edge of the active area to the edge was 500 mm. More details on the processing technology and the performance of the described BC-BJ solar cells are presented in Ref. [5]. The solar cells with efficiencies in the range of 19.7–20.2% and with two different diffusion profiles of the front n + diffusion and without the front n + diffusion were selected for the analysis of the performance stability under UV exposure. 3.3. Exposure of the lifetime samples and the solar cells to the UV illumination The stability of the described surface passivation schemes of the lifetime samples and the BC-BJ solar cells with and without the FSF diffusion under UV light was examined by the exposure to a xenon arc lamp with the total irradiation of 1000 kWh/m2. The spectrum of the Xe arc lamp applied in the UV exposure test was measured in the plane of the tested samples. This spectrum together with a standard AM1.5 spectrum is shown in Fig. 4. In the UV-spectral range a good agreement between both spectrums can be observed. The temperature of the lifetime samples and solar cells during the exposure equals 50 1C. The international standard IEC 61215 [16] defines the UV-test of the crystalline silicon photovoltaic modules as a UV exposure in the spectral range between 280 and 385 nm with the irradiation of 15 kW h/m2. The module temperature should equal 60 75 1C. The UV irradiation in this spectral range equals roughly 5% of the global solar irradiation. For a geographical location with the global solar irradiation of 1000 kW h/m2 (e.g. central Europe), the yearly UV irradiation equals 50 kWh/m2. In central Europe the UV exposure defined in the IEC standard (15 kW h/m2) corresponds to roughly 3.5 months of the out-door irradiation.
Results of the surface saturation current density of the test lifetime samples under the UV exposure tests are shown in Fig. 5. J0e of lifetime samples without FSF increases significantly already after the first few hours of UV exposure. J0e increases from initial value of 48–446 fA/cm2 after 55 h of UV light exposure and the value of J0e seems to saturate after that exposure time. A drastic increase of J0e to 446 fA/cm2, which for base resistivity of 1 O cm corresponds to S0,front of 139 cm/s, is expected to cause a significant performance degradation of the back-contact backjunction solar cells.
60 40
1000
20
500 400
600 800 Wavelength [nm]
1000
0 1200
Fig. 4. Spectrum of the xenon arc lamp, which was applied for the UV exposure of the lifetime samples and solar cells. The standard AM1.5 spectrum is shown in comparison [12]. The irradiation of both spectrums equals 1000 kWh/m2. Both spectrums are in a good agreement in the UV-spectral range (up to 400 nm). Transmission measurements of the solar module glass and of EVA encapsulant [22] are shown as well.
FSF-deep FSF-shallow no FSF
1000
J0e [fA/cm2]
Standard AM1.5 Spectrum Xe arc lamp
1500
0 200
4.1. Lifetime test structures under UV exposure
Surface Saturation Current Density
80
Transmission [%]
2000
Module glass EVA
Spectral irradiance [W/μm m2]
8000
2500
4. Results
100
10000
6000
The lifetime structures and solar cells were not covered with a module glass or any other intermediate layer during the exposure. Therefore the illumination spectrum absorbed by the test structures was rich in the high energy UV photons. The analyzed samples were exposed to UV illumination for 60 h. The resulting UV dosage equals 2.2 kW h/m2. For a location in central Europe such a UV dosage corresponds to roughly 15 days of outdoor exposure. However since in the normal operation mode, the solar cells are covered with solar glass and EVA encapsulant the UV irradiation, which reaches solar cell in a module, is strongly reduced. Given the transmission of the solar glass and of the EVA encapsulant, as shown in Fig. 4, only 21% of the UV dose defined by the IEC standard reaches the solar cell in a module. When taking the transmission losses of the UV irradiation in the solar module, the corresponding UV dose would equal 9.5 kW h/m2, which corresponds to 70 days of the outdoor exposure and 63% of the UV dosage defined by the IEC standard. The solar cells were illuminated from the front side, and the lifetime samples were illuminated from one side only. The UV illumination was performed in time steps, between which the characterization of these samples was performed. In case of the lifetime structures the minority carrier lifetime was measured using the quasi-steady-state photo-conductance (QSSPC) measurement set-up [17] and the generalized analysis method was applied [18] after each exposure step. Then the front surface recombination velocity S0,front and the surface saturation current density J0e were determined as described in Refs. [19] and [3], respectively. For the BC-BJ solar cells, the light current–voltage (I–V) characteristics, as well as external quantum efficiency (EQE) were measured after each UV exposure step.
ρsheet = 150 Ω/sq ρsheet = 350 Ω/sq
100
10 0.1
1 10 exposure time t [h]
100
Fig. 5. Surface saturation current density of textured test lifetime samples for two different front surface field diffusion profiles and for samples without FSF during 55 h of exposure to UV light.
F. Granek, C. Reichel / Solar Energy Materials & Solar Cells 94 (2010) 1734–1740
On the other hand, the test structures with either FSF diffusion profiles exhibit a significant increase of J0e during the exposure tests. J0e of the test structure with the ‘‘FSF-shallow’’ diffusion profile with sheet resistance of 350 O/sq increased from 29 initially to 36 fA/cm2 after the 55 h of exposure. The structures with the ‘‘FSF-deep’’ diffusion profile with sheet resistance of 150 O/sq increase from 34 to 45 fA/cm2 after the exposure. This proves that passivation using a FSF is significantly more stable under UV illumination and therefore appropriate for industrial applications, in contrast to unstable passivation without FSF. Similar results were also obtained by Gruenbaum et al. [9]. It is therefore expected that the BC-BJ solar cells with the applied FSF diffusion will be stabile under UV illumination. 4.2. Solar cell results Solar cells results of the analyzed BC-BJ solar cells under UV illumination are summarized in Tables 1–3. The initial solar cell performance (exposure time of 0 h) is given in the first row of the tables. In the last row of these tables (‘‘After FGA’’), the solar cell results after annealing in a forming gas atmosphere (FGA) at a temperature of 425 1C for 25 min are presented. The annealing process was performed after completing the UV exposure tests of all solar cells. The performance change of the solar cells after the final FGA process step will be discussed in Section 4.5. The BC-BJ solar cell without front surface field diffusion (Table 1) shows a major efficiency reduction after the UV illumination. Efficiency was reduced from 19.8 to 14.3%, mainly due to Table 1 I–V parameters of the BC-BJ solar cell without the FSF diffusion (cell no. BC47-22b) under UV illumination. Results in table are designated cell area measurements (2 2 cm2). Exposure time (h)
VOC (mV)
JSC (mA/cm2)
FF (%)
Z (%)
0 1 3 48 60
661 652 633 629 625
37.3 37.3 31.1 28.9 28.7
80.1 79.3 79.4 79.5 79.7
19.8 19.3 15.6 14.4 14.3
After FGA
659
36.9
80.4
19.6
Table 2 I–V parameters of the BC-BJ solar cell with the ‘‘FSF-deep’’ diffusion profile (cell no. BC47-17b) under UV illumination. Results in table are designated cell area measurements (2 2 cm2). Exposure time (h)
VOC (mV)
JSC (mA/cm2)
FF (%)
Z (%)
0 48 60
652 651 651
38.8 38.1 38.4
80.1 79.9 79.5
20.2 19.8 19.9
After FGA
661
38.3
80.4
20.4
Table 3 I–V parameters of the BC-BJ solar cell with the ‘‘FSF-shallow’’ diffusion profile (cell no. BC47-10b) under UV illumination. Results in table are designated cell area measurements (2 2 cm2). Exposure time (h)
VOC (mV)
JSC (mA/cm2)
FF (%)
Z (%)
0 48 60
653 651 651
39.2 38.1 38.4
76.9 79.9 79.5
19.7 19.8 19.9
After FGA
661
39.0
80.5
20.8
1737
loss in short-circuit current of 8.6 mA/cm2. This loss is manifested as a strong reduction of the EQE in the whole solar spectrum, as will be shown in the next section. Also the open-circuit voltage decreased significantly after the UV illumination. This can be explained by the strong increase in the total saturation current density of the solar cell, due to the increase of the front surface saturation current. In the case of solar cells with both FSF diffusion profiles (see Tables 2 and 3), the performance degradation after the UV illumination is only minimal, proving that the application of the FSF diffusion indeed strongly increases the radiation resistance of the BC-BJ solar cells to the high energy UV photons. Note the initial lower fill factor of the solar cell BC47-10b is believed to be caused by the contacting problems during the solar cell measurement. It is therefore not believed that the UV illumination affected the FF in a positive way, as might be interpreted from the results in Table 3. 4.3. External quantum efficiency External quantum efficiency of the analyzed solar cells before and after UV illumination is presented in Fig. 6. EQE of the solar cell without FSF degrades rapidly from 92% to 75% at wavelength of 500 nm already after 3 h of exposure to UV light. After 60 h of UV exposure the EQE of the solar cell without FSF decreases even further to 65%. Thus, an EQE decrease of nearly 30% absolute was caused by 60 h of exposure to UV light, and is responsible for the reduction of the short-circuit current presented in Table 1. EQE of the solar cells with both FSF diffusion profiles was reduced only by 2% absolute after the UV illumination. For the solar cell with ‘‘FSF-deep’’ after 60 h of exposure, the EQE dropped from 92% to 90% at a wavelength of 500 nm. For the solar cell with ‘‘FSF-shallow’’ diffusion profile the EQE dropped from 90% to 88%. 4.4. Device simulations The influence of the UV exposure on the EQE of the solar cell without FSF was compared with the results of the lifetime test structures and modelled using PC1D [20] simulations. First, the S0,front of the textured lifetime samples was measured before and after UV exposure as described in Section 4.1. Secondly, the measured EQEs were fitted using the PC1D model of a backjunction solar cell. The measured reflectance spectrum of the actual solar cell was used in the simulations. In order to fit the EQE results in the short wavelength range, the S0,front was varied. The experimentally determined and modelled EQEs are presented in Fig. 7. The PC1D fit in the long wavelength range is not in a very good agreement with measured data due to strong influence of the two-dimensional effects, which could not be described by the one-dimensional PC1D simulation. Please mind that the analyzed solar cells have the pitch of 1.8 mm; these devices are therefore strongly two-dimensional structures. This effect will have an impact on the value of JSC determined by the PC1D simulations. However, the effect of the increased front surface recombination rate affects the quantum efficiency of the solar cells mainly in the short to medium-length wavelength of the solar spectrum. In the short and medium-length wavelength spectrum the measured and modelled EQE are in a good agreement. The results of PC1D fitting and S0,front determined from the lifetime measurement of the n + nn + test structures are summarized in Table 4. A very good agreement between the S0,front determined by analysis of the lifetime samples and by fitting of the EQE of the solar cells was obtained. This proves that the
1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1
BC47-22b no FSF diffusion before UV exposure after 3 h exposure after 48 h exposure after 60 h exposure
0.0 300 400 500 600 700 800 900 1000 11001200
External Quantum Efficiency EQE [-]
F. Granek, C. Reichel / Solar Energy Materials & Solar Cells 94 (2010) 1734–1740
External Quantum Efficiency EQE [-]
1738
1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1
BC47-17b with FSF-deep diffusion before UV exposure after 60 h exposure
0.0 300 400 500 600 700 800 900 1000 11001200
Wavelength λ [nm]
External Quantum Efficiency EQE [-]
Wavelength λ [nm] 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1
BC47-10b with FSF-shallow diffusion before UV exposure after 60 h exposure
0.0 300 400 500 600 700 800 900 1000 11001200
Wavelength λ [nm]
External Quantum Efficiency EQE [-]
Fig. 6. External quantum efficiency of the BC-BJ solar cells without FSF diffusion and with two different FSF diffusion profiles before and after exposure to UV light. Duration of the UV exposure is shown in the graph.
1.0
Table 5 PC1D simulation of the illuminated I–V parameters of the n-type back-junction solar cell without the FSF diffusion for the S0,front parameters as determined in Table 4.
0.9 0.8 0.7 0.6 0.5 0.4
BC47-22b no FSF diffusion before UV-exposure after UV-exposure
0.3
PC1D Simulation
0.2
(S0,front=145 cm/s)
0.1
VOC (mV)
JSC (mA/cm2)
FF (%)
Z (%)
18 145
661 635
38.9 29.1
80.5 80.1
20.7 14.8
PC1D Simulation (S0,front=18 cm/s)
0.0 300 400 500 600 700 800 900 1000 1100 1200
Wavelength λ [nm] Fig. 7. External quantum efficiency of the BC-BJ solar cell without the front surface field measured before and after 60 h of exposure to UV light (symbols). PC1D simulations of the measured EQE with fitted S0,front are shown as well (thin lines).
Table 4 Front surface recombination velocity before and after exposure to UV light determined by analysis of the lifetime structures exposed to UV light and by fitting of the measured EQE using the PC1D model of the back-junction solar cell.
before UV exposure after 60 h of UV exposure
S0,front (cm/s)
S0,front determined using lifetime samples (cm/s)
S0,front determined by PC1D fitting of the measured EQE of the solar cell (cm/s)
15 139
18 145
increase of S0,front during the UV exposure is responsible for the degradation of the solar cell performance. The S0,front values determined by PC1D fitting of the measured EQEs of the solar cells without FSF, were applied to simulate the I–V parameters of the back-junction solar cells. These results are summarized in Table 5. Comparison of the results of the measured solar cells before and after UV exposure shows a good agreement in short-circuit current density. The PC1D simulation results in higher JSC values than the real cells. As previously discussed, this effect is caused by not accurate PC1D fit of the EQE data in the long wavelength range, due to two-dimensional structure of the analyzed solar cells. Initial open-circuit voltage of the measured and simulated solar cell is in a good agreement. The value of VOC of the measured solar cell after UV exposure is 10 mV lower then the simulated value, which results from the increased front surface recombination rate. The strongly increased front surface recombination rate is not only attributed to the UV exposure. This effect is believed to be caused by the additional damage of the solar cell caused by the multiple (10 times) measurement of the quantum efficiency, which requires placing of the solar cell under shadow mask. In the measurement set-up of the back-contact
F. Granek, C. Reichel / Solar Energy Materials & Solar Cells 94 (2010) 1734–1740
UV exposure
20 Efficiency [%]
solar cell available at the time of the present investigations, was not optimally designed and each mechanical positioning of the solar cell under the shading mask was causing scratches at the edges of the soalr cells. The local defects on the front cell side lead to lower VOC values. On the other hand the local scratches are not influencing the quantum effieciy measurement, because the quantum efficiency is measured only locally in the middle of the solar cell. Therefore, the mentioned edge effects are affecting the open-circuit voltage but not the quantum efficiency of the measured solar cells.
1739
15
10
5
UV exposure
PC1D model
Forming Gas Anneal
experimental data no FSF with FSF-shallow, ρsheet =350 Ω/sq with FSF-deep, ρsheet =150 Ω/sq
4.5. Regeneration of the UV-degraded lifetime samples and solar cells under forming gas anneal
External Quantum Efficiency EQE [-]
After the degradation under the UV illumination, the lifetime sample and solar cell without FSF were exposed to the anneal process in the forming gas (N2H2) atmosphere at the temperature of 425 1C and process duration time of 25 min. The elevated temperature and the hydrogen-rich atmosphere leads of the FGA process lead to strong reduction of the interface states of the Si/SiO2 interface [21]. In this way, the interface states at the front surface, which were created by the UV exposure (as discussed in Section 2), can be passivated. This leads to the reduction of the surface recombination velocity and to the regeneration of the performance of the BC-BJ solar cell. The surface saturation current density of the lifetime sample without FSF decreases from 446 fA/cm2 after UV illumination to 111 fA/cm2 after the FGA process. This corresponds to the reduction of the front surface recombination velocity from 139 to 35 cm/s. The impact of the anneal process on the BC-BJ solar cell without FSF degraded under UV exposure is shown in Fig. 8. Here the external quantum efficiency of the BC-BJ solar cell without the FSF is shown before and after UV exposure and after FGA process. The FGA process performed after the UV exposure almost completely removes the detrimental effect of the UV illumination, and regenerates the front surface passivation to the same as before the exposure. The efficiency of this solar cell increases from 14.3% directly after degradation to 19.6% after the FGA process. Therefore the efficiency loss due to the UV exposure could be almost fully recovered (see Table 1). The degradation effect of the UV exposure is not permanent and can be fully reversed by the passivation of the created interface states during the FGA process. The effect of the efficiency degradation under UV exposure and the subsequent regeneration under FGA process is additionally shown schematically in Fig. 9. As discussed before, the S0,front of all solar cells was determined 1.0 0.9 0.8 0.7
UV exposure
Forming Gas Anneal
0.6 0.5 0.4 0.3 0.2 0.1
BC47-22b no FSF diffusion before UV exposure after 60h UV exposure FGA (425 °C, 25 min.)
0.0 300 400 500 600 700 800 900 1000 1100 1200
Wavelength λ [nm] Fig. 8. External quantum efficiency of the BC-BJ solar cell without the front surface field phosphorus diffusion before and after exposure to UV light and after regeneration of the UV damage by a forming gas anneal step.
0 0 1 2 3 4 5 10 10 10 10 10 10 Front surface recombination velocity S0,front [cm/s] Fig. 9. Efficiency of the back-junction solar cells as a function of the front surface recombination velocity simulated with PC1D for a solar cell with and without FSF diffusion profiles (thin lines). Open symbols represent the efficiency of the BC-BJ solar cells measured after UV exposure steps, as marked with arrows. S0,front was determined by analyzing the lifetime test samples with and without FSF, respectively. The closed symbol represents the efficiency of the solar cell without FSF, which was regenerated in forming gas anneal process after the degradation under UV illumination.
using the lifetime test structures before and after UV exposure. The efficiency of the solar cells with and without FSF was measured after each UV exposure step and is plotted together with the S0,front determined by analyzing the lifetime samples. The lines in Fig. 9 represent PC1D simulations of the efficiency of the n-type Si back-junction solar cells as a function of the front surface recombination velocity, for the case of solar cells with and without FSF diffusion. For the solar cell without the FSF the efficiency degradation due to UV exposure follows the PC1D simulation line. After the FGA the S0,front decreases and the efficiency increases, as marked with a closed symbol. The solar cells with both FSF diffusions do not show significant performance degradation after UV exposure, as marked with open symbols (circles and squares). The slight efficiency increase of the solar cell with ‘‘FSF-shallow’’ diffusion profile after the UV exposure is caused by the increased contact quality during the measurement, which positively affected the fill factor of this cell.
5. Conclusion The front surface passivation quality is one of the most critical parameters for achieving high efficiencies with back-contact back-junction solar cells. The quality of the front surface passivation of the BC-BJ solar cells under the illumination with high energy UV photons was analyzed in this work. Additionally, the impact of the phosphorus doped front surface field layer on the passivation quality and its radiation resistance to UV illumination was investigated. Two different FSF diffusion profiles were analyzed. A deeply diffused Gaussian doping profile with low phosphorus surface concentration and a shallow errorfunction diffusion profile with high surface concentration of phosphorus atoms were applied in this study. Lifetime samples and solar cells without the front surface field showed a significant performance reduction when exposed to ultraviolet light spectrum at one sun intensity for 55 h. The surface saturation current density increased from 48 to 446 fA/cm2 after the UV exposure. At the same time, due to such a drastic increase in the front surface saturation current density, the efficiency of the BC-BJ solar cells without the FSF diffusion reduced from 19.8% to 14.3%.
1740
F. Granek, C. Reichel / Solar Energy Materials & Solar Cells 94 (2010) 1734–1740
In contrast to the lifetime samples and solar cells without the FSF diffusion, the tested n + nn + structures and the BC–BJ solar cells with both applied FSF diffusion profiles are significanly less affected by the UV exposure, i.e. the surface saturation current density increased only by factor of 25% and the efficiency of these cells decreased only 0.3%abs after the UV illumination. It was therefore found that the presence of the n + diffused layer on the front cell side drastically improves the stability of the solar cell under the UV exposure. Moreover no difference between two different applied FSF diffusion profiles on the radiation resistance of the surface passivation to the UV exposure was observed. Besides other positive aspects of the application of the FSF diffusion, it is strongly recommended to apply a FSF layer for the design of n-type BC-BJ solar cells, in order to assure their high efficiency under the long term field operation. Fitting of the measured quantum efficiency of the BC-BJ solar cells before and after UV exposure with a 1-dimensional backjunction solar cell model resulted in front surface recombination velocity values, which are in very good agreement with S0,front results obtained in the experimental study of the n + nn + structures. It can be therefore stated that the observed degradation of the performance of the solar cells without FSF under UV illumination was caused only by the increase in the front surface recombination velocity at the silicon surface S0,front. The increase of the front surface recombination rate is due to the increase in the density of the surface states of the Si/SiO2 interface, which could be caused by the photon injection of the electrons from the conduction band of the silicon to the conduction band of oxide. Finally it was shown that the reduced surface passivation quality, caused by the UV illumination of the samples without FSF, could be improved during the forming gas anneal. At the elevated temperature and in a hydrogen-rich atmosphere a strong reduction of the interface states of the Si/SiO2 interface occurs. The performance of the UV-degraded BC-BJ could in this way be almost fully recovered from 14.3% to 19.6% after the FGA process.
Acknowledgements This work was supported by the German Federal Ministry for the Environment, Nature Conservation and Nuclear Safety in the frame of the project QUEBEC (FKZ 0329988). The authors thank O. Schultz-Wittmann, K.-A. Weiss and J. Benick for valuable discussions, T. Leimenstoll, S. Seitz and A. Herbolzheimer for ¨ sample processing and E. Schaffer for solar cell measurements. References [1] D. De Ceuster, P. Cousins, D. Rose, D. Vicente, P. Tipones, W. Mulligan, Low cost, high volume production of 422% efficiency silicon solar cells, in: Proceedings of the 22nd European Photovoltaic Solar Energy Conference, Milan, Italy, 2007, pp. 816–819.
[2] R.M. Swanson, Device physics for backside-contact solar cells, in: Proceedings of the 33rd IEEE Photovoltaic Specialists Conference, San Diego, USA, 2008. [3] F. Granek, C. Reichel, M. Hermle, D.M. Huljic, O. Schultz, S.W. Glunz, Front surface passivation of n-type high-efficiency back-junction silicon solar cells using front surface field, in: Proceedings of the 22nd European Photovoltaic Solar Energy Conference Milan, Italy, 2007, pp. 1454–1457. [4] F. Granek, M. Hermle, S.W. Glunz, Analysis of the current linearity at low illumination of high-efficiency back-junction back-contact silicon solar cells, Physica Status Solidi RRL 2 (4) (2008) 151–153. [5] F. Granek, M. Hermle, D. Huljic, O. Schultz-Wittmann, S.W. Glunz, Enhanced lateral current transport via the front N + diffused layer of n-type highefficiency back-junction back-contact silicon solar cells, Progress in Photovoltaics: Research and Applications 17 (1) (2009) 47–56. [6] W.P. Lee, Y.L. Khong, W.S. Seow, Modelling the ultraviolet irradiation effect on the effective minority carrier recombination lifetime of silicon solar cells, Journal of Applied Physics 85 (2) (1999) 994–1001. [7] L. Zhong, S.F., Interface traps creation by sub-band gap irradiation in silicon dioxide on silicon without applied electric field, Journal of Applied Physics 79 (5) (1996) 2509–2512. [8] P.E. Gruenbaum, R.A. Sinton, R.M. Swanson, Light-induced degradation at the silicon/silicon dioxide interface, Applied Physics Letters 52 (17) (1988) 1407–1409. [9] P.E. Gruenbaum, R.A. Sinton, R.M. Swanson, Stability problems in point contact solar cells, in: Proceedings of the 20th IEEE Photovoltaic Specialists Conference, Las Vegas, Nevada, USA, 1998, pp. 423–428. [10] R.A. Sinton, Y. Kwark, J.Y. Gan, R.M. Swanson, 27.5-percent silicon concentrator solar cells, IEEE Electron Device Letters, EDL- 7 (10) (1986) 567–569. [11] P.E. Gruenbaum, R.R. King, R.M. Swanson, Photoinjected hot-electron damage in silicon point-contact solar cells, Journal of Applied Physics 66 (12) (1989) 6110–6114. [12] IEC, Photovoltaic devices–Part 3: measurement principles for terrestrial photovoltaic (PV) solar devices with reference spectral irradiance data, International Standard IEC60904-3, 2008. [13] P.E. Gruenbaum, J.Y. Gan, R.R. King, R.M. Swanson, Stable passivations for high-efficiency silicon solar cells, in: Proceedings of the 21st IEEE Photovoltaic Specialists Conference, Kissimmee, FL, USA, 1990, pp. 317–322. [14] D.S. Ruby, W.K. Schubert, The effects of concentrated ultraviolet light on high-efficiency silicon solar cells, in: Proceedings of the 22nd IEEE Photovoltaic Specialists Conference, Las Vegas, Nevada, USA, 1991, pp. 111–117. [15] G. Masetti, M. Severi, S. Solmi, Modeling of carrier mobility against carrier concentration in arsenic-, phosphorus-, and boron-doped silicon, IEEE Transactions on Electron Devices 30 (7) (1983) 764–769. [16] IEC, Crystalline silicon terrestrial photovoltaic (PV) modules – Design qualification and type approval, International Standard 61215 (Ed. 2), 2005. [17] R.A. Sinton, A. Cuevas, and M. Stuckings, Quasi-steady-state photoconductance, a new method for solar cell material and device characterization, in: Proceedings of the 25th IEEE Photovoltaic Specialists Conference, Washington DC, USA, 1996, pp. 457–460. [18] H. Nagel, C. Berge, A.G. Aberle, Generalized analysis of quasi-steady-state and quasi-transient measurements of carrier lifetimes in semiconductors, Journal of Applied Physics 86 (11) (1999) 6218–6221. [19] A. Cuevas, G. Giroult-Matlakowski, P.A. Basore, C. DuBois, R.R. King, Extraction of the surface recombination velocity of passivated phosphorusdoped silicon emitters, in: Proceedings of the First World Conference on Photovoltaic Energy Conversion—WCPEC, vol. 2, Waikoloa, Hawaii, USA, 1994, pp. 1446–1449. [20] D.A. Clugston, P.A. Basore, PC1D version 5: 32 bit solar cell modelling on personal computers, in: Proceedings of the 26th IEEE Photovoltaic Specialists Conference, Anaheim, California, USA, 1997 pp. 207–210. [21] H. Nagayoshi, et al., Effect of hydrogen-radical annealing for SiO2 passivation, Japanese Journal of Applied Physics 35 (8B) (1996) L1047–L1049. [22] H. Schmidhuber, Verkapselung von kristallinen Silizium-Solarzellen, Dissertation, Gesamthochschule Hagen, 2003.
Contents lists available at ScienceDirect
Solar Energy Materials & Solar Cells journal homepage: www.elsevier.com/locate/solmat
Back-contact back-junction silicon solar cells under UV illumination F. Granek n, C. Reichel Fraunhofer Institute for Solar Energy Systems, Heidenhofstr. 2, D-79110 Freiburg, Germany
a r t i c l e in f o
a b s t r a c t
Article history: Received 8 December 2009 Received in revised form 4 May 2010 Accepted 14 May 2010
The performance of n-type Si back-contact back-junction (BC-BJ) solar cells under illumination with high energy ultraviolet (UV) photons was investigated. The impact of the phosphorus doped front surface field (FSF) layer on the stability of the front surface passivation under UV illumination was investigated. Lifetime samples and solar cells without the front surface field showed a significant performance reduction when exposed to ultraviolet light. The surface saturation current density (J0e) increased from 48 to 446 fA/cm2 after the UV exposure. At the same time the efficiency of the BC-BJ solar cells without the FSF diffusion reduced from 19.8% to 14.3%. In contrast to the lifetime samples and solar cells without the FSF diffusion, the tested n + nn + structures and the BC-BJ solar cells with applied FSF diffusion profiles were significantly more stable under UV exposure, i.e. J0e increased only by a factor of 25% and the efficiency of these cells decreased only 0.3%abs by the UV illumination. Finally it was shown that the performance of the UV-degraded solar cells without FSF could be improved during a forming gas anneal (FGA). Due to application of FGA the efficiency almost fully recovered from 14.3% to 19.6%. & 2010 Elsevier B.V. All rights reserved.
Keywords: c-Si Back-contact UV-stability Front surface field
1. Introduction Back-contact back-junction (BC-BJ) silicon solar cells represent an attractive high-efficiency cell structure. In mass production, BC-BJ solar cells achieve average efficiencies of 22.4% as presented by De Ceuster et al. [1]. For comparison, standard silicon solar cells currently have efficiency in the range of 16–18%. Recently a new record efficiency of 23.4% for a large area (149 cm2) BC-BJ solar cell was announced by Swanson [2]. Due to the fact that the collecting p–n junction is placed on the rear surface of BC-BJ solar cells, and that most of the photogeneration takes place close to the front side, the requirements on the front surface passivation quality are very high. Thus, a low front surface recombination rate is one of the critical factors influencing the efficiency of the back-junction solar cells. The front surface passivation scheme needs not only to be of very high quality. It is also essential that the applied front surface passivation scheme is stable under solar cell operating conditions. Especially the stability of the surface passivation under the high energy ultraviolet (UV) part of the solar spectrum is of main importance in order to maintain high device performance during the long-term field operation of the photovoltaic modules. In our previous work [3] the front surface passivation scheme using a phosphorus doped front surface field (FSF) and a stack system of a thermal oxide and a PECVD silicon nitride for the
n
Corresponding author. Tel.: + 49 761 4588 5355; fax: + 49 761 4588 9250. E-mail address: fi[email protected] (F. Granek).
0927-0248/$ - see front matter & 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.solmat.2010.05.038
BC-BJ solar cells was developed and studied. Also other positive effects of the FSF on the performance of the BC-BJ solar cells were investigated, such as improvement of the low-illumination performance [4] and improvement of the lateral transport of the majority carriers [5]. In the present work the influence of the exposure of the front surfaces of the BC-BJ solar cells to UV light is investigated. Additionally the influence of the FSF diffusion on the UV stability of lifetime samples and on the performance of the BC-BJ solar cells is presented.
2. Influence of the UV light on the front surface passivation of BC-BJ solar cells The impact of the UV illumination of the oxide passivated silicon surfaces was investigated by many authors in the microelectronic field [6,7]. In the field of the silicon solar cells the work of Gruenbaum was pioneering. Already in 1988 Gruenbaum et al. [8,9] reported that the efficiency of some of the point-contact concentrator solar cells, developed at the Stanford University by Sinton et al. [10], decreased after exposure to concentrated sunlight. The decrease in solar cell performance was caused by an increase in the front surface recombination velocity. The studies of Gruenbaum et al. showed that the ultraviolet component of the incident light spectrum caused damage on the front surface.
F. Granek, C. Reichel / Solar Energy Materials & Solar Cells 94 (2010) 1734–1740
Gruenbaum et al. [11] performed the UV exposure and photoinjection experiments. It was discovered that the UV light of energy greater than 3.1 eV causes an increase in the surface recombination velocity (S0,front) and increases interface state densities of the surfaces passivated with oxide. The energy of 3.1 eV corresponds to light with wavelength shorter than 400 nm. In the terrestrial solar spectrum there is a significant amount of photons with such energy [12]. The absorption of the UV photons with wavelength shorter than 400 nm could inject electrons from the conduction band in silicon into the conduction band of silicon oxide. This photoinjection would then create defects at the Si/SiO2 interface. However, Gruenbaum et al. [13] and Ruby and Schubert [14] showed that not all solar cell structures are prone to degradation under UV light. The formation of the diffused phosphorus region on the front side creates a high field region. This field reduces the concentration of the minority charge carriers close to the recombination centres at the front silicon surface. Even if S0,front increases substantially, the effective surface recombination velocity (Seff) may increase minimally, leading to just a small reduction in efficiency of the solar cell.
3. Experimental Symmetrical lifetime test structures, as well as high-efficiency BC-BJ solar cells were processed and analyzed in order to investigate the impact of the high energy UV photons on front surface saturation current density (J0e) and on the solar cell performance.
1735
temperature drive-in oxidation process at the temperature of 1050 1C. The high oxidation temperature caused redistribution of the phosphorus atoms and transition from the error-function dopant profile (profile: FSF-shallow) to Gaussian profile (profile: FSF-deep). The silicon oxide layer was then etched back in HF solution. Sheet resistance (shown in Fig. 2) of the investigated front surface field phosphorus profiles was calculated by integrating the dopant profiles using the mobility model of Masetti et al. [15]. Next, all samples were passivated with a thin (10 nm) thermal SiO2 layer and an antireflection-SiNx coating (thickness of 60 nm). Finally, the samples were annealed in a forming gas atmosphere at a temperature of 425 1C for 15 min. 3.2. Back-contact back-junction Si solar cells The schematic cross-section of the BC–BJ solar cell analyzed in this work is shown in Fig. 3. The cells were also fabricated from n-type FZ–Si wafers with the resistivity of 1 O cm and a thickness of around 160 mm. The front cell side has the same structure as both surfaces of the lifetime structures. Thus, the front side of the BC–BJ solar cells is textured with random pyramids. For selected samples, the front surface was additionally passivated with phosphorus n + front surface field diffusion. The same phosphorus diffusion profiles (FSF-deep and FSF-shallow) as discussed in Section 3.1 were applied to the solar cells. The front surface was then passivated
1020 ρ
sheet
= 150 Ω/sq
FSF-shallow ρ
sheet
= 350 Ω/sq
Symmetrical n + nn + test structures for the minority carrier lifetime measurements were processed on n-type FZ–Si wafers with a thickness of 250 mm and specific resistivity of 1 O cm. Due to very high bulk lifetime of the minority carriers in the applied ntype FZ–Si wafers, the effective minority carrier lifetime is almost entirely limited by the surface recombination. In this way an accurate determination of the surface recombination is possible. The lifetime structures shown in Fig. 1 were processed on textured wafers (random pyramids texture). For selected samples, directly after the texturization process, both surfaces were additionally passivated by the application of the phosphorus doped front surface field. FSF doping profiles were formed by diffusion from a liquid POCl3 source in a tube furnace. Two different FSF doping profiles investigated in this study are shown in Fig. 2. Both profiles were formed at the same diffusion temperature. After diffusion the phosphorus glass was fully etched back in a buffered HF solution. The deep diffused (‘‘FSF-deep’’ in Fig. 2) profile was formed by an additional high
Phosphorus concentration ND [cm-3]
FSF-deep 3.1. Lifetime structures for determination of the surface saturation current density
1019
1018
1017
0.0
0.2
0.6
0.8
1.0
1.2
1.4
Depth [μm] Fig. 2. Secondary ion mass spectroscopy (SIMS) profiles of the studied phosphorus dopant profiles measured after all high temperature processing steps. Sheet resistance of both diffusion profiles is shown in the graph.
AR SiNX
AR SiNX
0.4
SiO 2
SiO2 n+ FSF (optional) n+ BSF
n++ FSF FSF(optional) (optional) n
n-Si
p+ emitter
n-Si n-Si metal fingers pitch
Fig. 1. Textured symmetrical structures used for the lifetime measurements and for the determination of the surface recombination current density.
SiO 2
Fig. 3. Schematic cross-section of the n-type high-efficiency back-contact backjunction silicon solar cell investigated in this work. Note that sketch is not to scale.
1736
F. Granek, C. Reichel / Solar Energy Materials & Solar Cells 94 (2010) 1734–1740
with a thin (10 nm) thermally grown silicon dioxide layer. PECVD SiNx antireflection coating layer (60 nm) was deposited on top of the oxide layer. The rear side was patterned using industrially feasible technology (e.g. screen-printing and laser ablation). These technologies lead to a pitch of the finished solar cells of 1.8 mm. Both emitter p++ and back surface field n++ line diffusions are separated by an undiffused gap, the rear structure forms an interdigitated grid. The rear cell surface is passivated with thermally grown silicon oxide. Metal fingers are contacted to the diffused regions through local openings in the thermal oxide layer. The size of the active cell area is 2 2 cm2. Solar cells were cut from the host wafer with a laser, where the distance from the edge of the active area to the edge was 500 mm. More details on the processing technology and the performance of the described BC-BJ solar cells are presented in Ref. [5]. The solar cells with efficiencies in the range of 19.7–20.2% and with two different diffusion profiles of the front n + diffusion and without the front n + diffusion were selected for the analysis of the performance stability under UV exposure. 3.3. Exposure of the lifetime samples and the solar cells to the UV illumination The stability of the described surface passivation schemes of the lifetime samples and the BC-BJ solar cells with and without the FSF diffusion under UV light was examined by the exposure to a xenon arc lamp with the total irradiation of 1000 kWh/m2. The spectrum of the Xe arc lamp applied in the UV exposure test was measured in the plane of the tested samples. This spectrum together with a standard AM1.5 spectrum is shown in Fig. 4. In the UV-spectral range a good agreement between both spectrums can be observed. The temperature of the lifetime samples and solar cells during the exposure equals 50 1C. The international standard IEC 61215 [16] defines the UV-test of the crystalline silicon photovoltaic modules as a UV exposure in the spectral range between 280 and 385 nm with the irradiation of 15 kW h/m2. The module temperature should equal 60 75 1C. The UV irradiation in this spectral range equals roughly 5% of the global solar irradiation. For a geographical location with the global solar irradiation of 1000 kW h/m2 (e.g. central Europe), the yearly UV irradiation equals 50 kWh/m2. In central Europe the UV exposure defined in the IEC standard (15 kW h/m2) corresponds to roughly 3.5 months of the out-door irradiation.
Results of the surface saturation current density of the test lifetime samples under the UV exposure tests are shown in Fig. 5. J0e of lifetime samples without FSF increases significantly already after the first few hours of UV exposure. J0e increases from initial value of 48–446 fA/cm2 after 55 h of UV light exposure and the value of J0e seems to saturate after that exposure time. A drastic increase of J0e to 446 fA/cm2, which for base resistivity of 1 O cm corresponds to S0,front of 139 cm/s, is expected to cause a significant performance degradation of the back-contact backjunction solar cells.
60 40
1000
20
500 400
600 800 Wavelength [nm]
1000
0 1200
Fig. 4. Spectrum of the xenon arc lamp, which was applied for the UV exposure of the lifetime samples and solar cells. The standard AM1.5 spectrum is shown in comparison [12]. The irradiation of both spectrums equals 1000 kWh/m2. Both spectrums are in a good agreement in the UV-spectral range (up to 400 nm). Transmission measurements of the solar module glass and of EVA encapsulant [22] are shown as well.
FSF-deep FSF-shallow no FSF
1000
J0e [fA/cm2]
Standard AM1.5 Spectrum Xe arc lamp
1500
0 200
4.1. Lifetime test structures under UV exposure
Surface Saturation Current Density
80
Transmission [%]
2000
Module glass EVA
Spectral irradiance [W/μm m2]
8000
2500
4. Results
100
10000
6000
The lifetime structures and solar cells were not covered with a module glass or any other intermediate layer during the exposure. Therefore the illumination spectrum absorbed by the test structures was rich in the high energy UV photons. The analyzed samples were exposed to UV illumination for 60 h. The resulting UV dosage equals 2.2 kW h/m2. For a location in central Europe such a UV dosage corresponds to roughly 15 days of outdoor exposure. However since in the normal operation mode, the solar cells are covered with solar glass and EVA encapsulant the UV irradiation, which reaches solar cell in a module, is strongly reduced. Given the transmission of the solar glass and of the EVA encapsulant, as shown in Fig. 4, only 21% of the UV dose defined by the IEC standard reaches the solar cell in a module. When taking the transmission losses of the UV irradiation in the solar module, the corresponding UV dose would equal 9.5 kW h/m2, which corresponds to 70 days of the outdoor exposure and 63% of the UV dosage defined by the IEC standard. The solar cells were illuminated from the front side, and the lifetime samples were illuminated from one side only. The UV illumination was performed in time steps, between which the characterization of these samples was performed. In case of the lifetime structures the minority carrier lifetime was measured using the quasi-steady-state photo-conductance (QSSPC) measurement set-up [17] and the generalized analysis method was applied [18] after each exposure step. Then the front surface recombination velocity S0,front and the surface saturation current density J0e were determined as described in Refs. [19] and [3], respectively. For the BC-BJ solar cells, the light current–voltage (I–V) characteristics, as well as external quantum efficiency (EQE) were measured after each UV exposure step.
ρsheet = 150 Ω/sq ρsheet = 350 Ω/sq
100
10 0.1
1 10 exposure time t [h]
100
Fig. 5. Surface saturation current density of textured test lifetime samples for two different front surface field diffusion profiles and for samples without FSF during 55 h of exposure to UV light.
F. Granek, C. Reichel / Solar Energy Materials & Solar Cells 94 (2010) 1734–1740
On the other hand, the test structures with either FSF diffusion profiles exhibit a significant increase of J0e during the exposure tests. J0e of the test structure with the ‘‘FSF-shallow’’ diffusion profile with sheet resistance of 350 O/sq increased from 29 initially to 36 fA/cm2 after the 55 h of exposure. The structures with the ‘‘FSF-deep’’ diffusion profile with sheet resistance of 150 O/sq increase from 34 to 45 fA/cm2 after the exposure. This proves that passivation using a FSF is significantly more stable under UV illumination and therefore appropriate for industrial applications, in contrast to unstable passivation without FSF. Similar results were also obtained by Gruenbaum et al. [9]. It is therefore expected that the BC-BJ solar cells with the applied FSF diffusion will be stabile under UV illumination. 4.2. Solar cell results Solar cells results of the analyzed BC-BJ solar cells under UV illumination are summarized in Tables 1–3. The initial solar cell performance (exposure time of 0 h) is given in the first row of the tables. In the last row of these tables (‘‘After FGA’’), the solar cell results after annealing in a forming gas atmosphere (FGA) at a temperature of 425 1C for 25 min are presented. The annealing process was performed after completing the UV exposure tests of all solar cells. The performance change of the solar cells after the final FGA process step will be discussed in Section 4.5. The BC-BJ solar cell without front surface field diffusion (Table 1) shows a major efficiency reduction after the UV illumination. Efficiency was reduced from 19.8 to 14.3%, mainly due to Table 1 I–V parameters of the BC-BJ solar cell without the FSF diffusion (cell no. BC47-22b) under UV illumination. Results in table are designated cell area measurements (2 2 cm2). Exposure time (h)
VOC (mV)
JSC (mA/cm2)
FF (%)
Z (%)
0 1 3 48 60
661 652 633 629 625
37.3 37.3 31.1 28.9 28.7
80.1 79.3 79.4 79.5 79.7
19.8 19.3 15.6 14.4 14.3
After FGA
659
36.9
80.4
19.6
Table 2 I–V parameters of the BC-BJ solar cell with the ‘‘FSF-deep’’ diffusion profile (cell no. BC47-17b) under UV illumination. Results in table are designated cell area measurements (2 2 cm2). Exposure time (h)
VOC (mV)
JSC (mA/cm2)
FF (%)
Z (%)
0 48 60
652 651 651
38.8 38.1 38.4
80.1 79.9 79.5
20.2 19.8 19.9
After FGA
661
38.3
80.4
20.4
Table 3 I–V parameters of the BC-BJ solar cell with the ‘‘FSF-shallow’’ diffusion profile (cell no. BC47-10b) under UV illumination. Results in table are designated cell area measurements (2 2 cm2). Exposure time (h)
VOC (mV)
JSC (mA/cm2)
FF (%)
Z (%)
0 48 60
653 651 651
39.2 38.1 38.4
76.9 79.9 79.5
19.7 19.8 19.9
After FGA
661
39.0
80.5
20.8
1737
loss in short-circuit current of 8.6 mA/cm2. This loss is manifested as a strong reduction of the EQE in the whole solar spectrum, as will be shown in the next section. Also the open-circuit voltage decreased significantly after the UV illumination. This can be explained by the strong increase in the total saturation current density of the solar cell, due to the increase of the front surface saturation current. In the case of solar cells with both FSF diffusion profiles (see Tables 2 and 3), the performance degradation after the UV illumination is only minimal, proving that the application of the FSF diffusion indeed strongly increases the radiation resistance of the BC-BJ solar cells to the high energy UV photons. Note the initial lower fill factor of the solar cell BC47-10b is believed to be caused by the contacting problems during the solar cell measurement. It is therefore not believed that the UV illumination affected the FF in a positive way, as might be interpreted from the results in Table 3. 4.3. External quantum efficiency External quantum efficiency of the analyzed solar cells before and after UV illumination is presented in Fig. 6. EQE of the solar cell without FSF degrades rapidly from 92% to 75% at wavelength of 500 nm already after 3 h of exposure to UV light. After 60 h of UV exposure the EQE of the solar cell without FSF decreases even further to 65%. Thus, an EQE decrease of nearly 30% absolute was caused by 60 h of exposure to UV light, and is responsible for the reduction of the short-circuit current presented in Table 1. EQE of the solar cells with both FSF diffusion profiles was reduced only by 2% absolute after the UV illumination. For the solar cell with ‘‘FSF-deep’’ after 60 h of exposure, the EQE dropped from 92% to 90% at a wavelength of 500 nm. For the solar cell with ‘‘FSF-shallow’’ diffusion profile the EQE dropped from 90% to 88%. 4.4. Device simulations The influence of the UV exposure on the EQE of the solar cell without FSF was compared with the results of the lifetime test structures and modelled using PC1D [20] simulations. First, the S0,front of the textured lifetime samples was measured before and after UV exposure as described in Section 4.1. Secondly, the measured EQEs were fitted using the PC1D model of a backjunction solar cell. The measured reflectance spectrum of the actual solar cell was used in the simulations. In order to fit the EQE results in the short wavelength range, the S0,front was varied. The experimentally determined and modelled EQEs are presented in Fig. 7. The PC1D fit in the long wavelength range is not in a very good agreement with measured data due to strong influence of the two-dimensional effects, which could not be described by the one-dimensional PC1D simulation. Please mind that the analyzed solar cells have the pitch of 1.8 mm; these devices are therefore strongly two-dimensional structures. This effect will have an impact on the value of JSC determined by the PC1D simulations. However, the effect of the increased front surface recombination rate affects the quantum efficiency of the solar cells mainly in the short to medium-length wavelength of the solar spectrum. In the short and medium-length wavelength spectrum the measured and modelled EQE are in a good agreement. The results of PC1D fitting and S0,front determined from the lifetime measurement of the n + nn + test structures are summarized in Table 4. A very good agreement between the S0,front determined by analysis of the lifetime samples and by fitting of the EQE of the solar cells was obtained. This proves that the
1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1
BC47-22b no FSF diffusion before UV exposure after 3 h exposure after 48 h exposure after 60 h exposure
0.0 300 400 500 600 700 800 900 1000 11001200
External Quantum Efficiency EQE [-]
F. Granek, C. Reichel / Solar Energy Materials & Solar Cells 94 (2010) 1734–1740
External Quantum Efficiency EQE [-]
1738
1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1
BC47-17b with FSF-deep diffusion before UV exposure after 60 h exposure
0.0 300 400 500 600 700 800 900 1000 11001200
Wavelength λ [nm]
External Quantum Efficiency EQE [-]
Wavelength λ [nm] 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1
BC47-10b with FSF-shallow diffusion before UV exposure after 60 h exposure
0.0 300 400 500 600 700 800 900 1000 11001200
Wavelength λ [nm]
External Quantum Efficiency EQE [-]
Fig. 6. External quantum efficiency of the BC-BJ solar cells without FSF diffusion and with two different FSF diffusion profiles before and after exposure to UV light. Duration of the UV exposure is shown in the graph.
1.0
Table 5 PC1D simulation of the illuminated I–V parameters of the n-type back-junction solar cell without the FSF diffusion for the S0,front parameters as determined in Table 4.
0.9 0.8 0.7 0.6 0.5 0.4
BC47-22b no FSF diffusion before UV-exposure after UV-exposure
0.3
PC1D Simulation
0.2
(S0,front=145 cm/s)
0.1
VOC (mV)
JSC (mA/cm2)
FF (%)
Z (%)
18 145
661 635
38.9 29.1
80.5 80.1
20.7 14.8
PC1D Simulation (S0,front=18 cm/s)
0.0 300 400 500 600 700 800 900 1000 1100 1200
Wavelength λ [nm] Fig. 7. External quantum efficiency of the BC-BJ solar cell without the front surface field measured before and after 60 h of exposure to UV light (symbols). PC1D simulations of the measured EQE with fitted S0,front are shown as well (thin lines).
Table 4 Front surface recombination velocity before and after exposure to UV light determined by analysis of the lifetime structures exposed to UV light and by fitting of the measured EQE using the PC1D model of the back-junction solar cell.
before UV exposure after 60 h of UV exposure
S0,front (cm/s)
S0,front determined using lifetime samples (cm/s)
S0,front determined by PC1D fitting of the measured EQE of the solar cell (cm/s)
15 139
18 145
increase of S0,front during the UV exposure is responsible for the degradation of the solar cell performance. The S0,front values determined by PC1D fitting of the measured EQEs of the solar cells without FSF, were applied to simulate the I–V parameters of the back-junction solar cells. These results are summarized in Table 5. Comparison of the results of the measured solar cells before and after UV exposure shows a good agreement in short-circuit current density. The PC1D simulation results in higher JSC values than the real cells. As previously discussed, this effect is caused by not accurate PC1D fit of the EQE data in the long wavelength range, due to two-dimensional structure of the analyzed solar cells. Initial open-circuit voltage of the measured and simulated solar cell is in a good agreement. The value of VOC of the measured solar cell after UV exposure is 10 mV lower then the simulated value, which results from the increased front surface recombination rate. The strongly increased front surface recombination rate is not only attributed to the UV exposure. This effect is believed to be caused by the additional damage of the solar cell caused by the multiple (10 times) measurement of the quantum efficiency, which requires placing of the solar cell under shadow mask. In the measurement set-up of the back-contact
F. Granek, C. Reichel / Solar Energy Materials & Solar Cells 94 (2010) 1734–1740
UV exposure
20 Efficiency [%]
solar cell available at the time of the present investigations, was not optimally designed and each mechanical positioning of the solar cell under the shading mask was causing scratches at the edges of the soalr cells. The local defects on the front cell side lead to lower VOC values. On the other hand the local scratches are not influencing the quantum effieciy measurement, because the quantum efficiency is measured only locally in the middle of the solar cell. Therefore, the mentioned edge effects are affecting the open-circuit voltage but not the quantum efficiency of the measured solar cells.
1739
15
10
5
UV exposure
PC1D model
Forming Gas Anneal
experimental data no FSF with FSF-shallow, ρsheet =350 Ω/sq with FSF-deep, ρsheet =150 Ω/sq
4.5. Regeneration of the UV-degraded lifetime samples and solar cells under forming gas anneal
External Quantum Efficiency EQE [-]
After the degradation under the UV illumination, the lifetime sample and solar cell without FSF were exposed to the anneal process in the forming gas (N2H2) atmosphere at the temperature of 425 1C and process duration time of 25 min. The elevated temperature and the hydrogen-rich atmosphere leads of the FGA process lead to strong reduction of the interface states of the Si/SiO2 interface [21]. In this way, the interface states at the front surface, which were created by the UV exposure (as discussed in Section 2), can be passivated. This leads to the reduction of the surface recombination velocity and to the regeneration of the performance of the BC-BJ solar cell. The surface saturation current density of the lifetime sample without FSF decreases from 446 fA/cm2 after UV illumination to 111 fA/cm2 after the FGA process. This corresponds to the reduction of the front surface recombination velocity from 139 to 35 cm/s. The impact of the anneal process on the BC-BJ solar cell without FSF degraded under UV exposure is shown in Fig. 8. Here the external quantum efficiency of the BC-BJ solar cell without the FSF is shown before and after UV exposure and after FGA process. The FGA process performed after the UV exposure almost completely removes the detrimental effect of the UV illumination, and regenerates the front surface passivation to the same as before the exposure. The efficiency of this solar cell increases from 14.3% directly after degradation to 19.6% after the FGA process. Therefore the efficiency loss due to the UV exposure could be almost fully recovered (see Table 1). The degradation effect of the UV exposure is not permanent and can be fully reversed by the passivation of the created interface states during the FGA process. The effect of the efficiency degradation under UV exposure and the subsequent regeneration under FGA process is additionally shown schematically in Fig. 9. As discussed before, the S0,front of all solar cells was determined 1.0 0.9 0.8 0.7
UV exposure
Forming Gas Anneal
0.6 0.5 0.4 0.3 0.2 0.1
BC47-22b no FSF diffusion before UV exposure after 60h UV exposure FGA (425 °C, 25 min.)
0.0 300 400 500 600 700 800 900 1000 1100 1200
Wavelength λ [nm] Fig. 8. External quantum efficiency of the BC-BJ solar cell without the front surface field phosphorus diffusion before and after exposure to UV light and after regeneration of the UV damage by a forming gas anneal step.
0 0 1 2 3 4 5 10 10 10 10 10 10 Front surface recombination velocity S0,front [cm/s] Fig. 9. Efficiency of the back-junction solar cells as a function of the front surface recombination velocity simulated with PC1D for a solar cell with and without FSF diffusion profiles (thin lines). Open symbols represent the efficiency of the BC-BJ solar cells measured after UV exposure steps, as marked with arrows. S0,front was determined by analyzing the lifetime test samples with and without FSF, respectively. The closed symbol represents the efficiency of the solar cell without FSF, which was regenerated in forming gas anneal process after the degradation under UV illumination.
using the lifetime test structures before and after UV exposure. The efficiency of the solar cells with and without FSF was measured after each UV exposure step and is plotted together with the S0,front determined by analyzing the lifetime samples. The lines in Fig. 9 represent PC1D simulations of the efficiency of the n-type Si back-junction solar cells as a function of the front surface recombination velocity, for the case of solar cells with and without FSF diffusion. For the solar cell without the FSF the efficiency degradation due to UV exposure follows the PC1D simulation line. After the FGA the S0,front decreases and the efficiency increases, as marked with a closed symbol. The solar cells with both FSF diffusions do not show significant performance degradation after UV exposure, as marked with open symbols (circles and squares). The slight efficiency increase of the solar cell with ‘‘FSF-shallow’’ diffusion profile after the UV exposure is caused by the increased contact quality during the measurement, which positively affected the fill factor of this cell.
5. Conclusion The front surface passivation quality is one of the most critical parameters for achieving high efficiencies with back-contact back-junction solar cells. The quality of the front surface passivation of the BC-BJ solar cells under the illumination with high energy UV photons was analyzed in this work. Additionally, the impact of the phosphorus doped front surface field layer on the passivation quality and its radiation resistance to UV illumination was investigated. Two different FSF diffusion profiles were analyzed. A deeply diffused Gaussian doping profile with low phosphorus surface concentration and a shallow errorfunction diffusion profile with high surface concentration of phosphorus atoms were applied in this study. Lifetime samples and solar cells without the front surface field showed a significant performance reduction when exposed to ultraviolet light spectrum at one sun intensity for 55 h. The surface saturation current density increased from 48 to 446 fA/cm2 after the UV exposure. At the same time, due to such a drastic increase in the front surface saturation current density, the efficiency of the BC-BJ solar cells without the FSF diffusion reduced from 19.8% to 14.3%.
1740
F. Granek, C. Reichel / Solar Energy Materials & Solar Cells 94 (2010) 1734–1740
In contrast to the lifetime samples and solar cells without the FSF diffusion, the tested n + nn + structures and the BC–BJ solar cells with both applied FSF diffusion profiles are significanly less affected by the UV exposure, i.e. the surface saturation current density increased only by factor of 25% and the efficiency of these cells decreased only 0.3%abs after the UV illumination. It was therefore found that the presence of the n + diffused layer on the front cell side drastically improves the stability of the solar cell under the UV exposure. Moreover no difference between two different applied FSF diffusion profiles on the radiation resistance of the surface passivation to the UV exposure was observed. Besides other positive aspects of the application of the FSF diffusion, it is strongly recommended to apply a FSF layer for the design of n-type BC-BJ solar cells, in order to assure their high efficiency under the long term field operation. Fitting of the measured quantum efficiency of the BC-BJ solar cells before and after UV exposure with a 1-dimensional backjunction solar cell model resulted in front surface recombination velocity values, which are in very good agreement with S0,front results obtained in the experimental study of the n + nn + structures. It can be therefore stated that the observed degradation of the performance of the solar cells without FSF under UV illumination was caused only by the increase in the front surface recombination velocity at the silicon surface S0,front. The increase of the front surface recombination rate is due to the increase in the density of the surface states of the Si/SiO2 interface, which could be caused by the photon injection of the electrons from the conduction band of the silicon to the conduction band of oxide. Finally it was shown that the reduced surface passivation quality, caused by the UV illumination of the samples without FSF, could be improved during the forming gas anneal. At the elevated temperature and in a hydrogen-rich atmosphere a strong reduction of the interface states of the Si/SiO2 interface occurs. The performance of the UV-degraded BC-BJ could in this way be almost fully recovered from 14.3% to 19.6% after the FGA process.
Acknowledgements This work was supported by the German Federal Ministry for the Environment, Nature Conservation and Nuclear Safety in the frame of the project QUEBEC (FKZ 0329988). The authors thank O. Schultz-Wittmann, K.-A. Weiss and J. Benick for valuable discussions, T. Leimenstoll, S. Seitz and A. Herbolzheimer for ¨ sample processing and E. Schaffer for solar cell measurements. References [1] D. De Ceuster, P. Cousins, D. Rose, D. Vicente, P. Tipones, W. Mulligan, Low cost, high volume production of 422% efficiency silicon solar cells, in: Proceedings of the 22nd European Photovoltaic Solar Energy Conference, Milan, Italy, 2007, pp. 816–819.
[2] R.M. Swanson, Device physics for backside-contact solar cells, in: Proceedings of the 33rd IEEE Photovoltaic Specialists Conference, San Diego, USA, 2008. [3] F. Granek, C. Reichel, M. Hermle, D.M. Huljic, O. Schultz, S.W. Glunz, Front surface passivation of n-type high-efficiency back-junction silicon solar cells using front surface field, in: Proceedings of the 22nd European Photovoltaic Solar Energy Conference Milan, Italy, 2007, pp. 1454–1457. [4] F. Granek, M. Hermle, S.W. Glunz, Analysis of the current linearity at low illumination of high-efficiency back-junction back-contact silicon solar cells, Physica Status Solidi RRL 2 (4) (2008) 151–153. [5] F. Granek, M. Hermle, D. Huljic, O. Schultz-Wittmann, S.W. Glunz, Enhanced lateral current transport via the front N + diffused layer of n-type highefficiency back-junction back-contact silicon solar cells, Progress in Photovoltaics: Research and Applications 17 (1) (2009) 47–56. [6] W.P. Lee, Y.L. Khong, W.S. Seow, Modelling the ultraviolet irradiation effect on the effective minority carrier recombination lifetime of silicon solar cells, Journal of Applied Physics 85 (2) (1999) 994–1001. [7] L. Zhong, S.F., Interface traps creation by sub-band gap irradiation in silicon dioxide on silicon without applied electric field, Journal of Applied Physics 79 (5) (1996) 2509–2512. [8] P.E. Gruenbaum, R.A. Sinton, R.M. Swanson, Light-induced degradation at the silicon/silicon dioxide interface, Applied Physics Letters 52 (17) (1988) 1407–1409. [9] P.E. Gruenbaum, R.A. Sinton, R.M. Swanson, Stability problems in point contact solar cells, in: Proceedings of the 20th IEEE Photovoltaic Specialists Conference, Las Vegas, Nevada, USA, 1998, pp. 423–428. [10] R.A. Sinton, Y. Kwark, J.Y. Gan, R.M. Swanson, 27.5-percent silicon concentrator solar cells, IEEE Electron Device Letters, EDL- 7 (10) (1986) 567–569. [11] P.E. Gruenbaum, R.R. King, R.M. Swanson, Photoinjected hot-electron damage in silicon point-contact solar cells, Journal of Applied Physics 66 (12) (1989) 6110–6114. [12] IEC, Photovoltaic devices–Part 3: measurement principles for terrestrial photovoltaic (PV) solar devices with reference spectral irradiance data, International Standard IEC60904-3, 2008. [13] P.E. Gruenbaum, J.Y. Gan, R.R. King, R.M. Swanson, Stable passivations for high-efficiency silicon solar cells, in: Proceedings of the 21st IEEE Photovoltaic Specialists Conference, Kissimmee, FL, USA, 1990, pp. 317–322. [14] D.S. Ruby, W.K. Schubert, The effects of concentrated ultraviolet light on high-efficiency silicon solar cells, in: Proceedings of the 22nd IEEE Photovoltaic Specialists Conference, Las Vegas, Nevada, USA, 1991, pp. 111–117. [15] G. Masetti, M. Severi, S. Solmi, Modeling of carrier mobility against carrier concentration in arsenic-, phosphorus-, and boron-doped silicon, IEEE Transactions on Electron Devices 30 (7) (1983) 764–769. [16] IEC, Crystalline silicon terrestrial photovoltaic (PV) modules – Design qualification and type approval, International Standard 61215 (Ed. 2), 2005. [17] R.A. Sinton, A. Cuevas, and M. Stuckings, Quasi-steady-state photoconductance, a new method for solar cell material and device characterization, in: Proceedings of the 25th IEEE Photovoltaic Specialists Conference, Washington DC, USA, 1996, pp. 457–460. [18] H. Nagel, C. Berge, A.G. Aberle, Generalized analysis of quasi-steady-state and quasi-transient measurements of carrier lifetimes in semiconductors, Journal of Applied Physics 86 (11) (1999) 6218–6221. [19] A. Cuevas, G. Giroult-Matlakowski, P.A. Basore, C. DuBois, R.R. King, Extraction of the surface recombination velocity of passivated phosphorusdoped silicon emitters, in: Proceedings of the First World Conference on Photovoltaic Energy Conversion—WCPEC, vol. 2, Waikoloa, Hawaii, USA, 1994, pp. 1446–1449. [20] D.A. Clugston, P.A. Basore, PC1D version 5: 32 bit solar cell modelling on personal computers, in: Proceedings of the 26th IEEE Photovoltaic Specialists Conference, Anaheim, California, USA, 1997 pp. 207–210. [21] H. Nagayoshi, et al., Effect of hydrogen-radical annealing for SiO2 passivation, Japanese Journal of Applied Physics 35 (8B) (1996) L1047–L1049. [22] H. Schmidhuber, Verkapselung von kristallinen Silizium-Solarzellen, Dissertation, Gesamthochschule Hagen, 2003.