An industrial solution to light-induced degradation of crystalline silicon solar cells

Abstract

Boron-oxygen defects can cause serious light-induced degradation (LID) of commercial solar cells based on the boron-doped crystalline silicon (c-Si), which are formed under the injection of excess carriers induced either by illumination or applying forward bias. In this contribution, we have demonstrated that the passivation process of boron-oxygen defects can be induced by applying forward bias for a large quantity of solar cells, which is much more economic than light illumination. We have used this strategy to trigger the passivation process of batches of aluminum back surface field (Al-BSF) solar cells and passivated emitter and rear contact (PERC) solar cells. Both kinds of the treated solar cells show high stability in efficiency and suffer from very little LID under further illumination at room temperature. This technology is of significance for the suppression of LID of c-Si solar cells for the industrial manufacture.

Keywords

Boron-oxygen defects / c-Si solar cells / light-induced degradation / passivation / forward bias

Introduction

Commercial solar cells based on boron-doped crystalline silicon (c-Si) usually suffer from severe light-induced degradation (LID) with an efficiency loss up to 2%–3%, which has been proved to be caused by the formation of boron-oxygen defects [ 1– 3]. Composed of one substitutional boron atom (B_s) and one interstitial oxygen dimer (O_2i), boron-oxygen defects are formed under the injection of excess carriers induced either by illumination or applied forward bias, which act as the strong carrier recombination centers that can reduce the minority carriers lifetime and thus the efficiency of solar cells [ 2, 4– 6].When annealed at 200°C in dark for 20 min, the boron-oxygen defects can reverse from recombination active form (degradation state) to inactive form (annealing state). They can turn to recombination active form again when illuminated at room temperature subsequently [ 7– 8].

Methods toward suppressing LID have been widely studied, such as low oxygen content silicon (FZ-Si) [ 9], boron free doped silicon [ 7, 10– 11], and germanium or carbon co-doped silicon [ 12– 13]. Recent researches have shown that boron-oxygen defects could be permanent deactivated under excess carriers injection induced either by illumination or by applying forward bias at slightly elevated temperature (70°C–200°C), which is called passivation process. In this process the boron-oxygen defects can be reversed from recombination active form (degradation state) to the stable inactive form (passivation state), holding permanently the full lifetime of minority carriers and hence the full efficiency of the solar cells [ 8, 14– 15]. The passivation treatment has attracted extensive interest, but almost always the illumination is used to trigger the passivation process because of the simple setup of the experiments, which consists only of a lamp and a heating stage [ 16– 19]. However, this method is limited only in laboratory use and can hardly be applied to industrial manufacture. Because applying this passivation treatment to manufacture process of the solar cells means that each cell should be illuminated at elevated temperature individually, which need extra enormous space, a huge light source and a huge heating stage. All of these issues imply extra tremendous cost, which makes it impossible to be applied to industrial solar cell production [ 20].

In this manuscript, the degradation process and the passivation process triggered by both illumination and applying forward bias are demonstrated. A promising way to solve the LID problem for industrial manufacture is investigated. The passivation treatment triggered by applying forward bias has been used for a stack of 300 solar cells simultaneously and the stability of treated samples is also studied.

Experiment details

The degradation and passivation experiments induced by illumination were carried out on three 4 cm $$\begin{gathered} \\ \times \end{gathered}$$ 4 cm p-type<1 0 0>oriented c-Si samples (1–3 W·cm boron-doped, 200 mm thick), which was cut from the same 15.6 cm \times 15.6 cm commercial solar cell-using wafer. The interstitial oxygen concentration was determined as [O_i] = 7 \times 10¹⁷ cm⁻³ by Fourier transform infrared spectroscopy (FTIR) at room temperature. After an acidic damage etching and a standard RCA cleaning for 15 min, the wafers were annealed in Ar ambient for 30 min at 650°C to eliminate thermal donors. The three wafers were marked as S₁, S₂, S₃ for the subsequent experiments after a double-side passivation with 80 nm-thick PECVD deposited SiN_x:H films. S₁ and S₂ were illuminated with the light intensity of 70 MW·cm⁻² at 50°C for 1600 min. The S₁ was subsequently illuminated with the light intensity of 70 MW·cm⁻² at an elevated temperature of 150°C for 100 min, while the S₂ was annealed at 200°C in dark for 20 min. S₃ was used as a reference. Then, the three samples were illuminated with the light intensity of 70 MW•cm⁻² at 50°C. Lifetimes measurements were carried out by quasi-steady-state photo conductance (QSSPC) technique at room temperature with a fixed injection level of 5.5 \times 10¹⁴ cm⁻³.

Anti-LID equipment made in Changzhou Shichuang Energy Technology corporation was used for the degradation and passivation experiments triggered by applying forward bias, which can treat as many as 2400 solar cells simultaneously. Figure 1 shows the schematic diagram of the equipment, and the heating temperature as well as the forward current intensity can be adjusted. The stacks of numerous solar cells were connected in series and each solar cell was in contact with the adjacent ones by the opposite polarity. The samples to be investigated were standard aluminum back surface field (Al-BSF) solar cells and passivated emitter and rear contact (PERC) solar cells (15.6 cm \times 15.6 cm, 200 mm-thick c-Si, 1–3 W·cm boron-doped, 50 W/sq), which were made by the standard celling process. First, one Al-BSF solar cell was chosen to study the degradation and passivation processes triggered by applying forward bias with a small forward current injection level of 1 A at 50°C and 150°C, respectively. Then, a stack of 300 Al-BSF solar cells were treated simultaneously for the passivation process under a forward current injection level of 30 A at 220°C for 60 min, which is called “special current injecting passivation treatment (SCIPT)” here. Then the SCIPT was applied to a stack of 300 PERC solar cells. Samples from the top, middle and bottom parts of the two kinds of treated solar cells stacks were illuminated under one sun at room temperature for a month. The efficiency and open-circuit voltage of the solar cells were characterized by the I-V and suns-V_oc measurements under an injection level of 1 sun at room temperature.

Results and discussion

Figure 2 depicts the illumination time dependence of lifetime t and the normalized boron-oxygen defects concentration N_t* of sample S₁ under an illumination intensity of 70 MW·cm⁻² at 50°C. N_t^* is given by the measured lifetime t using the equation below,

(1) {{\displaystyle N}}_t^{*}=\mathrm{1}/\tau (t)-\mathrm{1}/\tau (\mathrm{0}),

where t(t) is the effective lifetime measured at the illumination time t, and t(0) is the initial effective lifetime before illumination. It can be seen that the effective lifetime decreases sharply in the first few minutes, which is the so-called fast degradation. Afterwards, the sample undergoes a slow degradation with a small lifetime decreasing rate and finally reaches a saturation value after 1000 min [ 21– 22]. Correspondingly, the normalized boron-oxygen defects concentration N_t* derived from Eq. (1) increases with the illumination time and saturates 1000 min later.

Figure 3 demonstrates the value variation of lifetime t and N_t* of sample S₁ as a function of illumination time. After the full degradation shown in Fig. 2, the sample was illuminated with the injection intensity of 70 MW·cm⁻² at an elevated temperature of 150°C. It is shown that the effective lifetime recovers quickly with the illumination time, which implies that the boron-oxygen defects are deactivated quickly at the elevated temperature. After 40 min, the effective lifetime saturates at a value of 4.83 ms. This value is a little smaller than the initial lifetime value of 5.14 ms. It may be caused by the metal contamination like copper and nickel, which can form the permanent active recombination centers with the injection of excess carriers [ 23– 25].

The stability of effective lifetime for various samples after the passivation process is demonstrated in Fig. 4. Note that the lifetime stability is tested by the effective lifetime evolution with the illumination time at 50°C. It can be seen that the effective lifetime of S₁ almost keeps constant, which indicates that the boron-oxygen defects are permanently deactivated. However, both S₂ and S₃ suffer from light-induced degradation again due to the formation of the boron-oxygen defects. The results prove that the passivation treatment is an efficient way to suppress the LID caused by the boron-oxygen defects.

The degradation and passivation processes triggered by the applying forward bias are shown in Fig. 5. It demonstrates the time dependence of the open-circuit voltage of the Cz-Si Al-BSF solar cell under small forward current injection intensity of 1 A at different temperatures of 50°C and 150°C. One can see that the solar cell suffers from a current-induced degradation at 50°C, and the open-circuit voltage decreases with time. However, the passivation can be triggered under the same current intensity of 1 A at elevated temperature of 150°C and the open-circuit voltage recovers to the initial value. It strongly proved that passivation can be induced not only by illumination but also by applying forward bias [ 20– 26]. The application of forward bias is another choice to be applied for the passivation treatment of boron-oxygen defects in c-Si.

Figure 6 depicts the parameter variation of the solar cells under the illumination intensity of 1 sun at room temperature for 24 h. The samples are from the top, middle and bottom parts of the stack of 300 Al-BSF solar cells, which have undergone the SCIPT simultaneously. An untreated sample is chosen as a reference. It can be seen that the reference sample suffers from the light-induced degradation with an efficiency loss of 1.2%. The corresponding loss of open-circuit voltage (V_oc), short circuit current (J_sc) and fill factor (FF) are 0.34%, 0.42% and 0.41%, respectively. However, the treated samples from the top, middle and bottom parts of the stack show much better stability, which suffer from much less LID after illumination for 24 h, with an efficiency loss of 0.10% for the top treated sample, 0.15% for the middle treated sample and 0.2% for the bottom treated sample, respectively. These results imply that all solar cells in the treated stack have been well passivated and suffer from little LID. It means that we can simultaneously treat a large quantity of solar cells for once to passivate the boron-oxygen defects. Thus, we can reasonably say the SCIPT is an effective way to solve the LID problem of Al-BSF solar cells for numerous solar cells in the industrial scale.

The efficiency stability of the treated Al-BSF and PERC solar cells is demonstrated in Fig. 7, which depicts the illumination time dependence of the efficiency loss of the treated samples under the illumination intensity of 1 sun at room temperature for 30 days. Note that the treated Al-BSF and PERC samples are from the middle part of the stacks, both of which have undergone the SCIPT. In addition, the untreated Al-BSF and PERC samples are chosen as references as well. One can see that the efficiency loss is less than 0.22% for the treated Al-BSF solar cell after illumination for 30 days, while it is more than 1.5% for the untreated Al-BSF solar cell. For the PERC samples, these two values are 1.01% and 5.01%, respectively. It is noteworthy that the PERC solar cells are reported to suffer from much severer LID than Al-BSF solar cells, which may be caused by the unstable aluminum oxide layer on the back or the gettering defects during aluminum back surface field formation [ 27]. Both of the treated Al-BSF and PERC solar cells suffer from much smaller efficiency losses than the untreated ones during illumination, which prove that the SCIPT is an effective way to suppress the LID of both Al-BSF and the PERC solar cells.

Conclusions

In summary, we have demonstrated that both of the degradation process and the passivation process of boron-oxygen defects can be triggered either by illumination or by applying forward bias. By combining heat treatment with forward bias, an SCIPT method have been applied to stacks of c-Si Al-BSF and PERC solar cells to remove the recombination-active boron-oxygen defects. The illumination stability experiments show that not only the treated Al-BSF solar cells but also the treated PERC solar cells have good illumination stability and suffer from much smaller LID than the untreated ones. These results means the SCIPT is an effective way to solve the LID problem for a large quantity of solar cells for industrial application.

References

[1] Bothe K, Hezel R, Schmidt J. Recombination-enhanced formation of the meta stable boron–oxygen complex in crystalline silicon. Applied Physics Letters, 2003, 83(6): 1125–1127

[2] Fischer H, Pschunder W. Investigation of photon and thermal induced changes in silicon solar cells. In: Proceedings of the 10th IEEE Photovoltaic Specialists Conference. Alto P, Calif, 1974, 404–411

[3] Lim B, Hermann S, Bothe K, Schmidt J, Brendel R. Solar cells on low-resistivity boron-doped Czochralski-grown silicon with stabilized efficiencies of 20%. Applied Physics Letters, 2008, 93(16): 162102

[4] Bothe K, Sinton R, Schmidt J. Fundamental boron-oxygen-related carrier lifetime limit in mono and multicrystalline silicon. Progress in Photovoltaics: Research and Applications, 2005, 4(13): 287–296

[5] Schmidt J, Bothe K. Structure and transformation of the meta stable boron-and oxygen-related defect center in crystalline silicon. Physical Review B: Condensed Matter and Materials Physics, 2004, 69(2): 024107

[6] Voronkov V V, Falster R. Latent complexes of interstitial boron and oxygen dimers as a reason for degradation of silicon-based solar cells. Journal of Applied Physics, 2010, 107(5): 053509

[7] Schmidt J, Aberle A G, Hezel R. Investigation of carrier lifetime instabilities in Cz-grown silicon. In:Proceedings of the 26th IEEE Photovoltaic Specialists Conference . Anaheim, 1997

[8] Herguth A, Schubert G, Kaes M, Hahn G. Investigations on the long time behavior of the metastable boron-oxygen complex in crystalline silicon. Progress in Photovoltaics: Research and Applications, 2008, 16(2): 135–140

[9] Glunz S W, Rein S, Warta W, Knobloch J, Wettling W. Degradation of carrier lifetime in Cz silicon solar cells. Solar Energy Materials and Solar Cells, 2001, 65(1–4): 219–229

[10] Yoshida T, Kitagawara Y. Bulk lifetime decreasing phenomena induced by light-illumination in high-purity p-type CZ-Si crystals. In: Proceedings of the 4th International Symposium on High Purity Silicon IV, 1996, 450–454

[11] Glunz S W, Rein S, Knobloch J, Wettling W, Abe T. Comparison of boron and gallium doped p-type Czochralski silicon for photovoltaic application. Progress in Photovoltaics: Research and Applications, 1999, 7(6): 463–469

[12] Yu X, Wang P, Chen P, Li X, Yang D. Suppression of boron-oxygen defects in p-type Czochralski silicon by germanium doping. Applied Physics Letters, 2010, 97(5): 051903

[13] Wu Y, Yu X, He H, Chen P, Yang D. Suppression of boron-oxygen defects in Czochralski silicon by carbon co-doping. Applied Physics Letters, 2015, 106(10): 102105

[14] Herguth A, Schubert G, Kaes M, Hahn G. Avoiding boron-oxygen related degradation in highly boron doped Cz silicon. In: Proceedings of the 21st European Photovoltaic Solar Energy Conference, 2006, 530–537

[15] Lim B, Bothe K, Schmidt J. Deactivation of the boron-oxygen recombination center in silicon by illumination at elevated temperature. Physica Status Solidi (RRL)-Rapid Research Letters, 2008, 2(3): 93–95

[16] Lim B, Bothe K, Schmidt J. Impact of oxygen on the permanent deactivation of boron-oxygen-related recombination centers in crystalline silicon. Journal of Applied Physics, 2010, 107(12): 123707

[17] Lim B, Liu A, Macdonald D, Bothe K, Schmidt J. Impact of dopant compensation on the deactivation of boron-oxygen recombination centers in crystalline silicon. Applied Physics Letters, 2009, 95(95): 232109

[18] Wilking S, Herguth A, Hahn G. Influence of hydrogen on the regeneration of boron-oxygen related defects in crystalline silicon. Journal of Applied Physics, 2013, 113(19): 194503

[19] Lim B, Bothe K, Schmidt J. Accelerated deactivation of the boron-oxygen-related recombination centre in crystalline silicon. Semiconductor Science and Technology, 2011, 26(9): 95009–95011(3)

[20] Herguth A, Hahn G. Towards a high throughput solution for boron-oxygen related regeneration. In: 28th European Photovoltaic Solar Energy Conference and Exhibition, 2013, 1507–1511

[21] Bothe K, Schmidt J. Fast-forming boron-oxygen-related recombination center in crystalline silicon. Applied Physics Letters, 2005, 87(26): 262108

[22] Hashigami H, Dhamrin M, Saitoh T. Characterization of the initial rapid decay on light-induced carrier lifetime and cell performance degradation of Czochralski-grown silicon. Japanese Journal of Applied Physics, 2003, 42 (Part 1, No. 5A): 2564–2568

[23] Inglese A, Lindroos J, Savin H. Accelerated light-induced degradation for detecting copper contamination in p-type silicon. Applied Physics Letters, 2015, 107(1): 41–46

[24] Lindroos J, Yli-Koski M, Haarahiltunen A, Savin H. Room-temperature method for minimizing light-induced degradation in crystalline silicon. Applied Physics Letters, 2012, 101(23): 232108

[25] Lindroos J, Savin H. Formation kinetics of copper-related light-induced degradation in crystalline silicon. Journal of Applied Physics, 2014, 116(23): 234901

[26] Bothe K, Schmidt J. Electronically activated boron-oxygen-related recombination centers in crystalline silicon. Journal of Applied Physics, 2006, 99(1): 013701

[27] Fertig F, Krauß K, Rein S. Light-induced degradation of PECVD aluminium oxide passivated silicon solar cells. Physica Status Solidi (RRL)-Rapid Research Letters, 2015, 9(1): 41–46

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag Berlin Heidelberg