Light-induced degradation of crystalline silicon solar cells is a frequently observed phenomenon. Two main causes of such degradation effects have been identified in the past, both of them being electronically driven and both related to the most common acceptor element, boron, in silicon: (i) the dissociation of iron-boron pairs and (ii) the formation of recombination-active boron-oxygen complexes. While the first mechanism is particularly relevant in metal-contaminated solar-grade multicrystalline silicon materials, the latter process is important in monocrystalline Czochralski-grown silicon, rich in oxygen. This paper starts with a short review of the characteristic features of the two processes. We then briefly address the effect of iron-boron dissociation on solar cell parameters. Regarding the boron-oxygen-related degradation, the current status of the physical understanding of the defect formation process and the defect structure are presented. Finally, we discuss different strategies for effectively avoiding the degradation.

The substitutional boron-vacancy B_sV complex in silicon is investigated using the local density functional theory. These theoretical results give an explanation of the experimentally reported, well established metastability of the boron-related defect observed in p-type silicon irradiated at low temperature and of the two hole transitions that are observed to be associated with one of the configurations of the metastable defect.   B_sV is found to have several stable configurations, depending on charge state.  In the positive charge state the second nearest neighbor configuration with C₁ symmetry is almost degenerate with the second nearest neighbor configuration that has C_1h symmetry since the bond reconstruction is weakened by the removal of electrons from the center. A third nearest neighbor configuration of B_sV has the lowest energy in the negative charge state.  An assignment of the three energy levels associated with B_sV is made.  The experimentally observed E_v + 0.31 eV and E_v + 0.37 eV levels are related to the donor levels of second nearest neighbor B_sV with C₁ and C_1h symmetry respectively. The observed E_v + 0.11 eV level is assigned to the vertical donor level of the third nearest neighbor configuration. The boron-divacancy complex B_sV₂ is also studied and is found to be stable with a binding energy between V₂ and B_s of around 0.2 eV. Its energy levels lie close to those of the V₂.  However, the defect is likely to be an important defect only in heavily doped material.

Carrier lifetime degradation in crystalline silicon solar cells under illumination with white light is a frequently observed phenomenon. Two main causes of such degradation effects have been identified in the past, both of them being electronically driven and both are related to the most common acceptor element, boron, in silicon: (i) the dissociation of interstitial ironsubstitutional boron (Fe_iB_s) pairs and (ii) the formation of recombination-active boron-oxygen complexes. In solar-grade multicrystalline silicon (mc-Si), the first mechanism is most relevant. This well-known process, which is linked to the degree of iron contamination in the material, can also be observed in single- crystalline iron-contaminated B-doped float-zone (FZ) and Czochralski (Cz) silicon and is not restricted to mc-Si. The second carrier lifetime degradation effect can be observed in metal-impurity-free B-doped Cz-Si rich in oxygen. This effect is attributed to the simultaneous presence of B_s and interstitial oxygen (O_i). Interestingly, as for the Fe_iB_s dissociation, this degradation effect also occurs in the dark when minority-carriers are injected (e.g., by a forward-biased pn junction), leading to the conclusion that the degradation is caused by the presence of minority-carriers and that photons are not directly involved. However, in contrast to the Fe_iB_s-related lifetime degradation, which also occurs during annealing above ~100°C, the latter degradation effect is fully reversible by annealing above ~200°C, i.e., the degraded lifetime recovers during low temperature annealing, making it relatively easy to distinguish between the two effects. Recently, much research has been devoted to the boron-oxygen-related degradation problem, which is presently the main obstacle for making single-crystalline Cz-Si an ideal cost-saving material for high-efficiency solar cells. This contribution reviews the present physical understanding of both degradation effects and discusses different approaches for reducing or even completely avoiding them. Special attention is paid to a recently proposed defect reaction model [J. Schmidt and K. Bothe, Phys. Rev. B 69, 024107 (2004)] of the boron-oxygen degradation, in which a fast-diffusing oxygen dimer (O_2i) is trapped by B_s to form a B_sO_2i complex, acting as a highly effective recombination center. Results of theoretical calculations using density functional theory show that B_sO_2i is a bistable defect with a donor level in the upper half of the silicon band gap, in good agreement with the results of temperature- and injection- dependent lifetime measurements. Calculated activation energies for the dissociation and association of the B_sO_2i complex are also in excellent agreement with the barrier energies determined experimentally on lifetime samples and solar cells.

Light-induced degradation of crystalline silicon solar cells is a frequently observed phenomenon. Two main causes of such degradation effects have been identified in the past, both of them being electronically driven and both are related to the most common acceptor element, boron, in silicon: (i) the dissociation of iron-boron pairs and (ii) the formation of recombination-active boron-oxygen complexes. While the first mechanism is particularly relevant in metal-contaminated solar-grade multicrystalline silicon materials, the latter process is important in monocrystalline Czochralski-grown silicon rich in oxygen. This paper starts with a short review of the characteristic features of both processes. We then briefly address the effect of iron-boron dissociation on solar cell parameters. Regarding the boron-oxygen-related degradation, the current status of the physical understanding of the defect formation process and the defect structure are presented. Finally, we discuss different strategies for effectively avoiding the degradation

The formation mechanism and properties of the boron-oxygen center responsible for the degradation of Czochralski-grown Si(B) solar cells during operation is investigated using density functional calculations.We find that boron traps an oxygen dimer to form a bistable defect with a donor level in the upper half of the band gap. The activation energy for its dissociation is found to be 1.2 eV. The formation of the defect from mobile oxygen dimers, which are shown to migrate by a Bourgoin mechanism under minority carrier injection, has a calculated activation energy of 0.3 eV. These energies and the dependence of the generation rate of the recombination center on boron concentration are in good agreement with observations.
See also Erratum (printing error in a figure caption) in Phys Rev Lett 93 (2004) 169904

The nature and the anomalous emission characteristics reported for an electron trap PE5 induced by proton implantation in n-GaAs have been investigated. Using conventional majority-carrier Deep Level Transient Spectroscopy (DLTS), the signature of the defect level was determined after 1.0 MeV proton and 2.0 MeV helium ion irradiations respectively. No evidence of the reported anomalous emission property was observed. In contrast, we found that the defect exhibited the usual emission characteristics with apparent electronic energy level at (0.19±0.02) eV and a capture cross-section of (1.93±0.3)xl0^-13 cm². Our analyses further suggested that the defect is not necessarily hydrogen-impurity related as previously proposed and that it seems likely to arise from initial atomic displacement.

By the use of Deep Level Transient Spectroscopy (DLTS) and capacitance-voltage profiling we have studied the production rate of the well known electron trap E3 in 1.0 MeV proton- irradiated n-GaAs Schottky diodes fabricated from VPE and MBE grown layers. In VPE-grown layers irradiated at room temperature to cumulative doses of 5.55 x 10¹¹ cm^-2, the trap E3 production rate was found to be 302±30cm^-1, while in the MBE-grown layer irradiated to a total dose of 8.0 x 10¹² cm^-2, the defect production rate was evaluated to be 285±15 cm^-1. A comparison of these results with those obtained elsewhere in proton-irradiated LPE grown layers showed no significant sample dependence of E3 production rate; consistent with published work in electron-irradiated n-GaAs. The implication of these results concerning the nature of the E3 defect is also discussed.