The Role of Silicon Interstitials in the Formation of Boron-Oxygen Defects in Crystalline Silicon Daniel Macdonald 1,a , P.N.K. Deenapanray 2,b , A. Cuevas 1,c , S. Diez 3,d and S.W. Glunz 3,e 1 Department of Engineering, The Australian National University, Canberra 0200 ACT, Australia 2 Centre for Sustainable Energy Systems, The Australian National University, Canberra 0200 ACT, Australia 3 Fraunhofer Institute for Solar Energy Systems (ISE), Heidenhofstrasse 2, D-79110 Freiburg, Germany a daniel.macdonald@anu.edu.au, b prakash.deenapanray@anu.edu.au, c andres.cuevas@anu.edu.au, d stdiez@ise.fhg.de, e glunz@ise.fhg.de Keywords: crystalline silicon, Czochralski, boron-oxygen defect, silicon interstitial Abstract. Oxygen-rich crystalline silicon materials doped with boron are plagued by the presence of a well-known carrier-induced defect, usually triggered by illumination. Despite its importance in photovoltaic materials, the chemical make-up of the defect remains unclear. In this paper we examine whether the presence of excess silicon self-interstitials, introduced by ion-implantation, affects the formation of the defects under illumination. The results reveal that there is no discernible change in the carrier-induced defect concentration, although there is evidence for other defects caused by interactions between interstitials and oxygen. The insensitivity of the carrier-induced defect formation to the presence of silicon interstitials suggests that neither interstitials themselves, nor species heavily affected by their presence (such as interstitial boron), are likely to be involved in the defect structure, consistent with recent theoretical modelling. Introduction Solar cells made with any boron-doped, oxygen-rich crystalline silicon material, such as Czochralski silicon (Cz-Si), suffer from a well-known carrier-induced defect which acts as a recombination centre [1-3]. The defect is known to occur only in the presence of both boron and oxygen, and is activated by the presence of excess carriers, arising either through illumination or carrier injection. A characteristic feature of the defect is that it may then be de-activated by annealing above 200°C. While the defect has been well characterised in terms of its electronic properties [4,5], its exact chemical make-up remains unclear. An early model was based on interstitial boron - interstitial oxygen pairs (B i O i ) [6], while a more recent model suggests the defect may be composed of substitutional boron complexed with an interstitial oxygen dimer B s O 2i [7]. This latter model can account for the approximately quadratic dependence of the relative defect concentration on the interstitial oxygen content [7,8]. While models involving interstitial boron have been promoted on the basis of molecular dynamics calculations [9], more recent density functional calculations have given strong support for the oxygen dimer model [10]. For any model involving interstitial boron, defect generation is likely to depend very strongly on the presence of intrinsic point defects, i.e. silicon self-interstitials and vacancies. An excess of silicon interstitials will tend to increase the interstitial boron concentration via a ‘kick-out’ mechanism, and hence increase the defect concentration, while an excess of vacancies will suppress it. It is also possible that silicon interstitials themselves directly form part of the defect. Recently, Rein et al. [8] showed that the presence of excess vacancies does not result in any significant change in the relative defect concentration N t * , at least for vacancy concentrations in the 10 12 cm -3 range. These results suggest that vacancies are not likely to be involved in the defect, and