Ti-H and Ni-H interactions in Si: first principles theory
- PDF / 58,379 Bytes
- 6 Pages / 612 x 792 pts (letter) Page_size
- 107 Downloads / 259 Views
1268-EE05-01
Ti-H and Ni-H interactions in Si: first-principles theory D.J. Backlund and S.K. Estreicher Physics Department, Texas Tech University, Lubbock, TX 79409-1051, USA ABSTRACT Hydrogen is commonly used to remove (or at least reduce) the electrical activity of numerous defects and impurities in Si. Although hydrogenation works quite well for many defects, it has generally been unsuccessful with transition metal (TM) impurities. A number of {TM,Hn} complexes have been detected using optical or electrical techniques. Even though the gap levels of the isolated TM shift upon hydrogenation, many {TM,Hn} complexes remain electrically active. The nature of the complexes responsible for specific DLTS lines is generally not known, and the number of H interstitials in a given complex is assumed. We have performed systematic first-principles calculations involving Ti-H and Ni-H interactions in Si, assuming both interstitial and substitutional sites for the TM. The equilibrium configurations, binding energies, and approximate gap levels of all the {Ti,Hn} and {Ni,Hn} complexes are calculated. INTRODUCTION Transition metal (TM) impurities from the 3d series such as Ti or Ni are undesirable contaminants in semiconductor devices and photovoltaic (PV) Si materials because they introduce deep levels in the band gap or form electrically-active precipitates [1]. TM impurities are sometimes present in the source material or are inadvertently introduced during processing. Removing such impurities by gettering is impractical as it involves long anneals at high temperatures. The hydrogenation of Si reduces the electrical activity of many defects, but does not always passivate TM impurities [2]. Several authors have investigated the interactions of H with TMs [3-8] and found that while some levels in the gap are passivated, new levels simultaneously appear. Singh et al. [9] studied the effects of hydrogenation on Ti impurities by low-energy ion bombardment and concluded that the deep levels of Ti are not passivated by atomic hydrogen. Ab initio minimal basis set Hartree-Fock calculations in small H-saturated clusters [10] predict that H binds to substitutional or interstitial Ti, but the impact on the gap levels was not calculated. Deep-level transient spectroscopy (DLTS) experiments confirmed the formation of {Ti,H} complexes in Si. Jost and Weber [11] introduced H into Ti-contaminated n-type and ptype samples by a combination of wet chemical etching and exposure to a remote hydrogen plasma, and then annealed the samples at 470K. The DLTS spectrum of the etched n-type samples shows four peaks. Two of them are associated with the single acceptor at Ec−0.09eV and single donor at Ec−0.27eV (= Ev+0.89eV) levels of isolated interstitial Ti (Tii). The other two, at Ec−0.31eV and Ec−0.57eV, are associated with near-surface defects and anneal out after three hours at 570K. A depth profile provides evidence that the concentration of the Ec−0.57eV level is greatest near the surface and decreases towards the bulk. The depth profile of the E
Data Loading...