Towards the Modeling of Impurity-Related Defects in Irradiated n -Type Germanium: a Challenge to Theory

  • PDF / 642,538 Bytes
  • 7 Pages / 612 x 792 pts (letter) Page_size
  • 99 Downloads / 193 Views

DOWNLOAD

REPORT


LECTRONIC PROPERTIES OF SEMICONDUCTORS

Towards the Modeling of Impurity-Related Defects in Irradiated n-Type Germanium: a Challenge to Theory V. V. Emtseva,* and G. A. Oganesyana a

Ioffe Institute, St. Petersburg, 194021 Russia *e-mail: [email protected]

Received June 28, 2020; revised July 13, 2020; accepted July 13, 2020

Abstract—Electrical measurements on heavily doped n-type germanium subjected to gamma-irradiation show that the features of impurity-related defect formation before n–p conversion of conductivity type are the same as those previously observed in lightly and moderately doped materials, thus extending the range of doping from ≈1014 to ≈1016 cm–3. It is clear now that the presently adopted model of the dominant impurityrelated defects as simple vacancy-impurity pairs in irradiated n-Ge, in analogy to such defects reliably identified in irradiated n-Si, appears to be inconsistent with the experimental information collected so far. As a consequence, the impurity diffusion simulations in heavily doped Ge based on this model need to be reconsidered. The requirements to be met while modeling impurity-related defects in irradiated n-Ge in accordance with the reliable experimental data are established. Keywords: germanium, irradiation, impurity-related defects DOI: 10.1134/S106378262011007X

1. INTRODUCTION The longstanding problem of how the well-known models of various complexes of native point defects with technologically important dopant impurities in silicon may also be extended to those in germanium remains a critical unanswered question. As a result, the fundamental challenges of understanding impurity diffusion processes as well as changes of electrical parameters of germanium under irradiation conditions also remain unanswered. In fact, even in the simple case of doping of Si with the group-V and groupIII shallow impurities their ionization energies can feature strong variations from the effective mass theory if the sizes of impurity and host atoms mismatch considerably like in n-Si:Bi, p-Si:Ga, and p-Si:In. This is in sharp contrast to n- and p-Ge where such “size” effects for all the shallow impurities are found to be only marginal; see for instance [1]. Looking at impurity diffusion processes mediated by native point defects in Si and Ge a marked difference in the behavior appears to be much more pronounced as well; see for instance [2]. Some concerted attempts to refine the existing diffusion models in Ge can be found in the literature [3–5]. In contrast to native point defects and impurityrelated complexes in Ge, the structural and electrical properties of many point defects in Si have been established by various experimental techniques successfully, first of all by means of EPR that allows one a searching look into the atomic structures of the

defects. Unfortunately, this is not the case for Ge. Nonetheless, many attempts at defect modeling in Ge have often been made assuming close similarities of the important features of point defects in Si and Ge. It can be misleading, however, as i