• PDF / 623,552 Bytes
• 10 Pages / 420.48 x 639 pts Page_size
• 64 Downloads / 219 Views

DOWNLOAD

REPORT

---

ON THE KINETICS OF OXYGEN CLUSTERING AND THERMAL DONOR FORMATION IN CZOCHRALSKI SILICON

T. Y. TAN IBM Thomas J. Watson Research Center, Yorktown Heights, NY

10598

R. KLEINHENZ and C. P. SCHNEIDER IBM General Technology Division, Hopewell Junction, NY

12533

ABSTRACT 0

We report the results of an experiment of annealing Czochralski silicon at 450 C for up to 500 Concentrations of oxygen atoms (Ci) and thermal donors (TD) have both been monitored. hrs. Analyses of the oxygen concentration data yielded the apparent interpretation that the overall kinetics is dominated by the formation of small clusters (dimers and trimers). This cannot account for TI) formation, since they are supposed to be larger clusters. On the other hand, analyses of existing TD kinetic models did not yield calculated Ci values to satisfactorily account for the present Ci data. We believe that a satisfactory TD model is not yet available.

I. INTRODUCTION 1

In 1954 Fuller et al. discovered the generation of thermal donors (TD) at -450"C in Czochralski (CZ) silicon containing a large amount of interstitial oxygen atoms (01). In 1958 Kaiser, Frisch 2 and Reiss (KFR) worked out a model to explain the TD formation kinetics by assigning the cluster containing four oxygen atoms (presumably covalently bonded to the silicon lattice) as the electrically active TD centers. This is arrived at by noticing that the TD initial formation rate is proportional to 3 the fourth power of the oxygen content (Cio) in the materials. In 1977 Helmireich and Sirtl noticed that the KFR model requires a rather high oxygen diffusivity value and that in materials with high 3 carbon contents the fourth power dependency of TD initial formation rate on Cio is not satisfied. ,4 5 They proposed that TD are OV centers (V: silicon lattice vacancy). In 1982 Gosele and Tan found that the KFR model is internally inconsistent and proposed a molecular 04 (TD) model to rectify the situation. The 04 is resulting from initial formation of molecular 02 from Oi. Since that time many 3 14 other models on the nature and on the kinetics of TD formation have been proposed.6-1 Bourret has recently summarized the salient features of most of these models. 2 3 The existence of at least eleven models, ,3,5-1 with nine of them proposed within the most 5 13 recent three to four years of time, " seems to indicate that the problem of TD formation is so complicated that it is not yet satisfactorily resolved. Of relevance to this paper, we pay special 2 7 attentions to the KFR model, the Oehrlein model, and the Ourmazd-Schroter-Bourret (OSB) 12 model, since these are the only models for which the TD formation kinetics have been calculated. In principle, the model of Gosele and Tans could also have yielded some numerical results, but this has not been done at the time. By now most people believe that there is no evidence for the existence of a molecule-like species of oxygen in silicon. The Oehrdein model maintains that the dominant TID cluster is 03. Among other features, the OSB model maintains that