Transient Structural Relaxation and Melting Temperature of Amorphous Silicon
- PDF / 533,619 Bytes
- 6 Pages / 420.48 x 639 pts Page_size
- 87 Downloads / 229 Views
TRANSIENT STRUCTURAL RELAXATION AND MELTING TEMPERATURE OF AMORPHOUS SILICON
W.C. SINKE*), T. WARABISAKO, M. MIYAO and T. TOKUYAMA**), Central Research Laboratory, Hitachi Ltd., P.O. Box 2, Kokubunji, Tokyo 185, Japan.
S. ROORDA and F.W. SARIS, FOM-Institute for Atomic- and Molecular Physics, Kruislaan 407, 1098 SJ Amsterdam, The Netherlands.
ABSTRACT The structure of ion-implanted a-Si has been studied using Raman spectroscopy. It is shown that the average bond-angle distortion and the strain energy stored in the random network vary drastically upon annealing. Relaxation is enhanced in the case of heating by laser irradiation. The results imply that the apparent melting temperature of a-Si may vary from one experiment to another, depending on the heating rate and -method.
INTRODUCTION The behaviour of amorphous silicon (a-Si) upon heating, i.e. crystallization and melting, has been studied intensively over the past decade. Considerable progress has been made in understanding important processes like solid-phase epitaxial regrowth and pulsed-laser induced melting and resolidification of a-Si on single-crystal silicon (c-Si). Nevertheless some important questions remain to be solved. Among these, the question concerning the melting temperature (Tm) of a-Si is one of the most intriguing. On the basis of thermodynamical calculations, Bagley and Chen [1] and Spaepen and Turnbull [2] predicted in 1978 that the Tm of a-Si should be considerably lower (> 300 K) than that of c-Si (being 1412'C). Their conclusion was based on a difference in free energy between the crystalline and the amorphous state. At that time, however, no experimental results were available to support their theory. The first indication of a melting point lowering was found by Baeri et al. [3] from experiments using pulsed electron beam heating of a-Si. In contrast to this, Kokorowski et al. [4] did not find any evidence for a significant difference in T. in their experiments with cw laser heating of a-Si. Subsequently, important new data was provided by Thompson et al. [5], who studied nanosecond pulsed-laser melting of a-Si and concluded that a-Si melts at a temperature =200 K below that for melting of c-Si. Their interpretation has been widely accepted since, but seems difficult to be reconciled with results obtained by Olson et al. [6,7] who performed detailed and elaborate experiments using cw and microsecond pulsed-laser heating and were able to heat a-Si to temperatures well in excess of 1300"C without melting (superheating effects could influence the apparent T , but the temperature would be expected to increase with heating rate, in contrast to the findings). Wn extra complication comes from the observation by Olson et al. [7] that the Tm of a-Si under jps laser irradiation seems to vary with film thickness. In the discussion described above it is assumed that the free energy of a-Si does not vary much with the preparation method or thermal treatments. Following the reasoning in references [I ] and [2], this implies that the Tm of a-Si has
Data Loading...
