• PDF / 1,239,315 Bytes
• 8 Pages / 414 x 635.4 pts Page_size
• 12 Downloads / 220 Views

DOWNLOAD

REPORT

---

PULSED-LASER ANNEALING

231

OF ION-IMPLANTED GaAs:

THEORY AND EXERIMENT

R. F. WOOD, Douglas H. Lowndes, and W. H. Christie* Solid State Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37830, USA

P.

0. Box X,

ABSTRACT It is shown that, in the pulsed-laser irradiation of crystalline or lightly damaged GaAs, good agreement is obtained between measured and calculated thresholds for melting, for catastrophic damage due to vaporization, and for the duration of surface melting at various energy densities. Good agreement between theory and experiment is also obtained for dopant profile spreading during pulsed-laser annealing.

INTRODUCTION In the past few years, numerous experimental and theoretical studies of pulsed-laser and pulsed-electron beam annealing and recrystallization of semiconductors have been reported [1,2]. The vast majority of this work has been concerned with Si, although a number of studies of compound semiconductors, particularly GaAs, have also appeared. In this paper, we present the results of recent experimental and theoretical studies of pulsed-laser annealing of GaAs. The purpose of these studies was to evaluate the agreement which could be obtained between experimental results on dopant profile redistributions and the duration of surface melting, as measured by timeresolved reflectivity (TRR), and the predictions of the melting model of pulsed-laser annealing. Despite the outstanding success the melting model has had in explaining virtually all of the experimental data on pulsed-laser annealing, controversy concerning the detailed physical mechanisms involved in pulsed annealing

persists. The melting model [3-61 assumes that most of the energy of the absorbed laser radiation is transferred from the electronic system to the lattice in a time less than or comparable to the pulse duration time. Subsequent processes are assumed to involve primarily conventional thermal melting, with penetration of the melt front to the underlying perfect crystal substrate from which recrystallization by liquid phase epitaxy proceeds. An alternative model supposes that pulsed-laser annealing is a primarily nonthermal effect, resulting from formation of a dense electron-hole plasma which persists in a well-localized region for times of the order of hundreds of nanoseconds [7]. One of a number of fundamental theoretical barriers to acceptance of a longlived plasma model has been that of understanding how normally very rapid 1 1 12 carrier-lattice relaxation rates (- 10 -10 /sec at low carrier densities) [81 might be slowed as a result of formation of a high-density plasma. Thus, the importance of time-resolved measurements of physical properties to determine the time scale for transfer of energy to the lattice during the pulsed annealing process is clear. Auston et al. [91 were the first to use the TRR technique to quantitatively interpret the duration of surface melting during pulsed-laser *Analytical Chemistry Division,

ORNL.

Operated by Union Carbide Corporation under contract W-7405-eng-26.

for