Neutron Activation Measurements of As and Ga Loss during Transient Annealing of GaAs
- PDF / 301,392 Bytes
- 7 Pages / 420.48 x 639 pts Page_size
- 17 Downloads / 244 Views
Vic.
E.M. LAWSON Australian Atomic Energy Commission,
N.S.W.
2232.
Australia
3000.
L.H.R.L.
N.S.W.
2232. Australia
ABSTRACT
Significant dissociation is normally detected under non-optimised transient annealing of GaAs. We have utilised neutron activation to measure As and Ga loss from virgin and implanted material annealed under various transient conditions. Complementary RBS data are reported. In particular, surface dissociation has been measured as a function of pulsed ruby laser power and for several combinations of time and surface temperature using an incoherent light source and a vitreous carbon strip heater. The results indicate that neutron activation analysis offers a powerful tool to identify the conditions required to minimise GaAs dissociation during annealing. For examplq rubý laser pulses of energy 0.29-1.38 J cm- 2 caused 4 As loss of - 90 x 10 cm-
INTRODUCTION
Transient annealing methods for the removal of the radiation damage accompanying ion implantation of semiconductors have been widely reported [1,2]. For silicon, the annealing conditions required for the restoration of single crystallinity and electrical activation of the doped layer are reasonably well established. The situation for GaAs is less satisfactory [3]. Initial enthusiasm for transient annealing, based on the expectation that reduced dissociation would accompany such methods and allow uncapped processing, has been tempered by the understanding that Ga and As loss still occur and that the optimum conditions for successful annealing need to be much more closely defined than is the case for silicon. A method for the accurate measurement of this loss is a desirable prerequisite to the definition of optimum annealing conditions and subsequent electrical characterisation. In this paper we report data which suggests that neutron activation analysis can meet this requirement.
EXPERIMENTAL
Virgin (100) oriented GaAs was primarily used in this work. Some samples were implanted with Te under conditions as detailed in Tables I and II. As and Ga loss associated with three transient methods were measured; pulsed ruby laser, incoherent quartz halogen light and a vitreous carbon strip heater. In each case, As and Ga lost during the anneal were deposited on quartz catcher slides located above and close to, but not touching, the GaAs samples [4]. These slides were then used in the neutron activation and Rutherford backscattering analyses (NAA and RBS) for As and Ga. Mat. Res.
Soc. Symp. Proc. Vol. 13 (1983)
)Elsevier Science Publishing Co.,
Inc.
634
TABLE I Arsenic and gallium loss measured for various laser energies incident on GaAs samples
x 1015 cm-2,
As:Ga atom ratio 8.2 2.5 3.4 1.9 3.3 1.4 2.1 2.0 1.6 1.0 1.8 1.7
0.9 1.6 2.8 17.0 2.7 16.0 13.0 21.0 23.0 86.0 3.1 3.8
7.5 4.0 9.5 33.0 8.9 23.0 27.0 43.0 36.0 90.0 5.6 6.6
0.29 0.36 0.42 0.57 0.58 0.74 0.83 1.08 1.10 1.38 0.64a) 0.70a) a)3
Number Ga 2 atoms cm(x 1014)
Number As atoms cm-2 (x 1014)
Laser energy (joules)
100 keV Te
TABLE II Arsenic and gallium loss measur
Data Loading...
