I. Thurzo et al., PROTON IMPLANTATION-INDUCED DAMAGE TO HEAVILY-DOPED N-GAAS AS ENVISAGED BY CHARGE DEEP-LEVEL TRANSIENT SPECTROSCOPY, Physica status solidi. a, Applied research, 162(2), 1997, pp. 547-557
Heavily doped (1 to 3 x 10(18) cm(-3)) GaAs:Si wafers were exposed to
high doses (10(14), 10(15) cm(-2)) of 150 keV protons. When applying t
he lower dose, the related charge DLTS spectra a of Al/GaAs diodes com
prise a peak of dielectric relaxation (DR) and a trap-limited peak (MG
) due to the emission from a midgap level. The former peak can be attr
ibuted to a linear Debye-type polarization connected with a hopping tr
ansport of electrons within the damaged region. If taking samples from
different positions on the original wafer, the position of tile DR pe
ak on the temperature axis T-m for a selected rate window has been fou
nd to vary when passing from one sample to another, a shift toward hig
her temperature has been accompanied by a corresponding increase in th
e activation energy (0.1 to 0.22 eV). The MG level emission rate is qu
ite sensitive to the electric field intensity in the semiconductor, as
manifested by shifting T-m to lower temperatures via higher reverse b
iases. By contrast, the ultimate dose of 10(15) protons/cm(2) introduc
ed defect levels at 0.07 and 0.31 eV below E-c, respectively, a result
that can be reconciled with previous reports relevant to lower doses.
Concluding, tile DR peak occurrence may be considered as a criterion
for a successful GaAs isolation after tile impact of protons, tile opt
imum dose lying closely to the 10(14) cm(-2) level.