Key points are not available for this paper at this time.
A new dislocation climb model is proposed for compound semiconductors with the zinc blende structure. Within the model a supersaturation of only one type of point defect is needed for the dislocation climb process. Consideration of the forces involved suggest that in compound semiconductors grown by liquid-phase epitaxy, interstitial climb is favored. A trail of vacancy defects is then expected to result from the climb process. Experimental observations in support of this model are presented for the case of the defect structures produced during degradation of Ga1−xAlxAs/GaAs (DH) lasers and GaP green-light-emitting diodes.
Petroff et al. (Fri,) studied this question.